Your enhancer-insertion library is a useful tool for probing carotenoid regulation, and the CMRP1 follow-up is both convincing and compelling. The long-range activity of the enhancer in your top hit is intriguing and raises a few questions about how ERGO should be interpreted.

If an insertion can influence genes across ~2 Mb, then many nearby loci are plausible targets. How confident are you that CMRP1 is the primary driver rather than one member of a broader set of co-activated genes? More generally, because NHEJ insertions favor open and insertion-tolerant regions, regulators positioned in less permissive chromatin may never be sampled. Insertions that disrupt essential genes, or essential neighboring genes, would also eliminate the corresponding clones before screening, also impacting sampling. Do you have a sense of how much of the genome is protected in this way? Along those lines, have you looked at whether enhancer effectiveness varies with chromatin context, and whether some genomic regions tend to dampen or block enhancer activity?

Did you characterize the expression of neighboring genes at all to distinguish between CMRP1-driven changes and insertion-related ones? Given that many insertions are tandem or structurally complex, did you assess whether enhancer copy number, truncation, or orientation contributed to the expression patterns or phenotypes you observed?

Finally, the use of ERGO here implements a pigment phenotype in the yellow-in-the-dark background. Do you envision pairing the enhancer library with non-colorimetric reporters or selectable screens to expand beyond carotenoid metabolism in the future?

Your integration of physiology, proteomics, and metabolomics across fluctuating-light and CO₂ conditions gives a clear, comprehensive picture of how C. reinhardtii shifts its energy budget when light and carbon availability vary. Ambient CO₂ drives cells into an ATP-limited, CCM-dependent state, while elevated CO₂ produces a more energetically permissive one; chloroplast-to-mitochondria electron flow becomes the dominant ATP-support pathway under carbon limitation, more than cyclic electron flow or flavodiiron activity. This opens up a few questions:

Because dissolved CO₂ is naturally low in aquatic habitats, ambient CO₂ in the lab recreates Chlamydomonas’ typical carbon environment, but it also forces continuous CCM activity and high ATP demand. How do you define the appropriate “baseline physiology” in this context? Should the CCM-on, low-CO₂ state be treated as the reference, or is the high-CO₂, CCM-off state a more useful baseline for interpreting metabolic and proteomic differences?

Your data also raise questions about standard Chlamydomonas culturing practices. Most labs grow cells in constant light at ambient CO₂, which your results suggest enforces a high-ATP-demand, CCM-dominated state. Do you think culturing norms should shift toward supplemented CO₂ or specific fluctuating-light regimes for baseline studies? If ambient CO₂ is the ecologically relevant state, how should we interpret findings generated under 2% CO₂, where major ATP sinks are artificially suppressed?

Lastly, given the strong mitochondrial contribution under carbon limitation, did you examine mitochondrial positioning or morphology under LL, HL, and FL conditions? It would also be useful to know whether different fluctuating-light pulse lengths shift reliance on malate-based chloroplast–mitochondria shuttling. Prior work shows CO₂-dependent mitochondrial repositioning (doi: 10.1111/tpj.70601), suggesting that both structural changes and fluctuation frequency could help distinguish productive from non-productive shuttling states.

The combination of targeted chromatin editors and quantitative modeling provides a compelling framework for dissecting how DNA methylation and H3K9me3 cooperate to enforce transgene silencing. The work makes a strong case that DNA methylation is the primary heritable silencing mark in hiPSCs and that TET1-based demethylation can act as a modular anti-silencing tool.

I have a few questions about the CHO vs hiPSC comparison and the generality of the proposed feedback loop:

The key contrast hinges on KRAB installing H3K9me3 with or without subsequent DNA methylation. It’s hard to tell whether the absence of a feedback loop in CHO reflects a cell-type effect, a species effect, or CHO-specific epigenetic drift. Have you tested KRAB-mediated silencing and methylation in any additional mammalian cell types (such as human somatic lines or another rodent line), or mined existing datasets to see whether H3K9me3–DNMT3A coupling is generally weaker outside pluripotent cells?

You show that DNMT3A recruitment produces stable silencing and that TET1 can collapse both DNA methylation and H3K9me3 at the reporter. This is a powerful tool regardless, but it would be interesting to know whether the same bistable logic applies to endogenous loci. Have you examined KRAB- or DNMT3A-targeted repression at endogenous promoters, or compared H3K9me3 and DNA methylation dynamics at native loci in the same cells?

The combination of MaSIF-mimicry and GluePCA provides a powerful framework for mapping the latent CRBN-MGD interactome, and the scale at which you identify both ZF and non-ZF binders is impressive. The work makes a strong case that CRBN’s binding landscape is far broader than its known degradome.

I have a few questions about the bonsai library, its expression in yeast, and the interpretation of the latent interactome:

Because all bonsai constructs are expressed in yeast, some human ZFs and non-ZF fragments may misfold, fail to coordinate Zn²⁺, or depend on mammalian PTMs or chaperones. These would appear as GluePCA negatives even if they are true binders in human cells. Have you assessed folding in yeast, estimated how much of the “non-binder” space may reflect yeast-specific false negatives, or tested any GluePCA-negative constructs in a mammalian binding assay?

GluePCA reports proximity-driven binding, but degradation requires a very specific geometry relative to the CRL4^CRBN ubiquitination machinery. Many binders may be sterically dead: capable of engaging CRBN-MGD but oriented such that lysines are inaccessible to the E2. Have you explored whether certain ZF orientations, linker patterns, lysine positions, or accessory elements distinguish productive binders from non-productive ones? And do you think the dataset is sufficient to computationally separate these classes? Such an analysis could help predict productively oriented binders that are more amenable to rational MGD design.