Thank you for this fascinating work. Using a young clade with many independent trait transitions is a powerful way to link genes to function. One thought: the PFAM enrichment tests could be strengthened by accounting for phylogenetic non-independence, since a single gene loss in one ancestor can appear as many independent associations. Phylogenetic comparative methods (e.g., Pagel’s correlation test or phylogenetic logistic regression on the 205-gene tree) could separate gene families that repeatedly co-vary with a trait from those inherited together. At Arcadia, we used a similar approach to link gene families to a parasitic trait across chelicerates (https://doi.org/10.57844/arcadia-4e3b-bbea). Personally, I am curious whether the ice-binding proteins in Glacidusochromonas look like good candidates for heterologous expression. I am interested in exploring Nature's solutions to the cryopreservation of Chlamydomonas.

Congratulations on this rigorous and impressive work.

The MDG1 reporter's behavior is a little mysterious to me. MDG1 is described as functionally irrelevant to repair, yet its promoter responds to [MIX+] status specifically. Is this a one-off transcriptional quirk of that locus, or does [MIX+] shift chromatin/expression more broadly at any similar sequence? Given your own earlier work on a 'prion-driven' active chromatin state, a genome-wide look (CUT&RUN plus RNA-seq comparing [mix-] and [MIX+]) seems like it could either generalize or contain this result. Perhaps if repeated for the other two prions of this class, you could explain more about what makes GAR, MOD, and MIX different from other prions.

Another thing to consider: combining high-activity CREs with binding sites for heterologous transcriptional activators to create chimeric promoters that are strong, minimally disruptive, and chemically inducible.

Lee et al. (2018) demonstrated ethanol-inducible expression in C. reinhardtii using the two-component AlcR-PalcA system from Aspergillus nidulans, but induction was modest (~1.74-fold) due largely to the weak transcriptional output of native PalcA in algae. Augmenting PalcA with the high-activity CREs from this study — particularly given the authors' finding that tandem CRE copies amplify output — could yield a robust, titratable system where gene expression is switched on simply by adding ethanol to the culture medium.

Reference: Lee et al. (2018) Journal of Applied Phycology. https://doi.org/10.1007/s10811-018-1480-8

We can't be sure that the albedo of the different pigments is what is driving the melt rate, since excreted or surface-presenting compounds are not quantified here. One could simply perform a titration of the pigments themselves in ice crystals as a control.