How can users create their own customized databases? - Find any reference to this in the methods/suppl

Note de synthèse détaillée : Vie affective et sexuelle des jeunes - Prévenir les risques et promouvoir les pratiques positives

Cette synthèse reprend les points clés et les idées principales des interventions présentées lors de la matinale sur la vie affective et sexuelle des jeunes, organisée par l'IREPS Occitanie et le DRAPPS, avec la participation du projet Sexpert (INSERM), du SIMS de l'Université de Toulouse, et de l'Académie de Toulouse.

L'objectif général est de "prévenir les risques et promouvoir les pratiques positives" en matière de santé sexuelle chez les jeunes.

1. Contexte général et définitions (Delphine Soulignac - IREPS/DRAPPS)

La matinale s'inscrit dans un cycle "Santé des jeunes" initié en avril 2023.

Elle met en lumière l'importance de la santé sexuelle comme thématique de santé publique centrale.

Définition de la Santé Sexuelle (OMS, 2006) : Bien plus qu'une simple absence de maladie ou de dysfonctionnement, la santé sexuelle est définie comme "un état de bien-être physique, émotionnel, mental et social".

Elle inclut "l'absence de toute forme de coercition, de discrimination et de violence".

Cette approche est "résolument positive" et s'appuie sur les "ressources des personnes, leur capacité à agir" et l'encouragement de "la multiplicité des leviers d'action possibles qu'il soient individuels, relationnels, organisationnels ou encore sociaux".

Politiques Publiques : La thématique est soutenue par la Stratégie Nationale de Santé Sexuelle (2017-2030) et ses déclinaisons régionales, comme le Projet Régional de Santé en Occitanie.

Déterminants de Santé : Promouvoir la santé sexuelle, c'est aussi "agir sur les déterminants de santé", notamment le "développement des compétences psychosociales", souvent considérées comme les "déterminants des déterminants".

Le DRAPPS (Dispositif Régional d'Appui en Prévention et Promotion de la Santé), porté par l'IREPS, vise à "s'informer, se former et échanger" et est fondé sur les besoins remontés par les acteurs de terrain.

L'événement de ce matin participe à la "production et diffusion de connaissance" et à la "diffusion d'outils", en favorisant la rencontre entre équipes de recherche et acteurs de terrain.

2. Le projet Sexpert : Comprendre et agir avec les jeunes (Philippe Martin - INSERM)

Le projet Sexpert, issu d'une thèse et financé par l'ANR, se concentre sur la santé sexuelle et reproductive des 15-24 ans. Il vise à développer, mettre en œuvre et évaluer une action d'éducation à la sexualité et de promotion de la santé sexuelle.

Enjeux et préoccupations des jeunes : La jeunesse est une période marquée par l'autonomisation, la socialisation et des "enjeux de santé sexuelle avec des questionnements, des intérêts, des besoins d'information".

La santé sexuelle "englobe de manière globale les questions de vie affective, sexuelle, relationnelle et reproductive incluant la non-discrimination, le plaisir sexuel ou encore les identités sexuelles et de genre qui sont bien évidemment des sujets centraux dans la vie des jeunes."

Au-delà des risques : Si les IST, les violences ou les grossesses non désirées sont des préoccupations, la santé sexuelle doit aussi être pensée sous l'angle des "comportements favorables de santé", comme l'usage de protections, le consentement, et les "relations positives".

• Approche participative : Les stratégies internationales et nationales prônent une "approche participative" et l'utilisation de "différents leviers éducatifs".

Le projet Sexpert a "questionné finalement les points de vue des jeunes sur leur préoccupation générale de santé sexuelle mais aussi sur leurs propositions concrètes en terme d'action d'éducation qu'on pourrait mettre en place en éducation à la sexualité sur Internet et par les nouveaux outils de communication."

• Perceptions des jeunes :Période complexe : La période adolescente et jeune adulte est "peu simple" et marquée par de "nombreux enjeux de vie qui dépassaient strictement le cadre professionnel".

• Recherche d'information : Les jeunes se tournent vers leurs "proches, leurs familles, leurs amis" et les "professionnels". Internet n'est pas toujours le premier recours, sauf pour des questions immédiates (IST, contraception).

• Difficulté à formuler les questions : La "gêne" est une "vraie barrière à la communication sur des questions et des problématiques de santé sexuelle". Richie (24 ans) exprime : "j'aurais trouvé gênant de matérialiser les questions que je me posais que ce soit sur Internet ou dans la vraie vie".

• Priorités diverses : Certains jeunes ne considèrent pas toujours la vie affective et sexuelle comme une priorité immédiate, comme Els (16 ans) : "Je ne me pose pas forcément de questions, je prends les choses comme elles viennent et puis je ne me prends pas trop la tête avec ça".

• Besoin d'intervenants spécialisés : Les jeunes expriment un besoin d'intervenants spécialisés sur les "dimensions émotionnelles, les IST, les violences ou encore les questions de contraception".

• Importance du climat de confiance : Pour aborder des sujets intimes, "un climat de confiance" est essentiel, ce qui a été observé lors des focus groups. Initialement centrés sur IST/contraception, les discussions se sont élargies à la "déconstruction des normes de genre, la question de plaisir et également des questions liées à la communication" une fois la confiance établie. Exemples de sujets importants pour les jeunes : "Tu es un garçon, les règles ça te concerne aussi", "Le plaisir, c'est être libre", "Le consentement, c'est communiquer".

• Propositions concrètes des jeunes pour l'éducation sexuelle numérique :Contenus éducatifs : Informations "valides, crédibles et bien classées", couvrant la santé sexuelle de manière globale mais aussi des sujets plus généraux liés à la vie des jeunes (santé mentale, autonomisation). Le contenu ne doit pas être "trop institutionnel".

• Plateforme : Doit être "sécurisée, modérée et personnalisée", avec des "espaces de réponse à leurs questions".

• Ressources : Offrir des "ressources géolocalisées" pour cibler les services proches des jeunes.

• Dispositif Sexpert (fin 2023) : Intégrera "des espaces informationnels, des espaces interactifs et des activités ludiques" dans un "espace sécurisé, modéré et également participatif et interactif". Une évaluation mesurera l'amélioration des "connaissances, attitudes positives et comportements favorables de santé sexuelle".

3. Actions du SIMS de l'Université de Toulouse (Benedict Allard & Pascal Faget)

Le Service de Santé des Étudiants (SIMS) de l'Université de Toulouse, obligatoire dans toutes les universités françaises, intervient auprès de plus de 100 000 étudiants.

Consultations de sexologie : Mises en place depuis 2020.

Le Dr Faget, médecin généraliste et sexologue, pose systématiquement la question * des violences ("Est-ce que vous avez été victime ou témoin de violence physique ou sexuelle ?") et * de la sexualité ("Est-ce que vous avez des questions en lien avec la sexualité ?") * à tous les patients, quel que soit le motif de consultation, pour "ouvrir une porte".

Outils pédagogiques : Utilisation de modèles 3D du clitoris, livres ("le petit manuel de sex education") et podcasts pour faciliter la conversation.

Actions de prévention systématiques :Vidéo "La tasse de thé" : Diffusée "systématiquement lors des rentrées" pour expliquer le consentement de manière claire et non violente.

"Si vous comprenez à quel point il est ridicule de forcer quelqu'un à boire du thé quand il n'en veut pas et que vous êtes capable de comprendre quand quelqu'un ne veut pas de thé, alors pourquoi est-il si difficile de le comprendre quand on parle de sexe ? Qu'il s'agisse de thé ou de sexe, le consentement est tout ce qui compte."

Pièce interactive "Le procès de King Kong" : Créée pour les écoles d'ingénieurs pour aborder les violences sexuelles en milieu étudiant et encourager l'intelligence collective.

Forums de santé sexuelle : Organisés sur les campus et dans les "villes d'équilibre" (Albi, Castres, Montauban, Tarbes, Cahors, Rodez), regroupant divers partenaires associatifs et institutionnels, permettant aux étudiants de s'informer de stand en stand.

Consultations en visio : Disponibles pour les étudiants éloignés de Toulouse.

Implication des pairs :Étudiants Relais Santé (ERS) : Environ 20 étudiants sont formés chaque année à la prévention sur toutes les thématiques de santé étudiante, y compris la santé sexuelle. Ils "nous aident à mettre en place des actions qui soient adaptées au public dans le langage du public dans les us et coutumes du public". Ils créent des quizz et des concours de dessins (ex: le tyrannosaure avec le préservatif).

Cafés Sexo : Des moments d'échanges animés par le Dr Faget à la demande des associations étudiantes ou des BDE, abordant des sujets comme le consentement, l'identité de genre, l'orientation sexuelle, la pornographie, le désir et le plaisir. L'objectif est de parler de "sexualité et non de sa sexualité", créant un espace de "participation" et de "respect".

Diffusion de ressources :Podcasts sur Campus FM : Enregistrements de 10 minutes sur des thèmes variés (les premières fois, le consentement, etc.), accessibles en ligne.

Outils visuels : Affiches comme le post d'Hélène Pouille sur le consentement ("être à l'écoute de ce qu'on ressent et demander à l'autre", "une seule hésitation, c'est non") et "la Licorne du genre" pour expliquer le sexe assigné à la naissance, l'identité de genre, l'expression du genre et l'attirance, afin de lutter contre l'homophobie et la transphobie.

Formation des professionnels : Information des collègues (médecins, psychologues, infirmiers, assistants sociaux) sur la légitimité à parler de sexualité avec les jeunes et l'utilisation d'outils, afin que le sujet ne soit "plus un tabou au SIMS de Toulouse".

Service Sanitaire en Santé : Participation à la formation des étudiants en santé (médecine, pharmacie, sage-femme, infirmier) au sein de ce dispositif pour les préparer à ouvrir le dialogue sur la santé sexuelle.

Alignement avec la Charte d'Ottawa : Les actions du SIMS s'inscrivent dans une approche globale de promotion de la santé, visant à "réorienter les services de santé" vers la promotion, "créer des milieux favorables" (gratuité des préservatifs et dépistages), "développer les aptitudes personnelles" (cafés sexo), "renforcer l'action communautaire" (ERS) et contribuer à "établir des politiques publiques saines".

4. Formation des personnels de l'Éducation Nationale (Ingrid Desplat - Infirmière conseillère technique auprès du recteur de l'académie de Toulouse)

L'Académie de Toulouse s'engage dans la formation des personnels à l'éducation à la sexualité, soulignant que cette éducation est un "droit" pour les élèves (trois séances par an et par groupe d'âge homogène, fixées par la circulaire de 2018).

Formateurs : L'académie compte huit formateurs aux profils variés (infirmiers, CPE, professeurs des écoles, professeurs spécialisés), certains ayant des diplômes universitaires en sexologie ou éducation à la sexualité, d'autres une grande expérience de terrain.

Public formé : Environ 200 à 250 personnels volontaires sont formés chaque année. Parcours de formation : Un parcours structuré est mis en place :

Niveau 1 (3 jours) : Abord généraliste, circulaire de 2018, compétences psychosociales, homosexualité, orientation sexuelle, et surtout "les besoins des adolescents" pour comprendre certains comportements (ex: masturbation en classe) et "apporter des éléments de réponse". L'accent est mis sur la posture de l'intervenant.

Niveau 2 (3 jours) : Approfondissement sur des sujets plus complexes comme le genre, qui "suscites le plus d'émotion" et "questionne le plus les adultes formés".

Un an d'écart est recommandé entre les deux niveaux pour permettre une "réflexion" personnelle sur la posture.

Approche pédagogique : L'éducation à la sexualité doit être une "approche globale, positive et bienveillante".

Elle doit intégrer des éléments "psycho-émotionnels", "juridiques et sociaux", et "biologiques" (ce qui marque un changement par rapport aux approches antérieures centrées uniquement sur la biologie et la prévention des IST/grossesses).

Compétences des intervenants : La formation vise à développer des compétences telles que la capacité à "réguler la parole", "savoir rebondir", "prendre en compte les représentations culturelles", "respecter le silence", adopter des "attitudes chaleureuses" et une "posture non moralisatrice", être "à l'aise avec la sexualité" et "s'adresser aux filles et aux garçons".

Protection des jeunes : L'éducation à la sexualité est un moyen de "protéger les jeunes", notamment face aux violences intrafamiliales et à l'inceste. "Si on parle jamais sexualité, le jeune qui en est victime, à qui se confie ?"

Modalités d'accès à la formation : Les personnels volontaires peuvent s'inscrire via le Plan Académique de Formation (PRAF) dans la rubrique "second degré". Des formations d'initiative locale sont également possibles à la demande des chefs d'établissement.

Réticences : Face aux réticences (souvent liées à l'idée que la sexualité est "du domaine de l'intime"), l'Académie rappelle que des règles sociales comme le consentement sexuel doivent être enseignées pour "éviter les violences sexuelles".

Le personnel formé est légitime à intervenir sur ces sujets, dans un cadre réglementaire qui permet une intervention sereine.

5. Thèmes transversaux et idées fortes

Plusieurs idées clés émergent de l'ensemble des interventions :

Approche Positive et Globale : La santé sexuelle est bien plus que la prévention des risques. Elle englobe le bien-être physique, émotionnel, mental et social, le plaisir, les relations positives, le consentement, et les identités de genre. Il s'agit de promouvoir une sexualité "épanouie".

Importance du Consentement : Le consentement est un pilier fondamental de l'éducation sexuelle, expliqué de manière claire et adaptée (ex: vidéo "La tasse de thé", post d'Hélène Pouille). Il doit être "clair, libre, éclairé, actualisé".

Participation et Implication des Jeunes : Essentielle pour des actions pertinentes. Les jeunes sont co-concepteurs des contenus (projet Sexpert) et acteurs de la prévention (Étudiants Relais Santé). Leurs préoccupations et leurs manières de s'informer doivent être prises en compte.

Rôle des Pairs : Les discussions entre pairs et la possibilité de se reconnaître en l'autre sont cruciales pour les jeunes (ex: Jonathan sur la recherche d'autres personnes gay, les cafés sexo).

Lever les Tabous et la Gêne : La difficulté à formuler des questions et la gêne sont des obstacles majeurs. Les professionnels doivent oser poser les questions et créer un climat de confiance pour libérer la parole.

Diversité des Besoins et des Publics : Les actions doivent être adaptées à la "diversité des préoccupations des jeunes", incluant les identités de genre et les orientations sexuelles, et toucher les publics minorisés (personnes en situation de handicap, de maladie chronique).

Accès à l'Information Validée et Crédible : Les jeunes recherchent des informations fiables, non "trop institutionnelles". Les outils numériques (podcasts, plateformes interactives) sont des compléments aux actions de terrain.

Formation et Posture des Professionnels : Cruciales pour une éducation sexuelle efficace. Les intervenants doivent être à l'aise avec le sujet, adopter une posture non moralisatrice, et être formés pour aborder les dimensions psycho-émotionnelles, juridiques et biologiques.

La légitimité de parler de sexualité doit être affirmée.

Partenariats et Rôle des Réseaux : La collaboration entre différentes structures (IREPS, INSERM, universités, associations, Éducation Nationale) est essentielle pour une approche globale et cohérente sur l'ensemble du territoire, y compris dans les zones éloignées.

Les ressources géolocalisées facilitent l'accès aux services.

Au-delà de la Connaissance Scientifique : Les besoins des jeunes vont au-delà de la simple information biologique et incluent des aspects relationnels, émotionnels et de communication.

Note de synthèse : Promotion des bonnes pratiques solaires

Cette note de synthèse a pour objectif de résumer les principaux thèmes, idées et faits importants concernant la promotion des bonnes pratiques solaires, en se basant sur les extraits de la "Matinale du Drapps : Promouvoir des bonnes pratiques solaires - Enjeux et pistes d’actions".

Elle inclura des citations pertinentes issues des sources originales.

Introduction :

Contexte et enjeux de la prévention solaire La Direction Générale de la Santé a souligné en 2020 la nécessité de recommandations sanitaires pour les indices UV, en particulier pour les populations à risque.

Malgré les effets positifs du soleil (notamment la synthèse de vitamine D), une exposition excessive aux UV est nocive et peut entraîner des cancers de la peau et un vieillissement prématuré.

Il est crucial de trouver un équilibre entre une exposition optimale pour la vitamine D et une protection adéquate.

Chiffres clés et impacts sur la santé :

80% des dommages liés aux UV sont provoqués avant 18 ans, les enfants étant plus exposés que les adultes. Un cancer sur trois est un cancer de la peau. Le coût de traitement du mélanome est estimé à 225 millions d'euros par l'INCA.

20 millions de personnes sont aveugles dans le monde à cause de la cataracte, l'exposition solaire étant un facteur de risque majeur.

Les effets sur la santé se manifestent à court, moyen et long terme :

• Court/Moyen terme : Brûlures, coups de soleil, atteintes aux yeux (photokératites, photoconjonctivites), allergies cutanées. Le bronzage lui-même est une réaction de défense de la peau, signalant une agression.
• Long terme : Cataracte, vieillissement prématuré de la peau (photo-vieillissement avec rides et taches), et cancers de la peau.
• Carcinomes : Environ 90% des cancers de la peau, très fréquents (130 000 à 230 000 cas/an, en augmentation). Leur pronostic est généralement bon, mais ils sont liés à une exposition prolongée (par exemple, professionnelle) et nécessitent de nombreuses chirurgies coûteuses.
• Mélanomes : 10% des cancers de la peau, plus rares mais plus graves (environ 18 000 cas/an et 2 000 décès). 83% d'entre eux sont liés à des expositions aux UV, notamment des expositions intenses et intermittentes, souvent durant l'enfance ou les vacances estivales. Ces cancers sont donc majoritairement évitables par une meilleure protection solaire.

Recommandations de protection solaire

Les recommandations actuelles de l'Institut National du Cancer (INCK) sont claires :

• Rechercher l'ombre au maximum, notamment entre 12h et 16h, période de rayonnement maximal.
• Privilégier la protection vestimentaire (t-shirt à manches, chapeau à large bord, lunettes de soleil). Ces moyens sont prioritaires sur la crème solaire.
• La crème solaire est un complément, à utiliser sur les zones découvertes. Il est souligné que "trop souvent la crème solaire est utilisée pour justement enlever les vêtements et pour augmenter son exposition et c'est quelque chose qui peut être assez néfaste."
• Se protéger même par temps nuageux, car le rayonnement UV n'est pas toujours associé à la chaleur.
• Le rayonnement est amplifié par la réverbération (eau, montagne).
• Vigilance maximale avec les enfants : limiter leur exposition au maximum, ne pas exposer un enfant de moins d'un an, et leur apprendre les bons réflexes.
• L'indice UV (disponible sur Météo France) est un outil éducatif. La protection est nécessaire à partir d'un indice 3, et doit être renforcée à partir de 8.
• Évolution de la représentation sociale du bronzage et comportements des Français

Historiquement, le bronzage a connu une inversion de sa signification sociale :

• Jusqu'à la fin du 19e siècle : Signe de prolétariat (travailleurs agricoles), la peau blanche étant associée à la noblesse.
• Début 20e siècle : Le bronzage est de moins en moins associé à la pauvreté.
• Années 30 (premiers congés payés) : Le bronzage devient "synonyme de succès, d'aisances financière, de loisirs."
• Années 80-90 : Apogée de la mode du bronzage, avec l'explosion des cabines UV et une forte médiatisation. Les premières recommandations sanitaires apparaissent à cette période.
• Aujourd'hui : Les dangers sont bien connus, mais la protection reste insuffisante. Le bronzage est toujours associé à la "bonne santé, de bien-être, de bonne mine, de beauté", véhiculant une image positive. Il y a une confusion avec la synthèse de vitamine D, nécessitant en réalité seulement "environ 15 minutes par jour" d'exposition, bien moins que le temps passé à bronzer.

Comportements des Français (Baromètre Cancer 2005, 2010, 2015) : * * Les moyens de protection sont insuffisamment utilisés, notamment la protection vestimentaire, en diminution. * Différenciation des comportements :Les femmes se protègent globalement mieux, utilisant plus l'ombre, la crème solaire et les lunettes. Les hommes privilégient les t-shirts et chapeaux. * Protection insuffisante chez les 15-24 ans (population sous-protégée et surexposée). * Un gradient social existe : les personnes moins diplômées sont globalement moins protégées. * Les phototypes plus foncés se protègent moins, bien que les lunettes de soleil restent indispensables pour tous. * Les résidents du nord ont moins le réflexe de se protéger aux heures critiques. * La protection des enfants s'est améliorée, les parents protégeant mieux leurs enfants qu'eux-mêmes. * Les connaissances augmentent, mais les fausses croyances persistent, notamment sur les conséquences des coups de soleil dans l'enfance. * Actions de prévention solaire : Exemples et initiatives * Plusieurs niveaux d'actions sont mis en œuvre :

1. Niveau International :

• Programme InterSun (OMS) : Vise à informer sur les effets des UV, encourager les pays à réduire les risques et fournir des conseils pratiques. Axes clés : surveillance et prédiction des risques (liés aux changements environnementaux et comportementaux), promotion de l'indice UV comme outil éducatif, et protection solaire des enfants via programmes éducatifs scolaires.
• Programme SunSmart (Australie) : Référence mondiale en prévention du cancer de la peau, pays à l'incidence très élevée. Objectifs : réduire l'incidence, éduquer le public via campagnes (notamment l'emblématique "Slip, Slop, Slap, Seek and Slide"), et mettre en œuvre des politiques de protection (écoles, lieux de travail).
• Impact : Environ 300 000 cancers de la peau évités, 1 000 vies sauvées, diminution de l'incidence chez les plus jeunes. Impact économique positif : "pour 1 dollar dépensé il y a eu un gain de 8,70 dans les 20 prochaines années."

2. Niveau National (France) :

• Campagne nationale annuelle (INCK) : Au printemps, informe la population sur les risques et conseils de prévention (ex: "l'ombre, c'est l'endroit le plus cool de l'été").
• Programmes d'éducation : "Vivre avec le soleil" (Sécurité Solaire), programme probant pour les écoles maternelles et primaires, basé sur une approche expérientielle et intégrable aux programmes scolaires (sciences).

3. Niveau Local/Régional :

• Projets d'urbanisme et de végétalisation : Visent à créer des zones d'ombre, réduisant les îlots de chaleur et l'exposition aux UV (ex: cours oasis dans les écoles). Ces projets peuvent bénéficier d'appels à projet comme "Zéro Expo" de l'INCK.
• Actions ciblées d'éducation à la santé :Étude PRISME (Santé Publique France Occitanie) : Recherche interventionnelle menée auprès de touristes en Occitanie (2019).
• Constat : Forte exposition intentionnelle ("sunbathing"), utilisation prioritaire de la crème solaire au détriment d'autres moyens, coups de soleil fréquents. Le littoral Occitan (200km, fort rayonnement UV, 8 millions de touristes/an) est un contexte à risque.
• Objectifs : Identifier les déterminants de la protection solaire (incluant les déterminants sociaux) et évaluer l'efficacité de deux interventions : une basée sur les messages d'apparence physique (photo-vieillissement) et une sur les messages sanitaires (cancer, yeux).
• Résultats : Protection insuffisante, notamment le port du t-shirt. 3/4 des touristes passent du temps à bronzer.
• Populations sous-protégées : 15-24 ans (et 12-14 ans pour les lunettes), hommes (crème solaire, lunettes), femmes (chapeau, t-shirt), peaux mates/moins sensibles, personnes à faible niveau d'étude, résidents du nord. Les connaissances et les fausses croyances, l'attirance pour le bronzage et le manque d'encouragement de l'entourage sont des déterminants.
• Efficacité des interventions :Intervention sanitaire : Augmente la protection à court terme, agit sur crème solaire et lunettes, touche les peaux sensibles et les niveaux d'études supérieurs.
• Intervention basée sur l'apparence physique : Augmente la protection à court et long terme, diminue l'exposition (évitement 12h-16h), touche les peaux sensibles et les niveaux d'études moins élevés (jusqu'au bac).
• Perspectives : Renouveler les campagnes, cibler les 15-24 ans et femmes exposées intentionnellement, les personnes socialement moins favorisées. Intégrer des messages sur le photo-vieillissement, apporter des connaissances, dénormaliser le bronzage, déployer des interventions dans divers milieux (scolaire, travail, touristique, sportif) et via différents canaux. Encourager les actions locales (zones d'ombre, accès aux moyens de protection, adaptation des horaires d'activités, sensibilisation des professionnels du tourisme, affichage de l'indice UV).
• Semaine des Épidories (Épidor, Institut du Cancer de Montpellier) : Programme éducatif pour enfants de 7 à 11 ans, basé sur le modèle COM-B (Capacité, Opportunité, Motivation, Comportement).
• Contenu : Quatre séances animées par les enseignants, abordant les effets du soleil, la sensibilité de la peau (phototype), les rayons UV et l'indice UV, et les moyens de protection (avec une montre UV). Des prolongements sont proposés en famille.
• Étude de faisabilité : Impact significatif (faible à important) sur le comportement de protection solaire, les attitudes des enfants et leur niveau de connaissance. Des analyses plus poussées et une évaluation à plus grande échelle et à long terme sont envisagées.
• Chiffres (2024) : 200 classes, 4569 élèves du CP au CM2.
• Journée des Épidories : Journée récompense pour 4 classes tirées au sort, avec des ateliers ludiques (professeur dermatologue, jeu de l'ombre, four solaire, carte du monde des protections, jeu de l'oie, constitution de sac à dos pour randonnée).
• Autres actions : Animations sur l'espace ludo-éducatif pour maternelles (3-6 ans), ateliers parents-enfants, formations et conférences pour étudiants et professionnels de la petite enfance.
• Déterminants structurels : L'approche COM-B permet de les considérer, mais les intervenants reconnaissent un impact limité sur ces déterminants, malgré une volonté de les prendre en compte (ex: projets de végétalisation des cours d'école).
• Tournée "Bon été, bon réflexe" (ARS Occitanie, C2DS, Primum non Nocere) : Opération itinérante de sensibilisation aux dangers estivaux depuis 2021.
• Objectifs : Diffuser des messages de prévention sur les dangers estivaux de manière positive, adapter le discours aux publics et aux spécificités territoriales (littoral vs. intérieur des terres), et former des "animateurs relais" pour démultiplier les actions.
• Thématiques des ateliers (participatifs et ludiques) : Prévention solaire, baignade sécurisée, consommation de sucre, moustiques tigres/tiques, sport et santé.
• Moyens : Matériel repérable (arche, tables, barnums, véhicule siglé), espace informatif, goodies (casquettes, chapeaux), deux animateurs dédiés. Des fiches "bons gestes" sont distribuées.
• Bilan (2024) : Plus de 2900 personnes touchées (+25% vs 2023). Les interventions ont lieu sur les marchés, zones de baignade (mers, lacs, rivières), parcs de loisirs, villes touristiques. Les animateurs relais ont démultiplié 59 ateliers, sensibilisant plus de 3500 personnes.
• Atelier soleil : Utilise une "nappe des dangers du soleil" interactive, permettant d'échanger sur les UV, l'indice UV, les effets à court et long terme. Adaptation du discours à l'âge du public. Interroge sur les situations de protection de personnages (incluant différents phototypes). Met en évidence l'exposition lors d'activités quotidiennes (promenade, vélo). Utilise un totem pour rappeler les messages clés (règle ABCDE pour les mélanomes).
• Discussion et perspectives
• Difficulté d'accès aux dermatologues : Soulignée comme un frein majeur au dépistage des mélanomes. Il est suggéré de passer par le médecin traitant et de pratiquer l'auto-diagnostic (règle ABCDE).
• Composition des crèmes solaires : Sujet de débat constant. Bien que certaines compositions aient été controversées (octocrylène), il est noté que les crèmes sont "plus propres aujourd'hui qu'elles ne l'étaient auparavant". La crème solaire reste un complément aux protections physiques, et son choix doit se faire sur sa capacité à protéger (SPF, UVA/UVB, résistance à l'eau) et sa composition environnementale.
• Impact de l'alimentation et autres comportements : Une carence en vitamine D peut provenir de l'alimentation, et l'exposition solaire n'est pas la seule solution.
• Réverbération des UV : Très importante sur le sable (+15%), l'écume de l'eau (+25%), et surtout la neige (+80%), amplifiant significativement l'exposition même à l'ombre ou dans l'eau.
• Rayonnement des écrans : Non abordé, mais pourrait faire l'objet de futures discussions.
• Pollution des crèmes solaires : La question de l'impact environnemental des crèmes solaires sur les milieux aquatiques est soulevée, et il est conseillé de choisir des produits avec des logos spécifiques.
• L'équilibre des moyens de protection : Il est complexe de jongler entre l'ombre, les vêtements et la crème solaire, mais en se tenant aux bases (ombre 12h-16h, vêtements, crème pour les zones découvertes), la protection devient plus simple.

En conclusion, la promotion des bonnes pratiques solaires est un enjeu de santé publique majeur, nécessitant une approche multiforme et adaptée aux différentes populations, environnements et représentations sociales.

Les initiatives présentées, qu'elles soient de recherche ou de terrain, montrent la volonté de cultiver une "culture commune [...] plus saine de nos pratiques solaires".

DOI: 10.1038/s41467-025-61071-y

Resource: RRID:Addgene_12254

Curator: @scibot

SciCrunch record: RRID:Addgene_12254

What is this?

DOI: 10.1038/s41467-025-61071-y

Resource: RRID:Addgene_48138

Curator: @scibot

SciCrunch record: RRID:Addgene_48138

What is this?

comprar, hacer, mover, almacenar y vender. recordar esto seimpre en P&L, es lo más fundamental.

sólo es hasta aquí cuando toda la cadena de suministro se junta para crear una sola cosa. esto está interesante y creo qeu aplica mucho a somos cacao. quizás deberíamos tratarnos como independienetes hasta lograr todos los pasos anteriores.

está interesante ver todos los posibles links que hay en una cadena de suministro. y tiene sentido reconocerlos para así facilitarlos y dar esa visibilidad que se menciona. si hay links no reconocidos (que aún se usan y si se usan son importantes) entonces su no mejora no va a lograr que la cadena de suministro trabaje bien en conjunto.

barreras que hayan dentro de la organización... cómo cuales. por ejemplo la distanci entre fábrica y tienda es una.

confirma lo dicho anterior, pero está interesante, porque el enfoque debe ser primero específico y luego sí ser excelentes en los links que hay entre lo específico y definene el conjutno y sus transacciones.

Damit deine Arbeiten keinen bias in Richtung Kritik bekommt, stelle auch dar, was die Hoffnungen und Versprechungen von KI für Klima, Meere, Biodiversität etc. sind.

Der Artikel von Vinuesa et al. (2020) ist hier die umfassendste Gegenüberstellung von Pro und Kontra KI für die Nachhaltigkeitsziele der UN, die mir bekannt ist.

esto no vale la pena mencionarlo al cierre, demasiado trivial quedarse en este tipo de detalles. Centrarse en hacia donde va la investigación en esta línea y que se podría hacer a futuro.

evitar determinismo tan radical

de dónde sale esto?

acá quedaría mejor ir sumando términos y no repetir

en esta descripción es confuso si la variable es ordinal o continua

faltan antecedentes y argumento para esta hipótesis; alternativamente se puede plantear como exploratoria, pero se recomienda señalarlo explícitamente en el texto.

ok, pero faltan los argumentos centrales que permitan orientar la pregunta.

este párrafo queda algo descontextualizado, tal vez es parte del anterior

Author response:

The following is the authors’ response to the current reviews.

Public Reviews:

Reviewer #1 (Public review):

Weaknesses:

The revised manuscript has gained much clarity and consistency. One previous criticism, however, has in my opinion not been properly addressed. I think the problem boils down to not clearly distinguishing between orthologs and paralogs/homologs. As this problem affects a main conclusion - the prevalence of deletions over insertions in the MTBC - it should be addressed, if not through additional analyses, then at least in the discussion.

Insertions and deletions are now distinguished in the following way: "Accessory regions were further classified as a deletion if present in over 50% of the 192 sub-lineages or an insertion/duplication if present in less than 50% of sub-lineages." The outcome of this classification is suspicious: not a single accessory region was classified as an insertion/duplication. As a check of sanity, I'd expect at least some insertions of IS6110 to show up, which has produced lineage- or sublineage-specific insertions (Roychowdhury et al. 2015, Shitikov et al. 2019). Why, for example, wouldn't IS6110 insertions in the single L8 strain show up here?

In a fully clonal organism, any insertion/duplication will be an insertion/duplication of an existing sequence, and thus produce a paralog. If I'm correctly understanding your methods section, paralogs are systematically excluded in the pangraph analysis. Genomic blocks are summarized at the sublineage levels as follows (l.184 ): "The DNA sequences from genomic blocks present in at least one sub-lineage but completely absent in others were extracted to look for long-term evolution patterns in the pangenome." I presume this is done using blastn, as in other steps of the analysis.

So a sublineage-specific copy of IS6110 would be excluded here, because IS6110 is present somewhere in the genome in all sublineages. However, the appropriate category of comparison, at least for the discussion of genome reduction, is orthology rather than homology: is the same, orthologous copy of IS6110, at the same position in the genome, present or absent in other sublineages? The same considerations apply to potential sublineage-specific duplicates of PE, PPE, and Esx genes. These gene families play important roles in host-pathogen interactions, so I'd argue that the neglect of paralogs is not a finicky detail, but could be of broader biological relevance.

Within the analysis we undertook we did look at paralogous blocks in pangraph, based on copy number per genome. However, this could have been clearer in the text and we will rectify this. We also focussed on duplicated/deleted blocks that were present in two of more sub-lineages. This is noted in figure 4 legend but we will make this clearer in other sections of the manuscript.

We agree that indeed the way paralogs are handled could still be optimised, and that gene duplicates of some genes could have biological importance. The reviewer is suggesting that a synteny analysis between genomes would be best for finding specific regions that are duplicated/deleted within a genome, and if those sections are duplicated/deleted in the same regions of the genome. Since Pangraph does not give such information readily, a larger amount of analysis would be required to confirm such genome position-specific duplications. While this is indeed important, we deem this to be out of scope for the current publication, but will note this as a limitation in the discussion. However, this does not fundamentally change the main conclusions of our analysis.

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

In this paper, Behruznia and colleagues use long-read sequencing data for 335 strains of the Mycobacterium tuberculosis complex to study genome evolution in this clonal bacterial pathogen. They use both a "classical" pangenome approach that looks at the presence and absence of genes, and a more general pangenome graph approach to investigate structural variants also in non-coding regions. The two main results of the study are that (1) the MTBC has a small pangenome with few accessory genes, and that (2) pangenome evolution is driven by deletions in sublineage-specific regions of difference. Combining the gene-based approach with a pangenome graph is innovative, and the former analysis is largely sound apart from a lack of information about the data set used. The graph part, however, requires more work and currently fails to support the second main result. Problems include the omission of important information and the confusing analysis of structural variants in terms of "regions of difference", which unnecessarily introduces reference bias. Overall, I very much like the direction taken in this article, but think that it needs more work: on the one hand by simply telling the reader what exactly was done, on the other by taking advantage of the information contained in the pangenome graph.

Strengths:

The authors put together a large data set of long-read assemblies representing most lineages of the Mycobacterium tuberculosis context, covering a large geographic area. State-of-the-art methods are used to analyze gene presence-absence polymorphisms (Panaroo) and to construct a pangenome graph (PanGraph). Additional analysis steps are performed to address known problems with misannotated or misassembled genes in pangenome analysis.

Weaknesses:

The study does not quite live up to the expectations raised in the introduction. Firstly, while the importance of using a curated data set is emphasized, little information is given about the data set apart from the geographic origin of the samples (Figure 1). A BUSCO analysis is conducted to filter for assembly quality, but no results are reported. It is also not clear whether the authors assembled genomes themselves in the cases where, according to Supplementary Table 1, only the reads were published but not the assemblies. In the end, we simply have to trust that single-contig assemblies based on long-reads are reliable.

We have now added a robust overview of the dataset to supplementary file 1. This is split into 3 sections: public genomes, which were assembled by others; sequenced genomes, which were created and assembled by us; the BUSCO information for all the genomes together. We did not assemble any public data ourselves but retrieved these from elsewhere. We have modified the text to be more specific on this (Line 114 onwards) and the supplementary file is updated to better outline the data.

One issue with long read assemblies could be that high rates of sequencing errors result in artificial indels when coverage is low, which in turn could affect gene annotation and pangenome inference (e.g. Watson & Warr 2019, https://doi.org/10.1038/s41587-018-0004-z). Some of the older long-read data used by the authors could well be problematic (PacBio RSII), but also their own Nanopore assemblies, six of which have a mean coverage below 50 (Wick et al. 2023 recommend 200x for ONT, https://doi.org/ 10.1371/journal.pcbi.1010905). Could the results be affected by such assembly errors? Are there lineages, for example, for which there is an increased proportion of RSII data? Given the large heterogeneity in data quality on the NCBI, I think more information about the reads and the assemblies should be provided.

We have now included an analysis where we looked to see if the sequencing platform influenced the resulting accessory genome size and the pseudogene count. The details of this are included in lines 207-219, and the results are outlined in lines 251-258. Essentially, we found no correlation between sequencing platform and genome characteristics, although less stringent cut-offs did suggest that PacBio SMRT-only assembled genomes may have larger accessory genomes. We do not believe this is enough to influence our larger inferences from this data. It should be noted that complete genomes, in general, give a better indication of pangenome size compared to draft genomes, as has been shown previously (e.g. Marin et al., 2024). Even with some small potential bias, this makes our analysis more robust than any previously published.

In relation to the sequencing depth of our own data, all genomes had coverage above 30x, which Sanderson et al. (2024) has shown to be sufficient for highly accurate sequence recovery. We fixed an issue with the L9 isolate from the previous submission, which resulted in a better BUSCO score and overall quality of that isolate and the overall dataset.

The part of the paper I struggled most with is the pangenome graph analysis and the interpretation of structural variants in terms of "regions of difference". To start with, the method section states that "multiple whole genomes were aligned into a graph using PanGraph" (l.159/160), without stating which genomes were for what reason. From Figure 5 I understand that you included all genomes, and that Figure 6 summarizes the information at the sublineage level. This should be stated clearly, at present the reader has to figure out what was done. It was also not clear to me why the authors focus on the sublineage level: a minority of accessory genes (107 of 506) are "specific to certain lineages or sublineages" (l. 240), so why conclude that the pangenome is "driven by sublineage-specific regions of difference", as the title states? What does "driven by" mean? Instead of cutting the phylogeny arbitrarily at the sublineage level, polymorphisms could be described more generally by their frequencies.

We apologise for the ambiguity in the methodology. All the isolates were inputted to Pangraph to create the pangenome using this method. This is now made clearer in lines 175-177. Standard pangenome statistics (size, genome fluidity, etc.) derived from this Pangraph output are now present in the results section as well (lines 301-320).

We then only looked at regions of difference at the sub-lineage level, meaning we grouped genomes by sub-lineage within the resulting graph and looked for blocks common between isolates of the same sub-lineage but absent from one or more other sub-lineages. We did this from both the Panaroo output and the Pangraph output and then retained only blocks found by both. The results of this are now outlined in lines 351-383.

We focussed on these sub-lineage-specific regions to focus on long-term evolution patterns and not be influenced by single-genome short-term changes. We do not have enough genomes of closely related isolates to truly look at very recent evolution, although the small accessory genome indicates this is not substantial in terms of gene presence/absence. We also did not want potential mis-annotations in a single genome to heavily influence our findings due to the potential issues pointed out by the reviewer above. We state this more clearly in the introduction (lines 106-108), methods (lines 184-186) and results (345-347), and we indicate the limitations in the Discussion, lines 452-457 and 471-473. We also changed the title to ‘shaped’ instead of ‘driven by’.

I fully agree that pangenome graphs are the way to go and that the non-coding part of the genome deserves as much attention as the coding part, as stated in the introduction. Here, however, the analysis of the pangenome graph consists of extracting variants from the graph and blasting them against the reference genome H37Rv in order to identify genes and "regions of difference" (RDs) that are variable. It is not clear what the authors do with structural variants that yield no blast hit against H37Rv. Are they ignored? Are they included as new "regions of difference"? How many of them are there? etc. The key advantage of pangenome graphs is that they allow a reference-free, full representation of genetic variation in a sample. Here reference bias is reintroduced in the first analysis step.

We apologise for the confusion here as indeed the RDs terminology is very MTBC-specific. Current RDs are always relevant to H37Rv, as that is how original discovery of these regions was done and that is how RDScan works. We clarify this in the introduction (lines 67-68). If we found a large sequence polymorphism (e.g. by Pangraph) and searched for known RDs using RDScan, we then assigned a current RD name to this LSP. This uses H37Rv as a reference. If we did not find a known RD, we then classified the LSP as a new RD if it is present in H37Rv, or left the designation as an LSP if not in H37Rv, thus expanding the analysis beyond the H37Rv-centric approaches used by others previously. This is hopefully now made clearer in the methods, lines 187-194.

Along similar lines, I find the interpretation of structural variants in terms of "regions of difference" confusing, and probably many people outside the TB field will do so. For one thing, it is not clear where these RDs and their names come from. Did the authors use an annotation of RDs in the reference genome H37Rv from previously published work (e.g. Bespiatykh et al. 2021)? This is important basic information, its lack makes it difficult to judge the validity of the results. The Bespiatykh et al. study uses a large short-read data (721 strains) set to characterize diversity in RDs and specifically focuses on the sublineage-specific variants. While the authors cite the paper, it would be relevant to compare the results of the two studies in more detail.

We have amended the introduction to explain this terminology better (lines 67-68). Naming of the RDs here came from using RDScan to assign current names to any accessory regions we found and if such a region was not a known RD, we gave it a lineage-related name, allowing for proper RD naming later (lines 187-194). Because the Bespiatyk paper is the basis for RDScan, our work implicitly compares to this throughout, as any RDs we find which were not picked up by RDScan are thus novel compared to that paper.

As far as I understand, "regions of difference" have been used in the tuberculosis field to describe structural variants relative to the reference genome H37Rv. Colloquially, regions present in H37Rv but absent in another strain have been called "deletions". Whether these polymorphisms have indeed originated through deletion or through insertion in H37Rv or its ancestors requires a comparison with additional strains. While the pangenome graph does contain this information, the authors do not attempt to categorize structural variants into insertions and deletions but simply seem to assume that "regions of difference" are deletions. This, as well as the neglect of paralogs in the "classical" pangenome analysis, puts a question mark behind their conclusion that deletion drives pangenome evolution in the MTBC.

We have now amended the analysis to specifically designate a structural variant as a deletion if present in the majority of strains and absent in a minority, or an insertion/duplication if present in a minority and absent in a majority (lines 191-192). We also ran Panaroo without merging paralogs to examine duplication in this output; Pangraph implicitly includes paralogs already.

From all these analyses we did not find any structural variants classed as insertions/duplications and did not find paralogs to be a major feature at the sub-lineage level (lines 377-383). While these features could be important on shorter timescales, we do not have enough closed genomes to confidently state this (limitation outlined in lines 452-457). Therefore, our assertion that deletions are a primary force shaping the long-term evolution in this group still holds.

Reviewer #2 (Public Review):

Summary:

The authors attempted to investigate the pangenome of MTBC by using a selection of state-of-the-art bioinformatic tools to analyse 324 complete and 11 new genomes representing all known lineages and sublineages. The aim of their work was to describe the total diversity of the MTBC and to investigate the driving evolutionary force. By using long read and hybrid approaches for genome assembly, an important attempt was made to understand why the MTBC pangenome size was reported to vary in size by previous reports.

Strengths:

A stand-out feature of this work is the inclusion of non-coding regions as opposed to only coding regions which was a focus of previous papers and analyses which investigated the MTBC pangenome. A unique feature of this work is that it highlights sublineage-specific regions of difference (RDs) that were previously unknown. Another major strength is the utilisation of long-read whole genomes sequences, in combination with short-read sequences when available. It is known that using only short reads for genome assembly has several pitfalls. The parallel approach of utilizing both Panaroo and Pangraph for pangenomic reconstruction illuminated the limitations of both tools while highlighting genomic features identified by both. This is important for any future work and perhaps alludes to the need for more MTBC-specific tools to be developed.

Weaknesses:

The only major weakness was the limited number of isolates from certain lineages and the over-representation others, which was also acknowledged by the authors. However, since the case is made that the MTBC has a closed pangenome, the inclusion of additional genomes would not result in the identification of any new genes. This is a strong statement without an illustration/statistical analysis to support this.

We have included a Heaps law and genome fluidity calculation for each pangenome estimation to demonstrate that the pangenome is closed. This is detailed in lines 225-228 with results shown in lines 274-278 and 316- 320 and Supplementary Figure 2. We agree that more closely related genomes would benefit a future version of this analysis and indicate we indicate the limitations in the Discussion, lines 452-457 and 471-473.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

Abstract

l. 24, "with distinct genomic features". I'm not sure what you are referring to here.

We refer to the differences in accessory genome and related functional profiles but did not want to bloat the abstract with such additional details

Introduction

l. 40, "L1 to L9". A lineage 10 has been described recently: https://doi.org/10.3201/eid3003.231466.

We have updated the text and the reference. Unfortunately, no closed genome for this lineage exists so we have not included it in the analyses. We note this in the results, like 232

l.62/3, "caused by the absence of horizontal gene transfer, plasmids, and recombination". Recombination is not absent in the MTBC, only horizontal gene transfer seems to be, which is what the cited studies show. Indeed a few sentences later homologous recombination is mentioned as a cause of deletions.

This has now been removed from the introduction

l. 67, "within lineage diversity is thought to be mostly driven by SNPs". Again I'm not sure what is meant here with "driven by". Point mutations are probably the most common mutational events, but duplications, insertions, deletions, and gene conversion also occur and can affect large regions and possibly important genes, as shown in a recent preprint (https://doi.org/10.1101/2024.03.08.584093).

We have changed the text to say ‘mostly composed of’. While indeed other SNVs may be contributing, the prevailing thought at lineage level is that SNPs are the primary source of diversity. The linked pre-print is looking at within transmission clusters and this has not been described at the lineage level, which could be done in a future work.

l. 100/1. "that can account for variations in virulence, metabolism, and antibiotic resistance". I would phrase this conservatively since the functional inferences in this study are speculative.

This has now been tempered to be less specific.

Methods

l. 108. That an assembly has a single contig does not mean that it is "closed". Many single contig assemblies on NCBI are reference-guided short-read assemblies, that is, fragments patched together rather than closed assemblies. The same could be true for long-read assemblies.

We specifically chose those listed as closed on NCBI so rely on their checks to ensure this is true. We have stated this better in the paper, line 117.

l. 111. From Supplementary Table 1 understand that for many genomes only the reads were available (no ASM number). Did you assemble these genomes? If yes, how? The assembly method is not indicated in the supplement, contrary to what is written here.

All public genomes were downloaded in their assembled forms from the various sources. This is specified better in the text (line 118) and the supplementary table 1 now lists the accessions for all the assemblies.

l. 113. How many assemblies passed this threshold? And is BUSCO actually useful to assess assembly quality in the MTBC? I assume the dynamic, repetitive gene families that cause problems for assembly and mapping in TB (PE, PPE, ESX) do not figure in the BUSCO list of single-copy orthologs.

All assemblies passed the BUSCO thresholds for high-quality genomes as laid out in Supplementary Table 1. While indeed this does not include multi-copy genes such as PE/PPE we focussed on regions of difference at the sub-lineage level where two or more genomes represent that sub-lineage. This means any assembly issues in a single genome would need to be exactly the same in another of the same sub-lineage to be included in our results. Through this, we aimed to buffer out issues in individual assemblies.

l. 147: Why is Panaroo used with -merge-paralogs? I understand that near-identical genes may not be too interesting from a functional perspective, but if the aim of the analysis is to make broad claims about processes driving genome evolution, paralogs should be considered.

We chose to do so with merged paralogs to look for larger patterns of diversity beyond within-genome paralogs. Additionally, this was required to build the core phylogenetic tree. However, as the reviewer points out, this may bias our findings towards deletions and away from duplications as a primary evolutionary force.

We repeated this without the merged paralogs option and indeed found a larger pangenome, as outlined in Table 1. However, at the sub-lineage level, this did not result in any new presence/absence patterns (lines 381-383). This means the paralogs tended to be in single genomes only. This still indicates that deletions are the primary force in the longer-term evolution of the complex but indeed on shorter spans this may be different.

l. 153: remove the comment in brackets.

This has been fixed and the proper URL placed in instead.

l. 159: which genomes, and why those?

This is now clarified to state all genomes were used for this analysis.

l. 161, "gene blocks": since this analysis is introduced as capturing the non-coding part of the genome, maybe just call them "blocks"?

All references to gene blocks are now changed to genomic blocks to be more specific.

l. 162: what happens with blocks that yield no hits against RvD1, TbD1, and H37Rv?

We named these with lineage-specific names (supplementary table 4) but did not assign RD names specifically.

l. 164: where does the information about the regions of difference come from? How exactly were these regions determined?

Awe have expanded this section to be more specific on the use of RDScan and new naming, along with how we determine if something is an RD/LSP.

Results

l. 185ff: This paragraph gives many details about the geographic origin of the samples, but what I'd expect here is a short description of assembly qualities, for example, the results of the BUSCO analysis, a description of your own Nanopore assemblies, or a small analysis of the number of indels/pseudogenes relative to sequencing technology or coverage (see comment in the public review).

This section (lines 231-258) has been expanded considerably to give a better overview of the dataset and any potential biases. Supplementary table 1 has also been expanded to include more information on each strain.

l. 187, "324 genomes published previously": 322 according to the methods section.

The number has been fixed throughout to the proper total of public genomes (329).

l. 201: define the soft core, shell, and cloud genes.

This is now defined on line 262

l. 228, "defined primarily by RD105 and RD207 deletions": this claim seems to come from the analysis of variable importance (Factoextra), which should be made clear here.

This has been clarified on line 333.

l. 237, "L8, serving as the ancestor of the MTBC": this is incorrect, equivalent to saying that the Chimpanzee is the ancestor of Homo sapiens.

We have changed this to basal to align with how it is described in the original paper.

l. 239, "The accessory genome of the MTBC". It is a bit confusing that the same term, 'accessory genome', is used here for the graph-based analysis, which is presented as a way to look at the non-coding part of the genome.

We have clarified the terminology on line 347 and improved consistency throughout.

l. 240/1, "specific to certain lineages and sublineages". What exactly do you mean by "specific" to? Present only in members of a certain lineage/sublineage? In all members of a certain lineage/sublineage? Maybe an additional panel in Figure 5, showing examples of lineage- and sublineage-specific variants, would help the reader grasp this key concept.

We have clarified this on line 349 and the legend of what is now figure 4.

l. 241/2, "82 lineage and sublineage-specific genomic regions ranging from 270 bp to 9.8 kb". Were "gene blocks" filtered for a minimum size, or why are there no variants smaller than 270 bp? A short description of all the blocks identified in the graph could be informative (their sizes, frequencies ...).

Yes, a minimum of 250bp was set for the blocks to only look at larger polymorphisms. This is clarified on line 177 and 304.

A second point: It is not entirely clear to me what Figure 6 is showing. Are you showing here a single representative strain per sublineage? Or have you somehow summarized the regions of difference shown in Figure 5 at the sublineage level? What is the tree on the left? This should be made clear in the legend and maybe also in the methods/results.

In figure 4 (which was figure 6), because each RD is common to all members of the same sub-lineage, we have placed a single branch for each sub-lineage. This is has been clarified in the legend.

l. 254, "this gene was classified as being in the core genome": why should a partially deleted gene not be in the core genome?

You are correct, we have removed that statement.

l. 258/259, "The Pangraph alignment approach identified partial gene deletion and non-coding regions of the DNA that were impacted by genomic deletion". I do not understand how you classify a structural variant identified in the pangenome graph as a deletion or an insertion.

This has been clarified as relative to H37Rv, as this is standard practice for RDs and general evolutionary analyses in MTBC, as outlined above.

l. 262/263 , "the accessory genome of the MTBC is small and is acquired vertically from a common ancestor within the lineage". If deletion is the main process involved here, "acquired" seems a bit strange.

We agree and changed the header to better reflect the discussion on mis-annotation issues

Figure 1: Good to know, but not directly relevant for the rest of the paper. Maybe move it to the supplement?

This has been moved to Supplementary figure 1

Figure 2: the y-axis is labeled 'Variable genome size', but from the text and the legend I figure it should be 'Number of accessory genes'?

This has been changed to ‘accessory genes’ in Figure 1 (which was figure 2 in previous version).

Figure 4: too small.

We will endeavour to ensure this is as large as possible in the final version.

Discussion

l. 271, "MTBC accessory genome is ... acquired vertically". See above.

Changed, as outlined above.

l. 292, "appeared to be fragmented genes caused by misassemblies". Is there a way to distinguish "true" pseudogenes from misassemblies? This could be a relevant issue for low-coverage long-read assemblies (see public review).

Not that we are currently aware of, but we do know other groups which are working on this issue.

l. 300/1, "the whole-genome approach could capture higher genetic variations". Do you mean the graph approach? I'm not sure that comparing the two approaches here makes sense, as they serve different purposes. A pangenome graph is a summary of all genetic variation, while the purpose of Panaroo is to study gene absence/presence. So by definition, the graph should capture more genetic variation.

This statement was specifically to state that much genetic variation in MTBC is outside the coding genes and so traditional “pangenome’ analyses are actually not looking at the full genomic variation.

l. 302/3, "this method identified non-coding regions of the genome that were affected by genomic deletions". See the comments above regarding deletions versus insertions. I'd say this method identifies coding and non-coding regions that were affected by genomic deletions and insertions.

We have undertaken additional analyses to be sure these are likely deletions, as outlined above.

l. 305: what are "lineage-independent deletions"?

We labelled these as convergent evolution, now clarified on line 443.

l. 329: How is RD105 "caused" by the insertion of IS6110? I did not find RD105 mentioned in the Alonso et al. paper. Similarly below, l. 331, how is RD207 "linked" to IS6110?

The RD105 connection was misattributed as IS6110 insertion is related to RD152, not RD105. This has now been removed.

RD207 is linked to IS6110 as its deletion is due to recombination between two such elements. This is now clarified on line 486.

l. 345, "the growth advantage gene group": not quite sure what this is.

We have fixed this on line 499 to state they are genes which confer growth advantages.

l. 373ff: The role of genetic drift in the evolution of the MTBC is an open question, other studies have come to different conclusions than Hershberg et al. (this has been recently reviewed: https://doi.org/10.24072/pcjournal.322).

We have outlined this debate better in lines 527-531

l. 375/6, "Gene loss, driven by genetic drift, is likely to be a key contributor to the observed genetic diversity within the MTBC." This sentence would need some elaboration to be intelligible. How does genetic drift drive gene loss?

We have removed this.

l. 395/6, "... predominantly driven by genome reduction. This observation underlines the importance of genomic deletions in the evolution of the MTBC." See comments above regarding deletions. I'm not convinced that your study really shows this, as it completely ignores paralogs and the processes counteracting reductive genome evolution: duplication and gene amplification.

As outlined above, we have undertaken additional analyses to more strongly support this statement.

l. 399, "the accessory genome of MTBC is a product of gene deletions, which can be classified into lineage-specific and independent deletions". Again, I'm not sure what is meant by lineage-independent deletions.

We have better defined this in the text, line 443, to be related to convergent evolution.

Reviewer #2 (Recommendations For The Authors):

Suggestions for improved or additional experiments, data, or analyses.

In lines 120-121, it is mentioned that TB-profiler v4.4.2 was used for lineage classification, but this version was released in February 2023. As I understand there have been some changes (inclusion/exclusion) of certain lineage markers. Would it not be appropriate to repeat lineage classification with a more recent version? This would of course require extensive re-analysis, so could the lineage marker database perhaps also be cited.

We have rerun all the genomes through TB-Profiler v6.5 and updated the text to state this; the exact database used is also now stated.

Could the authors perhaps include the sequencing summary or quality of the nanopore sequences? The L9 (Mtb8) sample had a relatively lower depth and resulted in two contigs. Yet one contig was the initial inclusion criteria. It is unclear whether these samples were excluded from some of the analyses. Mtb6 also has relatively low coverage. Was the sequencing quality adequate to accurately identify all the lineage markers, in particular those with a lower depth of coverage? Could a hybrid approach be an inexpensive way to polish these assemblies?

We reanalysed the L9 sample and, with some better cleaning, got it to a single contig with better depth and overall score. This is outlined in the Supplementary table 1 sheets. While depth is average, it is still above the recommended 30x, which is needed for good sequence recovery (Sanderson et al., 2024). We did indeed recover all lineage markers from these assemblies.

Recommendations for improving the writing and presentation.

The introduction is well-written and recent MTBC pangenomic studies have been incorporated, but I am curious as to why this paper was not referred to: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6922483/ I believe this was the first attempt to study the pangenome, albeit with a different research question. Nearly all previous analyses largely focused on utilizing the pangenome to investigate transmission.

Indeed this study did look at a pangenome of sorts, but specifically SNPs and not genes or regions. Since the latter is the main basis for pangenome work these days, we chose not to include this paper.

Minor corrections to the text and figures.

In line 129, it is explained that DNA was extracted to be suitable for PacBio sequencing, but ONT sequencing was used for the 11 new sequences. Is this a minor oversight or do the authors feel that DNA extracted for PacBio would be suitable for ONT sequencing? It is a fair assumption.

We apologise, this is a long-read extraction approach and not specific to PacBio. We have amended the text to state this.

In line 153, this should be removed: (Conor, could you please add the script to your GitHub page?).

This has been fixed now.

Note de synthèse détaillée : Collaboration et accompagnement des élèves en milieu éducatif

Introduction

Ce document de synthèse présente les thèmes principaux et les idées essentielles extraites de la table ronde "Journée d’étude CPE 2025".

Il met en lumière le rôle crucial de la collaboration entre les différents acteurs de l'éducation nationale et les partenaires extérieurs pour la réussite, l'épanouissement et l'insertion professionnelle des élèves, en particulier ceux à besoins éducatifs particuliers.

Les intervenants partagent leurs expériences et leurs perspectives sur les défis, les leviers et les recommandations pour favoriser un travail collaboratif efficace.

1. Missions des professionnels de l'éducation et de l'accompagnement Plusieurs intervenants ont détaillé leurs missions, soulignant la diversité des rôles et la complexité de l'écosystème éducatif :

2. Laura Renault (CPE en collège REP) : Les missions du CPE sont multiples et incluent l'organisation de l'espace et du temps de la vie scolaire, la gestion du service (recrutement et management des AED), la contribution à la réussite et au bien-être des élèves (accueil, accompagnement des projets personnels, orientation), et la participation à la politique de l'établissement comme conseiller du chef d'établissement.

Elle souligne l'importance de "l'organisation du service donc on pense au recrutement des assistants d'éducation aux objectifs aussi du service et à l'ambition qu'on veut pour notre service".

• Ludovique Brel (Directeur de l'IME Lérange) : En tant que directeur d'un IME accueillant des jeunes avec déficience intellectuelle et TSA, ses missions englobent la gestion RH, budgétaire, le projet d'établissement, et les relations avec les partenaires et les tutelles.

Il met en avant l'ouverture récente d'une unité d'autorégulation collège pour les jeunes avec troubles du neurodéveloppement (TND), qui vise à travailler sur les problématiques émotionnelles en collaboration avec les professionnels du collège.

Il précise que ces jeunes "sont dans leur classe d'âge leur classe de niveau c'est-à-dire qu'on parle pas de jeunes qui ont des particularités cognitives ou quoi que ce soit ce sont vraiment des jeunes qui sont en capacité classique".

• Nathalie Yayaten (Proviseure des lycées des métiers de l'automobile) : Ses fonctions consistent à administrer l'établissement, représenter l'État, et "lancer, impulser, animer le projet d'établissement qui amène une forme de gouvernance et de grandes de grands objectifs sur la conduite de la politique éducative et pédagogique de l'établissement".

Son expérience en collège REP et en cité éducative lui confère une perspective large sur les collaborations.

• Lise MERIS (Coordinatrice ULIS Lycée Professionnel et service école inclusive du rectorat) : Ses missions se concentrent sur la sécurisation du parcours des jeunes en situation de handicap, leur accompagnement vers le monde professionnel, et le développement de partenariats internes et externes (monde de l'entreprise, médico-social). Au rectorat, elle soutient la scolarité réussie et l'insertion professionnelle des jeunes, notamment lors des transitions entre le premier et le second degré.

Elle vise à "rendre l'élève acteur de son projet professionnel de l'accompagner du mieux possible au regard de ses points forts et de ses points d'effort".

• Stéphane Copman (Éducateur à l'IME Mermose) : Il accompagne 14 jeunes déficients intellectuels au collège, veillant à leur sécurité et à la réussite de leur projet de socialisation.

Contrairement aux enseignants, il n'a "aucune obligation de résultat" et s'adapte aux capacités de chaque jeune pour éviter l'échec scolaire.

Sa mission inclut également "de veiller à la circulation de l'information au sein du collège et en lien également avec mon établissement et les familles".

2. Le travail collaboratif : un impératif pour la réussite de l'élève

Les intervenants s'accordent sur l'importance fondamentale du travail collaboratif, non seulement au sein de l'établissement mais aussi avec l'environnement extérieur.

L'EPLE comme entité locale et ouverte : Nathalie Yayaten insiste sur le "L" de "Local" dans EPLE (Établissement Public Local d'Enseignement).

L'établissement ne doit pas se fermer à son environnement et doit s'adapter aux problématiques territoriales.

"L'établissement doit évoluer par rapport à son public doit s'adapter et doit aussi être très en en écoute et en réponse des problématiques territoriales sur lequel il est ancré".

Le CPE au carrefour des relations : Le CPE est perçu comme un pivot, "au carrefour des relations entre les différents collaborateurs au sein d'établissement", mais aussi "au carrefour de toutes ses relations à l'extérieur de l'établissement", ce qui "fait la richesse du métier".

Cohérence des discours et des actions : Pour que le jeune grandisse bien, il est essentiel que tous les acteurs (école, famille, clubs de sport, centres sociaux, professionnels paramédicaux) se connaissent et surtout "parlent d'une même voix ou où est en tout cas un discours très cohérent".

Les discours contradictoires, par exemple entre collège et centre social sur les devoirs, fragilisent le jeune et sa famille.

Exemples de collaborations réussies :

Dispositif "Devoirs faits" avec les centres sociaux : Permet aux enseignants d'expliquer des méthodes d'apprentissage et montre aux familles une collaboration "main dans la main".

Partenariat avec le PRE (Programme de Réussite Éducative) : Accueil des élèves exclus temporairement, offrant un lien et un accompagnement aux familles en difficulté.

Unités d'autorégulation (UAR) : Projet co-porté par le SMS (Service Médico-Social) et le collège, avec des décisions prises d'un commun accord entre les directions.

La formation commune de toutes les équipes (enseignants, CPE, infirmière) est un point clé pour créer une "culture commune" et dépasser les tensions liées aux différentes portes d'entrée professionnelles (enseigner vs prendre le jeune dans sa globalité).

Collaboration pour l'orientation professionnelle : Les entreprises sont invitées à présenter les métiers, offrant aux jeunes d'autres perspectives et sources d'information.

3. Les freins et les leviers du travail collaboratif

Malgré la volonté de collaborer, des difficultés peuvent apparaître. Les intervenants mettent l'accent sur les leviers pour les surmonter.

• Frustrations des CPE et des enseignants : Le CPE doit gérer des situations de décrochage ou de conflit dont il n'est pas à l'origine. Les enseignants sont confrontés à des publics hétérogènes et se sentent parfois démunis face aux "élèves à besoins éducatifs particuliers", pouvant s'épuiser.
• Lever les freins par la collaboration : La collaboration apporte un "autre regard" et permet de débloquer des situations complexes. Un éducateur parlant à un professeur, ou l'inverse, peut faire "s'ouvrir les chakras" et envisager d'autres approches.
• Le CPE comme levier central : Laura Renault affirme que "vous [les CPE] êtes le levier". Grâce à leur position transversale et leur liberté d'action, les CPE peuvent "créer ces collaborations qui vont permettre à un tel et un tel de poser un autre regard sur la situation".
• Importance des espaces d'échange formels : Les cellules de veille et les GPDs (Groupes de Prévention du Décrochage Scolaire) sont des moments clés pour le "regard croisé et d'échange et de partage d'information". Il est crucial de transmettre les informations nécessaires en respectant "l'éthique de travail" et la confiance du jeune et de sa famille.
• Patience et détermination : Ludovique Brel insiste sur ces deux qualités pour mener à bien les projets collaboratifs, reconnaissant que les choses ne se règlent pas du jour au lendemain. L'exemple de l'autorégulation montre une augmentation significative de la participation des enseignants à la formation, signe d'une "réelle volonté" de travailler ensemble.
• Le jeune au centre : L'objectif ultime est que "le jeune soit l'élément central". L'équipe SMS apporte un soutien par sa connaissance des problématiques spécifiques du jeune (ex: TND), permettant de faire converger la compétence pédagogique et la compétence médico-sociale pour le bien-être de l'élève.

4. Recommandations pour un travail collaboratif efficace

Plusieurs précautions et recommandations sont formulées pour favoriser des collaborations fructueuses :

• Chacun à sa place, dans le respect des missions : Lise MERIS souligne l'importance que "chacun reste à sa place" et "sache quelle est sa place". Il faut connaître les missions et les compétences de chacun pour savoir "qui fait quoi, à quel moment" et ainsi respecter les champs d'expertise.
• Non-jugement et bienveillance : Adopter une posture de non-jugement et de bienveillance envers le travail de l'autre est essentiel pour construire une collaboration "très horizontale".
• Postures d'ouverture avec les parents : Accueillir les parents "dans cette posture d'ouverture de collaboration" permet de changer la dynamique, de les faire baisser la garde et de construire des solutions avec eux, et non à leur place ou contre eux. "Le parent comprend que vous êtes en train de vous triturer l'esprit avec vos collègues à essayer de trouver une solution pour son jeune".
• Anticipation : La collaboration "s'anticipe". Il faut initier des démarches en amont pour être capable de travailler ensemble lorsque des difficultés surgissent.
• Valoriser les réussites : Ne pas se concentrer uniquement sur les situations de tension, mais aussi reconnaître les moments où "ça se passe bien" et les jeunes qui vont bien. "Il faut aussi se donner à notre travail une légitimité se dire que oui il y a quand même des choses qui vont et il faut aussi voilà se viscéralement en tous les cas y croire".
• Ne pas perdre la confiance du jeune et de la famille : La centralité du jeune et de sa confiance est primordiale. Il est important d'expliquer, de rassurer les familles et les jeunes.
• Vulgariser le langage : Face à l'abondance de sigles et de jargons professionnels, il est recommandé de "vulgariser au maximum" pour rendre les informations accessibles à tous, y compris aux familles et aux jeunes.

Un "glossaire" des sigles peut être utile.

Conclusion La table ronde a clairement démontré que la réussite de tous les élèves, notamment ceux à besoins éducatifs particuliers, repose sur une collaboration forte et structurée.

Le CPE, par sa position charnière, est un acteur clé de cette dynamique, en favorisant les échanges internes et externes.

Les défis liés aux cultures professionnelles différentes peuvent être surmontés par la patience, la détermination, la formation commune, le respect mutuel des compétences et une posture d'ouverture centrée sur le bien-être et la réussite du jeune.

Note de synthèse : Protection de l’Enfance et Scolarité(s) - L'école comme facteur et espace de protection et de résilience

Cette journée d'étude, troisième édition d'un événement à succès, met en lumière le rôle crucial de l'école dans la protection et la résilience des élèves bénéficiant de mesures de protection de l'enfance.

Elle vise à outiller les acteurs de terrain et à faciliter les rencontres entre partenaires clés (Rectorat de l'Académie de Lille, NPJJ de Roubaix, Départements du Nord et du Pas-de-Calais).

La thématique explore comment l'environnement scolaire peut devenir un espace de sécurité et de développement pour ces jeunes.

1. La Résilience : Concepts Clés et Application Scolaire

La résilience, loin d'être la force des super-héros, est « la force des faibles » (Martine Lanbeale), un processus qui survient après un trauma, non pas comme une résistance immédiate, mais comme un "néodéveloppement" après une sidération. Un élément fondamental est qu' « on ne résilie jamais tout seul ».

La présence de "tuteurs de résilience" est indispensable. Ces tuteurs, souvent des professionnels de l'éducation ou du social, ne se désignent jamais eux-mêmes ; c'est la personne en résilience qui leur attribue ce rôle, souvent a posteriori.

Une parole, un regard, une considération peuvent changer la vie d'un enfant, même si le tuteur n'en a jamais conscience.

Facteurs de protection de la résilience (selon Boris Cyrulnik) :

Attachement sécurisant : Un environnement où l'enfant se sent entouré et aimé, même si cet attachement initial est détruit par un trauma (comme l'histoire de Boris Cyrulnik lui-même).

Rébellion : La capacité à ne pas rester dans la sidération, à ne pas "faire une carrière de victime", bien que cela dépende de l'environnement proposé.

Capacité à raconter son histoire : L'expression du vécu, et notamment la lutte contre le sentiment de honte, qui est décrit comme "le poison de l'âme" et qui empêche de parler.

La spécificité de la résilience scolaire : L'école est un espace de protection parce qu'elle offre un cadre clair et sécurisant, avec des règles et des processus où "au moins ils savent comment fonctionner à l'école". Malgré le "cartable plus lourd" que portent ces enfants, l'école peut être un lieu où ils s'investissent "pour faire plaisir à leur enseignants en échange d'un regard positif".

L'école contribue au développement de compétences sociales, d'analyse, de compréhension, de lecture, de récit et d'expression artistique, toutes essentielles à la résilience.

Exemples concrets de résilience scolaire :

Le chauffeur de car : Un homme, enfant battu, a été "sauvé" par le sourire quotidien de son chauffeur de car.

Ce chauffeur, sans le savoir, a été un tuteur de résilience en donnant l'alerte un jour où l'enfant était plus gravement blessé.

Cette histoire souligne l'impact inattendu et souvent méconnu des tuteurs de résilience.

Philippe Maurice : Condamné à mort jeune, il a repris ses études en prison pour "ne pas devenir fou".

Il est passé d'un CAP à un doctorat en histoire médiévale, devenant finalement enseignant à l'ENS. L'école a été un moyen de sortir de la haine et de retrouver la maîtrise.

Estelle et Sophie : Ces deux jeunes filles, victimes d'abus, ont utilisé l'école différemment pour gérer leur trauma.

Estelle s'est réfugiée dans le travail scolaire pour s'isoler, tandis que Sophie a utilisé le théâtre à l'école pour rejouer son vécu. Ces cas illustrent la non-linéarité et la fluctuation des processus de résilience scolaire.

2. La Résilience Assistée : Postures Professionnelles et Ressources Locales

Devenir un potentiel tuteur de résilience est avant tout une question de posture plutôt que d'actions concrètes spécifiques. Il s'agit de « créer un lien sécure, une alliance ».

Des phrases simples comme "Tu sais si j'ai l'impression que c'est un peu compliqué pour toi, tu sais que je suis là pour toi" peuvent être vitales.

L'objectif n'est pas de juger, mais de considérer l'enfant dans la singularité de son histoire.

Ressources et Partenariats Locaux :

Points Accueil Écoute Jeunes (PAEJ) : Structures nationales (250 PAEJ), présentes notamment dans le Nord et le Pas-de-Calais, offrent un soutien psychologique et social, avec une approche de libre adhésion et une mission de prévention.

Maisons des Adolescents (MDA) : Structures départementales qui assurent l'accueil, l'accompagnement et l'orientation des adolescents, avec une équipe pluridisciplinaire (psychologues, infirmières, éducateurs, etc.). Elles mettent en place des expérimentations avec l'Éducation Nationale pour éviter de "perdre des jeunes en route".

Prévention Spécialisée (PS) : Intervient auprès des jeunes et familles en difficulté ou en rupture pour prévenir la marginalisation et faciliter l'insertion. Ses principes incluent le mandat non nominatif (approche territoriale), la libre adhésion et une approche globale des jeunes. Les équipes travaillent dans la rue pour maintenir le lien avec les familles.

Agents de Liaison Quartier Établissement (ALSCE) : Postes créés dans les années 90, déployés dans 74 collèges du Nord, ces agents font le lien entre l'établissement scolaire, le quartier et les familles. Ils sont "un pied à l'intérieur de l'établissement scolaire" et "un pied aussi dans le quartier", offrant une expertise locale et facilitant le contact avec les jeunes et leurs familles.

• Détection des signaux faibles et orientation : Les enseignants et professionnels doivent être attentifs aux changements de comportement récurrents, au désinvestissement scolaire (matériel manquant, devoirs non faits), au repli, à l'isolement, et à l'absentéisme perlé.

Le regard croisé des différents professionnels est crucial pour détecter précocement les situations. Les élèves eux-mêmes peuvent être des "messagers", informant les adultes des difficultés de leurs camarades.

Qui peut saisir ces structures ? ** * PAJ et MDA : Le jeune lui-même, sa famille, ses proches ou tout professionnel gravitant autour de l'adolescent peuvent les saisir. * Prévention Spécialisée : N'importe qui peut interpeller ces structures, y compris les directeurs d'école primaire. Leur présence dans le quartier et leur capacité à aller à domicile facilitent la prise de contact. * Services Sociaux du Département (Maison Nord Solidarité dans le Nord, MDS - Maison Départementale de la Solidarité - dans le Pas-de-Calais) : Représentent un interlocuteur essentiel, notamment en milieu rural où les ressources peuvent être plus limitées. Il est crucial de connaître ces partenaires "avant d'avoir un problème". * Service Social en Faveur des Élèves (ASDSN) : Les assistantes sociales scolaires sont des personnes ressources dans les établissements pour orienter vers les structures adaptées. Une expérimentation de "brigade du premier degré" est lancée dans le Nord pour mieux répertorier les partenaires et soutenir les équipes pédagogiques en primaire. * PMI (Protection Maternelle et Infantile)** : Service départemental qui intervient dans les écoles pour les bilans de santé et peut être un interlocuteur précieux pour les enseignants.

3. Conclusion

La journée d'étude souligne l'importance vitale d'un écosystème de soutien autour des enfants en situation de protection, avec l'école au cœur de ce dispositif.

La résilience est un processus complexe qui dépend fortement de la qualité des relations humaines et de la présence de tuteurs bienveillants, souvent sans le savoir.

La connaissance et la collaboration entre les professionnels de l'éducation et du social, ainsi que la capacité à identifier et orienter les élèves en difficulté, sont des piliers fondamentaux pour offrir à chaque enfant les conditions d'un "néodéveloppement" réussi.

Compte Rendu Détaillé : Le Conseiller Principal d'Éducation Accompagnant

Ce document explore le rôle du Conseiller Principal d'Éducation (CPE) en tant qu'accompagnant au sein de l'établissement scolaire, en insistant sur la dimension individuelle et collective de son travail.

Il s'appuie sur une approche socio-historique, praxéologique et didactique pour problématiser la question de l'accompagnement et de la collaboration.

1. Le Contexte Socio-Historique et la Problématique du Rôle du CPE

Le rôle du CPE en France, particulièrement dans le secondaire, est marqué par une dichotomie historique entre "l'instruction et l'éducation".

Les enseignants sont perçus comme détenteurs de l'instruction, tandis que les CPE, bien qu'ayant une mission éducative et pédagogique, souffrent d'un manque de légitimité à cet égard depuis leur création en 1970.

Cette tension persiste, et les CPE se battent toujours pour être reconnus comme pédagogues.

Dans un contexte de démocratisation scolaire ayant renforcé les inégalités, les CPE sont fortement attendus pour "accueillir [et] suivre les élèves de manière à pallier les insuffisances du système [et] les difficultés que connaissent les élèves".

Leur défi éducatif fondamental est d'"aider l'élève à se construire en tant que personne".

Cependant, ils sont également vus comme des garants de l'action pédagogique et de la régulation des désordres scolaires, un rôle souvent associé à la discipline plutôt qu'à l'éducation et à l'accompagnement.

La critique de l'autorité au XXe siècle, ou son "érosion" (Prerra), complexifie davantage le rôle du CPE, qui doit faire face à une confusion entre autoritarisme et autorité éducative nécessaire à la médiation entre l'individu et le collectif.

La "division du travail éducatif" (Tardif et Levasseur, Payet) est également une problématique majeure, le CPE se situant entre le cognitif/enseignement (enseignants) et le comportemental (personnels spécialisés comme les AESH).

Il est à la fois celui qui "aide les enseignants à enseigner, il aide les élèves à apprendre et en même temps il est très attendu pour réguler tous les désordres scolaires".

2. Les Missions du CPE et la Place de la Dimension Pédagogique

Les missions de la vie scolaire visent à placer les adolescents dans les meilleures conditions de vie individuelles et collectives, de réussite scolaire (depuis la circulaire de 2015) et d'épanouissement personnel.

La vie scolaire, selon Jean-Pierre Aubin, n'est pas la vie hors la classe, mais "l'ensemble des événements de la vie collective et puis aussi tout ce qui concerne la normalisation de la vie collective".

Le CPE, bien qu'en lien fort avec le pôle éducatif (problèmes sociaux et affectifs), est aussi connecté au pôle instruction.

La dimension pédagogique de son rôle est "très importante" car il est celui qui gère les problèmes personnels ou collectifs des élèves et qui consacre un temps significatif à l'accueil de la parole des élèves, notamment lors des entretiens.

Or, cette dimension est "assez peu connue encore" par de nombreux enseignants, sauf si elle fait partie de la culture de l'établissement.

Les textes officiels (circulaire de 2015, référentiel de 2013) légitiment le travail en équipe des CPE :

Promotion de la citoyenneté participative : Accompagnement des élèves dans leurs "apprentissages sociaux et citoyens". Suivi des élèves : Construction du parcours de l'élève sur les plans "pédagogiques et éducatifs", impliquant un travail individuel (entretiens) en lien avec les familles et le personnel d'éducation nationale. Organisation de la vie scolaire : L'équipe de vie scolaire, incluant les Assistants d'Éducation (AED), contribue également à l'accompagnement individuel et collectif.

3. L'Accompagnement comme Activité Pilier du Métier de CPE

L'entretien est l'"activité pilier du métier" de CPE, constituant la part la plus importante de leur temps et de leur activité, comme l'a montré une étude du CREC en 2007.

Cependant, les CPE ressentent souvent une "frustration" due à la "tyrannie de l'urgence" (Sylvie Condette) ou au "tourbillon professionnel" (Christine Fauquenois Simonet), qui les pousse à privilégier la gestion du collectif au détriment du suivi individuel et de la collaboration.

L'accompagnement est un "travail de l'ombre qui n'est pas quantifiable et surtout pas dans l'immédiat".

Les résultats ne sont pas instantanés, ce qui contraste avec les "attentes [d'un] coup de baguette magique" de la part des collègues ou de la direction.

C'est un processus de "coconstruction, un cheminement" sur le long terme, où le CPE doit être une "personne de confiance" pour l'élève.

L'accompagnement est défini comme une "activité" (et non un simple dispositif institutionnel) qui sous-tend des pratiques et une posture.

Il a une "dimension constructiviste", visant à aider l'élève à "construire son propre parcours" et à "se décider à agir, à s'assumer", sans agir à sa place.

Cela implique un "étayage" de "nature dialogique", une "médiation sociale" pour que l'élève se projette dans sa vie scolaire et sociale.

C'est une "relation disymétrique" mais "éducative" et "temporaire".

Le CPE doit concilier cette relation de confiance individuelle avec le travail collectif, car il "ne peut pas accompagner seul". L'information doit être échangée et le travail collaboratif mis en place.

L'accompagnement implique une "éthique de l'accompagnement" et de la "responsabilité" (méta-responsabilité), où le CPE, par son "autorité éducative" (au sens d'«augmenter l'autre»), permet à l'élève de grandir et de devenir autonome.

Ce rôle demande un "principe de retenue" et le "principe du doute", évitant de donner des solutions toutes faites.

Selon les travaux de Frédérique Lherbé Serini, l'accompagnant adopte trois figures posturales entremêlées :

Le Compagnon : À l'écoute, soucieux du bien-être de l'autre, proche de l'éthique du caresse, attentif à ce qui va bien et ce qui ne va pas. L'Accompagnateur : Dans la coconstruction, le cheminement avec l'élève. Le Guide : Intervient pour dire "stop" lorsque l'élève se met en danger ou que la construction n'est plus possible.

Les CPE ne sont pas uniquement des guides, mais adoptent majoritairement les postures de compagnon et d'accompagnateur.

4. Le Travail Collaboratif du CPE au sein de l'Établissement

Le travail coopératif est "prescrit" et est une "compétence partagée" entre enseignants et CPE. Cependant, sa réalité sur le terrain est peu documentée. Des facteurs de contexte peuvent défavoriser la collaboration :

L'intégration croissante des CPE dans "l'équipe de direction élargie" peut renforcer leur image administrative et créer des obstacles à la coopération avec les enseignants.

L'"intensification [et la] complexification" du métier d'enseignant, la remise en question des savoirs et la judiciarisation de l'institution peuvent générer des crispations entre les corps enseignants et les CPE.

Malgré cela, la collaboration CPE-enseignant a toujours été une constante dans les textes, évoluant de l'"échange d'information" (1982) à des "savoirs partagés [et une] collaboration étroite" (2015).

Les "collectifs de travail" se forment soit par "initiative libre" (besoins partagés face à des difficultés), soit par "assignation" (appartenance à une équipe prédéfinie, ex: équipe pédagogique d'un niveau).

Des "groupes de projets spécifiques" (devoir fait, café des parents) sont aussi des espaces de collaboration.

Le contexte de l'établissement (proximité des bureaux, culture d'établissement) et l'institutionnalisation de certains temps (heure de vie de classe) sont des facteurs clés pour la facilitation de la collaboration.

Le "pouvoir d'agir" du CPE, par ses initiatives, est important.

Un véritable collectif de travail, au-delà d'un simple groupe de professionnels, est caractérisé par :

L'adhésion à des "valeurs partagées".

L'élaboration commune de "règles de fonctionnement" et d'objectifs.

L'engagement volontaire et autonome des membres, conduisant à la création de "groupes de veille éducative" par exemple, qui s'inscrivent progressivement dans une "culture d'établissement".

Les pratiques de collaboration varient :

"Pratiques de collaboration minimum" : Échanges informels dans les couloirs ou à la salle des professeurs.

Le CPE doit organiser ces informations éparses.

Le "rayonnement d'un CPE" dans un établissement est lié à sa capacité à aller vers les autres et à initier ces rencontres.

5. Modalités de Travail du CPE et Collaboration

La "approche multidimensionnelle" est une compétence clé du CPE, lui permettant d'avoir une "vision globale et complexe" de la vie de l'élève (loisirs, difficultés, etc.) et de croiser les regards pour le bien-être et l'épanouissement de chacun.

Les CPE travaillent en étroite collaboration avec :

Les personnels d'éducation (AED) : Considérés comme une "mine d'or" d'informations grâce à leur proximité avec les élèves. Les réunions de service régulières sont cruciales pour la remontée de ces informations.

Les enseignants : Plus de 80% des CPE échangent régulièrement sur les entretiens avec les élèves, montrant que l'accompagnement ne se fait pas en vase clos.

La collaboration est souvent initiée par le CPE, ou par sollicitation d'autres membres de la communauté éducative, y compris les élèves eux-mêmes.

Le partage de savoirs sur les élèves et leurs difficultés permet au CPE d'"entrer dans la classe" différemment.

L'"éthique des collectifs de travail" est primordiale :

La plupart des CPE informent l'élève avant ou après le partage d'informations avec d'autres professionnels (professeur principal, infirmière, médecin scolaire), et un tiers le fait systématiquement avant.

C'est un "principe" pour certains CPE pour ne pas rompre la relation de confiance.

Le partage d'informations se fait avec "tact", en ne dévoilant pas tout le contenu de l'entretien mais uniquement ce qui est pertinent pour le suivi.

La question du "secret" (professionnel ou non) et du partage est récurrente et nécessite un travail important en formation initiale.

Les "entretiens tripartites" (avec l'élève, le CPE et un autre collègue) sont une pratique importante, permettant de confirmer une "cohérence" et une "cohésion d'équipe".

C'est une "prise de risque", car le résultat peut être inattendu, mais elle est jugée nécessaire car les CPE "ne peuvent pas travailler seul en entretien avec l'élève".

Le CPE sait aussi "relayer" l'accompagnement à une autre personne lorsque ses propres limites d'éducateur sont atteintes. C'est une compétence difficile mais essentielle, nécessitant de ne pas donner l'impression à l'élève d'être "abandonné".

Le CPE est souvent un "médiateur" entre les élèves et les adultes de l'établissement (enseignants, infirmière, etc.) lorsque la relation est coupée.

En Réseau d'Éducation Prioritaire (REP+), il existe une "culture commune de la manière d'accompagner les élèves" entre CPE, directeurs d'école et directeurs de SEGPA, avec une forte volonté de travailler sur la continuité école-collège concernant les savoirs sociaux et citoyens, et un travail commun sur les "rituels pour faire adhérer les élèves à la règle".

Conclusion

Le processus d'accompagnement par le CPE est caractérisé par des "effets très forts d'inter et de co et de rétroaction" entre les élèves et les CPE, qui s'étendent en "cercles concentriques" avec la formation de collectifs autour de la relation duale.

Ce travail en équipe nécessite une "vraie éthique", notamment le "tact", car les élèves sont sensibles à un "sentiment de loyauté".

Enfin, l'"expérience collective nourrit la reconnaissance individuelle", soulignant l'importance de développer des expériences de travail collaboratif dès la formation initiale des CPE et des enseignants.

Nassim - 30/06 - Il y a un intérêt mais son objectif est de pour l'instant rester indépedant capitalistiquement parlant, même quand même envoyé, j'attend un retour.

Author response:

The following is the authors’ response to the original reviews

Life Assessment

The authors use a synthetic approach to introduce synaptic ribbon proteins into HEK cells and analyze the ability of the resulting assemblies to cluster calcium channels at the active zone. The use of this ground-up approach is valuable as it establishes a system to study molecular interactions at the active zone. The work relies on a solid combination of super-resolution microscopy and electrophysiology, but would benefit from: (i) additional ultrastructural analysis to establish ribbon formation (in the absence of which the claim of these being synthetic ribbons might not be supported; (ii) data quantification (to confirm colocalization of different proteins); (iii) stronger validation of impact on Ca2+ function; (iv) in depth discussion of problems derived from the use of an over-expression approach.

We thank the editors and the reviewers for the constructive comments and appreciation of our work. Please find a detailed point-to-point response below. In response to the critique received, we have now (i) included an ultrastructural analysis of the SyRibbons using correlative light microscopy and cryo-electron tomography, (ii) performed quantifications to confirm the colocalisation of the various proteins, (iii) discussed and carefully rephrased our interpretation of the role of the ribbon in modulating Ca²⁺ channel function and (iv) discussed concerns regarding the use of an overexpression system. 

Public Reviews:

Reviewer #1 (Public Review):

We would like to thank the reviewer for the comments and advice to further improve our manuscript. We have completely overhauled the manuscript taking the suggestions of the reviewer into account.

(1) Are these truly "synthetic ribbons". The ribbon synapse is traditionally defined by its morphology at the EM level. To what extent these structures recapitulate ribbons is not shown. It has been previously shown that Ribeye forms aggregates on its own. Do these structures look any more ribbonlike than ribeye aggregates in the absence of its binding partners?

We thank reviewer 1 for their constructive feedback and critique of the work. 

We agree that traditionally, ribbon synapses have always been defined by the distinct morphology observed at the EM level. However, since the discovery of the core-components of ribbons (RIBEYE and Piccolino) confocal and super-resolution imaging of immunofluorescently labelled ribbons have gained importance for analysing ribbon synapses. A correspondence of RIBEYE immunofluorescent structures at the active zone to electron microscopy observations of ribbons has been established in numerous studies (Wong et al, 2014; Michanski et al, 2019, 2023; Maxeiner et al, 2016; Jean et al, 2018) even though direct correlative approaches have yet to be performed to our knowledge. We have now analysed SyRibbons using cryo-correlative electron-light microscopy. We observe that GFPpositive RIBEYE spots corresponded well with electron-dense structures, as is characteristic for synaptic ribbons (Robertis & Franchi, 1956; Smith & Sjöstrand, 1961; Matthews & Fuchs, 2010). We could also observe SyRibbons within 100 nm of the plasma membrane (see Fig. 3). We have now added this qualitative ultrastructural analysis of SyRibbons in the main manuscript (lines 272 - 294, Fig. 3 and Supplementary Fig. 3).

(2) No new biology is discovered here. The clustering of channels is accomplished by taking advantage of previously described interactions between RBP2, Ca channels and bassoon. The localization of Ribeye to bassoon takes advantage of a previously described interaction between the two. Even the membrane localization of the complexes required the introduction of a membraneanchoring motif.

We respectfully disagree with the overall assessment. Our study emphasizes the synthetic establishment of protein assemblies that mimic key aspects of ribbon-type active zone, defining minimum molecular requirements. Numerous previous studies have described the role of the synaptic ribbon in organising the spatial arrangement of Ca²⁺ channels, regulating their abundance and possibly also modulating their physiological properties (Maxeiner et al, 2016; Frank et al, 2010; Jean et al, 2018; Wong et al, 2014; Grabner & Moser, 2021; Lv et al, 2016). We would like to highlight that there remain major gaps between existing in vitro and in vivo data; for instance, no evidence for direct or indirect interactions between Ca²⁺ channels and RIBEYE have been demonstrated so far. While we do indeed take advantage of previously known interactions between RIBEYE and Bassoon (tom Dieck et al, 2005); between Bassoon, RBP2 and P/Q-type Ca²⁺ channels (Davydova et al, 2014); and between RBP2 and Ltype Ca²⁺ channels (Hibino et al, 2002), our study tries to bridge these gaps by establishing the indirect link between the synaptic ribbon (RIBEYE) and L-type CaV1.3 Ca²⁺ channels using a bottom-up approach, which has previously just been speculative. Our data shows how even in a synapse-naive heterologous expression system, ribbon synapse components assemble Ca²⁺ channel clusters and even show a partial localisation of Ca²⁺ signal. Moreover, we argue that the established reconstitution approach provides other interesting insights such as laying ground-up evidence supporting the anchoring of the synaptic ribbon by Bassoon. Finally, we expect that the established system will serve future studies aimed at deciphering the role of putative CaV1.3 or CaV1.4 interacting proteins in regulating Ca²⁺ channels of ribbon synapses by providing a more realistic Ca²⁺ channel assembly that has been available in heterologous expression systems used so far. In response to the reviewers comment we have augmented the discussion accordingly.  

(3) The only thing ribbon-specific about these "syn-ribbons" is the expression of ribeye and ribeye does not seem to participate in the localization of other proteins in these complexes. Bsn, Cav1.3 and RBP2 can be found in other neurons.

The synaptic ribbon made of RIBEYE is the key molecular difference in the molecular AZ ultrastructure of ribbon synapses in the eye and the ear. We hypothesize the ribbon to act as a superscaffold that enables AZ with large Ca²⁺ channel assemblies and readily releasable pools. In further support of this hypothesis, the present study on synthetic ribbons shows that CaV1.3 Ca²⁺ channel clusters are larger in the presence of SyRibbons compared to SyRibbon-less CaV1.3 Ca²⁺ channel clusters in tetratransfected HEK cells (Ca²⁺ channels, RBP, membrane-anchored Bassoon, and RIBEYE, Fig. 6). In response to the reviewers comment we now added an analysis of triple-transfected HEK cells (Ca²⁺ channels, RBP, membrane-anchored Bassoon), in which CaV1.3 Ca²⁺ channel clusters again are significantly smaller than at the SyRibbons and indistinguishable from SyRibbon-less CaV1.3 Ca²⁺ channel clusters (Fig. 6E, F).

(4) As the authors point out, RBP2 is not necessary for some Ca channel clustering in hair cells, yet seems to be essential for clustering to bassoon here.

Here we would like to clarify that RBP2 is indeed important in inner hair cells for promoting a larger complement of CaV1.3 and RBP2 KO mice show smaller CaV1.3 channel clusters and reduced whole cell and single-AZ Ca²⁺ influx amplitudes (Krinner et al, 2017). However, a key point of difference we emphasize on is that even though CaV1.3 clusters appeared smaller, they did not appear broken or fragmented as they do upon genetic perturbation of Bassoon (Frank et al, 2010), RIBEYE (Jean et al, 2018) or Piccolino (Michanski et al, 2023). This highlights how there may be a hierarchy in the spatial assembly of CaV1.3 channels at the inner hair cell ribbon synapse (also described in the discussion section “insights into presynaptic Ca²⁺ channel clustering and function”) with proteins like RBP2 regulating abundance of CaV1.3 channels at the synapse and organising them into smaller clusters – what we have termed as “nanoclustering”; while Bassoon and RIBEYE may serve as super-scaffolds further organizing these CaV1.3 nanoclusters into “microclusters”. Observations of fragmented Ca²⁺ channel clusters and broader spread of Ca²⁺ signal seen upon Ca²⁺ imaging in RIBEYE and Bassoon mutants (Jean et al, 2018; Frank et al, 2010; Neef et al, 2018), and the absence of such a phenotype in RBP2 mutants (Krinner et al, 2017) may be explained by such a differential role of these proteins in organising Ca²⁺ channel spatial assembly. The data of the present study on reconstituted ribbon containing AZs are in line with these observations in inner hair cells: RBP2 appears important to tether Ca²⁺ channels to Bassoon and these AZ-like assemblies are organised to their full extent by the presence of RIBEYE. As mentioned in the response to point 3 of the reviewer, we have now further strengthened this point by adding the analysis of SyRibbon-less CaV1.3 Ca²⁺ channel clusters in tripletransfected HEK cells (Ca²⁺ channels, RBP, membrane-anchored Bassoon, Fig. 6E, F). Moreover, we have revised the discussion accordingly. 

(5) The difference in Ca imaging between SyRibbons and other locations is extremely subtle.

We agree with the reviewer on the modest increase in Ca²⁺ signal amplitude seen in the presence of  SyRibbons and provide the following reasoning for this observation: 

(i) It is plausible that due to the overexpression approach, Ca²⁺ channels (along with RBP2 and PalmBassoon) still show considerably high expression throughout the membrane even in regions where SyRibbons are not localised. Indeed, this is evident in the images shown in the lower panel in Fig. 6B, where Ca²⁺ channel immunofluorescence is distributed across the plasma membrane with larger clusters formed underneath SyRibbons (for an opposing scenario, please see the cell in Fig. 6B upper panel with very localised CaV1.3 distribution underneath SyRibbons). This would of course diminish the difference in the Ca²⁺ signals between membrane regions with and without SyRibbons. We note that while the contrast is greater for native synapses, extrasynaptic Ca²⁺ channels have been described in numerous studies alone for hair cells (Roberts et al, 1990; Brandt, 2005; Zampini et al, 2010; Wong et al, 2014).

(ii) Nevertheless, we do not expect a remarkably big difference in Ca²⁺ influx due to the presence of SyRibbons in the first place. Ribbon-less AZs in inner hair cells of RIBEYE KO mice showed normal Ca²⁺ current amplitudes at the whole-cell and the single-AZ level (Jean et al, 2018). However, it was the spatial spread of the Ca2+ signal at the single-AZ level which appeared to be broader and more diffuse in these mutants in the absence of the ribbon, in contrast to the more confined Ca2+ hotspots seen in the wild-type controls. 

So, in agreement with these published observations – it appears that presence of SyRibbons helps in spatially confining the Ca²⁺ signal by super scaffolding nanoclusters into microclusters (see also our response to points 3 and 4 of the reviewer): this is evident from seeing some spatial confinement of Ca²⁺ signals near SyRibbons on top of the diffuse Ca²⁺ signal across the rest of the membrane as a result of overexpression in HEK cells. 

We have now carefully rephrased our interpretation throughout the manuscript and added further explanation in the discussion section.   

(6) The effect of the expression of palm-Bsn, RBP2 and the combination of the two on Ca-current is ambiguous. It appears that while the combination is larger than the control, it probably isn't significantly different from either of the other two alone (Fig 5). Moreover, expression of Ribeye + the other two showed no effect on Ca current (Figure 7). Also, why is the IV curve right shifted in Figure 7 vs Figure 5?

We agree with the reviewer that co-expression of palm-Bassoon and RBP2 seems to augment Ca²⁺ currents, while the additional expression of RIBEYE results in no change when compared to wild-type controls. We currently do not have an explanation for this observation and would refrain from making any claims without concrete evidence. As the reviewer also correctly pointed out, while the expression of the combination of palm-Bassoon and RBP2 raises Ca²⁺ currents, current amplitudes are not significantly different when compared to the individual expression of the two proteins (P > 0.05, Kruskal-Wallis test). In light of this, we have now carefully rephrased our MS. Moreover, we would like to thank reviewer 1 for pointing out the right shift in the IV curve which was due to an error in the values plotted on the x-axis. This has been corrected in the updated version of the manuscript. 

(7) While some of the IHC is quantified, some of it is simply shown as single images. EV2, EV3 and Figure 4a in particular (4b looks convincing enough on its own, but could also benefit from a larger sample size and quantification)

We have now added quantifications for the colocalisations of the various transfection combinations depicted in the above-mentioned figures collectively in Supplementary Figure 7 and added the corresponding results and methods accordingly. 

Reviewer #2 (Public Review):

We would like to thank the reviewer for the comments and advice to further improve our manuscript.

(1) Relies on over-expression, which almost certainly diminishes the experimentally-measured parameters (e.g. pre-synapse clustering, localization of Ca2+ entry).

We acknowledge this limitation highlighted by the reviewer arising from the use of an overexpression system and have carefully rephrased our interpretation and discussed possible caveats in the discussion section. 

(2) Are HEK cells the best model? HEK cells secrete substances and have a studied-endocytitic pathway, but they do not create neurosecretory vesicles. Why didn't the authors try to reconstitute a ribbon synapse in a cell that makes neurosecretory vesicles like a PC12 cell?

This is a valid point for discussion that we also had here extensively. We indeed did consider pheochromocytoma cells (PC12 cells) for reconstitution of ribbon-type AZs and also performed initial experiments with these in the initial stages of the project. PC12 cells offer the advantage of providing synaptic-like microvesicles and also endogenously express several components of the presynaptic machinery such as Bassoon, RIM2, ELKS etc (Inoue et al, 2006) such that overexpression of exogenous AZ proteins would have to be limited to RIBEYE only. 

However, a major drawback of PC12 cells as a model is the complex molecular background of these cells. We have also briefly described this in the discussion section (line 615 – 619). Naïve, undifferentiated PC12 cells show highly heterogeneous expression of various CaV channel types (Janigro et al, 1989); however, CaV1.3, the predominant type in ribbon synapses of the ear, does not seem to express in these cells (Liu et al, 1996). Furthermore, our attempts at performing immunostainings against CaV1.3 and at overexpressing CaV1.3 in PC12 cells did not prove successful and we decided on refraining from pursuing this further (data not shown). 

On the contrary, HEK293 cells being “synapse-naïve” provide the advantage of serving as a “blank canvas” for performing such reconstitutions, e.g. they lack voltage-gated Ca²⁺ channels and multidomain proteins of the active zone. Moreover, an important practical aspect for our choice was the availability of the HEK293 cell line with stable (and inducible) expression of the CaV1.3 Ca²⁺ channel complex. Finally, as described in lines 613 – 614 of the discussion section, even though HEK293 cells lack SVs and the molecular machinery required for their release, our work paves way for future studies which could employ delivery of SV machinery via co-expression (Park et al, 2021) which could then be analyzed by the correlative light and electron microscopy workflow we worked out and added during revision. 

(3) Related to 1 and 2: the Ca channel localization observed is significant but not so striking given the presence of Cav protein and measurements of Ca2+ influx distributed across the membrane. Presumably, this is the result of overexpression and an absence of pathways for pre-synaptic targeting of Ca channels. But, still, it was surprising that Ca channel localization was so diffuse. I suppose that the authors tried to reduce the effect of over-expression by using an inducible Cav1.3? Even so, the accessory subunits were constitutively over-expressed.

We agree with the reviewer on the modest increase in Ca²⁺ signal amplitude seen in the presence of SyRibbons. Yes, we employed inducible expression of the CaV1.3a subunit and tried to reduce the effect of overexpression by testing different induction times. However, we did not observe any major differences in expression and observed large variability in CaV1.3 expression across cells irrespective of induction duration. At all time points, there were cells with diffuse CaV1.3 localisation also in regions without SyRibbons which likely reduced the contrast of the Ca²⁺ signal we observe. We provide the following reasoning for this observation: 

(i) It is plausible that due to the overexpression approach, Ca²⁺ channels (along with RBP2 and PalmBassoon) still show considerable expression along the membrane also in regions where SyRibbons are not localised. Indeed, this is evident in the images shown in the lower panel in Fig. 6B where Ca²⁺ channel immunofluorescence is distributed across the plasma membrane with larger clusters formed underneath SyRibbons. This would of course diminish the difference in the Ca²⁺ signals between membrane regions with and without SyRibbons. We note that while the contrast is greater for native synapses, extrasynaptic Ca²⁺ channels have been described in numerous studies alone for hair cells (Roberts et al, 1990; Brandt, 2005; Zampini et al, 2010; Wong et al, 2014).

(ii) Nevertheless, we do not expect a striking difference in Ca²⁺ influx amplitude due to the presence of SyRibbons in the first place. Ribbon-less AZs in inner hair cells of RIBEYE KO mice showed normal Ca²⁺ current amplitudes at the whole-cell and the single-AZ level (Jean et al, 2018). Instead, it was the spatial spread of the Ca²⁺ signal at the single-AZ level which appeared to be broader and more diffuse in these mutants in the absence of the ribbon, in contrast to the more confined Ca²⁺ hotspots seen in the wildtype controls. 

So, in agreement with these published observations – it appears that presence of SyRibbons helps in spatially confining the Ca²⁺ signal by super scaffolding nanoclusters into microclusters: this is evident from seeing some spatial confinement of Ca²⁺ signals near SyRibbons on top of the diffuse Ca²⁺ signal across the rest of the membrane as a result of overexpression in HEK cells. 

We have now carefully rephrased our interpretation throughout the manuscript and added further explanation in the discussion section.   

Reviewer #3 (Public Review):

We would like to thank the reviewer for the comments and advice to further improve our manuscript.

(1) The results obtained in a heterologous system (HEK293 cells) need to be interpreted with caution. They will importantly speed the generation of models and hypothesis that will, however, require in vivo validation.

We acknowledge this limitation highlighted by Reviewer 3 arising from the use of an overexpression system and have carefully rephrased our interpretation and discussed possible caveats in the discussion section. We employed inducible expression of the CaV1.3a subunit and tried to reduce the effect of overexpression by testing different induction times. However, we did not observe any major differences in expression and observed large variability in CaV1.3 expression across cells irrespective of induction duration. At all time points, there were cells with diffuse CaV1.3 localisation, even in regions without SyRibbons and this could reduce the contrast of the Ca²⁺ signal we observe. We provide the following reasoning for this observation: 

(i) It is plausible that due to the overexpression approach, Ca²⁺ channels (along with RBP2 and PalmBassoon) still show considerable expression along the membrane also in regions where SyRibbons are not localised. Indeed, this is evident in the images shown in the lower panel in Fig. 6B where Ca²⁺ channel immunofluorescence is distributed across the plasma membrane with larger clusters formed underneath SyRibbons. This would of course diminish the difference in the Ca²⁺ signals between membrane regions with and without SyRibbons. We note that while the contrast is greater for native synapses, extrasynaptic Ca²⁺ channels have been described in numerous studies alone for hair cells (Roberts et al, 1990; Brandt, 2005; Zampini et al, 2010; Wong et al, 2014).

(ii) Nevertheless, we do not expect a striking difference in Ca²⁺ influx amplitude due to the presence of SyRibbons in the first place. Ribbon-less AZs in inner hair cells of RIBEYE KO mice showed normal Ca²⁺ current amplitudes at the whole-cell and the single-AZ level (Jean et al, 2018). Instead, it was the spatial spread of the Ca²⁺ signal at the single-AZ level which appeared to be broader and more diffuse in these mutants in the absence of the ribbon, in contrast to the more confined Ca²⁺ hotspots seen in the wildtype controls. 

So, in agreement with these published observations – it appears that presence of SyRibbons helps in spatially confining the Ca²⁺ signal by super scaffolding nanoclusters into microclusters: this is evident from seeing some spatial confinement of Ca²⁺ signals near SyRibbons on top of the diffuse Ca²⁺ signal across the rest of the membrane as a result of overexpression in HEK cells. 

(2) The authors analyzed the distribution of RIBEYE clusters in different membrane compartments and correctly conclude that RIBEYE clusters are not trapped in any of those compartments, but it is soluble instead. The authors, however, did not carry out a similar analysis for Palm-Bassoon. It is therefore unknown if Palm-Bassoon binds to other membrane compartments besides the plasma membrane. That could occur because in non-neuronal cells GAP43 has been described to be in internal membrane compartments. This should be investigated to document the existence of ectopic internal Synribbons beyond the plasma membrane because it might have implications for interpreting functional data in case Ca2+-channels become part of those internal Synribbons.

In response to this valid concern, we have now included the suggested experiment in Supplementary Figure 1. We investigated the subcellular localisation of Palm-Bassoon and did not find Palm-Bassoon puncta to colocalise with ER, Golgi, or lysosomal markers, suggesting against a possible binding with membrane compartments inside the cell. We have added the following sentence in the results section, line 145 : “Palm-Bassoon does not appear to localize in the ER, Golgi apparatus or lysosomes (Supplementary Fig 1 D, E and F).”

(3) The co-expression of RBP2 and Palm-Bassoon induces a rather minor but significant increase in Ca2+-currents (Figure 5). Such an increase does not occur upon expression of (1) Palm-Bassoon alone, (2) RBP2 alone or (3) RIBEYE alone (Figure 5). Intriguingly, the concomitant expression of PalmBassoon, RBP2 and RIBEYE does not translate into an increase of Ca2+-currents either (Figure 7).

We agree with the reviewer that co-expression of palm-Bassoon and RBP2 seems to augment Ca²⁺ currents, while the additional expression of RIBEYE results in no change when compared to wild-type controls. We currently do not have an explanation for this observation and would refrain from making any claims without concrete evidence. We also highlight that, while the expression of the combination of palm-Bassoon and RBP2 raises Ca²⁺ currents, current amplitudes are not significantly different when compared to the individual expression of the two proteins (P > 0.05, Kruskal-Wallis test). In light of this, we have now carefully rephrased our MS. 

(4) The authors claim that Ca2+-imaging reveals increased CA2+-signal intensity at synthetic ribbontype AZs. That claim is a subject of concern because the increase is rather small and it does not correlate with an increase in Ca2+-currents.

Thanks for the comment: please see our response to your first comment and the lines 585 – 610 in the discussion section.

Recommendations for the authors:  

Reviewer #2 (Recommendations For The Authors):

(1) The authors should have a better discussion of problems derived from over-expression.

Done. Please see above. 

(2) Ideally, the authors would repeat the study using a secretory cell line, but this is of course not possible. The idea could be brought forth, though.

As described above in our response to the public review of reviewer 2, we have discussed this idea in the discussion section (refer to lines 615 – 619), emphasizing on both the advantages and the limitations of using a secretory cell line (e.g. PC12 cells) instead of HEK293 cells as a model for performing such reconstitutions. 

Reviewer #3 (Recommendations For The Authors):

(1) There are several figures in which colocalization between different proteins is studied only displaying images but without any quantitative data. This should be corrected by providing such a quantitative analysis.

We have now added quantifications for the colocalisations of the various transfection combinations depicted in the above-mentioned figures collectively in Supplementary Figure 7 and added the corresponding results and methods accordingly. 

(2) The little increase in Ca2+-currents and Ca2+-influx associated to the clustering of Ca2+-channels to Synribbons is a concern. The authors should discuss if such a minor increase (found only when Palm-Bassoon and RBP2 ae co-expressed) would have or not physiological consequences in an actual synapse. They might discuss the comparison of those results and compare with results obtained in genetically modified mice in which Ca2+-currents are affected upon the removal of AZs proteins. On the other hand, they should explain why Ca2+-currents do not increase when the Synribbons are formed by RIBEYE, Palm-Bassoon and RBP2.

Done. Please see above. 

(3) The description of the patch-clamp experiments should be enriched by including representative currents. Did the authors measure tail currents?

We would like to thank the reviewer for the valuable suggestion and have now added representative currents to the figures (see Supplementary Figure 5B). We agree with the reviewer on the importance of further characterizing the Ca²⁺ currents in the presence and absence of SyRibbons by analysis of tail currents for counting the number of Ca²⁺ channels by non-stationary fluctuation analysis but consider this to be out of scope of the current study and an objective for future studies. 

(4) The current displayed in Figure 7 E should be explained better.

Previous studies have shown that Ca²⁺-binding proteins (CaBPs) compete with Calmodulin to reduce Ca²⁺-dependent inactivation (CDI) and promote sustained Ca²⁺ influx in Inner Hair Cells (Cui et al, 2007; Picher et al, 2017). In the absence of CaBPs, CaV1.3-mediated Ca²⁺ currents show more rapid CDI as in the case here upon heterologous expression in HEK cells ((Koschak et al, 2001), see also Picher et al 2017 where co-expression of CaBP2 with CaV1.3 inhibits CDI in HEK293 cells). The inactivation kinetics of CaV1.3 are also regulated by the subunit composition (Cui et al, 2007) along with the modulation via interaction partners and given the reconstitution here we do not find the currents very surprising. 

(5) Is the difference in Ca2+-influx still significantly higher upon the removal of the maximum value measured in positive Syribbons spots (Figure 7, panel K)?

Yes, on removing the maximum value, the P value increases from 0.01 to 0.03 but remains statistically significant. 

(6) In summary, although the approach pioneered by the authors is exciting and provides relevant results, there is a major concern regarding the interpretation of the modulation of Ca2+ channels.

We have now carefully rephrased our interpretation on the modulation of Ca²⁺ channels.  

References

Brandt A (2005) Few CaV1.3 Channels Regulate the Exocytosis of a Synaptic Vesicle at the Hair Cell Ribbon Synapse. Journal of Neuroscience 25: 11577–11585

Cui G, Meyer AC, Calin-Jageman I, Neef J, Haeseleer F, Moser T & Lee A (2007) Ca2+-binding proteins tune Ca2+-feedback to Cav1. 3 channels in mouse auditory hair cells. The Journal of Physiology 585: 791–803

Davydova D, Marini C, King C, Klueva J, Bischof F, Romorini S, Montenegro-Venegas C, Heine M, Schneider R, Schröder MS, et al (2014) Bassoon specifically controls presynaptic P/Q-type Ca(2+) channels via RIM-binding protein. Neuron 82: 181–194

tom Dieck S, Altrock WD, Kessels MM, Qualmann B, Regus H, Brauner D, Fejtová A, Bracko O, Gundelfinger ED & Brandstätter JH (2005) Molecular dissection of the photoreceptor ribbon synapse: physical interaction of Bassoon and RIBEYE is essential for the assembly of the ribbon complex. J Cell Biol 168: 825–836

Frank T, Rutherford MA, Strenzke N, Neef A, Pangršič T, Khimich D, Fejtova A, Gundelfinger ED, Liberman MC, Harke B, et al (2010) Bassoon and the synaptic ribbon organize Ca2+ channels and vesicles to add release sites and promote refilling. Neuron 68: 724–738

Grabner CP & Moser T (2021) The mammalian rod synaptic ribbon is essential for Cav channel facilitation and ultrafast synaptic vesicle fusion. eLife 10: e63844

Hibino H, Pironkova R, Onwumere O, Vologodskaia M, Hudspeth AJ & Lesage F (2002) RIM - binding proteins (RBPs) couple Rab3 - interacting molecules (RIMs) to voltage - gated Ca2+ channels. Neuron 34: 411–423

Inoue E, Deguchi-Tawarada M, Takao-Rikitsu E, Inoue M, Kitajima I, Ohtsuka T & Takai Y (2006) ELKS, a protein structurally related to the active zone protein CAST, is involved in Ca2+-dependent exocytosis from PC12 cells. Genes to Cells 11: 659–672

Janigro D, Maccaferri G & Meldolesi J (1989) Calcium channels in undifferentiated PC12 rat pheochromocytoma cells. FEBS Letters 255: 398–400

Jean P, Morena DL de la, Michanski S, Tobón LMJ, Chakrabarti R, Picher MM, Neef J, Jung S, Gültas M, Maxeiner S, et al (2018) The synaptic ribbon is critical for sound encoding at high rates and with temporal precision. Elife 7: e29275

Koschak A, Reimer D, Huber I, Grabner M, Glossmann H, Engel J & Striessnig J (2001) alpha 1D (Cav1.3) subunits can form l-type Ca2+ channels activating at negative voltages. J Biol Chem 276: 22100–22106

Krinner S, Butola T, Jung S, Wichmann C & Moser T (2017) RIM-Binding Protein 2 Promotes a Large Number of CaV1.3 Ca2+-Channels and Contributes to Fast Synaptic Vesicle Replenishment at Hair Cell Active Zones. Front Cell Neurosci 11: 334

Liu H, Felix R, Gurnett CA, De Waard M, Witcher DR & Campbell KP (1996) Expression and Subunit Interaction of Voltage-Dependent Ca2+ Channels in PC12 Cells. J Neurosci 16: 7557–7565

Lv C, Stewart WJ, Akanyeti O, Frederick C, Zhu J, Santos-Sacchi J, Sheets L, Liao JC & Zenisek D (2016) Synaptic Ribbons Require Ribeye for Electron Density, Proper Synaptic Localization, and Recruitment of Calcium Channels. Cell Reports 15: 2784–2795

Matthews G & Fuchs P (2010) The diverse roles of ribbon synapses in sensory neurotransmission. Nat Rev Neurosci 11: 812–822

Maxeiner S, Luo F, Tan A, Schmitz F & Südhof TC (2016) How to make a synaptic ribbon: RIBEYE deletion abolishes ribbons in retinal synapses and disrupts neurotransmitter release. The EMBO Journal 35: 1098–1114

Michanski S, Kapoor R, Steyer AM, Möbius W, Früholz I, Ackermann F, Gültas M, Garner CC, Hamra FK, Neef J, et al (2023) Piccolino is required for ribbon architecture at cochlear inner hair cell synapses and for hearing. EMBO Rep 24: e56702

Michanski S, Smaluch K, Steyer AM, Chakrabarti R, Setz C, Oestreicher D, Fischer C, Möbius W, Moser T, Vogl C, et al (2019) Mapping developmental maturation of inner hair cell ribbon synapses in the apical mouse cochlea. PNAS 116: 6415–6424

Neef J, Urban NT, Ohn T-L, Frank T, Jean P, Hell SW, Willig KI & Moser T (2018) Quantitative optical nanophysiology of Ca2+ signaling at inner hair cell active zones. Nat Commun 9: 290

Park D, Wu Y, Lee S-E, Kim G, Jeong S, Milovanovic D, Camilli PD & Chang S (2021) Cooperative function of synaptophysin and synapsin in the generation of synaptic vesicle-like clusters in non-neuronal cells. Nat Commun 12

Picher MM, Gehrt A, Meese S, Ivanovic A, Predoehl F, Jung S, Schrauwen I, Dragonetti AG, Colombo R, Camp GV, et al (2017) Ca2+-binding protein 2 inhibits Ca2+-channel inactivation in mouse inner hair cells. PNAS 114: E1717–E1726

Robertis ED & Franchi CM (1956) Electron Microscope Observations on Synaptic Vesicles in Synapses of the Retinal Rods and Cones. J Biophys Biochem Cytol 2: 307–318

Roberts WM, Jacobs RA & Hudspeth AJ (1990) Colocalization of ion channels involved in frequency selectivity and synaptic transmission at presynaptic active zones of hair cells. J Neurosci 10: 3664–3684

Smith CA & Sjöstrand FS (1961) A synaptic structure in the hair cells of the guinea pig cochlea. Journal of Ultrastructure Research 5: 184–192

Wong AB, Rutherford MA, Gabrielaitis M, Pangršič T, Göttfert F, Frank T, Michanski S, Hell S, Wolf F, Wichmann C, et al (2014) Developmental refinement of hair cell synapses tightens the coupling of Ca2+ influx to exocytosis. EMBO J 33: 247–264

Zampini V, Johnson SL, Franz C, Lawrence ND, Münkner S, Engel J, Knipper M, Magistretti J, Masetto S & Marcotti W (2010) Elementary properties of CaV1.3 Ca(2+) channels expressed in mouse cochlear inner hair cells. J Physiol 588: 187–199

Reviewer #3 (Public review):

Summary:

The paper studies learning rules in a simple sigmoidal recurrent neural network setting. The recurrent network has a single layer of 10 to 40 units. It is first confirmed that feedback alignment (FA) can learn a value function in this setting. Then so-called bio-plausible constraints are added: (1) when value weights (readout) is non-negative, (2) when the activity is non-negative (normal sigmoid rather than downscaled between -0.5 and 0.5), (3) when the feedback weights are non-negative, (4) when the learning rule is revised to be monotic: the weights are not downregulated. In the simple task considered all four biological features do not appear to impair totally the learning.

Strengths:

(1) The learning rules are implemented in a low-level fashion of the form: (pre-synaptic-activity) x (post-synaptic-activity) x feedback x RPE. Which is therefore interpretable in terms of measurable quantities in the wet-lab.

(2) I find that non-negative FA (FA with non negative c and w) is the most valuable theoretical insight of this paper: I understand why the alignment between w and c is automatically better at initialization.

(3) The task choice is relevant, since it connects with experimental settings of reward conditioning with possible plasticity measurements.

Weaknesses:

(4) The task is rather easy, so it's not clear that it really captures the computational gap that exists with FA (gradient-like learning) and simpler learning rule like a delta rule: RPE x (pre-synpatic) x (post-synaptic). To control if the task is not too trivial, I suggest adding a control where the vector c is constant c_i=1.

(5) Related to point 3), the main strength of this paper is to draw potential connection with experimental data. It would be good to highlight more concretely the prediction of the theory for experimental findings. (Ideally, what should be observed with non-negative FA that is not expected with FA or a delta rule (constant global feedback) ?).

(6a) Random feedback with RNN in RL have been studied in the past, so it is maybe worth giving some insights how the results and the analyzes compare to this previous line of work (for instance in this paper [1]). For instance, I am not very surprised that FA also works for value prediction with TD error. It is also expected from the literature that the RL + RNN + FA setting would scale to tasks that are more complex than the conditioning problem proposed here, so is there a more specific take-home message about non-negative FA? or benefits from this simpler toy task?

(6b) Related to task complexity, it is not clear to me if non-negative value and feedback weights would generally scale to harder tasks. If the task in so simple that a global RPE signal is sufficient to learn (see 4 and 5), then it could be good to extend the task to find a substantial gap between: global RPE, non-negative FA, FA, BP. For a well chosen task, I expect to see a performance gap between any pair of these four learning rules. In the context of the present paper, this would be particularly interesting to study the failure mode of non-negative FA and the cases where it does perform as well as FA.

(7) I find that the writing could be improved, it mostly feels more technical and difficult than it should. Here are some recommendations:<br /> 7a) For instance, the technical description of the task (CSC) is not fully described and requires background knowledge from other paper which is not desirable.<br /> 7b) Also the rationale for the added difficulty with the stochastic reward and new state is not well explained.<br /> 7c) In the technical description of the results I find that the text dives into descriptive comments of the figures but high-level take home messages would be helpful to guide the reader. I got a bit lost, although I feel that there is probably a lot of depth in these paragraphs.

(8) Related to the writing issue and 5), I wished that "bio-plausibility" was not the only reason to study positive feedback and value weights. Is it possible to develop a bit more specifically what and why this positivity is interesting? Is there an expected finding with non-negative FA both in the model capability? or maybe there is a simpler and crisp take-home message to communicate the experimental predictions to the community would be useful?

[1] https://www.nature.com/articles/s41467-020-17236-y

Comments on revisions:

Thank you for addressing all my comments in your reply.

Author response:

The following is the authors’ response to the original reviews

Summary of our revisions

(1) We have explained the reason why the untrained RNN with readout (value-weight) learning only could not well learn the simple task: it is because we trained the models continuously across trials with random inter-trial intervals rather than separately for each episodic trial and so it was not trivial for the models to recognize that cue presentation in different trials constitutes a same single state since the activities of untrained RNN upon cue presentation should differ from trial to trial (Line 177-185).

(2) We have shown that dimensionality was higher in the value-RNNs than in the untrained RNN (Fig. 2K,6H).

(3) We have shown that even when distractor cue was introduced, the value-RNNs could learn the task (Fig. 10).

(4) We have shown that extended value-RNNs incorporating excitatory and inhibitory units and conforming to the Dale's law could still learn the tasks (Fig. 9,10-right column).

(5) In the original manuscript, the non-negatively constrained value-RNN showed loose alignment of value-weight and random feedback from the beginning but did not show further alignment over trials. We have clarified its reason and found a way, introducing a slight decay (forgetting), to make further alignment occur (Fig. 8E,F).

(6) We have shown that the value-RNNs could learn the tasks with longer cue-reward delay (Fig. 2M,6J) or action selection (Fig. 11), and found cases where random feedback performed worse than symmetric feedback.

(7) We compared our value-RNNs with e-prop (Bellec et al., 2020, Nat Commun). While e-prop incorporates the effects of changes in RNN weights across distant times through "eligibility trace", our value-RNNs do not. The reason why our models can still learn the tasks with cue-reward delay is considered to be because our models use TD error and TD learning itself, even TD(0) without eligibility trace, is a solution for temporal credit assignment. In fact, TD error-based e-prop was also examined, but for that, result with symmetric feedback, but not with random feedback, was shown (their Fig. 4,5) while for another setup of reward-based e-prop without TD error, result with random feedback was shown (their SuppFig. 5). We have noted these in Line 695-711 (and also partly in Line 96-99).

(8) In the original manuscript, we emphasized only the spatial locality (random rather than symmetric feedback) of our learning rule. But we have now also emphasized the temporal locality (online learning) as it is also crucial for bio-plausibility and critically different from the original value-RNN with BPTT. We also changed the title.

(9) We have realized that our estimation of true state values was invalid (as detailed in page 34 of this document). Effects of this error on performance comparisons were small, but we apologize for this error.

Reviewer #1 (Public review):

Summary:

Can a plastic RNN serve as a basis function for learning to estimate value. In previous work this was shown to be the case, with a similar architecture to that proposed here. The learning rule in previous work was back-prop with an objective function that was the TD error function (delta) squared. Such a learning rule is non-local as the changes in weights within the RNN, and from inputs to the RNN depends on the weights from the RNN to the output, which estimates value. This is non-local, and in addition, these weights themselves change over learning. The main idea in this paper is to examine if replacing the values of these non-local changing weights, used for credit assignment, with random fixed weights can still produce similar results to those obtained with complete bp. This random feedback approach is motivated by a similar approach used for deep feed-forward neural networks.

This work shows that this random feedback in credit assignment performs well but is not as well as the precise gradient-based approach. When more constraints due to biological plausibility are imposed performance degrades. These results are not surprising given previous results on random feedback. This work is incomplete because the delay times used were only a few time steps, and it is not clear how well random feedback would operate with longer delays. Additionally, the examples simulated with a single cue and a single reward are overly simplistic and the field should move beyond these exceptionally simple examples.

Strengths:

• The authors show that random feedback can approximate well a model trained with detailed credit assignment.

• The authors simulate several experiments including some with probabilistic reward schedules and show results similar to those obtained with detailed credit assignments as well as in experiments.

• The paper examines the impact of more biologically realistic learning rules and the results are still quite similar to the detailed back-prop model.

Weaknesses:

*please note that we numbered your public review comments and recommendations for the authors as Pub1 and Rec1 etc so that we can refer to them in our replies to other comments.

Pub1. The authors also show that an untrained RNN does not perform as well as the trained RNN. However, they never explain what they mean by an untrained RNN. It should be clearly explained.

These results are actually surprising. An untrained RNN with enough units and sufficiently large variance of recurrent weights can have a high-dimensionality and generate a complete or nearly complete basis, though not orthonormal (e.g: Rajan&Abbott 2006). It should be possible to use such a basis to learn this simple classical conditioning paradigm. It would be useful to measure the dimensionality of network dynamics, in both trained and untrained RNN's.

We have added an explanation of untrained RNN in Line 144-147:

“As a negative control, we also conducted simulations in which these connections were not updated from initial values, referring to as the case with "untrained (fixed) RNN". Notably, the value weights w (i.e., connection weights from the RNN to the striatal value unit) were still trained in the models with untrained RNN.”

We have also analyzed the dimensionality of network dynamic by calculating the contribution ratios of each principal component of the trajectory of RNN activities. It was revealed that the contribution ratios of later principal components were smaller in the cases with untrained RNN than in the cases with trained value RNN. We have added these results in Fig. 2K and Line 210-220 (for our original models without non-negative constraint):

“In order to examine the dimensionality of RNN dynamics, we conducted principal component analysis (PCA) of the time series (for 1000 trials) of RNN activities and calculated the contribution ratios of PCs in the cases of oVRNNbp, oVRNNrf, and untrained RNN with 20 RNN units. Figure 2K shows a log of contribution ratios of 20 PCs in each case. Compared with the case of untrained RNN, in oVRNNbp and oVRNNrf, initial component(s) had smaller contributions (PC1 (t-test p = 0.00018 in oVRNNbp; p = 0.0058 in oVRNNrf) and PC2 (p = 0.080 in oVRNNbp; p = 0.0026 in oVRNNrf)) while later components had larger contributions (PC3~10,15~20 p < 0.041 in oVRNNbp; PC5~20 p < 0.0017 in oVRNNrf) on average, and this is considered to underlie their superior learning performance. We noticed that late components had larger contributions in oVRNNrf than in oVRNNbp, although these two models with 20 RNN units were comparable in terms of cue~reward state values (Fig. 2J-left).”

and Fig. 6H and Line 412-416 (for our extended models with non-negative constraint):

“Figure 6H shows contribution ratios of PCs of the time series of RNN activities in each model with 20 RNN units. Compared with the cases with naive/shuffled untrained RNN, in oVRNNbp-rev and oVRNNrf-bio, later components had relatively high contributions (PC5~20 p < 1.4×10,sup>−6</sup> (t-test vs naive) or < 0.014 (vs shuffled) in oVRNNbp-rev; PC6~20 p < 2.0×10⁻⁷ (vs naive) or PC7~20 p < 5.9×10⁻¹⁴ (vs shuffled) in oVRNNrf-bio), explaining their superior value-learning performance.”

Regarding the poor performance of the model with untrained RNN, we would like to add a note. It is sure that untrained RNN with sufficient dimensions should be able to well represent just <10 different states, and state values should be able to be well learned through TD learning regardless of whatever representation is used. However, a difficulty (nontriviality) lies in that because we modeled the tasks in a continuous way, rather than in an episodic way, the activity of untrained RNN upon cue presentation should generally differ from trial to trial. Therefore, it was not trivial for RNN to know that cue presentation in different trials, even after random lengths of inter-trial interval, should constitute a same single state. We have added this note in Line 177-185:

“This inferiority of untrained RNN may sound odd because there were only four states from cue to reward while random RNN with enough units is expected to be able to represent many different states (c.f., [49]) and the effectiveness of training of only the readout weights has been shown in reservoir computing studies [50-53]. However, there was a difficulty stemming from the continuous training across trials (rather than episodic training of separate trials): the activity of untrained RNN upon cue presentation generally differed from trial to trial, and so it is non-trivial that cue presentation in different trials should be regarded as the same single state, even if it could eventually be dealt with at the readout level if the number of units increases.”

The original value RNN study (Hennig et al., 2023, PLoS Comput Biol) also modeled tasks in a continuous way (though using backprop-through-time (BPTT) for training) and their model with untrained RNN also showed considerably larger RPE error than the value RNN even when the number of RNN units was 100 (the maximum number plotted in their Fig. 6A).

Pub2. The impact of the article is limited by using a network with discrete time-steps, and only a small number of time steps from stimulus to reward. What is the length of each time step? If it's on the order of the membrane time constant, then a few time steps are only tens of ms. In the classical conditioning experiments typical delays are of the order to hundreds of milliseconds to seconds. Authors should test if random feedback weights work as well for larger time spans. This can be done by simply using a much larger number of time steps.

In the revised manuscript, we examined the cases in which the cue-reward delay (originally 3 time steps) was elongated to 4, 5, or 6 time-steps. Our online value RNN models with random feedback could still achieve better performance (smaller squared value error) than the models with untrained RNN, although the performance degraded as the cue-reward delay increased. We have added these results in Fig. 2M and Line 223-228 (for our original models without non-negative constraint)

“We further examined the cases with longer cue-reward delays. As shown in Fig. 2M, as the delay increased, the mean squared error of state values (at 3000-th trial) increased, but the relative superiority of oVRNNbp and oVRNNrf over the model with untrained RNN remained to hold, except for cases with small number of RNN units (5) and long delay (5 or 6) (p < 0.0025 in Wilcoxon rank sum test for oVRNNbp or oVRNNrf vs untrained for each number of RNN units for each delay).”

and Fig. 6J and Line 422-429 (for our extended models with non-negative constraint):

“Figure 6J shows the cases with longer cue-reward delays, with default or halved learning rates. As the delay increased, the mean squared error of state values (at 3000-th trial) increased, but the relative superiority of oVRNNbp-rev and oVRNNrf-bio over the models with untrained RNN remained to hold, except for a few cases with 5 RNN units (5 delay oVRNNrf-bio vs shuffled with default learning rate, 6 delay oVRNNrf-bio vs naive or shuffled with halved learning rate) (p < 0.047 in Wilcoxon rank sum test for oVRNNbp-rev or oVRNNrf-bio vs naive or shuffled untrained for each number of RNN units for each delay).”

Also, we have added the note about our assumption and consideration on the time-step that we described in our provisional reply in Line 136-142:

“We assumed that a single RNN unit corresponds to a small population of neurons that intrinsically share inputs and outputs, for genetic or developmental reasons, and the activity of each unit represents the (relative) firing rate of the population. Cortical population activity is suggested to be sustained not only by fast synaptic transmission and spiking but also, even predominantly, by slower synaptic neurochemical dynamics [46] such as short-term facilitation, whose time constant can be around 500 milliseconds [47]. Therefore, we assumed that single time-step of our rate-based (rather than spike-based) model corresponds to 500 milliseconds.”

Pub3. In the section with more biologically constrained learning rules, while the output weights are restricted to only be positive (as well as the random feedback weights), the recurrent weights and weights from input to RNN are still bi-polar and can change signs during learning. Why is the constraint imposed only on the output weights? It seems reasonable that the whole setup will fail if the recurrent weights were only positive as in such a case most neurons will have very similar dynamics, and the network dimensionality would be very low. However, it is possible that only negative weights might work. It is unclear to me how to justify that bipolar weights that change sign are appropriate for the recurrent connections and inappropriate for the output connections. On the other hand, an RNN with excitatory and inhibitory neurons in which weight signs do not change could possibly work.

We examined extended models that incorporated inhibitory and excitatory units and followed Dale's law with certain assumptions, and found that these models could still learn the tasks. We have added these results in Fig. 9 and subsection “4.1 Models with excitatory and inhibitory units” and described the details of the extended models in Line 844-862:

Pub4. Like most papers in the field this work assumes a world composed of a single cue. In the real world there many more cues than rewards, some cues are not associated with any rewards, and some are associated with other rewards or even punishments. In the simplest case, it would be useful to show that this network could actually work if there are additional distractor cues that appear at random either before the CS, or between the CS and US. There are good reasons to believe such distractor cues will be fatal for an untrained RNN, but might work with a trained RNN, either using BPPT or random feedback. Although this assumption is a common flaw in most work in the field, we should no longer ignore these slightly more realistic scenarios.

We examined the performance of the models in a task in which distractor cue randomly appeared. As a result, our model with random feedback, as well as the model with backprop, could still learn the state values much better than the models with untrained RNN. We have added these results in Fig. 10 and subsection “4.2 Task with distractor cue”

Reviewer #1 (Recommendations for the authors):

Detailed comments to authors

Rec1. Are the untrained RNNs discussed in methods? It seems quite good in estimating value but has a strong dopamine response at time of reward. Is nothing trained in the untrained RNN or are the W values trained. Untrained RNN are not bad at estimating value, but not as good as the two other options. It would seem reasonable that an untrained RNN (if I understand what it is) will be sufficient for such simple Pavlovian conditioning paradigms. This is provided that the RNN generates a complete, or nearly complete basis. Random RNN's provided that the random weights are chosen properly can indeed generate a nearly complete basis. Once there is a nearly complete temporal basis, it seems that a powerful enough learning rule will be able to learn the very simple Pavlovian conditioning. Since there are only 3 time-steps from cue to reward, an RNN dimensionality of 3 would be sufficient. A failure to get a good approximation can also arise from the failure of the learning algorithm for the output weights (W).

As we mentioned in our reply to your public comment Pub1 (page 3-5), we have added an explanation of "untrained RNN" (in which the value weights were still learnt) (Line 144-147). We also analyzed the dimensionality of network dynamics by calculating the contribution ratios of principal components of the trajectory of RNN activities, showing that the contribution ratios of later principal components were smaller in the cases with untrained RNN than in the cases with trained value RNN (Fig. 2K/Line 210-220, Fig.6H/Line 412-416). Moreover, also as we mentioned in our reply to your public comment Pub1, we have added a note that even learning of a small number of states was not trivially easy because we considered continuous learning across trials rather than episodic learning of separate trials and thus it was not trivial for the model to know that cue presentation in different trials after random lengths of inter-trial interval should still be regarded as a same single state (Line 177-185).

Rec2. For all cases, it will be useful to estimate the dimensionality of the RNN. Is the dimensionality of the untrained RNN smaller than in the trained cases? If this is the case, this might depend on the choice of the initial random (I assume) recurrent connectivity matrix.

As mentioned above, we have analyzed the dimensionality of the network dynamics, and as you said, the dimensionality of the model with untrained RNN (which was indeed the initial random matrix as you said, as we mentioned above) was on average smaller than the trained value RNN models (Fig. 2K/Line 210-220, Fig.6H/Line 412-416).

Rec3. It is surprising that the error starts increasing for more RNN units above ~15. See discussion. This might indicate a failure to adjust the learning parameters of the network rather than a true and interesting finding.

Thank you very much for this insightful comment. In the original manuscript, we set the learning rate to a fixed value (0.1), without normalization by the squared norm of feature vector (as we mentioned in Line 656-7 of the original manuscript) because we thought such a normalization could not be locally (biologically) implemented. However, we have realized that the lack of normalization resulted in excessively large learning rate when the number of RNN units was large and it could cause instability and error increase as you suggested. Therefore, in the revised manuscript, we have implemented a normalization of learning rate (of value weights) that does not require non-local computations, specifically, division by the number of RNN units. As a result, the error now monotonically decreased, as the number of RNN units increased, in the non-negatively constrained models (Fig. 6E-left) and also largely in the unconstrained model with random feedback, although still not in the unconstrained model with backprop or untrained RNN (Fig. 2J-left)

Rec4. Not numbering equations is a problem. For example, the explanations of feedback alignment (lines 194-206) rely on equations in the methods section which are not numbered. This makes it hard to read these explanations. Indeed, it will also be better to include a detailed derivation of the explanation in these lines in a mathematical appendix. Key equations should be numbered.

We have added numbers to key equations in the Methods, and references to the numbers of corresponding equations in the main text. Detailed derivations are included in the Methods.

Rec5. What is shown in Figure 3C? - an equation will help.

We have added an explanation using equations in the main text (Line 256-259).

Rec6. The explanation of why alignment occurs is not satisfactory, but neither is it in previous work on feedforward networks. The least that should be done though

Regarding why alignment occurs, what remained mysterious (to us) was that in the case of nonnegatively constrained model, while the angle between value weight vector (w) and the random feedback vector (c) was relatively close (loosely aligned) from the beginning, it appeared (as mentioned in the manuscript) that there was no further alignment over trials, despite that the same mechanism for feedback alignment that we derived for the model without non-negative constraint was expected to operate also under the non-negative constraint. We have now clarified the reason for this, and found a way, introduction of slight decay (forgetting) of value weights, by which feedback alignment came to occur in the non-negatively constraint model. We have added these in the revised manuscript (Line 463-477):

“As mentioned above, while the angle between w and c was on average smaller than 90° from the beginning, there was no further alignment over trials. This seemed mysterious because the mechanism for feedback alignment that we derived for the models without non-negative constraint was expected to work also for the models with non-negative constraint. As a possible reason for the non-occurrence of feedback alignment, we guessed that one or a few element(s) of w grew prominently during learning, and so w became close to an edge or boundary of the non-negative quadrant and thereby angle between w and other vector became generally large (as illustrated in Fig. 8D). Figure 8Ea shows the mean±SEM of the elements of w ordered from the largest to smallest ones after 1500 trials. As conjectured above, a few elements indeed grew prominently.

We considered that if a slight decay (forgetting) of value weights (c.f., [59-61]) was assumed, such a prominent growth of a few elements of w may be mitigated and alignment of w to c, beyond the initial loose alignment because of the non-negative constraint, may occur. These conjectures were indeed confirmed by simulations (Fig. 8Eb,c and Fig. 8F). The mean squared value error slightly increased when the value-weightdecay was assumed (Fig. 8G), however, presumably reflecting a decrease in developed values and a deterioration of learning because of the decay.”

Rec7. I don't understand the qualitative difference between 4G and 4H. The difference seems to be smaller but there is still an apparent difference. Can this be quantified?

We have added pointers indicating which were compared and statistical significance on Fig. 4D-H, and also Fig. 7 and Fig. 9C.

Rec8. More biologically realistic constraints.

Are the weights allowed to become negative? - No.

Figure 6C - untrained RNN with non-negative x_i. Again - it was not explained what untrained RNN is. However, given my previous assumption, this is probably because the units developed in an untrained RNN is much further from representing a complete basis function. This cannot be done with only positive values. It would be useful to see network dynamics of units for untrained RNN. It might also be useful in all cases to estimate the dimensionality of the RNN. For 3 time-steps, it needs to be at least 3, and for more time steps as in Figure 4, larger.

As we mentioned in our reply to your public comment Pub3 (page 6-8), in the revised manuscript we examined models that incorporated inhibitory and excitatory units and followed Dale's law, which could still learn the tasks (Fig. 9, Line 479-520). We have also analyzed the dimensionality of network dynamics as we mentioned in our replies to your public comment Pub1 and recommendations Rec1 and Rec2.

Rec9. A new type of untrained RNN is introduced (Fig 6D) this is the first time an explanation of of the untrained RNN is given. Indeed, the dimensionality of the second type of untrained RNN should be similar to the bioVRNNrf. The results are still not good.

In the model with the new type of untrained RNN whose elements were shuffled from trained bioVRNNrf, contribution ratios of later principal components of the trajectory of RNN activities (Fig. 6H gray dotted line) were indeed larger than those in the model with native untrained RNN (gray solid line) but still much smaller than those in the trained value RNN models with backprop (red line) or random feedback (blue line). It is considered that in value RNN, RNN connections were trained to realize high-dimensional trajectory, and shuffling did not generally preserve such an ability.

Rec10. The discussion is too long and verbose. This is not a review paper.

We have made the original discussion much more compact (from 1686 words to 940 words). We have added new discussion, in response to the review comments, but the total length remains to be shorter than before (1589 words).

Reviewer #2 (Public review):

Summary:

Tsurumi et al. show that recurrent neural networks can learn state and value representations in simple reinforcement learning tasks when trained with random feedback weights. The traditional method of learning for recurrent network in such tasks (backpropagation through time) requires feedback weights which are a transposed copy of the feed-forward weights, a biologically implausible assumption. This manuscript builds on previous work regarding "random feedback alignment" and "value-RNNs", and extends them to a reinforcement learning context. The authors also demonstrate that certain nonnegative constraints can enforce a "loose alignment" of feedback weights. The author's results suggest that random feedback may be a powerful tool of learning in biological networks, even in reinforcement learning tasks.

Strengths:

The authors describe well the issues regarding biologically plausible learning in recurrent networks and in reinforcement learning tasks. They take care to propose networks which might be implemented in biological systems and compare their proposed learning rules to those already existing in literature. Further, they use small networks on relatively simple tasks, which allows for easier intuition into the learning dynamics.

Weaknesses:

The principles discovered by the authors in these smaller networks are not applied to deeper networks or more complicated tasks, so it remains unclear to what degree these methods can scale up, or can be used more generally.

We have examined extended models that incorporated inhibitory and excitatory units and followed Dale's law with certain assumptions, and found that these models could still learn the tasks. We have added these results in Fig. 9 and subsection “4.1 Models with excitatory and inhibitory units”.

We have also examined the performance of the models in a task in which distractor cue randomly appeared, finding that our models could still learn the state values much better than the models with untrained RNN. We have added these result in Fig. 10 and subsection “4.2 Task with distractor cue”.

Regarding the depth, we continue to think about it but have not yet come up with concrete ideas.

Reviewer #2 (Recommendations for the authors):

(1) I think the work would greatly benefit from more proofreading. There are language errors/oddities throughout the paper, I will list just a few examples from the introduction:

Thank you for pointing this out. We have made revisions throughout the paper.

line 63: "simultaneously learnt in the downstream of RNN". Simultaneously learnt in networks downstream of the RNN? Simulatenously learn in a downstream RNN? The meaning is not clear in the original sentence.

We have revised it to "simultaneously learnt in connections downstream of the RNN" (Line 67-68).

starting in line 65: " A major problem, among others.... value-encoding unit" is a run-on sentence and would more readable if split into multiple sentences.

We have extensively revised this part, which now consists of short sentences (Line 70-75).

line 77: "in supervised learning of feed-forward network" should be either "in supervised learning of a feed-forward network" or "in supervised learning of feed-forward networks".

We have changed "feed-forward network" to "feed-forward networks" (Line 83).

(2) Under what conditions can you use an online learning rule which only considers the influence of the previous timestep? It's not clear to me how your networks solve the temporal credit assignment problem when the cue-reward delay in your tasks is 3-5ish time steps. How far can you stretch this delay before your networks stop learning correctly because of this one-step assumption? Further, how much does feedback alignment constrain your ability to learn long timescales, such as in Murray, J.M. (2019)?

The reason why our models can solve the temporal credit assignment problem at least to a certain extent is considered to be because temporal-difference (TD) learning, which we adopted, itself has a power to resolve temporal credit assignment, as exemplified in that TD(0) algorithms without eligibility trance can still learn the value of distant rewards. We have added a discussion on this in Line 702-705:

“…our models do not have "eligibility trace" (nor memorable/gated unit, different from the original value-RNN [26]), but could still solve temporal credit assignment to a certain extent because TD learning is by itself a solution for it (notably, recent work showed that combination of TD(0) and model-based RL well explained rat's choice and DA patterns [132]).”

We have also examined the cases in which the cue-reward delay (originally 3 time steps) was elongated to 4, 5, or 6 time-steps, and our models with random feedback could still achieve better performance than the models with untrained RNN although the performance degraded as the cue-reward delay increased. We have added these results in Fig. 2M and Line 223-228 (for our original models without non-negative constraint)

“We further examined the cases with longer cue-reward delays. As shown in Fig. 2M, as the delay increased, the mean squared error of state values (at 3000-th trial) increased, but the relative superiority of oVRNNbp and oVRNNrf over the model with untrained RNN remained to hold, except for cases with small number of RNN units (5) and long delay (5 or 6) (p < 0.0025 in Wilcoxon rank sum test for oVRNNbp or oVRNNrf vs untrained for each number of RNN units for each delay).”

and Fig. 6J and Line 422-429 (for our extended models with non-negative constraint):

“Figure 6J shows the cases with longer cue-reward delays, with default or halved learning rates. As the delay increased, the mean squared error of state values (at 3000-th trial) increased, but the relative superiority of oVRNNbp-rev and oVRNNrf-bio over the models with untrained RNN remained to hold, except for a few cases with 5 RNN units (5 delay oVRNNrf-bio vs shuffled with default learning rate, 6 delay oVRNNrf-bio vs naive or shuffled with halved learning rate) (p < 0.047 in Wilcoxon rank sum test for oVRNNbp-rev or oVRNNrf-bio vs naive or shuffled untrained for each number of RNN units for each delay).”

As for the difficulty due to random feedback compared to backprop, there appeared to be little difference in the models without non-negative constraint (Fig. 2M), whereas in the models with nonnegative constraint, when the cue-reward delay was elongated to 6 time-steps, the model with random feedback performed worse than the model with backprop (Fig. 6J bottom-left panel).

(3) Line 150: Were the RNN methods trained with continuation between trials?

Yes, we have added

“The oVRNN models, and the model with untrained RNN, were continuously trained across trials in each task, because we considered that it was ecologically more plausible than episodic training of separate trials.” in Line 147-150. This is considered to make learning of even the simple cue-reward association task nontrivial, as we describe in our reply to your comment 9 below.

(4) Figure 2I, J: indicate the statistical significance of the difference between the three methods for each of these measures.

We have added statistical information for Fig. 2J (Line 198-203):

“As shown in the left panel of Fig. 2J, on average across simulations, oVRNNbp and oVRNNrf exhibited largely comparable performance and always outperformed the untrained RNN (p < 0.00022 in Wilcoxon rank sum test for oVRNNbp or oVRNNrf vs untrained for each number of RNN units), although oVRNNbp somewhat outperformed or underperformed oVRNNrf when the number of RNN units was small (≤10 (p < 0.049)) or large (≥25 (p < 0.045)), respectively.”

and also Fig. 6E (for non-negative models) (Line 385-390):

“As shown in the left panel of Fig. 6E, oVRNNbp-rev and oVRNNrf-bio exhibited largely comparable performance and always outperformed the models with untrained RNN (p < 2.5×10⁻¹² in Wilcoxon rank sum test for oVRNNbp-rev or oVRNNrf-bio vs naive or shuffled untrained for each number of RNN units), although oVRNNbp-rev somewhat outperformed or underperformed oVRNNrf-bio when the number of RNN units was small (≤10 (p < 0.00029)) or large (≥25 (p < 3.7×10⁻⁶)), respectively…”

Fig. 2I shows distributions, whose means are plotted in Fig. 2J, and we did not add statistics to Fig. 2I itself.

(5) Line 178: Has learning reached a steady state after 1000 trials for each of these networks? Can you show a plot of error vs. trial number?

We have added a plot of error vs trial number for original models (Fig. 2L, Line 221-223):

“We examined how learning proceeded across trials in the models with 20 RNN units. As shown in Fig. 2L, learning became largely converged by 1000-th trial, although slight improvement continued afterward.”

and non-negatively constrained models (Fig. 6I, Line 417-422):

“Figure 6I shows how learning proceeded across trials in the models with 20 RNN units. While oVRNNbp-rev and oVRNNrf-bio eventually reached a comparable level of errors, oVRNNrf-bio outperformed oVRNNbp-rev in early trials (at 200, 300, 400, or 500 trials; p < 0.049 in Wilcoxon rank sum test for each). This is presumably because the value weights did not develop well in early trials and so the backprop-type feedback, which was the same as the value weights, did not work well, while the non-negative fixed random feedback worked finely from the beginning.”

As shown in these figures, learning became largely steady at 1000 trials, but still slightly continued, and we have added simulations with 3000 trials (Fig. 2M and Fig. 6J).

(6) Line 191: Put these regression values in the figure caption, as well as on the plot in Figure 3B.

We have added the regression values in Fig. 3B and its caption.

(7) Line 199: This idea of being in the same quadrant is interesting, but I think the term "relatively close angle" is too vague. Is there another more quantatative way to describe this what you mean by this?

We have revised this (Line 252-254) to “a vector that is in a relatively close angle with c , or more specifically, is in the same quadrant as (and thus within at maximum 90° from) c (for example, [c₁  c₂  c₃]^T and [0.5c₁ 1.2c₂ 0.8c₃]T) “

(8) Line 275: I'd like to see this measure directly in a plot, along with the statistical significance.

We have added pointers indicating which were compared and statistical significance on Fig. 4D-H, and also Fig. 7 and Fig. 9C.

(9) Line 280: Surely the untrained RNN should be able to solve the task if the reservoir is big enough, no? Maybe much bigger than 50 units, but still.

We think this is not sure. A difficulty lies in that because we modeled the tasks in a continuous way rather than in an episodic way (as we mentioned in our reply to your comment 3), the activity of untrained RNN upon cue presentation should generally differ from trial to trial. Therefore, it was not trivial for RNN to know that cue presentation in different trials, even after random lengths of inter-trial interval, should constitute a same single state. We have added this note in Line 177-185:

“This inferiority of untrained RNN may sound odd because there were only four states from cue to reward while random RNN with enough units is expected to be able to represent many different states (c.f., [49]) and the effectiveness of training of only the readout weights has been shown in reservoir computing studies [50-53]. However, there was a difficulty stemming from the continuous training across trials (rather than episodic training of separate trials): the activity of untrained RNN upon cue presentation generally differed from trial to trial, and so it is non-trivial that cue presentation in different trials should be regarded as the same single state, even if it could eventually be dealt with at the readout level if the number of units increases.”

The original value RNN study (Hennig et al., 2023, PLoS Comput Biol) also modeled tasks in a continuous way (though using BPTT for training) and their model with untrained RNN also showed considerably larger RPE error than the value RNN even when the number of RNN units was 100 (the maximum number plotted in their Fig. 6A).

(10) It's a bit confusing to compare Figure 4C to Figure 4D-H because there are also many features of D-H which do not match those of C (response to cue, response to late reward in task 1). It would make sense to address this in some way. Is there another way to calculate the true values of the states (e.g., maybe you only start from the time of the cue) which better approximates what the networks are doing?

As we mentioned in our replies to your comments 3 and 9, our models with RNN were trained continuously across trials rather than separately for each episodic trial, and whether the models could still learn the state representation is a key issue. Therefore, starting learning from the time of cue would not be an appropriate way to compare the models, and instead we have made statistical comparison regarding key features, specifically, TD-RPEs at early and late rewards, as indicated in Fig. 4D-H.

(11) Line 309: Can you explain why this non-monotic feature exists? Why do you believe it would be more biologically plausible to assume monotonic dependence? It doesn't seem so straightforward to me, I can imagine that competing LTP/LTD mechanisms may produce plasticity which would have a non-monotic dependence on post-synaptic activity.

Thank you for this insightful comment. As you suggested, non-monotonic dependence on the postsynaptic activity (BCM rule) has been proposed for unsupervised learning (cortical self-organization) (Bienenstock et al., 1982 J Neurosci), and there were suggestions that triplet-based STDP could be reduced to a BCM-like rule and additional components (Gjorgjieva et al., 2011 PNAS; Shouval, 2011 PNAS). However, the non-monotonicity appeared in our model, derived from the backprop rule, is maximized at the middle and thus opposite from the BCM rule, which is minimized at the middle (i.e., initially decrease and thereafter increase). Therefore we consider that such an increase-then-decreasetype non-monotonicity would be less plausible than a monotonic increase, which could approximate an extreme case (with a minimum dip) of the BCM rule. We have added a note on this point in Line 355-358:

“…the dependence on the post-synaptic activity was non-monotonic, maximized at the middle of the range of activity. It would be more biologically plausible to assume a monotonic increase (while an opposite shape of nonmonotonicity, once decrease and thereafter increase, called the BCM (Bienenstock-Cooper-Munro) rule has actually been suggested [56-58]).”

(12) Line 363: This is the most exciting part of the paper (for me). I want to learn way more about this! Don't hide this in a few sentences. I want to know all about loose vs. feedback alignment. Show visualizations in 3D space of the idea of loose alignment (starting in the same quadrant), and compare it to how feedback alignment develops (ending in the same quadrant). Does this "loose" alignment idea give us an idea why the random feedback seems to settle at 45 degree angle? it just needs to get the signs right (same quadrant) for each element?

In reply to this encouraging comment, we have made further analyses of the loose alignment. By the term "loose alignment", we meant that the value weight vector w and the feedback vector c are in the same (non-negative) quadrant, as you said. But what remained mysterious (to us) was while the angle between w and c was relatively close (loosely aligned) from the beginning, it appeared (as mentioned in the manuscript) that there was no further alignment over trials (and the angle actually settled at somewhat larger than 45°), despite that the same mechanism for feedback alignment that we derived for the model without non-negative constraint was expected to operate also under the nonnegative constraint. We have now clarified the reason for this, and found a way, introduction of slight decay (forgetting) of value weights, by which feedback alignment came to occur in the non-negatively constraint model. We have added this in Line 463-477:

“As mentioned above, while the angle between w and c was on average smaller than 90° from the beginning, there was no further alignment over trials. This seemed mysterious because the mechanism for feedback alignment that we derived for the models without non-negative constraint was expected to work also for the models with non-negative constraint. As a possible reason for the non-occurrence of feedback alignment, we guessed that one or a few element(s) of w grew prominently during learning, and so w became close to an edge or boundary of the non-negative quadrant and thereby angle between w and other vector became generally large (as illustrated in Fig. 8D). Figure 8Ea shows the mean±SEM of the elements of w ordered from the largest to smallest ones after 1500 trials. As conjectured above, a few elements indeed grew prominently.

We considered that if a slight decay (forgetting) of value weights (c.f., [59-61]) was assumed, such a prominent growth of a few elements of w may be mitigated and alignment of w to c, beyond the initial loose alignment because of the non-negative constraint, may occur. These conjectures were indeed confirmed by simulations (Fig. 8Eb,c and Fig. 8F). The mean squared value error slightly increased when the value-weightdecay was assumed (Fig. 8G), however, presumably reflecting a decrease in developed values and a deterioration of learning because of the decay.”

As for visualization, because the model's dimension was high such as 12, we could not come up with better ways of visualization than the trial versus angle plot (Fig. 3A, 8A,F). Nevertheless, we would expect that the abovementioned additional analyses of loose alignment (with graphs) are useful to understand what are going on.

(13) Line 426: how does this compare to some of the reward modulated hebbian rules proposed in other RNNs? See Hoerzer, G. M., Legenstein, R., & Maass, W. (2014). Put another way, you arrived at this from a top-down approach (gradient descent->BP->approximated by RF->non-negativity constraint>leads to DA dependent modulation of Hebbian plasticity). How might this compare to a bottom up approach (i.e. starting from the principle of Hebbian learning, and adding in reward modulation)

The study of Hoerzer et al. 2014 used a stochastic perturbation, which we did not assume but can potentially be integrated. On the other hand, Hoerzer et al. trained the readout of untrained RNN, whereas we trained both RNN and its readout. We have added discussion to compare our model with Hoerzer et al. and other works that also used perturbation methods, as well as other top-down approximation method, in Line 685-711 (reference 128 is Hoerzer et al. 2014 Cereb Cortex):

“As an alternative to backprop in hierarchical network, aside from feedback alignment [36], Associative Reward-Penalty (A_R-P) algorithm has been proposed [124-126]. In A_R-P, the hidden units behave stochastically, allowing the gradient to be estimated via stochastic sampling. Recent work [127] has proposed Phaseless Alignment Learning (PAL), in which high-frequency noise-induced learning of feedback projections proceeds simultaneously with learning of forward projections using the feedback in a lower frequency. Noise-induced learning of the weights on readout neurons from untrained RNN by reward-modulated Hebbian plasticity has also been demonstrated [128]. Such noise- or perturbation-based [40] mechanisms are biologically plausible because neurons and neural networks can exhibit noisy or chaotic behavior [129-131], and might improve the performance of value-RNN if implemented.

Regarding learning of RNN, "e-prop" [35] was proposed as a locally learnable online approximation of BPTT [27], which was used in the original value RNN 26. In e-prop, neuron-specific learning signal is combined with weight-specific locally-updatable "eligibility trace". Reward-based e-prop was also shown to work [35], both in a setup not introducing TD-RPE with symmetric or random feedback (their Supplementary Figure 5) and in another setup introducing TD-RPE with symmetric feedback (their Figure 4 and 5). Compared to these, our models differ in multiple ways.

First, we have shown that alignment to random feedback occurs in the models driven by TD-RPE. Second, our models do not have "eligibility trace" (nor memorable/gated unit, different from the original valueRNN [26]), but could still solve temporal credit assignment to a certain extent because TD learning is by itself a solution for it (notably, recent work showed that combination of TD(0) and model-based RL well explained rat's choice and DA patterns [132]). However, as mentioned before, single time-step in our models was assumed to correspond to hundreds of milliseconds, incorporating slow synaptic dynamics, whereas e-prop is an algorithm for spiking neuron models with a much finer time scale. From this aspect, our models could be seen as a coarsetime-scale approximation of e-prop. On top of these, our results point to a potential computational benefit of biological non-negative constraint, which could effectively limit the parameter space and promote learning.”

Related to your latter point (and also replying to other reviewer's comment), we also examined the cases where the random feedback in our model was replaced with uniform feedback, which corresponds to a simple bottom-up reward-modulated triplet plasticity rule. As a result, the model with uniform feedback showed largely comparable, but somewhat worse, performance than the model with random feedback. We have added the results in Fig. 2J-right and Line 206-209 (for our original models without non-negative constraint):

“The green line in Fig. 2J-right shows the performance of a special case where the random feedback in oVRNNrf was fixed to the direction of (1, 1, ..., 1)^T (i.e., uniform feedback) with a random coefficient, which was largely comparable to, but somewhat worse than, that for the general oVRNNrf (blue line).”

and Fig. 6E-right and Line 402-407 (for our extended models with non-negative constraint):

“The green and light blue lines in the right panels of Figure 6E and Figure 6F show the results for special cases where the random feedback in oVRNNrf-bio was fixed to the direction of (1, 1, ..., 1) ^T (i.e., uniform feedback) with a random non-negative magnitude (green line) or a fixed magnitude of 0.5 (light blue line). The performance of these special cases, especially the former (with random magnitude) was somewhat worse than that of oVRNNrf-bio, but still better than that of the models with untrained RNN. and also added a biological implication of the results in Line 644-652:

We have shown that oVRNNrf and oVRNNrf-bio could work even when the random feedback was uniform, i.e., fixed to the direction of (1, 1, ..., 1) ^T, although the performance was somewhat worse. This is reasonable because uniform feedback can still encode scalar TD-RPE that drives our models, in contrast to a previous study [45], which considered DA's encoding of vector error and thus regarded uniform feedback as a negative control. If oVRNNrf/oVRNNrf-bio-like mechanism indeed operates in the brain and the feedback is near uniform, alignment of the value weights w to near (1, 1, ..., 1) is expected to occur. This means that states are (learned to be) represented in such a way that simple summation of cortical neuronal activity approximates value, thereby potentially explaining why value is often correlated with regional activation (fMRI BOLD signal) of cortical regions [113].”

Reviewer #3 (Public review):

Summary:

The paper studies learning rules in a simple sigmoidal recurrent neural network setting. The recurrent network has a single layer of 10 to 40 units. It is first confirmed that feedback alignment (FA) can learn a value function in this setting. Then so-called bio-plausible constraints are added: (1) when value weights (readout) is non-negative, (2) when the activity is non-negative (normal sigmoid rather than downscaled between -0.5 and 0.5), (3) when the feedback weights are non-negative, (4) when the learning rule is revised to be monotic: the weights are not downregulated. In the simple task considered all four biological features do not appear to impair totally the learning.

Strengths:

(1) The learning rules are implemented in a low-level fashion of the form: (pre-synaptic-activity) x (post-synaptic-activity) x feedback x RPE. Which is therefore interpretable in terms of measurable quantities in the wet-lab.

(2) I find that non-negative FA (FA with non negative c and w) is the most valuable theoretical insight of this paper: I understand why the alignment between w and c is automatically better at initialization.

(3) The task choice is relevant since it connects with experimental settings of reward conditioning with possible plasticity measurements.

Weaknesses:

(4) The task is rather easy, so it's not clear that it really captures the computational gap that exists with FA (gradient-like learning) and simpler learning rule like a delta rule: RPE x (pre-synpatic) x (postsynaptic). To control if the task is not too trivial, I suggest adding a control where the vector c is constant c_i=1.

We have examined the cases where the feedback was uniform, i.e., in the direction of (1, 1, ..., 1) in both models without and with non-negative constraint. In both models, the models with uniform feedback performed somewhat worse than the original models with random feedback, but still better than the models with untrained RNN. We have added the results in Fig. 2J-right and Line 206-209 (for our original models without non-negative constraint):

“The green line in Fig. 2J-right shows the performance of a special case where the random feedback in oVRNNrf was fixed to the direction of (1, 1, ..., 1) ^T (i.e., uniform feedback) with a random coefficient, which was largely comparable to, but somewhat worse than, that for the general oVRNNrf (blue line).”

and Fig. 6E-right and Line 402-407 (for our extended models with non-negative constraint):

“The green and light blue lines in the right panels of Figure 6E and Figure 6F show the results for special cases where the random feedback in oVRNNrf-bio was fixed to the direction of (1, 1, ..., 1) ^T (i.e., uniform feedback) with a random non-negative magnitude (green line) or a fixed magnitude of 0.5 (light blue line). The performance of these special cases, especially the former (with random magnitude) was somewhat worse than that of oVRNNrf-bio, but still better than that of the models with untrained RNN.”

We have also added a discussion on the biological implication of the model with uniform feedback mentioned in our provisional reply in Line 644-652:

“We have shown that oVRNNrf and oVRNNrf-bio could work even when the random feedback was uniform, i.e., fixed to the direction of (1, 1, ..., 1) ^T, although the performance was somewhat worse. This is reasonable because uniform feedback can still encode scalar TD-RPE that drives our models, in contrast to a previous study [45], which considered DA's encoding of vector error and thus regarded uniform feedback as a negative control. If oVRNNrf/oVRNNrf-bio-like mechanism indeed operates in the brain and the feedback is near uniform, alignment of the value weights w to near (1, 1, ..., 1) is expected to occur. This means that states are (learned to be) represented in such a way that simple summation of cortical neuronal activity approximates value, thereby potentially explaining why value is often correlated with regional activation (fMRI BOLD signal) of cortical regions [113].”

In addition, while preparing the revised manuscript, we found a recent simulation study, which showed that uniform feedback coupled with positive forward weights was effective in supervised learning of one-dimensional output in feed-forward network (Konishi et al., 2023, Front Neurosci).

We have briefly discussed this work in Line 653-655:

“Notably, uniform feedback coupled with positive forward weights was shown to be effective also in supervised learning of one-dimensional output in feed-forward network [114], and we guess that loose alignment may underlie it.”

(5) Related to point 3), the main strength of this paper is to draw potential connection with experimental data. It would be good to highlight more concretely the prediction of the theory for experimental findings. (Ideally, what should be observed with non-negative FA that is not expected with FA or a delta rule (constant global feedback) ?).

We have added a discussion on the prediction of our models, mentioned in our provisional reply, in Line 627-638:

“oVRNNrf predicts that the feedback vector c and the value-weight vector w become gradually aligned, while oVRNNrf-bio predicts that c and w are loosely aligned from the beginning. Element of c could be measured as the magnitude of pyramidal cell's response to DA stimulation. Element of w corresponding to a given pyramidal cell could be measured, if striatal neuron that receives input from that pyramidal cell can be identified (although technically demanding), as the magnitude of response of the striatal neuron to activation of the pyramidal cell. Then, the abovementioned predictions could be tested by (i) identify cortical, striatal, and VTA regions that are connected, (ii) identify pairs of cortical pyramidal cells and striatal neurons that are connected, (iii) measure the responses of identified pyramidal cells to DA stimulation, as well as the responses of identified striatal neurons to activation of the connected pyramidal cells, and (iv) test whether DA→pyramidal responses and pyramidal→striatal responses are associated across pyramidal cells, and whether such associations develop through learning.”

Moreover, we have considered another (technically more doable) prediction of our model, and described it in Line 639-643:

“Testing this prediction, however, would be technically quite demanding, as mentioned above. An alternative way of testing our model is to manipulate the cortical DA feedback and see if it will cause (re-)alignment of value weights (i.e., cortical striatal strengths). Specifically, our model predicts that if DA projection to a particular cortical locus is silenced, effect of the activity of that locus on the value-encoding striatal activity will become diminished.”

(6a) Random feedback with RNN in RL have been studied in the past, so it is maybe worth giving some insights how the results and the analyzes compare to this previous line of work (for instance in this paper [1]). For instance, I am not very surprised that FA also works for value prediction with TD error. It is also expected from the literature that the RL + RNN + FA setting would scale to tasks that are more complex than the conditioning problem proposed here, so is there a more specific take-home message about non-negative FA? or benefits from this simpler toy task? [1] https://www.nature.com/articles/s41467-020-17236-y

As for a specific feature of non-negative models, we did not describe (actually did not well recognize) an intriguing result that the non-negative random feedback model performed generally better than the models without non-negative constraint with either backprop or random feedback (Fig. 2J-left versus Fig. 6E-left (please mind the difference in the vertical scales)). This suggests that the non-negative constraint effectively limited the parameter space and thereby learning became efficient. We have added this result in Line 392-395:

“Remarkably, oVRNNrf-bio generally achieved better performance than both oVRNNbp and oVRNNrf, which did not have the non-negative constraint (Wilcoxon rank sum test, vs oVRNNbp : p < 7.8×10,sup>−6</sup> for 5 or ≥25 RNN units; vs oVRNNrf: p < 0.021 for ≤10 or ≥20 RNN units).”

Also, in the models with non-negative constraint, the model with random feedback learned more rapidly than the model with backprop although they eventually reached a comparable level of errors, at least in the case with 20 RNN units. This is presumably because the value weights did not develop well in early trials and so the backprop-based feedback, which was the same as the value weights, did not work well, while the non-negative fixed random feedback worked finely from the beginning. We have added this result in Fig. 6I and Line 417-422:

“Figure 6I shows how learning proceeded across trials in the models with 20 RNN units. While oVRNNbp-rev and oVRNNrf-bio eventually reached a comparable level of errors, oVRNNrf-bio outperformed oVRNNbp-rev in early trials (at 200, 300, 400, or 500 trials; p < 0.049 in Wilcoxon rank sum test for each). This is presumably because the value weights did not develop well in early trials and so the backprop-type feedback, which was the same as the value weights, did not work well, while the non-negative fixed random feedback worked finely from the beginning.”

We have also added a discussion on how our model can be positioned in relation to other models including the study you mentioned (e-prop by Bellec, ..., Maass, 2020) in subsection “Comparison to other algorithms” of the Discussion):

Regarding the slightly better performance of the non-negative model with random feedback than that of the non-negative model with backprop when the number of RNN units was large (mentioned in our provisional reply), state values in the backprop model appeared underdeveloped than those in the random feedback model. Slightly better performance of random feedback than backprop held also in our extended model incorporating excitatory and inhibitory units (Fig. 9B).

(6b) Related to task complexity, it is not clear to me if non-negative value and feedback weights would generally scale to harder tasks. If the task in so simple that a global RPE signal is sufficient to learn (see 4 and 5), then it could be good to extend the task to find a substantial gap between: global RPE, non-negative FA, FA, BP. For a well chosen task, I expect to see a performance gap between any pair of these four learning rules. In the context of the present paper, this would be particularly interesting to study the failure mode of non-negative FA and the cases where it does perform as well as FA.

In the cue-reward association task with 3 time-steps delay, the non-negative model with random feedback performed largely comparably to the non-negative model with backprop, and this remained to hold in a task where distractor cue, which was not associated with reward, appeared in random timings. We have added the results in Fig. 10 and subsection “4.2 Task with distractor cue”.

We have also examined the cases where the cue-reward delay was elongated. In the case of longer cue-reward delay (6 time-steps), in the models without non-negative constraint, the model with random feedback performed comparably to (and slightly better than when the number of RNN units was large) the model with backprop (Fig. 2M). In contrast, in the models with non-negative constraint, the model with random feedback underperformed the model with backprop (Fig. 6J, left-bottom). This indicates a difference between the effect of non-negative random feedback and the effect of positive+negative random feedback.

We have further examined the performance of the models in terms of action selection, by extending the models to incorporate an actor-critic algorithm. In a task with inter-temporal choice (i.e., immediate small reward vs delayed large reward), the non-negative model with random feedback performed worse than the non-negative model with backprop when the number of RNN units was small. When the number of RNN increased, these models performed more comparably. These results are described in Fig. 11 and subsection “4.3 Incorporation of action selection”.

(7) I find that the writing could be improved, it mostly feels more technical and difficult than it should. Here are some recommendations:

7a) for instance the technical description of the task (CSC) is not fully described and requires background knowledge from other paper which is not desirable.

7b) Also the rationale for the added difficulty with the stochastic reward and new state is not well explained.

7c) In the technical description of the results I find that the text dives into descriptive comments of the figures but high-level take home messages would be helpful to guide the reader. I got a bit lost, although I feel that there is probably a lot of depth in these paragraphs.

As for 7a), 'CSC (complete serial compound)' was actually not the name of the task but the name of the 'punctate' state representation, in which each state (timing from cue) is represented in a punctate manner, i.e., by a one-hot vector such as (1, 0, ..., 0), (0, 1, ..., 0), ..., and (0, 0, ..., 1). As you pointed out, using the name of 'CSC' would make the text appearing more technical than it actually is, and so we have moved the reference to the name of 'CSC' to the Methods (Line 903-907):

“For the agents with punctate state representation, which is also referred to as the complete serial compound (CSC) representation [1, 48, 133], each timing from a cue in the tasks was represented by a 10-dimensional one-hot vector, starting from (1 0 0 ... 0)^T for the cue state, with the next state (0 1 0 ... 0) ^T and so on.”

and in the Results we have instead added a clearer explanation (Line 163-165):

“First, for comparison, we examined traditional TD-RL agent with punctate state representation (without using the RNN), in which each state (time-step from a cue) was represented in a punctate manner, i.e., by a one-hot vector such as (1, 0, ..., 0), (0, 1, ..., 0), and so on.”

As for 7b), we have added the rationale for our examination of the tasks with probabilistic structures (Line 282-294):

“Previous work [54] examined the response of DA neurons in cue-reward association tasks in which reward timing was probabilistically determined (early in some trials but late in other trials). There were two tasks, which were largely similar but there was a key difference that reward was given in all the trials in one task whereas reward was omitted in some randomly determined trials in another task. Starkweather et al. [54] found that the DA response to later reward was smaller than the response to earlier reward in the former task, presumably reflecting the animal's belief that delayed reward will surely come, but the opposite was the case in the latter task, presumably because the animal suspected that reward was omitted in that trial. Starkweather et al.[54] then showed that such response patterns could be explained if DA encoded TD-RPE under particular state representations that incorporated the probabilistic structures of the task (called the 'belief state'). In that study, such state representations were 'handcrafted' by the authors, but the subsequent work [26] showed that the original value-RNN with backprop (BPTT) could develop similar representations and reproduce the experimentally observed DA patterns.”

As for 7c), we have extensively revised the text of the results, adding high-level explanations while trying to reduce the lengthy low-level descriptions (e.g., Line 172-177 for Fig2E-G).

(8) Related to the writing issue and 5), I wished that "bio-plausibility" was not the only reason to study positive feedback and value weights. Is it possible to develop a bit more specifically what and why this positivity is interesting? Is there an expected finding with non-negative FA both in the model capability? or maybe there is a simpler and crisp take-home message to communicate the experimental predictions to the community would be useful?

There is actually an unexpected finding with non-negative model: the non-negative random feedback model performed generally better than the models without non-negative constraint with either backprop or random feedback (Fig. 2J-left versus Fig. 6E-left), presumably because the nonnegative constraint effectively limited the parameter space and thereby learning became efficient, as we mentioned in our reply to your point 6a above (we did not well recognize this at the time of original submission).

Another potential merit of our present work is the simplicity of the model and the task. This simplicity enabled us to derive an intuitive explanation on why feedback alignment could occur. Such an intuitive explanation was lacking in previous studies while more precise mathematical explanations did exist. Related to the mechanism of feedback alignment, one thing remained mysterious to us at the time of original submission. Specifically, in the non-negatively constraint random feedback model, while the angle between the value weight (w) and the random feedback (c) was relatively close (loosely aligned) from the beginning, it appeared (as mentioned in the manuscript) that there was no further alignment over trials (and the angle actually settled at somewhat larger than 45°), despite that the same mechanism for feedback alignment that we derived for the model without non-negative constraint was expected to operate also under the non-negative constraint. We have now clarified the reason for this, and found a way, introduction of slight decay (forgetting) of value weights, by which feedback alignment came to occur in the non-negatively constraint model. We have added this in Line 463-477:

“As mentioned above, while the angle between w and c was on average smaller than 90° from the beginning, there was no further alignment over trials. This seemed mysterious because the mechanism for feedback alignment that we derived for the models without non-negative constraint was expected to work also for the models with non-negative constraint. As a possible reason for the non-occurrence of feedback alignment, we guessed that one or a few element(s) of w grew prominently during learning, and so w became close to an edge or boundary of the non-negative quadrant and thereby angle between w and other vector became generally large (as illustrated in Fig. 8D). Figure 8Ea shows the mean±SEM of the elements of w ordered from the largest to smallest ones after 1500 trials. As conjectured above, a few elements indeed grew prominently.

We considered that if a slight decay (forgetting) of value weights (c.f., [59-61]) was assumed, such a prominent growth of a few elements of w may be mitigated and alignment of w to c, beyond the initial loose alignment because of the non-negative constraint, may occur. These conjectures were indeed confirmed by simulations (Fig. 8Eb,c and Fig. 8F). The mean squared value error slightly increased when the value-weightdecay was assumed (Fig. 8G), however, presumably reflecting a decrease in developed values and a deterioration of learning because of the decay.”

Correction of an error in the original manuscript

In addition to revising the manuscript according to your comments, we have made a correction on the way of estimating the true state values. Specifically, in the original manuscript, we defined states by relative time-steps from a reward and estimated their values by calculating the sums of discounted future rewards starting from them through simulations. However, we assumed variable inter-trial intervals (ITIs) (4, 5, 6, or 7 time-steps with equal probabilities), and so until receiving cue information, agent should not know when the next reward will come. Therefore, states for the timings up to the cue timing cannot be defined by the upcoming reward, but previously we did so (e.g., state of "one timestep before cue") without taking into account the ITI variability.

We have now corrected this issue, having defined the states of timings with respect to the previous (rather than upcoming) reward. For example, when ITI was 4 time-steps and agent existed in its last time-step, agent will in fact receive a cue at the next time-step, but agent should not know it until actually receiving the cue information and instead should assume that s/he was at the last time-step of ITI (if ITI was 4), last − 1 (if ITI was 5), last − 2 (if ITI was 6), or last − 3 (if ITI was 7) with equal probabilities (in a similar fashion to what we considered when thinking about state definition for the probabilistic tasks). We estimated the true values of states defined in this way through simulations. As a result, the corrected true value of the cue-timing has become slightly smaller than the value described in the original manuscript (reflecting the uncertainty about ITI length), and consequently small positive TD-RPE has now appeared at the cue timing.

Because we measured the performance of the models by squared errors in state values, this correction affected the results reporting the performance. Fortunately, the effects were relatively minor and did not largely alter the results of performance comparisons. However, we sincerely apologize for this error. In the revised manuscript, we have used the corrected true values throughout the manuscript, and we have described the ways of estimating these values in Line 919-976.

https://books.google.hu/books?hl=en&lr=&id=V2n4zG5OVUkC&oi=fnd&pg=PA393&dq=winograd+%22understanding+computers+and+cognition%22+%2Bpdf&ots=XiDBsDVg53&sig=hyhLZDygKu-su6K5nKLf_XLMsA0&redir_esc=y#v=onepage&q&f=false

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

1. General Statements [optional]

The authors wish to thank the reviewers for fair and constructive comments and Review Commons for facilitating the process.

2. Point-by-point description of the revisions

Point-by-point replies to reviewers' comments on the original submitted manuscript are below. Authors' responses are in plain font.

Reviewers' comments:

Reviewer #1

(Evidence, reproducibility and clarity (Required)):

Summary: The authors identify cancer-associated ERBB4 mutations that are selected for functional characterization. Utilizing the BaF3 and MCF10A models, the authors investigate the potential oncogenic role for 11 recurrent ERBB4 mutations. Three mutants (S303F, E452K and L798R) were strongly transforming with the ability to transform both cell models, S303F being unique in its ability to transform both models in the absence of NRG-1. The authors perform modeling to decipher potential mechanisms of action of the ERBB4 S303F, E452K and L798R mutations. The authors assess the ability of HER3 mutations to dimerize with other HER family members and demonstrate that ERBB4 S303F can mediate its activating functions by stabilizing homo- and heterodimers with other ERBB receptors and that the heterodimerization is likely cell/tissue context dependent. The authors demonstrate that transforming ERBB4 mutants are sensitive to pan-ERBB inhibitors and drive resistance to EGFR-targeted therapy in EGFR-mutant NSCLC cells.

Major comments:

Patient data analysis is performed in more than 15 months ago in January 2024. This analysis should be updated.

We thank the reviewer for pointing out the aspect of constantly expanding mutation data in clinical cancer sample databases. We reanalyzed the patient data in cBioPortal (data download 02 May 2025). In this new analysis, the distribution of mutations in ERBB4 did not change (Reviewer only Fig. 1A), and the 18 selected mutations were still the most recurrently mutated ERBB4 mutations (Reviewer only Fig. 1B). Reanalysis of updated patient data did not change the initial rationale of the study, or the conclusions in the submitted manuscript.

Reviewer only Figure 1. Comparison of patient data derived from cBioPortal on January 2024 (01/2024) or May 2025 (05/2025). A) Figure 1B of the original submitted manuscript. B) Supplementary Figure S1C of the original submitted manuscript.

The rationale for selecting the mutations to be studied is not entirely clear. There are no references to support studying mutations in Fig 1B red boxes.

We apologize for not being sufficiently clear on our rationale for selecting the mutations for analysis. The spectrum of mutations across the ERBB4 gene do not demonstrate clear hotspots as seen in for example EGFR, KRAS, or BRAF. However, we observed that there are regions (not necessarily individual amino acid changes) in ERBB4 that seem to accumulate more mutations than other regions. Looking more closely, we observed that these "hot regions" tend to be located in areas where activating mutations have been described for other oncogenic ERBB family members and/or target structurally important regions for receptor activation such as dimerization interfaces. We hypothesized that these characteristics would suggest functional relevance for the mutations in these "hot regions". In the revised manuscript (on page 11), we have revised the text describing the selection of mutations for further analysis, and added references to justify our selection:

"While the missense mutations were distributed across the 1,308 amino acid sequence of ERBB4, lacking obvious hotspot mutations such as observed for example in EGFR or KRAS, clusters of recurrent mutations could be identified (Fig. 1B). These clusters tended to be located in specific regions that are targeted by activating mutations in other oncogenic ERBB family members (Greulich et al., 2005, 2012; Lee et al., 2006; Bose et al., 2013; Jaiswal et al., 2013)and/or are important for receptor activation (Ferguson et al., 2003; Bouyain et al., 2005; Liu et al., 2012), suggesting functional relevance (red boxes in Fig. 1B). Some recurrent mutations were located in the unstructured C-terminal tail of ERBB4 (Fig. 1B). We selected in total 18 ERBB4 missense mutations (indicated in Fig. 1B) that were recurrent and/or located in the abovementioned regions of interest for functional characterization (indicated in Fig. 1B and Supplementary Fig. S1C) - hypothesizing that these mutations would be actionable. Of the different mutants at the same position of ERBB4 amino acid sequence, the most recurrent amino acid change was selected for characterization."

Cell proliferation should be shown for BaF3 cells for continuity in Figure 2 instead of doubling time.

We agree that it may cause confusion that the results for the Ba/F3 and MCF10a experiments in Fig. 2C and D (Fig. 2D and E in the revised manuscript) are reported using a different metric. The reason for this is that these assays measure different outputs: in the Ba/F3 assay, the emergence of proliferating cells under IL3 deprivation is measured, with repeated cell viability measurements over time. In the MCF10a experiment, the ability of ERBB4 mutations to sustain the proliferation of MCF10a cells in the absence of EGF is measured, using a fixed time point (8 days). Thus, doubling time, as an indicator for the time required for the emergence of proliferating cells, is more suitable metric to quantify the relative transforming capability of the different ERBB4 mutations in the Ba/F3 cells. In the case of MCF10a cells, the relevant metric is the cell viability (as a surrogate marker for the number of cells) at the endpoint measurement.

The relative expression of HER3 constructs must be shown for BaF3 and MCF10A cells in Figure 2.

We assume the reviewer is asking to demonstrate the expression levels of different ERBB4 mutants in the Ba/F3 and MCF10a cells used in experiments in Fig. 2C and D (Fig. 2D and E in the revised manuscript). We would like to thank the reviewer for this very relevant point. Western blots demonstrating the expression levels of different ERBB4 mutants in the Ba/F3 and MCF10a cells have now been added as a data new panel in the Figure 2 (Fig. 2B in the revised manuscript). No ERBB3 expression constructs were introduced into the cells.

Blots in Figure 4 must be quantified.

The blots in Figure 4 have now been quantified, and the relative signal intensities are shown below each blot. We thank the reviewer for suggesting this relevant analysis. The analysis revealed two issues that we have now revised:

1) in Fig. 4D, the dimerization of EGFR with ERBB4 S303F is not convincingly increased when compared to EGFR dimerization with wild-type ERBB4. Therefore, we have omitted that conclusion from the results section:

"Taking into account these expression level differences, ERBB4 S303F did indeed co-immunoprecipitate more efficiently than wild-type ERBB4 with ERBB2 and EGFR both in the presence or absence of NRG-1 (Fig. 4D), demonstrating that the S303F mutation promotes the formation of ERBB heterodimers."

Omitting this data does not change our final conclusion, that the ERBB4 S303F mutation leads to enhanced ERBB4 heterodimerization.

2) In Fig. 4C, the previously published ERBB4 D595V mutant, used as a control in the experiment, does not clearly demonstrate enhanced ERBB4 homodimerization after quantifying the blots. Therefore, we have cropped the lanes representing the ERBB4 D595V mutant from the blot, and omitted the part of the results text that discusses this ERBB4 mutant:

"ERBB4 homodimers were assessed by crosslinking cell surface proteins with a cell membrane impermeable BS3, enabling detection of ERBB4 dimers as high molecular weight species of ERBB4 in western blot. Another activating extracellular ERBB4 mutation, D595V, was used as a positive control, as we have previously demonstrated D595V to stabilize ERBB4 dimers using the same assay (Kurppa et al., 2016). As predicted by the structural analyses, S303F resulted in more abundant active, phosphorylated ERBB4 dimers than wild-type ERBB4 in the presence of NRG-1, while the activating intracellular domain mutation L798R, that served as a negative control for dimer stabilization, did not (Fig. 4C)."

Omitting these data does not change our final conclusion, that the ERBB4 S303F mutation leads to enhanced ERBB4 homodimerization.

There are major concerns with Supplemental files. It is imperative that the effectiveness of HER3 shRNA be shown in S Fig3. These data are not interpretable without this.

We apologize for confusion related to the supplemental files. The effectiveness of the ERBB3 (HER3) shRNA is shown in the Supplementary Figure S3B of the original submitted manuscript.

Lanes in S Fig 4 are not marked again making data not interpretable.

Some of the lanes in the Supplementary Figure 4B were not marked because the experiment contained other ERBB4 constructs in addition to the ones that are marked and discussed in the manuscript text. The reason for leaving the unmarked lanes in the final figure was to emphasize that the bands indicated come from the same membrane, blot and exposure. We understand how this may cause confusion, and thus have now cropped the blots to include only the lanes discussed in the manuscript text.

It's unclear why Table 1 is included as this is already published data. This previously published data should be summarized in the text.

We are happy to elaborate the novelty of the data in Table 1 of the original submitted manuscript. The data is from the SUMMIT trial (NCT01953926) (Hyman et al., 2018), the results of which have been published. However, the three patients in the top part of the table were enrolled to the SUMMIT trial based on the ERBB4mutation in their tumor, and the data for these patients have not previously been published. We received these data directly from Puma Biotechnology. In addition, while the ERBB4 mutation status for the patients in the lower part of the table has been published in the supplementary files of the Hyman and others publication, we feel that the patients' ERBB4 mutations merit discussion, and including these patient data in the table would complement the data on the three patients in the top part of the table. Due to these reasons, we feel that the table contains unpublished and relevant data for the study, and would like to keep the table in the manuscript by moving it into the Supplementary Data (Supplementary Table S2).

To clarify the sources of the patient data, we have modified the methods section related to the table as follows:

"Neratinib efficacy data, cancer types and co-alterations of patients harboring an ERBB4 alteration, enrolled in PUMA-NER-5201, the SUMMIT trial (NCT01953926), and treated with neratinib as a single agent (240 mg/day) were obtained from Puma Biotechnology (for patients enrolled based on an ERBB4 mutation - previously unpublished data) and cBioPortal (for patients with ERBB4 as a co-altered gene, enrolled based on an ERBB2 or ERBB3 mutation)."

This text is now moved to "Supplementary Methods" under a new section "Neratinib efficacy in patients" on page 9 of the revised Supplementary Data -file

There is a disconnect why the last two figures focus on a single model of NSCLC whereas the three most transforming mutations are found most commonly in breast, melanoma and GI tract cancers.

The reviewer is correct in that the most transforming ERBB4 mutations are indeed found most commonly in beast and esophagogastric cancers and in melanoma. However, in the context of targeted therapy resistance,mutations that confer resistance are often acquired during therapy, and may not represent the typical cancer type-specific mutational patterns. The strongest evidence for a potential role of mutant ERBB4 in therapy resistance comes from the context of EGFR-targeted therapies and lung cancer. As mentioned in the results and discussion sections of the submitted manuscript, ERBB4 mutations identified in patients who developed resistance to EGFR-targeted therapy (Cremolini et al., 2019; Jänne et al., 2022), include the same mutation or mutation in the same residue as analyzed in the current study: the strongly transforming S303F or L798I. In addition, a recent study showed that EGFR-mutant lung cancer patients with co-occurring ERBB4 mutations have shorter relapse-free survival on osimertinib treatment (Vokes et al., 2022). Therefore, we focused on EGFR-mutant lung cancer as the model system to assess, as proof-of-concept, whether activating, transforming ERBB4 mutations are able to confer resistance to EGFR-targeted therapy. To make the transition to cancer therapy resistance and the rationale for choosing the model context more clear, we have added text to the start of the "Activating ERBB4 mutations drive resistance to EGFR-targeted therapy in EGFR-mutant NSCLC cells" -chapter of the revised manuscript:

"There is emerging evidence associating ERBB4 with cancer therapy resistance across various cancer types and treatment regimens (Merimsky et al., 2001, 2002; Mendoza-Naranjo et al., 2013; Nafi et al., 2014; Saglam et al., 2017; Wege et al., 2018; Wang et al., 2019; Zhang et al., 2023; Debets et al., 2023; Albert et al., 2024; Arribas et al., 2024), including ERBB4 mutations that have been found in patient tumors after acquisition of therapy resistance (Cremolini et al., 2019; Jänne et al., 2022; Vokes et al., 2022; Yaeger et al., 2023; Yuan et al., 2023). Intriguingly, the ERBB4 mutations identified in patients who developed resistance to EGFR-targeted therapy (Cremolini et al., 2019; Jänne et al., 2022), include the same mutation or mutation in the same residue as analyzed in the current study: the strongly transforming S303F or L798I. In addition, co-occurring ERBB4 mutations in EGFR-mutant lung cancer patients have been shown to associate with shorter progression-free survival on EGFR inhibitor therapy (Vokes et al., 2022). These observations point to the possibility that mutant ERBB4 could promote resistance to targeted therapies."

What are the differences in the recurrent ERBB4 mutant tumors versus ERBB4 wild-type tumors described in Figure 7?

The reviewer points out a very relevant question. We suspect that in the tumors expressing mutant ERBB4, the activating ERBB4 mutants are able to compensate for the loss of EGFR signaling, particularly since the on-treatment cancer cells demonstrate elevated levels of ERBB4 ligands (Fig. 7C, D). This is analogous to accumulating evidence suggesting that ERBB4 independently and together with ERBB3 (and/or with increased availability of their ligands) compensate for survival and growth signaling upon ERBB2- or EGFR-targeted therapy (Carrión-Salip et al., 2012; Wilson et al., 2012; Nafi et al., 2014; Canfield et al., 2015; Yonesaka et al., 2015; Donoghue et al., 2018; Shi et al., 2018; Debets et al., 2023; Udagawa et al., 2023). Unfortunately, we are unable to approach this hypothesis using samples from the in vivo experiment in Fig.7. The treatment of the mice was stopped after 189 days of treatment in order to assess how many tumors grew back (i.e. how many mice were cured by the treatment). For this reason, we do not have the appropriate controls to analyze ERBB4 mutant-associated changes in on-treatment tumors.

Figure 7C, D should be moved to supplemental as this is from previously published data and not strictly relevant to data shown in Fig 7.

The data shown in Fig. 7C and D are a re-analysis of published single-cell RNA-seq data. While the single cell RNA-sequencing data set is previously published, the analysis of ERBB4 ligand expression performed, and shown in Fig. 7C and D has not been published before. We feel that these data provide evidence of a previously unrecognized upregulation of ERBB4 ligand expression in on-treatment EGFR-mutant NSCLC cells in vivo. Furthermore, as discussed in the results section of the original submitted manuscript (page 26; page 28 of the revised manuscript), the upregulation of ERBB4 ligands in the on-treatment tumors provides a plausible mechanism supporting mutant ERBB4 activation upon EGFR inhibitor treatment, as the transforming ERBB4 mutants seem to retain at least partly the dependency of ligand stimulation. Thus, we feel that these data are unpublished and relevant for the manuscript, and we would like to keep these data panels in the main Figure 7.

Limitations should include consideration of endogenous levels of ERBB4 in the model systems used and disparate expression levels of wt ERBB4 versus ERBB4 mutation.

We thank the reviewer for pointing out that we have not thoroughly disclosed the endogenous levels of ERBB4 expression in the used model systems. None of the used model systems (MCF10a, Ba/F3, COS-7, PC-9) express detectable levels of ERBB4 protein. This was mentioned in the original submitted manuscript for COS-7 (page 19; page 20 of the revised manuscript), Ba/F3 cells (page 18; page 19 of the revised manuscript), and PC-9 cells (page 24; page 24 of the revised manuscript), but not for MCF10a cells. We have now made this point more clear, and added a sentence "Neither of these models express detectable levels of ERBB4" in the results section under the chapter "Majority of the recurrent ERBB4 mutations are transforming in Ba/F3 or MCF10a cells" (page 12-13 of the revised manuscript), as well as to the discussion section (page 30 of the revised manuscript).

Regarding the expression levels of different ERBB4 mutants versus ERBB4 wild-type, we have now added the new Figure 2B, showing the expression of all ERBB4 mutants and ERBB4 wild-type in Ba/F3 and MCF10a cells. We have also included the following text describing the expression levels of ERBB4 mutants in the results section under "Majority of the recurrent ERBB4 mutations are transforming in Ba/F3 or MCF10a cells" (page 13 of the revised manuscript):

"The different ERBB4 mutants demonstrated similar expression levels compared to wild-type ERBB4 in both model systems with the exception of R106C and G907E mutants that were expressed predominantly as immature receptor forms in both models, suggesting defective receptor maturation. Also, the R1304W mutant demonstrated lower expression levels in the Ba/F3 cells, and could not be expressed at all in the MCF10a cells (Fig. 2B)."

Minor comments:

Fig1B lists ERBB3 V104V mutation?

Thank you for noticing this mistake. This has now been corrected in the revised Figure 1B.

List frequency of ERBB4 mutations in the introduction

We thank the reviewer for the suggestion and have revised the introduction to include an example of the high frequency of ERBB4 missense mutations in cancer as follows:

"Yet, despite the high frequency of ERBB4 missense mutations in various cancer types (up to 30% in non-melanoma skin cancer, Supplementary Fig. S1A, B) and characterization of several potentially oncogenic ERBB4 mutations (Prickett et al. 2009; Nakamura et al. 2016; Chakroborty et al. 2022; Kurppa et al. 2016; Tvorogov et al. 2009), the rationale for clinically targeting ERBB4 in cancer has not been fully developed."

Clarification throughout if cells are serum-starved (how long) if stimulated with NRG-1

We thank the reviewer for the thoughtful suggestion and have revised the main text and figure legends accordingly; in the revised manuscript on pages 6, 8, 9, 13, 17, 20, 25 and 26 "(10% serum)", on page 25 "following short-term stimulation with NRG-1 after overnight serum starvation (Fig. 6A).", as well as figure legends of Fig. 2, 4, 5, 6, S2, and S3.

Reviewer #1 (Significance (Required)):

General assessment: This work fills a gap in cancer research understanding if ERBB4 mutations could be targeted. Concerns and comments need to be addressed before definitive conclusions can be made.

The authors wish to thank the reviewer for the positive assessment.

Reviewer #2

(Evidence, reproducibility and clarity (Required)):

Ojala et al. report a very extensive exploration of the functional relevance of somatic mutations occurring in the ERBB4 gene. The Authors demonstrate that 11 out of 18 mutations they studied have oncogenic potential, with some of them actionable using clinically available ERBB inhibitors, while giving resistance to EGFR inhibitors.

A very minor comment. At the beginning of page 21, I'd not define PD as the best respone. The Authors can write that all four patients progressed under treatment.

We would like to thank the reviewer for the comment. We agree with the reviewer, and have now revised the sentence in question as follows:

"Two of the three patients that were qualified for the SUMMIT trial due to a mutation in ERBB4, with no other qualifying mutations in ERBB family genes, had an ERBB4 mutation characterized in this study to be transforming (R544W and V840I) (Supplementary Table S2). Yet, neither of these patients, nor the patient with an ERBB4 VUS N465K, responded to neratinib and progressed under treatment (Supplementary Table S2)."

Reviewer #2 (Significance (Required)):

The work by Ojala et al. is the most detailed study of mutations occurring in ERBB4. Since these are relatively rare, they have not been properly studied up to now. The study is very well done.

The authors wish to thank the reviewer for the very positive statement.

Reviewer #3

(Evidence, reproducibility and clarity (Required)):

Summary - This work has mined cBioPortal to identify candidate cancer driver mutations in the gene encoding the ERBB4 receptor tyrosine kinase (Figure 1). These ERBB4 mutations occurred in clusters that are paralogous to activating mutations in other ERBB receptor genes or in clusters predicted to serve as dimerization interfaces of ERBB4. Eighteen such ERBB4 mutations were selected for characterization.

• These mutants were tested in BaF3 and MCF-10A cells in the context of the ERBB4 JM-a CYT-2 isoform (Figure 2). Several of these ERBB4 mutants exhibited greater agonist-dependent coupling to cell proliferation than wild-type ERBB4. Moreover, some of the mutants exhibited greater agonist-independent coupling to cell proliferation than wild-type ERBB4. Five ERBB4 mutants (S303F, E452K, L798R, R992C, S1289A) exhibited greater activity in the BaF3 cells, whereas nine ERBB4 mutants (S303F, R393W, E452K, R544W, R711C, S774G, L798R, V840I, G870R) exhibited greater activity in the MCF10A cells. Thus, eleven of the ERBB4 mutants (S303F, R393W, E452K, R544W, R711C, S774G, L798R, V840I, G870R, R992C, S1289A) exhibited a gain-of-function phenotype. It should be noted that several of the ERBB4 gain-of-function mutants (R393W, R544W, R711C, V840I, G870R, R992C, S1289A) exhibited cell type specificity.

• PyMol was used to "model" the effect of the most potent (S303F, E452K, and L798R) gain-of-function mutations on the structure of ERBB4 (Figure 3). These three mutations are predicted to cause increased ERBB4 dimerization.

• When expressed in MCF-10A cells, the most potent (S303F, E452K, and L798R) gain-of-function ERBB4 mutants exhibited elevated ligand-dependent and ligand-independent tyrosine phosphorylation. This was accompanied by elevated EGFR, ERBB2, and ERBB4 tyrosine phosphorylation and elevated signaling by canonical effector proteins (Figure 4).

• The homo- and heterodimerization of the most potent ERBB4 mutant (S303F) was studied following transient transfection of COS-7 cells (Figure 4). As predicted, the S303F mutant exhibited greater ERBB4 homodimerization and greater heterodimerization with EGFR and ERBB2, but not with ERBB3.

• The data from the clinical trial NCT01953926 was mined to evaluate whether the presence of an ERBB4 activating mutation found in this work is associated with sensitivity to the pan-ERBB inhibitor neratinib (Table 1). Surprisingly, a compelling association was NOT found. In contrast, the proliferation of BaF3 cells that express gain-of-function ERBB4 mutants is sensitive to the irreversible pan-ERBB inhibitors neratinib, afatinib, and dacomitinib (Figure 5).

• Mining the cBioPortal, AACR GENIE, and COSMIC datasets indicates that the three most potent ERBB4 gain-of-function mutants (S303F, E452K, and L798R) exhibit tissue specificity (Supplementary Figure S5). Moreover, the S303F mutation is coincident with a mutation in another ERBB receptor to a much lesser degree than other gain-of-function ERBB4 mutants, particularly E452K. This too is suggestive of differences in the mechanism of action among the gain-of-function ERBB4 mutants (Supplementary Figure S5).

• To test the effect of ERBB4 gain-of-function mutants on resistance to EGFR inhibitors, PC-9 NSCLC cells (which contain an endogenous gain-of-function EGFR mutant but do not endogenously express ERBB4) were transduced with ERBB4 gain-of-function mutants. In these cells the S303F and L715K mutants exhibited elevated ERBB4 signaling, but the L798R and K935I mutants did not. Nonetheless, the S303F, E715K, and K935I mutants promoted osimertinib resistance upon long-term treatment in vitro, whereas the L798R mutant did not (Figure 6). Moreover, the E715K and S303F mutants caused osimertinib resistance in vivo.

• Overall, this is an impressive body of work. The experiments have been carefully performed and the data are clearly presented. However, the breadth of this work makes it a bit unfocused and difficult to digest.

The authors wish to thank the reviewer for the positive statement.

Major Issues Affecting the Conclusions

The COS-7 data in Figure 4 are probably generated using supraphysiological levels of ERBB4 expression, raising concerns about the ability to draw general conclusions from these data. This issue should be addressed.

We appreciate the reviewer's insight on the details concerning experimentation in COS-7 cells. We acknowledge the drawbacks in experiments performed using transient overexpression of proteins in COS-7 cells using vectors with strong viral promoters. To mitigate these drawbacks, we routinely perform transient overexpression in COS-7 cells using the retroviral pBABE-vectors, which have a weak promoter and produce relatively moderate protein expression level. We have included here a reviewer-only figure (Reviewer-only Figure 2) that demonstrates the ERBB4 expression level derived from the pBABE-vector, compared to endogenous expression level of ERBB4 in T47D and MCF7 cells, as well as to ERBB4 expression derived from pcDNA3.1 vector that harbors a strong viral CMV promoter. With this, we hope to convince the reviewer that the ERBB4 expression levels in our COS-7 cell experiments are not supraphysiological.

Reviewer-only Figure 2. The expression level of ERBB4 in T47D and MCF7 cells, as well as in COS-7 cells transiently transfected with equal amounts of pBABE-puro-gateway-ERBB4JM-aCYT-2 plasmid, or pcDNA3.1.-ERBB4JM-aCYT-2 plasmid.

The inhibitor data shown in Figure 5 may be over-interpreted. The affinity of neratinib, afatinib, and dacomitinib for EGFR is reportedly higher than the affinity of these drugs for ERBB4. Thus, the failure of ERBB4 gain-of-function mutants to cause resistance to these inhibitors may be because the inhibitors bind to endogenous EGFR and therefore fail to bind to ERBB4.

We thank the reviewer for the insightful comments. The experiments in Figure 5 were performed in Ba/F3 cells, which do not express endogenous EGFR, or other kinase competent ERBB receptors (Riese et al., 1995). Therefore, it is unlikely that the observed cellular responses to neratinib, afatinib, or dacomitinib are affected by the drugs' preferable binding to EGFR.

Moreover, the conclusion that the gain-of-function ERBB4 mutants are targetable with these inhibitors appears to be an overreach.

We have revised our conclusion into that ERBB4 mutants are "sensitive to" these inhibitors, as supported by our data in Figure 5. This revision has been made in the abstract (page 2), introduction section (page 4), results section (page 23), and in the discussion (page 31) of the revised manuscript.

The inhibitor data shown in Figure 6 demonstrates that activating ERBB4 mutations are sufficient to drive inhibitor resistance. However, these data do not demonstrate that the mutations are necessary to drive inhibitor resistance. Thus, these data are of less value than represented in this work. Knockout or silencing (CRISPR or siRNA) experiments would be more definitive.

We agree with the reviewer that performing knock-out or silencing experiments to demonstrate the necessity of mutant ERBB4 for inhibitor resistance would strengthen the conclusions. However, the PC-9 cells (or any other EGFR-mutant NSCLC cell lines) do not express endogenous ERBB4, and do not have endogenous ERBB4 mutations. Therefore, knock-out or silencing experiments are unfortunately not possible in this setting.

Minor Issues That Can Confidently Be Addressed

In Figure 2, the MCF10A data are more compelling than the BaF3 data. Thus, an argument can be made that the BaF3 data belong in a supplemental figure. However, the combination of data from both cell lines illustrate the fact that ERBB4 mutants appear to exhibit cell type specificity. If this point is emphasized in the text, then Figure 2 should remain as currently presented.

We agree with the reviewer that our data suggest that the ERBB4 mutants demonstrate a level of context-specificity. This was mentioned in the results section of the original submitted manuscript (page 20; page 21 of the revised manuscript) as well as discussed in the discussion section (page 29; page 29 of the revised manuscript). To emphasize this further, we have revised our conclusions at the end of the "Majority of the recurrent ERBB4 mutations are transforming in Ba/F3 or MCF10a cells" -section as follows:

"Taken together, these analyses indicate a potential oncogenic role for 11 recurrent ERBB4 mutations. Eight of the mutations were transforming in only one of the models used, suggesting context-specificity. Three mutants (S303F, E452K and L798R) were strongly transforming with the ability to transform both cell models, S303F being unique in its ability to transform both models in the absence of NRG-1."

The modeling data shown in Figure 3 are a bit under-interpreted. It would appear that the S303F, E452K, and L798R mutants would cause increased ERBB4 signaling by (1) shifting the equilibrium of ERBB4 monomers between the tethered (inactive) state and the extended (active) state or by (2) directly fostering receptor dimerization. The modeling data should be interpreted in the context of these two paradigms.

We thank the reviewer again for an insightful observation. We have now revised the text describing the modeling data based on the reviewer's suggestions (please see the revised manuscript, under "Structural analysis of the transforming ERBB4 mutations").

The mechanistic data shown in Figure 4 are also a bit under-interpreted. The data from Figure 2 suggest that ERBB4 gain-of-function mutants are more likely to promote ERBB4 heterodimerization than ERBB4 homodimerization. Do the data from Figure 4 support this hypothesis?

The authors agree with the reviewer in that the activating ERBB4 mutations lead to increased activation of other ERBB family members (Fig. 4A), supporting a hypothesis that activating ERBB4 mutations lead to increased heterodimerization. We have discussed this throughout the original submitted manuscript, for example making these conclusions:

Results section, page 16 (page 18 of the revised manuscript): "In summary, these data indicate that S303F, E452K and L798R are activating, gain-of-function ERBB4 mutations that may co-operate with other ERBB receptors in malignant transformation.", page 19 (page 20 of the revised manuscript): "Together, these data suggest that while ERBB4 can be transforming in the absence of other ERBB receptors, mutant ERBB4 co-operates with ERBB3 to promote ligand-independent cell transformation.".

Discussion section, page 30 (page 31 of the revised manuscript: "Together, these findings imply that ERBB4 heterodimers with other ERBB receptors can contribute to cell transformation and growth, supporting the rationale for pan-ERBB inhibition approach in targeting mutant ERBB4 in cancer."

Reviewer #3 (Significance (Required)):

General Assessment: Strengths and Limitations

• This work makes a significant contribution to the hypothesis that ERBB4 gain-of-function mutants drive multiple human malignancies. However, this work dances around two issues. (1) Is heterodimerization of EGFR or ERBB2 with ERBB4 required for the transforming activity of these ERBB4 mutants? (2) Are these ERBB4 mutants found in the context of the JM-a/CYT-2 isoform or some other isoform? Are these ERBB4 mutants active in the context of isoforms other than JM-a/CYT-2?

We thank the reviewer for the very positive assessment and insight on specific ERBB4 biology that could affect the functional effect of mutations in ERBB4. We would like to comment on these insights:

1) Since the strongly transforming ERBB mutations all promoted the activation of EGFR, ERBB2, and ERBB3 (Fig. 4A), it is possible that heterodimerization plays a role in the transforming activity of these ERBB4 mutants. However, our data suggests that EGFR and ERBB2 are not necessary for transformation, since the Ba/F3 cells, where transformation by ERBB4 mutants was observed (Fig. 2D), do not express EGFR or ERBB2. We did see a consistent upregulation of endogenous ERBB3 upon IL3 deprivation in the ERBB4 S303F -expressing Ba/F3 cells (Fig. 4B), which contributed to the ERBB4 S303F -driven, IL3-independent transformation (Supplementary Fig. S3C-D).

2) None of the analyzed ERBB4 mutations are located in the JM- or CYT-regions of ERBB4, and thus could hypothetically be expressed in the context of any of the four ERBB4 isoforms. However, cancer tissues almost exclusively express the JM-a isoforms of ERBB4, with roughly similar ratios of CYT-1 and CYT-2 isoforms. We chose to use the JM-a CYT-2 isoform in this study, based on our previous work that has implicated the JM-a CYT-2 isoform as being more oncogenic than JM-a CYT-1 isoform, as elaborated in the original submitted manuscript: "The ERBB4 JM-a CYT-2 isoform was used in the studies based on previous findings suggesting that JM-a CYT-2 is the more oncogenic ERBB4 isoform of the cancer-associated isoforms (Veikkolainen et al., 2011) in hematopoietic cell contexts (relevant for the Ba/F3 cell model) (Määttä et al., 2006; Chakroborty et al., 2022)". We do agree with the reviewer that future studies should determine the relative contribution of JM-a CYT-1 and JM-a CYT-2 isoforms in the ability of mutant ERBB4 to drive cancer growth.

Advance: How Does This Work Advance the Field

• This work will undoubtedly reinvigorate the ERBB4 field.

Audience:

• Those with an interest in the role that ERBB receptors play in human tumors.

My Expertise:

• 30+ years of experience studying ERBB receptors.

REFERENCES

Albert, V. et al. (2024) 'HER4 Affects Sensitivity to Tamoxifen and Abemaciclib in Luminal Breast Cancer Cells and Restricts Tumor Growth in MCF-7-Based Humanized Tumor Mice', International Journal of Molecular Sciences. Multidisciplinary Digital Publishing Institute, 25(13), p. 7475. doi: 10.3390/ijms25137475.

Arribas, A. J. et al. (2024) 'ERBB4-mediated signaling is a mediator of resistance to BTK and PI3K inhibitors in B cell lymphoid neoplasms', Molecular Cancer Therapeutics, 23(3), pp. 368-380. doi: https://doi.org/10.1158/1535-7163.MCT-23-0068.

Bose, R. et al. (2013) 'Activating HER2 mutations in HER2 gene amplification negative breast cancer', Cancer Discovery, 3(2), pp. 224-237. doi: 10.1158/2159-8290.CD-12-0349.

Bouyain, S. et al. (2005) 'The extracellular region of ErbB4 adopts a tethered conformation in the absence of ligand', Proceedings of the National Academy of Sciences. National Academy of Sciences, 102(42), pp. 15024-15029. doi: 10.1073/pnas.0507591102.

Canfield, K. et al. (2015) 'Receptor tyrosine kinase ERBB4 mediates acquired resistance to ERBB2 inhibitors in breast cancer cells', Cell Cycle. Taylor & Francis, 14(4), pp. 648-655. doi: 10.4161/15384101.2014.994966.

Carrión-Salip, D. et al. (2012) 'Androgen-independent prostate cancer cells circumvent EGFR inhibition by overexpression of alternative HER receptors and ligands', International Journal of Oncology. Int J Oncol, 41(3), pp. 1128-1138. doi: 10.3892/ijo.2012.1509.

Chakroborty, D. et al. (2022) 'An Unbiased Functional Genetics Screen Identifies Rare Activating ERBB4 Mutations', Cancer Research Communications, 2(1), pp. 10-27. doi: 10.1158/2767-9764.crc-21-0021.

Cremolini, C. et al. (2019) 'Benefit from anti-EGFRs in RAS and BRAF wild-type metastatic transverse colon cancer: A clinical and molecular proof of concept study', ESMO Open, 4(2), p. e000489. doi: 10.1136/esmoopen-2019-000489.

Debets, D. O. et al. (2023) 'Deep (phospho)proteomics profiling of pre- treatment needle biopsies identifies signatures of treatment resistance in HER2+ breast cancer', Cell Reports Medicine. Elsevier, 4(10), p. 101203. doi: 10.1016/j.xcrm.2023.101203.

Donoghue, J. F. et al. (2018) 'Activation of ERBB4 in glioblastoma can contribute to increased tumorigenicity and influence therapeutic response', Cancers. Multidisciplinary Digital Publishing Institute (MDPI), 10(8), p. 243. doi: 10.3390/cancers10080243.

Ferguson, K. M. et al. (2003) 'EGF Activates Its Receptor by Removing Interactions that Autoinhibit Ectodomain Dimerization', Molecular Cell, 11(2), pp. 507-517. doi: 10.1016/S1097-2765(03)00047-9.

Greulich, H. et al. (2005) 'Oncogenic Transformation by Inhibitor-Sensitive and -Resistant EGFR Mutants', PLOS Medicine. Public Library of Science, 2(11), p. e313. doi: 10.1371/JOURNAL.PMED.0020313.

Greulich, H. et al. (2012) 'Functional analysis of receptor tyrosine kinase mutations in lung cancer identifies oncogenic extracellular domain mutations of ERBB2', Proceedings of the National Academy of Sciences. National Academy of Sciences, 109(36), pp. 14476-14481. doi: 10.1073/pnas.1203201109.

Hyman, D. M. et al. (2018) 'HER kinase inhibition in patients with HER2-and HER3-mutant cancers', Nature. Nature Publishing Group, 554(7691), pp. 189-194. doi: 10.1038/nature25475.

Jaiswal, B. S. S. et al. (2013) 'Oncogenic ERBB3 Mutations in Human Cancers', Cancer Cell, 23(5), pp. 603-617. doi: 10.1016/j.ccr.2013.04.012.

Jänne, P. A. et al. (2022) 'Efficacy and Safety of Patritumab Deruxtecan (HER3-DXd) in EGFR Inhibitor- Resistant, EGFR-Mutated Non-Small Cell Lung Cancer', Cancer Discovery. American Association for Cancer Research, 12(1), pp. 74-89. doi: 10.1158/2159-8290.CD-21-0715.

Kurppa, K. J. et al. (2016) 'Activating ERBB4 mutations in non-small cell lung cancer', Oncogene. Nature Publishing Group, 35(10), pp. 1283-1291. doi: 10.1038/onc.2015.185.

Lee, J. C. et al. (2006) 'Epidermal Growth Factor Receptor Activation in Glioblastoma through Novel Missense Mutations in the Extracellular Domain', PLoS Medicine | www, 3. doi: 10.1371/journal.

Liu, P. et al. (2012) 'A single ligand is sufficient to activate EGFR dimers', Proceedings of the National Academy of Sciences, 109(27), pp. 10861-10866. doi: 10.1073/pnas.1201114109.

Määttä, J. A. et al. (2006) 'Proteolytic cleavage and phosphorylation of a tumor-associated ErbB4 isoform promote ligand-independent survival and cancer cell growth', Molecular Biology of the Cell, 17(1), pp. 67-79. doi: 10.1091/mbc.E05-05-0402.

Mendoza-Naranjo, A. et al. (2013) 'ERBB4 confers metastatic capacity in Ewing sarcoma', EMBO Molecular Medicine. Nature Publishing Group, 5(7), pp. 1019-1034. doi: 10.1002/emmm.201202343.

Merimsky, O. et al. (2001) 'Correlation between c-erbB-4 receptor expression and response to gemcitabine-cisplatin chemotherapy in non-small-cell lung cancer', Annals of Oncology. Ann Oncol, 12(8), pp. 1127-1131. doi: 10.1023/A:1011665216616.

Merimsky, O. et al. (2002) 'ErbB-4 expression in limb soft-tissue sarcoma: Correlation with the results of neoadjuvant chemotherapy', European Journal of Cancer. Eur J Cancer, 38(10), pp. 1335-1342. doi: 10.1016/S0959-8049(02)00075-8.

Nafi, S. N. M. et al. (2014) 'Nuclear HER4 mediates acquired resistance to trastuzumab and is associated with poor outcome in HER2 positive breast cancer', Oncotarget. Impact Journals, LLC, 5(15), pp. 5934-5949. doi: 10.18632/oncotarget.1904.

Nakamura, Y. et al. (2016) 'Afatinib against esophageal or head-and-neck squamous cell carcinoma: Significance of activating oncogenic HER4 mutations in HNSCC', Molecular Cancer Therapeutics. American Association for Cancer Research, 15(8), pp. 1988-1997. doi: 10.1158/1535-7163.MCT-15-0737.

Prickett, T. D. et al. (2009) 'Analysis of the tyrosine kinome in melanoma reveals recurrent mutations in ERBB4.', Nature genetics, 41(10), pp. 1127-1132. doi: 10.1038/ng.438.

Riese, D. J. et al. (1995) 'The cellular response to neuregulins is governed by complex interactions of the erbB receptor family', Molecular and Cellular Biology, 15(10), pp. 5770-5776. doi: 10.1128/mcb.15.10.5770.

Saglam, O. et al. (2017) 'ERBB4 Expression in Ovarian Serous Carcinoma Resistant to Platinum-Based Therapy', Cancer control. NIH Public Access, 24(1), p. 89. doi: 10.1177/107327481702400115.

Shi, J. et al. (2018) 'The HER4-YAP1 axis promotes trastuzumab resistance in HER2-positive gastric cancer by inducing epithelial and mesenchymal transition', Oncogene, 37(22), pp. 3022-3038. doi: 10.1038/s41388-018-0204-5.

Tvorogov, D. et al. (2009) 'Somatic mutations of ErbB4: Selective loss-of-function phenotype affecting signal transduction pathways in cancer', Journal of Biological Chemistry, 284(9), pp. 5582-5591. doi: 10.1074/jbc.M805438200.

Udagawa, H. et al. (2023) 'HER4 and EGFR Activate Cell Signaling in NRG1 Fusion-Driven Cancers: Implications for HER2-HER3-specific Versus Pan-HER Targeting Strategies', Journal of Thoracic Oncology. Elsevier Inc, 19(1), pp. 106-118. doi: 10.1016/j.jtho.2023.08.034.

Veikkolainen, V. et al. (2011) 'Function of ERBB4 is determined by alternative splicing', Cell Cycle, 10(16), pp. 2647-2657. doi: 10.4161/cc.10.16.17194.

Vokes, N. I. et al. (2022) 'Concurrent TP53 mutations facilitate resistance evolution in EGFR mutant lung adenocarcinoma', Journal of Thoracic Oncology. International Association for the Study of Lung Cancer, 17(6), pp. 779-792. doi: 10.1016/j.jtho.2022.02.011.

Wang, D. S. et al. (2019) 'Liquid biopsies to track trastuzumab resistance in metastatic HER2-positive gastric cancer', Gut. BMJ Publishing Group, 68(7), pp. 1152-1161. doi: 10.1136/gutjnl-2018-316522.

Wege, A. K. et al. (2018) 'HER4 expression in estrogen receptor-positive breast cancer is associated with decreased sensitivity to tamoxifen treatment and reduced overall survival of postmenopausal women', Breast Cancer Research. Breast Cancer Res, 20(1). doi: 10.1186/s13058-018-1072-1.

Wilson, T. R. et al. (2012) 'Widespread potential for growth-factor-driven resistance to anticancer kinase inhibitors', Nature. Nature Publishing Group, 487(7408), pp. 505-509. doi: 10.1038/nature11249.

Yaeger, R. et al. (2023) 'Molecular Characterization of Acquired Resistance to KRASG12C-EGFR Inhibition in Colorectal Cancer', Cancer Discovery, 13(1), pp. 41-55. doi: 10.1158/2159-8290.CD-22-0405.

Yonesaka, K. et al. (2015) 'The pan-HER family tyrosine kinase inhibitor afatinib overcomes HER3 ligand heregulin-mediated resistance to EGFR inhibitors in non-small cell lung cancer.', Oncotarget. Oncotarget, 6(32), pp. 33602-11. doi: 10.18632/oncotarget.5286.

Yuan, S. Q. et al. (2023) 'Residual circulating tumor DNA after adjuvant chemotherapy effectively predicts recurrence of stage II-III gastric cancer', Cancer Communications. John Wiley & Sons, Ltd, 43(12), pp. 1312-1325. doi: 10.1002/cac2.12494.

Zhang, J. et al. (2023) 'Tracking of trastuzumab resistance in patients with HER2-positive metastatic gastric cancer by CTC liquid biopsy.', American journal of cancer research. e-Century Publishing Corporation, 13(11), pp. 5684-5697. Available at: http://www.ncbi.nlm.nih.gov/pubmed/38058840 (Accessed: 16 April 2024).

Synthèse détaillée des thèmes et faits marquants de la présentation de Valérie Masson-Delmotte

• Cette synthèse reprend les points essentiels de l'intervention de Valérie Masson-Delmotte sur les attaques contre les sciences du climat et de la durabilité, en contextualisant ces attaques par rapport aux faits scientifiques établis et en explorant les liens avec la vie démocratique et l'action climatique.

I. Les attaques contre les sciences du climat et de la durabilité : Une nouvelle phase d'obstruction systémique

• Valérie Masson-Delmotte introduit son intervention en situant sa propre expérience de recherche, notamment sur la géoingénierie polaire, comme une "distraction ou une diversion coûteuse" qui s'inscrit dans un contexte plus large d'attaques.

Elle décrit une "nouvelle phase" dans les politiques favorables aux énergies fossiles, caractérisée non plus par un simple "rideau de fumée" de désinformation, mais par une "obstruction tous azimuts".

A. Le Projet 2025 et ses objectifs

Cette nouvelle phase, pensée par le Projet 2025, est motivée par la clarté des connaissances scientifiques et la capacité croissante à répondre aux besoins humains sans énergies fossiles.

L'objectif est de "détruire la capacité à produire ces faits scientifiques".

Les actions concrètes de l'administration Trump (si elle était réélue, nommée "Trump 2" dans le texte) incluent :

• Affirmation d'une "urgence énergétique" et obstruction à la transition énergétique.
• Suppression des étiquettes énergie des politiques favorables aux énergies fossiles.
• Arrêt du plan Biden Inflation Reduction Act, favorable aux technologies propres.
• Destruction des "pierres angulaires du droit de l'environnement", notamment le suivi des espèces menacées.
• Ciblage de la décision de l'EPA (Endangerment Finding) classant les émissions de six gaz à effet de serre comme "dangereux pour le bien-être et la santé humaine".

B. Méthodes d'intimidation et de démantèlement

Les attaques se traduisent par :

• Intimidation des organisations scientifiques, menaces de coupes budgétaires et leur réalisation.
• Censure par mots clés et "licenciements massifs".
• Suppression de contenu lié au changement climatique sur les sites ministériels.
• Interdictions de formation intégrant les aspects liés au changement climatique.
• Démantèlement de la capacité à produire des connaissances, notamment sur la santé environnementale à l'EPA.
• Arrêt de financements de programmes de recherche sur le changement climatique et la santé pour le ministère de la Défense.
• Sabotage de la capacité de suivi de l'état de l'atmosphère, entraînant la dégradation des prévisions météorologiques et la fermeture d'agences d'alerte.
• Affaiblissement des capacités de suivi des lois environnementales (Endangered Species Act, Marine Mammal Protection Act).
• Destruction de la capacité d'évaluation nationale sur le changement climatique (US Global Change Research Program).
• Démantèlement de l'Office of Global Change au Département d'État, affectant la participation américaine aux instances internationales (IPCC).
• Coupes budgétaires massives par mots clés (NSF, NOAA, NASA) visant à éliminer les laboratoires de recherche sur l'atmosphère, l'océan, l'Antarctique, et les instituts coopératifs.
• Harcèlement (ex: arrêt du paiement du bail pour le laboratoire commun Columbia NASA).
• Décimation des centres d'adaptation au changement climatique, d'information sur l'énergie et des centres sur les sciences de l'eau.
• Volonté de démanteler l'agence de réponse rapide en cas de catastrophe.
• Ciblage du suivi des émissions de gaz à effet de serre, des centres de modélisation et des aspects liés aux injustices environnementales.
• Masson-Delmotte souligne l'interdiction explicite de mots clés tels que "atténuation", "gaz à effet de serre", "changement climatique", "science du climat", "risque climatique", "résilience", "justice environnementale", "pollution", "vulnérabilités", "evidence-based" et "science-based". Cette méthode, déjà observée sous Trump 1 à l'EPA, est désormais "généralisée", avec des leçons tirées pour une action "à une échelle plus large".

II. Le constat scientifique, "trop clair" et "très clair"

• La présentation met en lumière l'état actuel et l'évolution des indicateurs clés du système climatique, soulignant l'urgence et l'ampleur de l'influence humaine.

*** A. Émissions et concentrations de gaz à effet de serre**

• Émissions mondiales : Continuent d'augmenter, mais "moins vite sur la dernière décennie que précédemment". 24 pays ont réduit leurs émissions grâce à des politiques publiques efficaces.
• Chine : Les émissions n'ont pas augmenté depuis environ un an, grâce à l'électrification des mobilités et le déploiement des énergies renouvelables.
• Gaz à effet de serre fluorés : Baisse due à l'effet du Protocole de Montréal.
• Composés soufrés : Leur réduction (effet refroidissant) contribue à dévoiler l'effet réchauffant des gaz à effet de serre.
• Concentrations atmosphériques : Continuent d'augmenter, particulièrement vite pour le CO2 en 2023-2024, en partie à cause de l'effet El Niño qui a "fragilisé le puits de carbone terrestre". Le Global Carbon Project estime que l'effet climat a sapé "environ 27 % de l'augmentation du puits de carbone terrestre".
• Forçage radiatif : L'influence humaine atteint environ "3 W par mètre carré", avec un rythme d'augmentation sur la dernière décennie "encore plus fort que la décennie précédente".

B. Réchauffement climatique et ses conséquences

• Déséquilibre énergétique de la Terre : A augmenté pour atteindre "1 W par mètre carré".
• Réchauffement observé :
• Moyenne planétaire : "1,24°C" sur la dernière décennie.
• Océan : "1°C".
• Continents : "environ 1,8°C".
• Part attribuable aux activités humaines : "1,22°C" sur les 1,24°C observés.
• Rythme de hausse : "0,27°C de plus par décennie".
• Réchauffement attribuable à l'influence humaine (extrapolé) : "1,36°C plus chaud".
• Dépassement des seuils : Atteindre "plus d'un degré et demi comme en 2024 pour la première fois, c'est une chance sur 6" ; avec El Niño et variabilité chaude Atlantique Nord, "c'est une chance sur deux".
• Montée du niveau de la mer : "Quasiment 23 cm depuis 1900".

Rythme sur les 10 dernières années : "plus de 4 mm par an" avec une accélération due à l'accumulation de chaleur dans l'océan et la contribution croissante du Groenland et de l'Antarctique.

C. Attribution des responsabilités

Les méthodes d'attribution, développées par Klaus Hasselman, "n'ont fait que s'affiner pour un constat qui est vraiment très clair maintenant".

• Responsabilité historique par pays (depuis 1851) :
• États-Unis : "premier responsable du réchauffement à ce jour".
• Chine : "2e rang", devant l'Union européenne à 27.
• Certaines analyses suggèrent que l'effort de décarbonation en Chine pourrait faire qu'elle "ne rattrape jamais le niveau de responsabilité des États-Unis".
• Tensions géopolitiques : Les pays en développement estiment que l'Accord de Paris "remet à zéro" la responsabilité historique, tandis que les pays industrialisés et exportateurs d'énergies fossiles sont plus à l'aise avec cette approche.
• Pays les moins développés : "6 % des de contribution au réchauffement historique", alors qu'ils sont "le milliard de personnes qui vit dans l'ensemble des pays les moins développés" et sont les plus vulnérables.
• Responsabilité par catégorie de revenus (étude de 2024, incluant consommation et investissements) :
• "2/3 du réchauffement est attribuable au top 10 %" des émetteurs mondiaux depuis 1990.
• "1/5 au top 1 %" des émetteurs.
• Si tout le monde avait le même train d'émissions que les 10 % les plus riches, le réchauffement aurait déjà été de "3 degrés au lieu de 0,6°C depuis 1990".
• Le top 10 % contribue "6 et demi fois plus" que la moyenne planétaire par personne.
• Les émissions du top 10 % en Chine ou aux États-Unis ont une responsabilité "deux à trois fois plus grande que la moyenne de la population mondiale sur l'augmentation des vagues de chaleur dans des régions particulièrement vulnérables".
• Responsabilité des acteurs économiques : Application de la méthode d'attribution "end-to-end" (émission -> réchauffement -> événement extrême -> coût des dommages) aux "plus grandes supermajors pétrolières".
• Les cinq plus grandes entreprises pétrolières et gazières mondiales sont responsables depuis 1991 d'"environ 2 000 milliards de dollars de pertes économiques uniquement liées à la chaleur extrême".
• Ces constats "dérangent bien sûr" et posent la question de la "responsabilité par rapport aux pertes et dommages".
• Droits des générations futures : L'exposition "sans précédent" aux extrêmes climatiques des jeunes générations. Pour les personnes nées en 2020 :
• Si le réchauffement est limité à 1°C, "52 % de cette génération sera exposée de manière inédite".
• Si le réchauffement atteint 3°C, "la quasi-totalité de cette génération" (plus de 90 %) sera exposée de manière inédite.
• Ces connaissances scientifiques "très claires" peuvent "nourrir... des contentieux juridiques, des délibérations politiques" et sont perçues comme "hostiles" par certains acteurs économiques ou catégories de la population qui "tirent partie de la situation actuelle".

III. Démocratie et action climatique : Des liens cruciaux et des défis

Valérie Masson-Delmotte explore les liens entre la gouvernance démocratique et la capacité à mettre en œuvre une action climatique efficace, soulignant les vulnérabilités et les leviers d'action.

A. Trajectoires socio-économiques et risques

Le scénario de

• "rivalité régionale",
• "résurgence du nationalisme",
• "préoccupation en matière de compétitivité sécurité",
• "perte d'une vision plus large",
• "baisse des investissements dans l'éducation et le développement technologique",
• "développement économique lent",
• "consommation à forte intensité matérielle",
• "aggravation d'inégalité",
• "faible priorité aux problèmes environnementaux",
• "forte dégradation de l'environnement" (SSP3)

est le scénario "à plus haut risque" pour l'atténuation et l'adaptation. * Il existe un "jeu d'acteur organisé puissant ancré dans le national conservatisme qui vise en fait à créer une bifurcation par rapport aux tendances récentes au focus mis sur la soutenabilité créer une bifcation vers ce type de trajectoire".

B. Démocratie comme levier d'action climatique

• Réduction des émissions : "24 pays qui l'ont déjà réduit les ont déjà réduites fortement dans la durée tous des démocraties".
• Politiques publiques efficaces : Une étude comparant 1500 politiques publiques montre que "60 des 63 exemples de combinaison de politique publique efficace sont des pays qui ont des indices de démocratie parmi les plus élevés". Les exceptions sont l'Arabie Saoudite et la Chine.
• Caractéristiques des démocraties efficaces : "Pluralisme, la liberté de la presse, le dynamisme de la société civile, contrôle de la corruption".
• Capacités de transformation (selon le GIEC) : Vont de pair avec la capacité de "délibérer, de construire des transitions justes qui tiennent compte des inégalités des vulnérabilités des droits humains des enjeux de cohésion sociale", l'engagement des parties prenantes (y compris la jeunesse), le renforcement des processus délibératifs (assemblées citoyennes), des rôles clairs pour les pouvoirs publics, et un suivi et évaluation.

C. Obstacles démocratiques à l'action climatique

• Clientélisme et corruption : Le développement démocratique n'est associé à la réduction des émissions que dans les pays "qui ont un faible niveau de clientélisme et de corruption".
• Rente des énergies fossiles : Les "effets de rente liés aux énergies fossiles sapent les institutions et la capacité à mettre en œuvre de manière efficace les lois et régulation environnementale". Les pays dépendants de l'extraction de pétrole et de gaz ont souvent des "difficultés de contrôle de la corruption" et maintiennent des "subventions élevées aux énergies fossiles".
• Inégalités de revenus : Sapent la capacité à agir efficacement. Dans les pays les plus inégalitaires, les personnes à hauts revenus (très émettrices) "s'opposent au renforcement de la régulation environnementale" et leur opposition est renforcée par la possession de médias privés et le contrôle des réseaux sociaux.

En conclusion, Valérie Masson-Delmotte souligne l'importance de "renouveler et vivifier la vie démocratique" face aux attaques des discours populistes et du contrôle de la désinformation.

Le "renforcement des délibérations qui s'appuient sur les connaissances et les faits est pour moi un levier d'action critique par rapport à la défense de la capacité de délibération des sociétés démocratiques vis-à-vis du renforcement de l'action pour le climat".

When verifying the general definition of informatics, it mostly aligns with Health Informatics but needs a few clarifications to fully fit this domain. Health Informatics is not just about transforming data — it’s about using information and technology to improve patient care, safety, and health outcomes. This work relies on tools like electronic health records (EHRs), clinical decision support systems, telehealth, and secure data sharing to make sure information is available and meaningful to clinicians, patients, and public health teams (Sumner et al., 2025). A good definition must also include how Health Informatics depends on standards like HL7 and FHIR to ensure systems can work together. It also involves designing user-friendly systems and following strict privacy rules like HIPAA, plus strong ethical practices around consent and data security (Valenta et al., 2018). It’s important to note the growing role of artificial intelligence and remote monitoring in today’s digital health systems too. Overall, Health Informatics clearly goes beyond general IT — it’s an interdisciplinary field that connects clinical care, data science, and technology to support safer, more efficient, and patient-centered healthcare (Mansurjonovich, 2023). When done well, Health Informatics empowers teams to deliver better evidence-based decisions. It also helps reduce errors and supports health equity by improving access to quality information and care.

Rationale: Including these specific elements makes the definition of Health Informatics more accurate because it shows that this domain is not only about handling data but also about applying it responsibly to real healthcare challenges. It highlights how Health Informatics connects people, technology, and policy to improve patient safety and outcomes, which goes beyond what general IT or computer science do on their own. By focusing on ethical, secure, and patient-centered data use, the definition makes it clear how Health Informatics stands apart as a unique and critical field in modern healthcare (Mansurjonovich, 2023; Sumner et al., 2025).

References:

Mansurjonovich, J. M. (2023). Designing an electronic didactic environment to ensure interdisciplinary integration in the teaching of" Informatics and information technologies" during professional education. Conference, 3(03), 78-82.

Sumner, B., Martin, R., Gladman, T., Wilkinson, T. J., & Grainger, R. (2025). Understanding the gap: A balanced multi-perspective approach to defining essential digital health competencies for medical graduates. BMC medical education, 25(1), 1-11.

Valenta, A. L., Berner, E. S., Boren, S. A., Deckard, G. J., Eldredge, C., Fridsma, D. B., Gadd, C., Gong, Y., Johnson, T., Jones, J., Manos, E. L., Phillips, K. T., Roderer, N. K., Rosendale, D., Turner, A. M., Tusch, G., Williamson, J. J., & Johnson, S. B. (2018). AMIA Board White Paper: AMIA 2017 core competencies for applied health informatics education at the master’s degree level. Journal of the American Medical Informatics Association, 25(12), 1657–1668. https://doi.org/10.1093/jamia/ocy132

Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

Learn more at Review Commons

Reply to the reviewers

Revision Plan

June 28, 2025

Manuscript number: RC-2025-02982

Corresponding author(s): Babita Madan, Nathan Harmston, David Virshup

General Statements In Wnt signaling, the relative contributions of ‘canonical (β-catenin dependent) and non- canonical (β-catenin independent) signaling remains unclear. Here, we exploited a unique and highly robust in vivo system to study this. Our study is therefore the first comprehensive analysis of the β-catenin independent arm of the Wnt signaling pathway in a cancer model and illustrates how a combination of cis-regulatory elements can determine Wnt-dependent gene regulation.

We are very pleased with the reviews; it appears we communicated our goal and our findings clearly, and in general the reviewers felt the study provided important information, was well planned and the results were “crystal clear”.

While more experiments could strengthen and extend the results, we feel our results are already very robust due to the use of multiple replicates in the in vivo system.

The Virshup lab in Singapore closed July 1, 2025 and so additional wet lab studies are not feasible.

1. Description of the planned revisions

Insert here a point-by-point reply that explains what revisions, additional experimentations and analyses are planned to address the points raised by the referees.

Below we address the points raised by the reviewers:

Reviewer #1 (Evidence, reproducibility and clarity (Required)):

…

The article has the merit of addressing a yet-unsolved question in the field (if beta-catenin can also repress genes) that only a limited number of studies has tried to tackle, and provides useful datasets for the community. The system employed is elegant, and the PORCN-inhibition bypassed by a ____constitutively active beta-catenin is clean and ingenious. The manuscript is clearly written.

We thank the reviewers for their kind comments on the importance of the data. Our orthotopic model provides the opportunity to exploit robust Wnt regulated gene expression in a more responsive microenvironment than can be achieved in cell culture and simple flank xenograft models.

Here we propose a series of thoughts and comments that, if addressed, would in our opinion improve the study and its description.

1) We wonder why a xenograft model is necessary to induce a robust WNT response in these cells.

The authors describe this set-up as a strength, as it is supposed to provide physiological relevance, yet it is not clear to us why this is the case.

We welcome the opportunity to expand on our choice of an orthotopic xenograft model. It has been long established that cancer cells behave differently in different in vivo locations (Killion et al., 1998). Building on this, we confirmed this in our system that identical pancreatic cancer cells treated with the same PORCN inhibitor had very different responses in vitro, in the flank and in their orthotopic environment (Madan et al., 2018). To quote from our prior paper, “Looking only at genes decreasing more than 1.5-fold at 56 hours, we would have missed 817/1867 (44%) genes using a subcutaneous or 939/1867 (50%) using an in vitro model. Thus, the overall response to Wnt inhibition was reduced in the subcutaneous model and further blunted in vitro. An orthotopic model more accurately represents real biology.

The reason for this is presumably the very different orthotopic microenvironment, including tissue appropriate stroma-tumor, vascular-tumor, lymphatic-tumor, and humoral interactions.

Moreover, as the authors homogenize the tumour to perform bulk RNA-seq, we wonder whether they are not only sequencing mRNA from the cancer cells but also from infiltrating immune cells and/or from the surrounding connective tissue.

In experiments generating RNA-seq data from xenograft models, the resulting sequences can originate from either human (graft) or mouse (host). In order to account for this, following standard practice, we filtered reads prior to alignment using Xenome (Conway et al., 2012). We have added additional text to the methods to highlight this step in our pipeline.

2) If, as the established view implies, Wnt/beta-catenin only leads to gene activation, pathway

inhibition would free up the transcriptional machinery - there is evidence that some of its constituents are rate-limiting. The free machinery could now activate some other genes: the net effect observed would be their increased transcription upon Wnt inhibition, irrespective of beta-catenin's presence. Could this be considered as an alternative explanation for the genes that go up in both control and bcat4A lines upon ETC-159 administration? This, we think, is in part corroborated by the absence of enrichment of biological pathways in this group of genes. The genes that are beta-catenin-dependent and downregulated (D&R) are obviously not affected by this alternative explanation.

This is an interesting suggestion, and we will incorporate this thought into our discussion of potential mechanisms.

3) The authors mention that HPAF-II are Wnt addicted. Do they die upon ETC-159 administration, and is this effect rescued by exogenous WNT addition?

We and several others have previously reported that Wnt-addicted cells differentiate and/or senesce upon Wnt withdrawal in vivo but not in vitro. This is related to the broader changes in gene expression in the orthotopic tumors. The effect of PORCN inhibition has been demonstrated by us and others and is rescued by Wnt addition, downstream activation of Wnt signaling by e.g. APC mutation, and, as we show here, stabilized β-catenin.

4) Line 120: the authors write about Figure 1C: "This demonstrates that the growth of β-cat4A cells in vitro largely requires Wnts to activate β-catenin signaling." The opposite is true: control cells require WNT and form less colony with ETC159, while β-cat4A are independent from Wnt secretion.

We appreciate the reviewer pointing out our mis-statement. This error has now been corrected in the revised manuscript.

5) Lines 226-229: "The β-catenin independent repressed genes were notably enriched for motifs bound by homeobox factors including GSC2, POU6F2, and MSGN1. This finding aligns with the known role of non-canonical Wnt signaling in embryonic development" This statement assumes that target genes, or at least the beta-catenin independent ones, are conserved across tissues, including developing organs. This contrasts with the view that target genes in addition to the usual suspects (e.g., AXIN2, SP5 etc.) are modulated tissue-specifically - a view that the authors (and in fact, these reviewers) appear to support in their introduction.

We agree with the reviewer that a majority of Wnt-regulated genes are tissue specific. Indeed, the β-catenin independent Wnt-repressed genes may also be tissue specific. In other tissues, we speculate that other β-catenin independent Wnt-repressed genes may also have homeobox factor binding sites as well and so the general concept remains valid. We do not have sufficient data in other tissues to resolve this issue.

7) The luciferase and mutagenesis work presented in Figure 5 are crystal-clear. One important aspect that remains to be clarified is whether beta-catenin and/or TCF7L2 directly bind to the NRE sites. Or do the authors hypothesize that another factor binds here? We suggest the authors to show TCF7L2 binding tracks at the NRE/WRE motifs in the main figures.

A major question of the reviewers was, can we provide additional evidence that the NRE is bound by LEF/TCF family members. Our initial analysis of more datasets indicates TCF7L2 peaks are enriched on NREs in Wnt-β-catenin responsive cell lines like HCT116 and PANC1. These analyses appear to further support the model that the NRE binds TCF7L2, but we fully agree these analyses can neither prove nor disprove the model.

In our revision, we will analyze additional cut and run datasets as suggested and look at the HEPG2 datasets suggested by reviewer 1. We are concerned about tissue specificity as some of the genes are not expressed in e.g. HEPG2 or HEK293 cells where datasets are available. However, our data continues to support a functional role for the NRE in the modulation of β-catenin regulated genes. The best analysis would be more ChIP-Seq or Cut and Run assays on tissues, not cells, but these studies are beyond what we can do.

What about other TCF/LEFs and beta-catenin? Are there relevant datasets that could be explored to test whether all these bind here during Wnt activation?

As above, We will analyze additional ChIP and Cut & Run datasets to address this question looking at β-catenin and other LEF/TCF family members. We also reflect on the fact that ChIP-Seq does not necessarily imply that the targeted factor (e.g.,TCF7L2) is bound in the target site in all the cells.

The repression might be mediated by beta-catenin partnering with other factors that bind the NRE even by competing with TCF7L2.

We appreciate the insightful comments and now incorporate this into our discussion.

8) In general, while we greatly appreciate the github page to replicate the analysis, we feel that the methods' description is lacking, both concerning analytical details (e.g., the cutoff used for MACS2 peak calling) or basic experimental planning (e.g, how the luciferase assays were performed).

We thank reviewers for the suggestions and will add further details regarding the analysis

and experimental planning in the method sections.

9) The paper might benefit from the addition of quality metrics on the RNA-seq. Interesting for example would be to see a PCA analysis - as a more unbiased approach - rather than the kmeans clustering.

We have this data and will add it to the revised manuscript.

10) It seems that in Figure 3A the clusters are mislabelled as compared to Figure 3B and Figure 1. Here the repressor clusters are labelled DR5, DR6 and DN7 whereas in the rest of the paper they are labelled DR1, DR2 and DN1.

Thank you for pointing out this issue. This has now been corrected in Figure 3.

11) The siCTNNB1 in Figure 5E is described to be a significant effect in the text whereas in Figure 5E this has a p value of 0.075.

Thank you for pointing out the p value did not cross the 0.05 threshold. We have modified the text to remove the word ‘significant’.

12) Line 396: 'Here we confirm and extend the identification of a TCF-dependent negative regulatory element (NRE), where beta-catenin interacts with TCF to repress gene expression'. We suggest caution in stating that beta-catenin and TCF directly repress gene expression by binding to NRE. In the current state the authors do not show that TCF & beta-catenin bind to these elements. See our previous point 7.

We appreciate the suggestion of the reviewers. We will be more cautious in our interpretation.

Further suggestions - or food for thoughts:

13) A frequently asked question in the field concerns the off-target effects of CHIR treatment as opposed to exposure to WNT ligands. CHIR treatment - in parallel to bcat4A overexpression - would allow the authors to delineate WNT independent effects of CHIR treatment and settle this debate.

We thank the reviewers for suggesting this interesting experiment to sort out the non- Wnt effects of GSK3 inhibition. Such a study would require a new set of animal experiments and a different analysis; we think this is beyond the scope of this manuscript.

14) We think that Figure 4C could be strengthened by adding more public TCF-related datasets (e.g., from ENCODE) to confirm the observation across datasets from different laboratories. In particular, the HEPG2 could possibly be improved as there is an excellent TCF7L2 dataset available by ENCODE.

Many more datasets are easily searchable through: https://www.factorbook.org/.

As above, we will analyze the HEPG2 dataset. We plan on updating Fig 4 with data from analysis from different datasets such as (Blauwkamp et al., 2008; Zambanini et al., 2022).

15) The authors show that there is no specific spacing between NREs and WREs. This implies that it is not likely that TCF7L2 recognizes both at the same time through the C-clamp. Do the authors think that there might be a pattern discernible when comparing the location of WRE and NRE in relation to the TCF7L2 ChIP-seq peak summit? This would allow inferring whether TCF7L2 more likely directly binds the WRE (presumably) and if the NRE is bound by a cofactor.

This is an interesting suggestion and we will conduct this analysis as suggested on available datasets (as the result may be different in different tissue types with varying degrees of Wnt/β-catenin signaling).

Reviewer #2 (Evidence, reproducibility and clarity (Required)):

Overall, the study provides a solid framework for understanding noncanonical transcriptional ____outputs of Wnt signaling in a cancer context. The majority of the conclusions are well supported by the data. However, there are a few substantive points that require clarification before the manuscript is ready for publication.

Major Comments

The authors' central claim-that their findings represent a comprehensive analysis of the β-catenin- independent arm of Wnt signaling and uncover a "cis-regulatory grammar" governing Wnt-dependent gene activation versus repression-is overstated based on the presented data.

We appreciate the reviewers concern and will temper our language.

Specifically:

• Figure 3B identifies TF-binding motifs enriched among different Wnt-responsive gene clusters, but the authors only functionally investigate the role of NRE in β-catenin-dependent repression, particularly in the context of TCF motif interaction.

• To support a broader claim regarding cis-regulatory grammar, additional analyses are required:

o What is the distribution of NREs across all clusters? Are they exclusive to β-catenin-dependent repressed clusters, or more broadly present?

The distribution of the NREs is a statistically significant enrichment; they are observed in the repressed clusters more frequently than expected by chance alone, but they are present elsewhere as well. We have tempered our language around the cis-regulatory grammar.

o Do NREs interact with other enriched motifs beyond TCF? Is this interaction specific to repression or also involved in activation?

This is an interesting question beyond the scope of this analysis. Our dataset uses multiple interventions; The NREs may interact with other motifs but we would need more transcriptional analysis data with biological intervention to assess this.

o A more comprehensive analysis of cis-element combinations is needed to draw conclusions about their collective influence on gene regulation across clusters.

We agree; This would be a great question if we had TCF binding data in our orthotopic xenograft model. It’s a dataset we do not have, nor do we have the resources to pursue this.

Other important clarifications:

• The use of the term "wild-type" to describe HPAF-II cells is potentially misleading. These cells are not genetically wild-type and harbor multiple oncogenic alterations.

Thank you for pointing this out. We will use the word “parental” in the text

• The manuscript does not clearly present the kinetics of Wnt target downregulation upon ETC-159 treatment of HPAF-II cells. Understanding whether repression mirrors activation dynamics (e.g., delay or persistence of Wnt effects) is essential to interpreting the system's temporal behavior.

We previously addressed the temporal dynamics of activation and repression in our more comprehensive time course papers (Harmston et al., 2020; Madan et al., 2018); there are differences in the dynamics that are difficult to tease out in this new dataset as the density of time points is less. Having said that, we will compare the time course and annotate the sets of genes identified in this current study with the data from our original study to provide more information on the temporal dynamics of this system.

Minor Comment

• The statement in Figure 1C (lines 119-120) that "growth of β-cat4A cells in vitro largely requires Wnts to activate β-catenin signaling" is inconsistent with the data. As the β-cat4A allele encodes a constitutively active form of β-catenin, Wnts should not be required. Please revise this conclusion for clarity.

We thank the reviewers for pointing out this mis-statement. We have corrected this.

Reviewer #2 (Significance (Required)):

This study offers a systematic classification of Wnt-responsive gene expression dynamics, differentiating between β-catenin-dependent and -independent mechanisms. The insights into temporal expression patterns and the potential role of the NRE element in transcriptional repression add depth to our understanding of Wnt signaling. These findings have relevance for developmental biology, stem cell biology, and cancer research-particularly in understanding how Wnt-mediated repression may influence tumor progression and therapeutic response.

Nice review; thank you.

Reviewer #3 (Evidence, reproducibility and clarity (Required)):

… The work advances understanding of Wnt mediated repression via cis regulatory grammar.

Major Concerns

1) Statistical thresholds and clustering - The criteria for classifying β catenin-dependent versus - independent genes rely on FDR cutoffs above or below 0.1. If the more stringent cutoff of 0.05 was used, how many genes would still be considered Wnt regulated?

We can readily address this in a revised manuscript.

2) Validation of selected β catenin-dependent and -independent Wnt target genes - While the authors identify β catenin-dependent and -independent Wnt target genes (4 selected genes from different clusters in Fig.2), RT-qPCR based validation of Axin2 has been performed in Fig. S3. Authors should also validate other 3 genes as well.

We had considered performing qPCR to re-validate some of our gene-expression changes but qPCR analyses is intrinsically more error prone than RNAseq, and we believe the literature shows that qPCR from the same samples will not add any extra utility. Previous studies that have examined this question have reported excellent correlation between the RNAseq and pPCR (Asmann et al., 2009; Griffith et al., 2010; Wu et al., 2014).

3) NRE mechanistic insight - The most important contribution of this manuscript is the extension of the importance of the NRE motif in Wnt regulated enhancers. But the mutagenesis data provided is insufficient to conclusively nail down that the NREs are responsible for the repression. The effects in the synthetic reporters in Fig. 4D are small - it's not clear that there is much activity in the MimRep to be repressed by the NREs. The data in Fig. 5 is a better context to test the importance of the NREs, but the authors use deletion analysis which is too imprecise and settle for single nucleotide mutants in individual NREs in the ABHD11-AS1 reporter. In the Axin2 report, they mutate sequences outside of the NRE. It's too inconsistent. They should mutate 3 or 4 positions within the NRE in BOTH motifs in the context of the ABHD11-AS1 reporter. Same for the Axin2 reporter.

We feel our analysis, coupled with the Kim paper (Kim et al., 2017), support the role of the NRE. We agree that more data is always desirable, but in our current circumstances are we cannot add additional wetlab experiments.

Regarding Figure 4D, this is a synthetic system lacking the endogenous elements in the promoter. We agree with the reviewer that the effect is small but we would also like to point out that adding the well-established 2WRE in front of the MinRep increased the transcription activity to 1.5 fold, which is of similar magnitude change of the 2NRE deceasing the transcriptional activity 1/1.5 = 0.6.

In Kim et al, it was shown that mutating the 11st nucleotide of the NRE motif showed the strongest effect, so we followed their lead in only mutated the 11st nucleotide in ABHD11- AS1 NRE.

As for the putative NRE sequence present in AXIN2 promoter, its downstream sequence is polyT (__GTGTTTTTTTT__TTTTTTTTTT), if we only mutate 11st nucleotide to G/C, we could create similar sequence to NRE, so we mutated sequences outside of the NRE to fully disrupt it.

4) Even if the mutagenesis is done more completely, the results simply replicate that of the Goentoro group. In Kim et al 2017, they provide suggestive (not convincing) evidence that TCFs directly bind to the NRE. The authors of this manuscript should explore that in more detail, e.g., can purified TCF bind to the NRE sequence? Can the authors design experiments to directly test whether beta-catenin is acting through the NRE - their data currently only demonstrates that the NRE provide a negative input to the reporters - that's an important mechanistic difference.

We point out that our minimal reporter studies with the NRE showed a repressive effect in HCT116 (colorectal cancer cells with stabilized β-catenin) but not HT1080 (sarcoma cells with low Wnt) supporting the importance of β-catenin acting through the NRE (Figs. 4D, 4E).

We fully agree with the reviewers that additional study of TCF interaction with the NRE would be of value. While EMSA and culture-based ChIP assays would be of some value, the best study should be done in vivo where the system is most robust. We are not in a position to do these studies, but we will add in a discussion of this as a limitation of the current study.

5) In vertebrates, some TCFs are more repressive than others and TLEs have been implicated in repressive. Exploring these factors in the context of the NRE would increase the value of this story.

This is an interesting idea but beyond the scope of the current manuscript. It is likely this would be dependent on tissue specific expression, local expression levels, and local binding of co-factors. As we look at other TCF members in other datasets we may be able to address this. Further wetlab experiments are beyond the scope of this work.

**Referees cross-commenting**

I respectfully disagree that the luciferase assays are sufficient. Using deletion analysis to understand the function of specific binding sites is insufficient and the more specific mutations of NREs are incomplete. Regarding this paper extending our knowledge of direct transcriptional repression by Wnt/bcat signaling, I don't agree that it adds much - there are numerous datasets where Wnt signaling activates and represses genes - the trick is determining whether any of the repressed genes are the result and direct regulation by TCF/bcat. They don't explore that. The main finding is an extension of the work by Lea Goentoro on the importance of the NRE motif, but they don't address whether TCF directly associates with this sequence. Goentoro argued in the 2017 paper that it does, but that data is unconvincing to me. Can purified TCF bind the NRE? Without that information (done carefully) this manuscript is very limited.

We respectfully disagree with the reviewer regarding the contribution of this manuscript. There are certainly many datasets looking at Wnt-regulated genes in tissue culture, but these cell-based studies are underpowered to really understand Wnt biology. There are only two papers, ours and Cantú’s, that address Wnt repressed genes in any depth. No prior papers have differentiated β-catenin dependent from β-catenin independent genes before, and certainly not in an orthotopic animal model.

A major impact of our study is the finding that only 10% of Wnt regulated genes are independent of β-catenin, at least in pancreatic cancer. We feel this is a major contribution. We further add to this analysis by re-enforcing/extend the prior evidence on the NRE in humans (and correct the motif sequence!) for Wnt-repressed genes. Our data supports the fine-tuning of the Wnt/β-catenin regulated genes by a cis-regulatory grammar.

Reviewer #3 (Significance (Required)):

Overall, this study advances our understanding of the dual roles of Wnt signaling in gene activation and repression, highlighting the role of the NRE motif. But this is an extension of the original NRE paper (Kim et al 2017) with no mechanistic advance beyond that original work. The transcriptomics in the first part of the manuscript have some value, but similar data sets already exist.

We respectfully but strongly disagree with the reviewer. First, our work examines the NRE in a large-scale in vivo transcriptome dataset, significantly extending the candidate gene approach of Kim et al. Secondly, we disagree with the comment that “similar data sets already exist.” Indeed, reviewer 1 (C. Cantú) specifically pointed out we had addressed an “yet-unsolved question in the field” on whether and how β-catenin repressed genes.

__3. __Description of the revisions that have already been incorporated in the transferred manuscript

To date we have only corrected several typographical errors.

1. Description of analyses that authors prefer not to carry out

We fully agree with the reviewers that additional study of TCF interaction with the NRE would be of value. While EMSA and cell culture-based ChIP assays would be of some modest value, they have already been done in vitro by Kim et al. (Kim et al., 2017) and the best next study should be done in vivo in Wnt-responsive cancers or tissues where the biology is most robust (Madan et al., 2018) . We are not in a position to do these studies, but we will add this into the discussion as a limitation of the current study. We also acknowledge that the NRE may interact with other currently unidentified factors.

Reviewer 1 asked about considering experiments to determine non-Wnt effects of GSK3 inhibitors like CHIR. Such a study, while interesting, would require a new set of animal experiments and a different analysis; we think this is beyond the scope of this manuscript.

Finally, we note that the Virshup lab at Duke-NUS Medical School in Singapore, where these in vivo studies were performed, has closed as of July 1, 2025 and the various lab members have moved on to new adventures. Because of this, we are unable to undertake new wet-lab studies.

Thank you for your consideration,

For the authors,

David Virshup

References:

Asmann YW, Klee EW, Thompson EA, Perez EA, Middha S, Oberg AL, Therneau TM, Smith DI,

Poland GA, Wieben ED, Kocher J-PA. 2009. 3’ tag digital gene expression profiling of human

brain and universal reference RNA using Illumina Genome Analyzer. BMC Genom 10:531–

1. doi:10.1186/1471-2164-10-531

Blauwkamp TA, Chang MV, Cadigan KM. 2008. Novel TCF-binding sites specify transcriptional

repression by Wnt signalling. The EMBO Journal 27:1436–1446. doi:10.1038/emboj.2008.80

Conway T, Wazny J, Bromage A, Tymms M, Sooraj D, Williams ED, Beresford-Smith B. 2012.

Xenome—a tool for classifying reads from xenograft samples. Bioinformatics 28:i172–i178.

doi:10.1093/bioinformatics/bts236

Griffith M, Griffith OL, Mwenifumbo J, Goya R, Morrissy AS, Morin RD, Corbett R, Tang MJ, Hou

Y-C, Pugh TJ, Robertson G, Chittaranjan S, Ally A, Asano JK, Chan SY, Li HI, McDonald H,

Teague K, Zhao Y, Zeng T, Delaney A, Hirst M, Morin GB, Jones SJM, Tai IT, Marra MA.

1. Alternative expression analysis by RNA sequencing. Nat Methods 7:843–847.

doi:10.1038/nmeth.1503

Harmston N, Lim JYS, Arqués O, Palmer HG, Petretto E, Virshup DM, Madan B. 2020.

Widespread Repression of Gene Expression in Cancer by a Wnt/β-Catenin/MAPK Pathway.

Cancer Res 81:464–475. doi:10.1158/0008-5472.can-20-2129

Killion JJ, Radinsky R, Fidler IJ. 1998. Orthotopic models are necessary to predict therapy of

transplantable tumors in mice. Cancer metastasis reviews 17:279–284.

Kim K, Cho J, Hilzinger TS, Nunns H, Liu A, Ryba BE, Goentoro L. 2017. Two-Element

Transcriptional Regulation in the Canonical Wnt Pathway. Current Biology 27:2357-2364.e5.

doi:10.1016/j.cub.2017.06.037

Madan B, Harmston N, Nallan G, Montoya A, Faull P, Petretto E, Virshup DM. 2018. Temporal

dynamics of Wnt-dependent transcriptome reveals an oncogenic Wnt/MYC/ribosome axis. J

Clin Invest 128:5620–5633. doi:10.1172/jci122383

Wu AR, Neff NF, Kalisky T, Dalerba P, Treutlein B, Rothenberg ME, Mburu FM, Mantalas GL,

Sim S, Clarke MF, Quake SR. 2014. Quantitative assessment of single-cell RNA-sequencing

methods. Nat Methods 11:41–46. doi:10.1038/nmeth.2694

Zambanini G, Nordin A, Jonasson M, Pagella P, Cantù C. 2022. A new cut&run low volume-

urea (LoV-U) protocol optimized for transcriptional co-factors uncovers Wnt/b-catenin tissue-

specific genomic targets. Development 149. doi:10.1242/dev.201124

Je m'interroge sur le choix de la graphie. Y a-t-il une raison pour laquelle on le note « symétrie - asymétrie » plutôt qu'avec une barre oblique?

Compte rendu : Le Recueil de la Parole des Enfants Victimes et le Défi des Enquêtes sur les Agressions Sexuelles

Ce document examine les méthodes et les défis rencontrés par la police, en particulier la brigade des mineurs, lors du recueil de la parole des enfants victimes de violences, notamment sexuelles.

Il met en lumière l'évolution des pratiques d'audition, l'importance des environnements adaptés et la complexité de l'enquête judiciaire face à des témoignages d'enfants.

1. L'Évolution du Recueil de la Parole des Enfants : Les "Salles Mélanie"

Historiquement, l'audition des enfants victimes se faisait dans des "bureaux ordinaires, des endroits où les conditions de confort et d'écoute n'étaient pas réunies pour avoir leur témoignage souvent si douloureux."

Conscients de cette lacune, de nouvelles approches ont été mises en place, notamment la création de "salles Mélanie".

• Concept et Aménagement : Les salles Mélanie, nommées d'après la première enfant auditionnée dans ces conditions, sont des "pièces spécialement aménagées avec du mobilier, des jeux pour enfants, des couleurs chaudes et qui permettent un petit peu comme un petit cocon de recueillir la parole de ces enfants de manière plus efficace et plus délicate."

L'objectif est de créer un environnement qui "ressemble à une école, ça ressemble à ce qu'ils ont d'habitude, ça ressemble à une chambre d'enfance, ça ressemble pas à un bureau", afin de les "sécuriser" et qu'ils se "sentent en sécurité ici."

• Outils Pédagogiques : Ces salles sont équipées de divers outils pour faciliter l'expression des enfants.

Les poupées anatomiques, par exemple, sont essentielles : "Il y a un garçon et une fille et on peut complètement les déshabiller de des chaussures jusqu'au vêtements...

Ce qui fait que ça permet à l'enfant de matérialiser s'il souhaite pas parler, ben finalement il peut nous expliquer à travers les les poupons en fait."

Des dessins ou l'utilisation d'objets comme des stylos peuvent également aider l'enfant à "montrer ce qu'il a demandé de faire".

• Impact sur le Témoignage : Les enquêteurs estiment que ces environnements permettent une parole plus "précise" et plus libre.

Un policier souligne que l'enfant "va révéler le traumatisme qu'il a pu vivre."

2. La Spécificité de l'Audition des Mineurs

L'audition d'un enfant diffère grandement de celle d'un adulte ou même d'un adolescent, nécessitant des compétences et des approches spécifiques.

• Difficulté d'Expression : Contrairement aux adolescents qui "vont tout nous expliquer", l'enfant "va falloir aller chercher un peu plus les éléments."

Ils peuvent bloquer, avoir peur, ou manquer de vocabulaire pour exprimer ce qu'ils ont vécu.

Certains "font que des signes de la tête qui veulent pas parler mais par contre qui vont nous répondre à travers ce gestuel surtout les petits parce qu'ils ont peut-être moins de vocabulaire moins la facilité de s'exprimer."

• Techniques d'Audition Modernes : L'affaire Outreau a marqué un tournant, amenant de "grands changements par rapport au recueil de la parole de l'enfant."

Auparavant, les questions étaient "dirigées" ("il a mis son zizi dans ton sexe"), ce qui pouvait influencer l'enfant.

Désormais, la technique consiste à aborder le sujet "du général" et d'y aller "petit à petit", pour que l'enfant "puisse dire ce qui s'est passé exactement mais de façon moins brutale et moins l'influencer aussi."

L'enquêteur doit s'adapter au "terme" propre à l'enfant pour désigner les parties intimes ou les actes.

La Vérité et le Mensonge : Une règle fondamentale est de "dire que la vérité et puis jamais de mensonge."

L'enfant est invité à donner un exemple de mensonge, comme "Il m'a rien fait" si un problème est survenu.

Cependant, la déstabilisation de l'enquêteur face à un enfant qui ne parle pas révèle la complexité de cette approche.

3. La Charge Émotionnelle et Psychologique des Enquêteurs

Le travail à la brigade des mineurs est humainement très exigeant et impacte profondément les policiers.

• Sensibilité du Travail : Les enquêteurs sont confrontés à des "affaires de mœurs, de viol" impliquant des enfants, ce qui est "pas forcément évident".

Un enquêteur confie que cela "nous bouleverse", car ils ne sont "pas dénué du tout de sentiments.

Bien au contraire." Ils doivent cependant "essayer de mettre de la distance parce que sinon c'est trop ça serait trop compliqué au quotidien."

• Mécanismes d'Adaptation : Avec le temps, les enquêteurs se "forge une sorte de carapace".

Le fait d'en "parler aussi, c'est aussi une façon de de se libérer quoi, de passer à autre chose, de pas garder tout pour nous."

L'humour peut parfois être utilisé comme "une façon de détourner en fait."

Certains dossiers restent cependant "ancrés dans un coin du cerveau et c'est vrai que parfois c'est difficile de faire le vide en rentrant à la maison."

• Motivation : Malgré la difficulté, les enquêteurs trouvent ce domaine "passionnant, intéressant" car "un enfant, il a pas moyen de défense.

Donc c'est vrai que c'est important de pouvoir l'aider à notre niveau si on si on peut essayer de comprendre et puis de lui trouver une solution."

4. Le Processus Judiciaire et ses Limites

Le recueil de la parole de l'enfant n'est que le début d'un processus judiciaire complexe où la preuve est primordiale.

• Le Doute et la Preuve : La crédibilité de la parole de l'enfant est au centre de l'enquête.

"C'est difficile de dire au départ si la parole de l'enfant est crédible ou pas."

L'enquête vise à "vérifier les paroles que l'enfant a pu dénoncer à travers des auditions de témoin, à travers des examens médicaux, à travers des choses comme ça.

Parce que justement, il faut bah il faut étayer les paroles de l'enfant."

• L'Absence de Preuves Matérielles : L'un des cas présentés illustre la difficulté lorsque le mis en cause "nie la totalité des faits" et qu'il n'y a "aucun autre élément matériel" pour corroborer la version de l'enfant, y compris le certificat médical. Dans ce cas, "c'est une parole, contre une autre" et "on n'a rien de plus."

• Le Classement Sans Suite : Face à l'insuffisance de preuves, le parquet peut décider un "classement 21" (classement sans suite).

Cette décision est une source de "frustration" pour les enquêteurs car, même si cela ne signifie pas que la parole de l'enfant est mise en doute, "on n'a pas assez d'éléments pour vous confronter." C'est une application du principe "le doute profite à l'accusé."

• Conséquences : Malgré la libération de l'accusé, le passage par les services de police et l'inscription dans les fichiers constituent un "précédent" pour le futur.

L'exemple de Zoé, toujours placée en famille d'accueil et ne vivant plus avec son beau-père, montre que des mesures de protection sont prises indépendamment de l'issue judiciaire.

En conclusion, si la création des salles Mélanie et l'évolution des techniques d'audition représentent une avancée majeure pour mieux recueillir la parole des enfants victimes, la phase d'enquête et la confrontation aux limites de la preuve matérielle restent un défi constant pour la justice et un lourd fardeau pour les enquêteurs.

Document d'information détaillé : Violence des mineurs et culture du couteau en France

Ce document analyse les thèmes principaux, les faits marquants et les idées essentielles concernant l'augmentation de la violence chez les mineurs en France, en particulier l'usage des armes blanches.

Il s'appuie sur les extraits de l'émission "Envoyé spécial - Coups de couteaux chez les ados".

Thèmes principaux

• L'augmentation alarmante de la violence à l'arme blanche chez les mineurs : Le reportage met en évidence une multiplication des agressions, souvent filmées et partagées sur les réseaux sociaux.
• La "culture du couteau" et son accessibilité : De nombreux adolescents portent des couteaux pour se défendre ou pour intimider, et l'acquisition de ces armes est étonnamment facile pour les mineurs.
• Les motivations derrière le port d'armes et les agressions : Peur des représailles, rivalités de quartiers, quête de réputation, influence des réseaux sociaux et de la musique.
• L'impact dévastateur sur les victimes et leurs familles : Traumatismes physiques et psychologiques, sentiment d'insécurité, et détresse des parents.
• Les réponses des autorités et des acteurs de la prévention : Contrôles policiers, actions de sensibilisation, et limites du système judiciaire et éducatif.

Faits et idées les plus importants

• Multiplication des drames : Le reportage débute en soulignant que "Depuis quelques mois, les drames se sont multipliés à l'extérieur et même à l'intérieur des lycées ou des collèges."
• Cas emblématiques :Élias, 14 ans : Tué à la machette à Paris en janvier pour un vol de téléphone portable.
• Sou, 17 ans : Poignardé à mort dans le dos en mars.
• Jeune fille de 15 ans à Nantes : Poignardée "à 57 reprises" par un élève de son établissement en avril.
• Victime à Perpignan : Un lycéen de 15 ans poignardé dans le dos par des agresseurs de quartier rivaux au sein même de son lycée. Il a subi une "perforation des intestins" et porte de graves cicatrices.

Le père de la victime témoigne de l'horreur : "Je jamais vu ça. Même moi quand j'allais quand j'étais jeune et j'allais ici à l'école non je suis bagarré à coup de point et tout.

Bon après le lendemain j'étais réglé on était pote. C'est bon mais jamais coup de de couteau."

Prévalence du port de couteaux :

• Ryan, 16 ans (Marseille) : "J'ai un opinel. [...] C'est pour me protéger. Je préfère toujours en avoir un sur moi que de pas en avoir. C'est toujours mieux."

Il ajoute : "Si le mec il a un couteau, je préfère moi aussi en avoir un. Je préfère le planter qui me plante. C'est logique ça."

• Il dissimule son couteau "dans les parties intimes soit dans la sacoche soit là je le mets dans la chaussette".

• Les couteaux Opinel sont "le couteau à la mode.

C'est plus facile d'acheter un couteau que d'acheter un filet au fich à Mcdo. Mettre moins de temps." * Chiffres alarmants : "Rien qu'en 2024, 10400 agressions à l'arme blanche ont été perpétrées en France." Cependant, il est "difficile de savoir précisément combien concernent les mineurs. Faute de statistique." * Facilité d'acquisition des couteaux pour les mineurs : * La loi est claire : "Un couteau à la fixe comme un couteau de bouché, un poignard ou un cran d'arrêt sont des armes de catégorie D.

Leur détention et leur transport sont formellement interdits aux mineurs." * Pourtant, un adolescent de 17 ans a pu acheter un couteau dans un supermarché discount et une quincaillerie sans vérification d'âge.

• Sur internet, l'achat est encore plus simple : "Il lui suffit de mentir sur son âge et en quelques clics, la commande est passée.

Aucune vérification de l'âge ou de l'identité de l'acheteur n'est effectuée. 48 heures plus tard, nous recevons le couteau à notre domicile. Bref, acheter un couteau, rien de plus facile pour un mineur."

Causes et facteurs aggravants :

• Rivalités de quartiers / "Match retour" : Plusieurs agressions sont liées à des vengeances inter-quartiers, comme l'agression à Perpignan.

"L'adolescent agressé serait une victime de ce que les bandes appellent un match retour."

• Réseaux sociaux : "Les agressions entre ados à l'arme blanche se multiplient en France. [...]

Ces drames font le tour des réseaux sociaux et la une des journaux."

Les jeunes se donnent rendez-vous pour se battre via des applications comme Snap, et les vidéos d'agressions sont partagées sur des boucles cryptées (Telegram) : "Avec les outils numériques, l'embrouille s'accélère."

• Quête de réputation : Selon le sociologue Marwan Mohamed, "C'est quoi le carburant des embrouilles des rivalités de quartier ? C'est la réputation.

C'est une compétition symbolique."

La rue offre "des gratifications immédiates et des gratifications puissantes" aux jeunes en échec scolaire ou sans emploi. * Musique : La "culture du couteau est présente dans de nombreux morceaux de rap comme celui du chanteur Jul paru il y a quelques années. dans lequel il fait l'apologie d'un célèbre couteau avec son manche en bois."

Réponse de la justice et de la prévention : * Contrôles policiers : Les forces de l'ordre effectuent des contrôles de sacs aux abords des lycées. "94 armes blanches ont été saisies à l'occasion de près de 1000 opérations de contrôle entre les mois de mars et d'avril."

Ces actions visent la "prévention" et "une présence dissuasive", mais sont limitées car les policiers "n'ont pas le droit d'effectuer des fouilles au corps".

• Justice des mineurs : La Présidente du tribunal pour enfants de Bobigny, Muriel Léglin, affirme ne pas constater une "flambée" mais une augmentation de 12% des saisines en 2024 par rapport à 2023, sans toutefois atteindre les niveaux de 2018-2019.

Elle insiste sur le rôle "nouveau" et problématique des réseaux sociaux.

• Non-laxité de la justice : Elle réfute l'idée d'une justice laxiste ou trop lente, expliquant la rapidité des procédures depuis la réforme de 2021 (saisine du juge entre 10 jours et 3 mois).

"À partir de 13 ans, on est responsable pénalement. Vous avez l'âge d'aller en prison."

• Manque de moyens : Le problème majeur réside dans l'exécution des mesures éducatives ou de contrôle judiciaire :

"nous avons 200 mesures éducatives ou mesures de contrôle judiciaire qui ont été confiées à la protection judiciaire de la jeunesse et qui ne sont pas exécutées faute d'effectif éducatif."

• Importance de la prévention : Elle souligne que la justice est souvent "la voiture balai de ce qui n'a pas fonctionné auparavant à savoir la prévention, la scolarité qui ne fonctionne pas.

On a une proportion de jeunes qui sont en échec scolaire et en désertion scolaire, en décrochage scolaire qui est extrêmement importante et les difficultés familiales globales".

Actions de sensibilisation :

• Adama Camara : Fondateur de l'association Sada Solidaire, il a perdu son jeune frère poignardé en 2011.

Il fait de la prévention dans les collèges et lycées, soulignant les conséquences de la violence : "les embrouilles s'amèent à deux choses, voire trois. La mort, la prison, le handicap."

Il insiste : "arrêter de marcher avec des couteaux. On se protège pas avec un couteau. [...] Quand tu marches avec un couteau dans la poche, tu marches avec l'arme du crime."

• Cours de self-défense : Certains parents inscrivent leurs enfants à des cours de "boxe de rue" pour apprendre à se défendre contre les agressions au couteau.

"Ils apprennent à se défendre, à se débrouiller surtout dans la vie au quotidien parce que ce sont des situations qui se passent au quotidien."

Conclusion

Le reportage brosse un tableau préoccupant de la violence à l'arme blanche chez les mineurs en France.

Il met en lumière non seulement la gravité des actes et leurs conséquences dramatiques, mais aussi la complexité des facteurs sous-jacents (rivalités de quartiers, influence des réseaux sociaux, facilité d'accès aux armes, quête de réputation).

Si la justice des mineurs se veut plus rapide et non laxiste, elle est freinée par un manque criant de moyens pour l'application des mesures éducatives et un système de prévention en amont qui semble défaillant.

Les initiatives individuelles de sensibilisation et de self-défense tentent de combler ces lacunes face à un phénomène qui bouleverse les familles et les communautés éducatives.

Synthèse du Briefing : PFAS - La Grande Intox de l'Industrie

• Ce briefing explore la problématique des PFAS (substances per- et polyfluoroalkylées), surnommés "polluants éternels", en se basant sur une enquête journalistique approfondie, notamment en France avec les cas de l'entreprise Tefal (groupe Seb) et Arkema, ainsi qu'aux États-Unis.

Il met en lumière la dangerosité de ces molécules, les efforts de l'industrie pour dissimuler leurs effets et retarder la réglementation, et les conséquences sanitaires et environnementales.

1. Les PFAS : Molécules "Miracles" devenues "Polluants Éternels"

Les PFAS sont des molécules chimiques synthétiques, plus de 10 000 recensées, créées dans les années 1930.

Elles ont révolutionné de nombreux secteurs grâce à leurs propriétés uniques : "ultra résistantes à la chaleur, imperméables à l'eau." Initialement utilisées pour la fabrication de la bombe atomique, leurs applications industrielles se sont multipliées : "poêles anti-adhésives, batteries électriques, vêtements imperméables, cosmétiques."

Le problème majeur des PFAS est leur persistance : "Indestructibles, ils s'accumulent dans l'environnement. On les appelle les polluants éternels."

On les retrouve "dans le corps des gens, dans la faune et la flore, dans l'air, dans l'eau. Toute notre planète est contaminée."

2. Dangers pour la Santé et l'Environnement

L'enquête révèle que la dangerosité de certaines de ces substances est connue depuis des décennies par les industriels, qui auraient "dissimulé pendant très longtemps ce qu'ils savaient sur ces molécules."

Le PFOA (acide perfluorooctanoïque), un PFAS particulièrement nocif utilisé par Tefal jusqu'en 2012, a été classé "cancérogène pour l'homme en 2023" par le Centre international de recherche sur le cancer (OMS).

L'exposition aux PFAS est associée à de nombreux effets sur la santé :

Cholestérol et diabète.

• Cancers : "cancer des testicules et du rein," "augmenterait aussi le risque de cancer du sein."
• Système immunitaire : Baisse de l'efficacité des vaccins et des antibiotiques, notamment chez les enfants exposés.
• Impact sur le fœtus : Le PFOA "modifierait le génome du fétus, augmentant sa probabilité de développer du diabète ou un futur cancer." Un expert déclare : "Nous transmettons désormais la toxicité à la prochaine génération qui paiera pour cela."
• Foie : Une exposition répétée peut provoquer des "effets néfastes sur le foie."

Un ancien salarié de Tefal témoigne de la détérioration de son foie avec des enzymes hépatiques "jusqu'à deux fois supérieures à la normale."

• Dans l'environnement, les PFAS contaminent "nos légumes, dans nos œufs, dans nos eaux potables." La contamination des nappes phréatiques peut être "pour l'éternité."

3. Dissimulation et Stratégies Industrielles

L'enquête dénonce les décennies d'efforts des industriels pour "tromper les pouvoirs publics, tromper la communauté scientifique à propos de leur impact réel sur la santé."

Tefal/Groupe Seb à Rumilly (Haute-Savoie) :

• Rejets historiques : Tefal a utilisé "plus d'une tonne par an" de PFOA jusqu'en 2012. L'usine a rejeté "des substances chimiques dans l'environnement" et des documents confidentiels révèlent que des "boues de traitement des effluants environ 30000 m³ ont été déposées dans une ancienne carrière" (le plan d'eau des Puys) et d'autres sites, souvent "à même le sol sans bâche de protection."
• Connaissance des risques : Dès 2002, une fiche de sécurité d'un fournisseur de Tefal précisait que le PFOA "est absorbé à travers les poumons, l'appareil gastro-intestinal et la peau et n'est éliminé que très lentement du sang. Une exposition répétée au-dessus de la limite d'exposition professionnelle peut provoquer des effets néfastes sur le foie."
• Déni et communication : Le directeur général de Seb affirme que "Le PFO n'est pas considéré comme nocif pour les ouvriers et nous n'avons eu aucune maladie professionnelle d à ça."

Pourtant, un email confidentiel de 2024 évoque des "campagnes de surveillance biologique" montrant une "corrélation entre la suppression du PFOA dans les matières premières et la diminution des valeurs biologiques mesurées," suggérant une présence antérieure dans le sang des ouvriers.

• Pollution de l'eau potable : En novembre 2022, l'eau potable de Rumilly a été "brusquement considérée impropre à la consommation" en raison d'une contamination aux PFAS.

Les habitants ont bu de l'eau contaminée pendant des décennies, certains affichant des taux de PFOA "cinq fois plus que la moyenne française" dans le sang.

Arkema (Sud de Lyon) :

• Connaissance précoce : Dès 2006, Arkema était "informé de la contamination des populations aux alentours des usines américaines," avec un lien clair entre les concentrations de PFA dans l'eau potable et le sang.
• Stratégie de défense : En 2009, des industriels, dont Arkema, réfléchissaient déjà à "une stratégie de défense" sur la question de la présence des PFAS dans l'eau potable "avant qu'il n'apparaisse dans les médias ou qu'il n'engendre des pressions réglementaires."
• Solution connue mais retardée : Dès 2013, Arkema avait validé en laboratoire une solution (charbons actifs) pour "réduire de manière très importante les émissions" de PFAS, mais n'a installé la station qu'en 2022, "presque 10 ans plus tard."
• Dissimulation de rapport : Une version tronquée d'un rapport d'analyse indépendant a été transmise à la justice par Arkema, supprimant une partie de la conclusion "détaillant pourquoi les molécules retrouvées dans la nappe pourrait en partie provenir des rejets de l'usine Archema."

Un journaliste dénonce : "Ce sont des industriels qui fabriquent en permanence ce doute et qui n'ont pas peur de tronquer leur rapport."

4. Lobbying et Influence Politique

Les industriels déploient des efforts considérables pour "empêcher leur interdiction" et "retarder, retarder, retarder pour repousser le plus loin possible toute réglementation qui pourrait affecter ses produits, ses procédés et au final ses profits."

• Tefal face à la loi française : Lors de l'examen d'une loi visant à interdire les polluants éternels, Tefal a mené une "vaste campagne de lobbying," allant jusqu'à "payer les salariés pour faire grève" et manifestants devant l'Assemblée nationale.

L'entreprise a fourni aux parlementaires des "éléments de langage" défendant les fluoropolymères, un type de PFAS, comme étant "peu préoccupant aux impacts insignifiants sur la santé et sur l'environnement" et "incapable de traverser la barrière gastro-intestinale."

L'OCDE réfute avoir "finalisé de telles conclusions," et le directeur de recherche au CNRS, Pierre Labadi, déclare que cette présentation est "tronquée" car "le caractère extrêmement persistant de l'ensemble d'épiface" est "avéré."

• Résultat du lobbying : Les ustensiles de cuisine ont été "exclus du champ d'application de la loi" en première lecture, un député écologique dénonçant une "grosse pression au niveau de Ber."

• Lobbying européen : Au niveau européen, l'industrie a inondé les autorités de "milliers de mails d'études socio-économiques, de rapports scientifiques" avec les mêmes arguments.

5. Coût de la Dépollution et Responsabilité

La dépollution des sites contaminés est un défi "titanesque."

Le coût estimé de la dépollution de toute l'Europe est de "100 milliards d'euros par an, plus de la moitié du budget annuel de l'Union européenne."

• Responsabilité industrielle : Les autorités locales et les associations estiment que les pollueurs devraient payer. La métropole de Lyon, confrontée à une dépollution de l'eau potable coûtant "entre 5 et 10 millions d'euros," a saisi la justice, refusant que "l'usager qui paie qui qui subit à la pollution et qui paye le coût de la dépollution."
• Déni de responsabilité : Tefal a versé 3600 € à la collectivité de Rumilly pour les filtres à charbon, mais le directeur général refuse d'admettre la pleine responsabilité, arguant qu'il n'y avait "pas de norme" à l'époque et que "il y a au moins cinq entreprises mobilisées par l'adréal" sur le site.
• Non-reconnaissance des maladies professionnelles : Tefal nie toute maladie professionnelle liée aux PFAS, malgré les témoignages d'anciens salariés.

Conclusion

Le scandale des PFAS est symptomatique d'une industrie chimique qui a sciemment dissimulé les dangers de ses produits pendant des décennies, privilégiant les profits.

Les conséquences sanitaires et environnementales sont massives et durables, avec un coût de dépollution faramineux qui soulève la question de la responsabilité des industriels.

La bataille pour une réglementation stricte et l'indemnisation des victimes est loin d'être terminée, face à un lobbying industriel puissant.

Comme le souligne un militant, "Chaque mois, chaque année qui passe, ça se compte en vie humaine.

Donc c'est absolument inacceptable."

DOI: 10.1208/s12248-025-01093-y

Resource: None

Curator: @sjvitug

SciCrunch record: RRID:IMSR_JAX:029720

What is this?

DOI: 10.1208/s12248-025-01093-y

Resource: RRID:IMSR_JAX:000664

Curator: @sjvitug

SciCrunch record: RRID:IMSR_JAX:000664

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 51064-2-AP, RRID:AB_11042321)

Curator: @sjvitug

SciCrunch record: RRID:AB_11042321

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# SA00001-2, RRID:AB_2722564)

Curator: @sjvitug

SciCrunch record: RRID:AB_2722564

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AE009, RRID:AB_2771925)

Curator: @sjvitug

SciCrunch record: RRID:AB_2771925

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# c17.9c6, RRID:AB_528410)

Curator: @sjvitug

SciCrunch record: RRID:AB_528410

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 66008-4-Ig, RRID:AB_2918475)

Curator: @sjvitug

SciCrunch record: RRID:AB_2918475

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AS003, RRID:AB_2769851)

Curator: @sjvitug

SciCrunch record: RRID:AB_2769851

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9578, RRID:AB_2721060)

Curator: @sjvitug

SciCrunch record: RRID:AB_2721060

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 4370, RRID:AB_2315112)

Curator: @sjvitug

SciCrunch record: RRID:AB_2315112

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 3724, RRID:AB_1549585)

Curator: @sjvitug

SciCrunch record: RRID:AB_1549585

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 8146, RRID:AB_10950495)

Curator: @sjvitug

SciCrunch record: RRID:AB_10950495

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# anti-spitz, RRID:AB_528474)

Curator: @sjvitug

SciCrunch record: RRID:AB_528474

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Rat-Elav-7E8A10 anti-elav, RRID:AB_528218)

Curator: @sjvitug

SciCrunch record: RRID:AB_528218

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2A1, RRID:AB_2109711)

Curator: @sjvitug

SciCrunch record: RRID:AB_2109711

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Smoothened (20C6), RRID:AB_528472)

Curator: @sjvitug

SciCrunch record: RRID:AB_528472

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Drosophila Ptc (Apa 1), RRID:AB_528441)

Curator: @sjvitug

SciCrunch record: RRID:AB_528441

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# N2 7A1 ARMADILLO, RRID:AB_528089)

Curator: @sjvitug

SciCrunch record: RRID:AB_528089

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: Adobe Photoshop (RRID:SCR_014199)

Curator: @sjvitug

SciCrunch record: RRID:SCR_014199

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: ImageJ (RRID:SCR_003070)

Curator: @sjvitug

SciCrunch record: RRID:SCR_003070

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: GraphPad Prism (RRID:SCR_002798)

Curator: @sjvitug

SciCrunch record: RRID:SCR_002798

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 4d4, RRID:AB_528512)

Curator: @sjvitug

SciCrunch record: RRID:AB_528512

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2b10, RRID:AB_528186)

Curator: @sjvitug

SciCrunch record: RRID:AB_528186

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 1g9, RRID:AB_528278)

Curator: @sjvitug

SciCrunch record: RRID:AB_528278

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32733, RRID:AB_2633282)

Curator: @sjvitug

SciCrunch record: RRID:AB_2633282

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32732, RRID:AB_2633281)

Curator: @sjvitug

SciCrunch record: RRID:AB_2633281

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# f461.3b, RRID:AB_528409)

Curator: @sjvitug

SciCrunch record: RRID:AB_528409

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 40-1a, RRID:AB_528100)

Curator: @sjvitug

SciCrunch record: RRID:AB_528100

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 5H7B11, RRID:AB_579779)

Curator: @sjvitug

SciCrunch record: RRID:AB_579779

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 3B8D12, RRID:AB_579781)

Curator: @sjvitug

SciCrunch record: RRID:AB_579781

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 3A6B4, RRID:AB_579780)

Curator: @sjvitug

SciCrunch record: RRID:AB_579780

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9701, RRID:AB_331535)

Curator: @sjvitug

SciCrunch record: RRID:AB_331535

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Molecular Probes Cat# A-21434, RRID:AB_141733)

Curator: @sjvitug

SciCrunch record: RRID:AB_141733

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32727, RRID:AB_2633276)

Curator: @sjvitug

SciCrunch record: RRID:AB_2633276

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32733, RRID:AB_2633282)

Curator: @scibot

SciCrunch record: RRID:AB_2633282

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Molecular Probes Cat# A-21434, RRID:AB_141733)

Curator: @scibot

SciCrunch record: RRID:AB_141733

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32727, RRID:AB_2633276)

Curator: @scibot

SciCrunch record: RRID:AB_2633276

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 66008-4-Ig, RRID:AB_2918475)

Curator: @scibot

SciCrunch record: RRID:AB_2918475

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AE009, RRID:AB_2771925)

Curator: @scibot

SciCrunch record: RRID:AB_2771925

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 51064-2-AP, RRID:AB_11042321)

Curator: @scibot

SciCrunch record: RRID:AB_11042321

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AS003, RRID:AB_2769851)

Curator: @scibot

SciCrunch record: RRID:AB_2769851

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# SA00001-2, RRID:AB_2722564)

Curator: @scibot

SciCrunch record: RRID:AB_2722564

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32732, RRID:AB_2633281)

Curator: @scibot

SciCrunch record: RRID:AB_2633281

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 8146, RRID:AB_10950495)

Curator: @scibot

SciCrunch record: RRID:AB_10950495

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 4370, RRID:AB_2315112)

Curator: @scibot

SciCrunch record: RRID:AB_2315112

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# anti-spitz, RRID:AB_528474)

Curator: @scibot

SciCrunch record: RRID:AB_528474

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 3724, RRID:AB_1549585)

Curator: @scibot

SciCrunch record: RRID:AB_1549585

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Rat-Elav-7E8A10 anti-elav, RRID:AB_528218)

Curator: @scibot

SciCrunch record: RRID:AB_528218

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9701, RRID:AB_331535)

Curator: @scibot

SciCrunch record: RRID:AB_331535

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9578, RRID:AB_2721060)

Curator: @scibot

SciCrunch record: RRID:AB_2721060

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# c17.9c6, RRID:AB_528410)

Curator: @scibot

SciCrunch record: RRID:AB_528410

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# N2 7A1 ARMADILLO, RRID:AB_528089)

Curator: @scibot

SciCrunch record: RRID:AB_528089

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 40-1a, RRID:AB_528100)

Curator: @scibot

SciCrunch record: RRID:AB_528100

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2b10, RRID:AB_528186)

Curator: @scibot

SciCrunch record: RRID:AB_528186

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Drosophila Ptc (Apa 1), RRID:AB_528441)

Curator: @scibot

SciCrunch record: RRID:AB_528441

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 3A6B4, RRID:AB_579780)

Curator: @scibot

SciCrunch record: RRID:AB_579780

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# f461.3b, RRID:AB_528409)

Curator: @scibot

SciCrunch record: RRID:AB_528409

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Smoothened (20C6), RRID:AB_528472)

Curator: @scibot

SciCrunch record: RRID:AB_528472

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: Adobe Photoshop (RRID:SCR_014199)

Curator: @scibot

SciCrunch record: RRID:SCR_014199

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2A1, RRID:AB_2109711)

Curator: @scibot

SciCrunch record: RRID:AB_2109711

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: ImageJ (RRID:SCR_003070)

Curator: @scibot

SciCrunch record: RRID:SCR_003070

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 4d4, RRID:AB_528512)

Curator: @scibot

SciCrunch record: RRID:AB_528512

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: GraphPad Prism (RRID:SCR_002798)

Curator: @scibot

SciCrunch record: RRID:SCR_002798

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 1g9, RRID:AB_528278)

Curator: @scibot

SciCrunch record: RRID:AB_528278

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 3B8D12, RRID:AB_579781)

Curator: @scibot

SciCrunch record: RRID:AB_579781

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 5H7B11, RRID:AB_579779)

Curator: @scibot

SciCrunch record: RRID:AB_579779

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (IMSR Cat# JAX_012886,RRID:IMSR_JAX:012886)

Curator: @sjvitug

SciCrunch record: RRID:IMSR_JAX:012886

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Abcam Cat# ab150116, RRID:AB_2650601)

Curator: @scibot

SciCrunch record: RRID:AB_2650601

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Abcam Cat# ab150077, RRID:AB_2630356)

Curator: @scibot

SciCrunch record: RRID:AB_2630356

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Cell Signaling Technology Cat# 3670, RRID:AB_561049)

Curator: @scibot

SciCrunch record: RRID:AB_561049

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Hangzhou HuaAn Biotechnology Cat# EM21002, RRID:AB_2819164)

Curator: @scibot

SciCrunch record: RRID:AB_2819164

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Novus Cat# NB110-39113, RRID:AB_790014)

Curator: @scibot

SciCrunch record: RRID:AB_790014

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: None

Curator: @scibot

SciCrunch record: RRID:AB_3069633

What is this?

DOI: 10.1007/s12035-025-05167-y

Resource: (Huabio Cat# ET1601-4, RRID:AB_3069615)

Curator: @scibot

SciCrunch record: RRID:AB_3069615

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: Adobe Photoshop (RRID:SCR_014199)

Curator: @scibot

SciCrunch record: RRID:SCR_014199

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AE009, RRID:AB_2771925)

Curator: @scibot

SciCrunch record: RRID:AB_2771925

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: ImageJ (RRID:SCR_003070)

Curator: @scibot

SciCrunch record: RRID:SCR_003070

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32733, RRID:AB_2633282)

Curator: @scibot

SciCrunch record: RRID:AB_2633282

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32727, RRID:AB_2633276)

Curator: @scibot

SciCrunch record: RRID:AB_2633276

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Molecular Probes Cat# A-21434, RRID:AB_141733)

Curator: @scibot

SciCrunch record: RRID:AB_141733

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 51064-2-AP, RRID:AB_11042321)

Curator: @scibot

SciCrunch record: RRID:AB_11042321

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Thermo Fisher Scientific Cat# A32732, RRID:AB_2633281)

Curator: @scibot

SciCrunch record: RRID:AB_2633281

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# SA00001-2, RRID:AB_2722564)

Curator: @scibot

SciCrunch record: RRID:AB_2722564

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: GraphPad Prism (RRID:SCR_002798)

Curator: @scibot

SciCrunch record: RRID:SCR_002798

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9701, RRID:AB_331535)

Curator: @scibot

SciCrunch record: RRID:AB_331535

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 3724, RRID:AB_1549585)

Curator: @scibot

SciCrunch record: RRID:AB_1549585

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 9578, RRID:AB_2721060)

Curator: @scibot

SciCrunch record: RRID:AB_2721060

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 8146, RRID:AB_10950495)

Curator: @scibot

SciCrunch record: RRID:AB_10950495

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Cell Signaling Technology Cat# 4370, RRID:AB_2315112)

Curator: @scibot

SciCrunch record: RRID:AB_2315112

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# anti-spitz, RRID:AB_528474)

Curator: @scibot

SciCrunch record: RRID:AB_528474

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (Proteintech Cat# 66008-4-Ig, RRID:AB_2918475)

Curator: @scibot

SciCrunch record: RRID:AB_2918475

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (ABclonal Cat# AS003, RRID:AB_2769851)

Curator: @scibot

SciCrunch record: RRID:AB_2769851

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# N2 7A1 ARMADILLO, RRID:AB_528089)

Curator: @scibot

SciCrunch record: RRID:AB_528089

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2A1, RRID:AB_2109711)

Curator: @scibot

SciCrunch record: RRID:AB_2109711

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Drosophila Ptc (Apa 1), RRID:AB_528441)

Curator: @scibot

SciCrunch record: RRID:AB_528441

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 4d4, RRID:AB_528512)

Curator: @scibot

SciCrunch record: RRID:AB_528512

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Smoothened (20C6), RRID:AB_528472)

Curator: @scibot

SciCrunch record: RRID:AB_528472

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 2b10, RRID:AB_528186)

Curator: @scibot

SciCrunch record: RRID:AB_528186

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# Rat-Elav-7E8A10 anti-elav, RRID:AB_528218)

Curator: @scibot

SciCrunch record: RRID:AB_528218

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# f461.3b, RRID:AB_528409)

Curator: @scibot

SciCrunch record: RRID:AB_528409

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 5H7B11, RRID:AB_579779)

Curator: @scibot

SciCrunch record: RRID:AB_579779

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 3A6B4, RRID:AB_579780)

Curator: @scibot

SciCrunch record: RRID:AB_579780

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# c17.9c6, RRID:AB_528410)

Curator: @scibot

SciCrunch record: RRID:AB_528410

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 40-1a, RRID:AB_528100)

Curator: @scibot

SciCrunch record: RRID:AB_528100

What is this?

DOI: 10.1038/s44318-025-00489-y

Resource: (DSHB Cat# 1g9, RRID:AB_528278)

Curator: @scibot

SciCrunch record: RRID:AB_528278

What is this?

DOI: 10.1016/j.medj.2025.100752

Resource: (BD Biosciences Cat# 562254, RRID:AB_11154419)

Curator: @scibot

SciCrunch record: RRID:AB_11154419

What is this?

DOI: 10.1016/j.ccell.2025.06.006

Resource: (Thermo Fisher Scientific Cat# 368-3697-80, RRID:AB_2896141)

Curator: @scibot

SciCrunch record: RRID:AB_2896141

What is this?

Dossier d'information détaillé : Mieux dans ma tête - Parlons santé mentale

Introduction

Ce dossier vise à synthétiser les thèmes principaux et les informations cruciales abordées lors de l'émission "Mieux dans ma tête - Parlons santé mentale".

L'objectif central de cette émission est de libérer la parole autour de la santé mentale et de démystifier les troubles psychiques, souvent entourés de tabous et de stigmatisation en France.

Comme le souligne Eddie Riner, parrain de cette initiative, "La santé mentale, c'est une chose dont tout le monde devrait [se] soucier en France."

1. La Santé Mentale, un Sujet Universel et encore Tabou

Une préoccupation pour tous : L'émission insiste sur le fait que la santé mentale concerne chacun d'entre nous.

"Au cours de notre vie, nous allons tous connaître des moments où notre santé mentale sera altérée. Ça peut commencer avec un simple stress, mais aussi des tâches quotidiennes qui nous dépassent avec la fameuse charge mentale."

Les troubles psychiques : Au-delà du stress quotidien, des troubles plus complexes comme la dépression, la bipolarité et la schizophrénie sont évoqués, pour lesquels "on ose moins parler".

Le tabou persistant : Un sondage Ipsos révèle que "près d'un Français sur deux" trouve qu'il est difficile de parler de santé mentale, démontrant à quel point le sujet reste un tabou dans le pays.

2. Démystification et Lutte contre la Stigmatisation

L'émission met en lumière des témoignages poignants pour changer la perception des troubles psychiques.

• Schizophrénie : Une maladie qui ne rime pas avec "débilité mentale".Florent, atteint de schizophrénie, partage son expérience :

"Au début de la maladie, j'associais la schizophrénie avec une forme de débilité mentale. C'est-à-dire que je m'interdisais de vivre puisque je me dis... de toutes les façons c'était vrai l'échec puisque j'étais plus bête que la moyenne."

• Le Professeur Antoine Pellissolo, psychiatre, déconstruit cette idée reçue :

"Absolument. Ça n'a rien à voir avec l'intelligence en fait. Les troubles psychiques, ce sont des maladies qui perturbent les émotions, la perception du monde... mais pas l'intelligence."

• L'exemple de "grands personnages" comme Einstein ou Léonard de Vinci, ayant connu des épisodes psychiques, a aidé Florent à déclencher une prise de conscience et à croire en une vie heureuse malgré la maladie.

• Bruno Guillon témoigne de son changement de perception après avoir rencontré des personnes atteintes de schizophrénie :

"Pour moi, quelqu'un qui était atteint de de schizophrénie, c'était un fou. C'est-à-dire c'était l'iconographie de la chambre capitonnée avec la camisole de force et quelqu'un qui a aucun discernement."

Son reportage à la radio "Les Antoonoirs" a prouvé le contraire.

• La radio comme outil thérapeutique : Le reportage sur "Les Antoonoirs" à Roubaix montre comment des personnes atteintes de troubles psychiques, dont la schizophrénie, animent une émission de radio sous prescription médicale.

Ce projet leur permet de "ne pas être dans la solitude parce que la solitude elle tue plus qu'une maladie" et de "diminuer le traitement médicamenteux".

• Le désir d'être vu comme des "êtres humains à part entière": Un participant de l'émission des "Antoonoirs" exprime :

"Moi, j'espère qu'une chose, c'est qu'on nous voit vraiment comme ce qu'on est, des êtres humain à part entière. Euh sur un même plan d'égalité, ça c'est important."

• Bipolarité : Un chemin de "rétablissement" continu.Noël, atteint de troubles bipolaires, insiste sur la distinction fondamentale : "on ne vous qualifie pas comme patient bipolaire mais quelqu'un qui souffre de trouble bipolaire car la distinction est très importante." Il préfère le terme "en rétablissement" car "c'est un chemin qui se poursuit et qui comporte ce long périple."

• Les tatouages de Noël symbolisent son parcours et sa volonté de vivre, comme le montre l'inscription "No way you kill yourself".

• Le rôle de la famille : La mère de Noël, Monique, exprime le soulagement d'avoir enfin un diagnostic : "D'avoir un mot, je pense que ça a aidé parce que jusqu'ici on est bon, on savait que Noël était pas bien...

Mais le fait qu'il y a un nom euh sur la maladie, c'est quand même quelque chose, c'est quand même positif parce que ça ça nous dit bah finalement depuis son enfance, depuis tout le temps, il était pas bien mais on sait pourquoi et c'était peut-être pas de notre faute, c'était comme ça quoi."

• L'hérédité : Astrid Chevance explique que les antécédents familiaux peuvent indiquer une vulnérabilité : "L'hérédité c'est pas seulement la génétique, c'est aussi le partage d'une vie commune, le milieu social et cetera."

3. Comprendre et Gérer la Dépression

La dépression est présentée comme la maladie psychiatrique la plus fréquente en France, touchant "12,5 % des Français âgés de 18 à 85 ans" chaque année.

• Distinction déprime/dépression : La déprime est passagère, tandis que la dépression est "beaucoup plus intense" et "nous terrasse".

• Symptômes de la dépression : Antoine Pellissolo décrit la dépression comme "une rupture avec sa vie habituelle", incluant désespoir, douleur et "dérèglements du corps" (perte d'appétit, troubles du sommeil, fatigue).

• Ce qui se passe dans le cerveau : Une animation explique la baisse des neurotransmetteurs (sérotonine, dopamine, noradrénaline) entraînant moins de plaisir et de motivation.

Le cortex préfrontal, responsable des décisions rationnelles, n'est plus en mesure de prendre du recul, et l'amygdale cérébrale "surinterprète des messages négatifs".

Une "tempête de molécules inflammatoires" (cytokines) aggrave les symptômes.

• Le rôle des aidants et la difficulté à parler : Marise, qui a vaincu la dépression, témoigne de "idées très sombres, noires".

Son mari, Michel, a été un aidant malgré lui, soulignant la difficulté de la famille à savoir comment aider et l'absence de soutien pour eux.

• Briser la culpabilité du patient : Marise a rencontré un psychologue avec lequel le courant ne passait pas et s'est dit "ils étaient tous pareils". Anne-Victoire Rousselet, psychologue, insiste sur le droit du patient de changer de thérapeute : "Bah, évidemment qu'on a le droit... c'est une histoire de rencontre."

4. Le Burnout : Une Maladie des "Forts"

Christophe partage son expérience du burnout, soulignant que "beaucoup de gens considèrent que c'est la [maladie] des faibles et au contraire, c'est plutôt la maladie des forts, des personnes qui se sont pas écoutées, qui se sont pas arrêtées à temps."

• Les signaux faibles : Douleurs physiques inexpliquées ("le corps parlait"), arrêt des activités plaisantes ("j'ai pas le temps d'aller au sport en ce moment").

• Le soutien de l'entourage : La femme de Christophe témoigne de son désarroi face à son mari "sur son canapé toute la journée à se morfondre", sans savoir comment agir. Il est souligné que les aidants souffrent aussi et ont besoin de soutien.

• Prévention en entreprise : L'importance de poser des questions sur le bien-être au travail et à la maison est mise en avant.

• La durée de l'arrêt de travail : Christophe n'a été arrêté que 3 mois, ce qui est considéré comme très court.

En Belgique, où l'épuisement professionnel est reconnu, la moyenne est de 14 mois. L'idée est qu'il faut "sortir du milieu hostile" et revenir "outillé et différent".

5. L'Anxiété Généralisée et les Crises d'Angoisse

Clara souffre de troubles anxieux généralisés, ayant connu de "terribles crises d'angoisse" et des difficultés à sortir de chez elle.

• Symptômes : Cœur qui bat à fond, pertes d'équilibre, sentiment de déréalisation ("on est plus ancré dans la réalité").

• Le cercle vicieux : La peur de refaire des crises d'angoisse alimente les crises elles-mêmes : "on a peur d'avoir peur et plus on a peur d'angoisser, plus on fait des crises d'angoisse."

• Thérapie et soutien : Clara a suivi une "thérapie comportementale et cognitive" pendant deux ans. Son compagnon l'a soutenue en étant à l'écoute et en contactant sa psychologue pour obtenir des outils.

• La reconquête de soi : Clara a réalisé son rêve de voyager seule pendant un mois, défi qu'elle ne pouvait pas relever auparavant, montrant ainsi son rétablissement.

6. Le Rôle des Professionnels de la Santé Mentale Une clarification est apportée sur les différents rôles :

• Médecin généraliste : Premier recours pour un diagnostic initial et une orientation.

• Psychiatre : Médecin spécialiste des troubles psychiques (8-9 ans d'études de médecine). Ils peuvent prescrire des médicaments et certains sont également formés aux psychothérapies. Leurs consultations sont remboursées.

• Psychologue : Ont un master de psychologie clinique et des stages. Ils sont décrits comme des "rééducateurs de la psychée". Depuis peu, "12 séances chez un psychologue sont remboursé[e]s par l'assurance maladie" sans prescription médicale.

• Psychothérapeute/Psychopraticien : Des distinctions sont nécessaires pour s'y retrouver.

• L'importance du bon "feeling" : Il est crucial de trouver un professionnel avec qui le courant passe. "C'est une histoire de rencontre."

7. L'Impact de l'Hygiène de Vie sur la Santé Mentale

L'émission met en évidence trois piliers essentiels pour le bien-être mental :

• Le Sommeil : "Chouchouter son sommeil" est crucial. "Quand on dort pas, on est fatigué, on a du mal à se concentrer sur les choses, à gérer ses émotions." La quantité (environ 7h) et la régularité sont importantes.

L'Activité Physique : Le sport a un impact direct sur le cerveau.

• Mécanismes : Libération d'endorphines (hormones du bonheur) apaisant la douleur et procurant un sentiment de bien-être.

Production de dopamine (motivation, plaisir) et de sérotonine (humeur), "le même que celui visé par certains antidépresseurs."

Stimulation de la "neurogénèse" (création de nouveaux neurones) dans l'hippocampe, essentielle contre la dépression.

• Témoignages : Marise a vaincu sa dépression grâce au football en rejoignant une équipe de femmes. Clara, Noël, Florent et Christophe témoignent tous des bienfaits de la danse, du vélo, de la musculation ou de la marche.

• La "surf thérapie" : Des psychiatres utilisent le surf comme outil thérapeutique pour des patients atteints de troubles psychiques, permettant une "resocialisation" et une reconquête de la confiance en soi.

Virginie témoigne : "Ça m'a apporté ce côté-là. Ouais. de reprendre confiance en moi, de retrouver une certaine fierté de ce que je fais parce qu'au départ, je ne pensais pas y arriver."

L'Alimentation : Le régime méditerranéen est fortement recommandé.

• Impact : Une étude montre qu'un régime de type méditerranéen a conduit à "une rémission pour un tiers des participants de la dépression en moins de 3 mois."

• Composition : Principalement végétal, local, de saison, coloré, riche en antioxydants, polyphénols, oléagineux, céréales complètes, bonnes matières grasses (huile d'olive) et poissons gras (oméga-3).

• Méfiance envers les aliments ultra-transformés : Ils "sont faits pour être irrésistibles" mais leur consommation excessive augmente le risque "d'anxiété, de dépression, de trouble du sommeil" en créant une inflammation "persistante" dans le corps.

• L'équilibre : Les plaisirs occasionnels ("junk food") sont acceptables, mais ne doivent pas devenir la norme.

8. L'Importance de la Prévention et du Soutien

• Le défi "Courir pour toi" : Louis, dont le frère Simon s'est suicidé après des années de lutte contre la dépression, a lancé un défi sportif (130 marathons en 150 jours) pour sensibiliser et lever des fonds pour la "Maison Perchée", une association qui aide les personnes confrontées à des troubles psychiques.

• Détecter les signaux et agir : En cas d'idées noires ou suicidaires, il est crucial de "saisir la perche", d'être à l'écoute et d'engager le dialogue. Il ne faut jamais banaliser ces propos.

• Appeler à l'aide : Si la situation est inquiétante, il faut contacter le médecin traitant ou, en urgence, le SAMU (15).

• Différences hommes/femmes face à la dépression :

Les femmes sont plus souvent diagnostiquées (2 pour 1 homme), peut-être parce que la société leur laisse plus d'espace pour exprimer la tristesse, tandis que les hommes peuvent avoir plus de mal à parler de leur souffrance.

• La "Maison Perchée" : Un lieu "sans jugement" qui sert de "sas entre l'hospitalisation et la le retour à la vraie vie", offrant des ateliers et un soutien par les pairs ("pair-aidance"). Flavie, bipolaire, y a trouvé "une pluralité dans l'être humain" et a pu "renaître".

Conclusion

• L'émission "Mieux dans ma tête - Parlons santé mentale" est un appel vibrant à la prise de conscience et à l'action.

Elle démontre que les troubles psychiques sont des maladies réelles et complexes, qui ne doivent plus être stigmatisées.

Grâce aux témoignages et aux éclaircissements des professionnels, elle offre de l'espoir en montrant que des solutions existent, qu'il s'agisse de thérapies, de médicaments ou d'une meilleure hygiène de vie.

Le message clé est clair : oser en parler, demander de l'aide et reconnaître l'importance d'une approche globale pour la santé mentale.

Comme le dit un participant des "Antoonoirs", il s'agit de voir les personnes atteintes de troubles psychiques comme des "êtres humains à part entière" sur un pied d'égalité.

Compte Rendu Détaillé : Sommes-nous tous racistes ?

Ce document synthétise les thèmes principaux, les idées essentielles et les faits marquants tirés de l'émission "Sommes-nous tous racistes ?".

Il met en lumière les mécanismes inconscients des préjugés et de la discrimination à travers diverses expériences scientifiques.

Introduction : Les Préjugés Universels et la Question du Racisme

L'émission s'ouvre sur une interrogation fondamentale : "Vous êtes raciste, vous et moi ?

Est-ce que je suis raciste ?" (Lucien Jean-Baptiste).

Elle pose l'idée que, quelles que soient nos origines ou caractéristiques, "nous avons tous des idées reçues, des a prioris, des préjugés sur tout ce qui ne nous ressemble pas, que nous ne connaissons pas."

L'objectif de l'émission est d'explorer ces mécanismes inconscients.

Pour ce faire, 50 volontaires participent à des "expériences étonnantes" sous le faux titre "Les mystères de notre cerveau", afin de ne pas biaiser leurs réactions.

Le psychosociologue Sylvain De Louvet, expert scientifique, décode les résultats des expériences.

Marie Drucker et Lucien Jean-Baptiste, réalisateur et comédien engagé, commentent les comportements observés.

L'émission révèle que le racisme, la misogynie, le sexisme, l'antisémitisme, l'homophobie et la grossophobie s'appuient sur les "mêmes mécanismes" inconscients et documentés scientifiquement.

Thèmes et Idées Clés : Les Mécanismes Inconscients des Préjugés

1. La Recherche de Similarité et ses Conséquences (Expérience de la Salle d'Attente)

Description de l'expérience : Des participants sont invités à s'asseoir dans une salle d'attente où deux chaises sont disponibles, une à côté d'un homme blanc et l'autre à côté d'un homme noir. La position des acteurs est inversée à mi-parcours.

Observations et conclusions :

• Les participants choisissent majoritairement de s'asseoir à côté de la personne blanche, quel que soit son emplacement.
• Sylvain De Louvet explique : "Ce n'est pas un comportement raciste en tant que tel.

Ce qui s'explique très facilement, c'est l'idée que on cherche la similarité. On va chercher les gens qui nous ressemblent." * Cette tendance est qualifiée de "reptilien[ne]", certains thèse évolutionnistes suggérant que "les tribus primitives déjà avaient tendance à se méfier de la différence de l'autre et à plutôt chercher la similitude, la similarité."

• Impact : Bien que non raciste en soi, ce mécanisme a des "conséquences quand on va chercher un emploi, l'accès au logement et cetera, c'est terrible." Un DRH, même tolérant, peut inconsciemment favoriser quelqu'un qui lui ressemble.

2. L'Influence des Préjugés sur le Jugement (Expérience du Jury)

Description de l'expérience : Les participants jouent le rôle de jurés et doivent attribuer une peine de prison à un accusé pour le même crime (coups et blessures volontaires ayant entraîné la mort).

Deux profils sont présentés : un homme blanc et un homme d'origine maghrébine.