Conserved nuclear hormone receptors controlling a novel plastic trait target fast-evolving genes expressed in a single cell

Environment shapes development through a phenomenon called developmental plasticity. Deciphering its genetic basis has potential to shed light on the origin of novel traits and adaptation to environmental change. However, molecular studies are scarce, and little is known about molecular mechanisms associated with plasticity. We investigated the gene regulatory network controlling predatory vs. non-predatory dimorphism in the nematode Pristionchus pacificus and found that it consists of genes of extremely different age classes. We isolated mutants in the conserved nuclear hormone receptor nhr-1 with previously unseen phenotypic effects. They disrupt mouth-form determination and result in animals combining features of both wild-type morphs. In contrast, mutants in another conserved nuclear hormone receptor nhr-40 display altered morph ratios, but no intermediate morphology. Despite divergent modes of control, NHR-1 and NHR-40 share transcriptional targets, which encode extracellular proteins that have no orthologs in Caenorhabditis elegans and result from lineage-specific expansions. An array of transcriptional reporters revealed co-expression of all tested targets in the same pharyngeal gland cell. Major morphological changes in this gland cell accompanied the evolution of teeth and predation, linking rapid gene turnover with morphological innovations. Thus, the origin of feeding plasticity involved novelty at the level of genes, cells and behavior.

(1) The authors raise the question of NHR-1 and NHR-40 perhaps working as a heterodimer, but do not address it further. Do they feel that the phenotypic differences between the two null mutants are large enough to preclude this possibility? If so, they should make the argument clearly. Have they tested interaction in the yeast two-hybrid system, or in another manner? If so, even experiments which do not completely resolve the question would be good to include here.
(2) Although the authors clearly demonstrate that nhr-1 null mutations are epistatic to nhr-40(gf) mutations, they do not appear to have tested whether they are also epistatic to nhr-40(null) alleles. This test should be simple and is an important way of probing the relationship between the two genes, and of further testing the model that nhr-1 mutants are the most downstream because they are involved in cell differentiation, rather than cell fate specification.
(3) The analysis of potential targets of nhr-1 and nhr-40 forms a critical part of the paper, and is largely solid. However, a few issues could be cleared up. First, it would be helpful for them to emphasize clearly throughout that the targets could be direct or indirect. Second, have they tested any of the transcriptional reporters in nhr-1 or nhr-40 backgrounds? Looking at a few of these could give a clearer impression of cell-by-cell regulation, as opposed to the whole body regulation of the RNAseq studies.
(4) Perhaps most important, the knockouts of potential target genes were beautifully done, but the lack of phenotypes is surprising. Have the authors considered ablating the cell they implicate through the expression studies, g1D? This might be a quick and easy test to confirm that the central topic of half their paper is indeed involved in controlling mouth morph development.
(5) The extreme redundancy the authors observe can also be problematic from an evolutionary perspective, and they should expand their discussion to address this point.
(6) Although we like the author's evolutionary model at the end, they state it a little too strongly. For example, the sentences: "We speculate that the striking co-expression of the target genes results from an ancient regulatory linkage between the NHRs and the promoters of the ancestral target genes. Such divergent evolutionary dynamics of transcription factors and their downstream targets might represent general features of GRNs" This summary neglects the possibility of these transcription factors capturing new promoters by mutation of target sites. Do we have any information about how complex NHR-1 or NHR-40 target sequences might be?
(7) Since all of the target genes are expressed in the pharyngeal gland cell, g1D, which has a long process that terminates into the buccal cavity, is it possible that the NHR target genes are involved in sensing environmental factors. For example, is the duodecuple Astacin mutant less sensitive to environmental signals.

Figure 2
FPKM should be defined in the legend Line 98 Change "and we are yet to find" to "and we had yet to find" Line 254 Change "NHR-40 and NHR-1 where more highly expressed" to "NHR-40 and NHR-1 were more highly expressed" Line 390 Change "On the contrary, we identified" to "By contrast, we showed "