Identifying the mechanisms behind neuronal fate specification are key to understanding normal neural development in addition to neurodevelopmental disorders such as autism and schizophrenia. In vivo cell fate specification is difficult to study in vertebrates. However, the nematode Caenorhabditis elegans, with its invariant cell lineage and simple nervous system of 302 neurons, is an ideal organism to explore the earliest stages of neural development. We used a comparative transcriptome approach to examine the role of cnd-1/NeuroD1 in C. elegans nervous system development and function. This basic helix-loop-helix transcription factor is deeply conserved across phyla and plays a crucial role in cell fate specification in both the vertebrate nervous system and pancreas. We find that cnd-1 controls expression of ceh-5, a Vax2-like homeobox class transcription factor, in the RME head motorneurons and PVQ tail interneurons. We also show that cnd-1 functions redundantly with the Hox gene ceh-13/labial in defining the fate of DD1 and DD2 embryonic ventral nerve cord motorneurons. These data highlight the utility of comparative transcriptomes for identifying transcription factor targets and understanding gene regulatory networks. 1684 gthree G3: Genes|Genomes|Genetics G3 (Bethesda) Oxford University Press PMC7466980 7466980 7466980 32601060 10.1534/g3.120.401515 cnd-1 /NeuroD1 Functions with the Homeobox Gene ceh-5 /Vax2 and Hox Gene ceh-13 /labial To Specify Aspects of RME and DD Neuron Fate in Caenorhabditis elegans Aquino-Nunez Wendy 1 Mielko Zachery E 1 Dunn Trae 1 Santorella Elise M 1 Hosea Ciara 1 Leitner Lauren 1 McCalla Derrica 1 Simms Claire 1 Verola Wendy M 1 Vijaykumar Sharanya 1 Hudson Martin L 1 7 1 Department of Molecular and Cellular Biology, Kennesaw State University, Kennesaw, GA 30144 7 Corresponding author: Department of Molecular and Cellular Biology, Kennesaw State University, Building 12, Room SC507, 370 Paulding Avenue, Kennesaw, GA 30144. However, the nematode Caenorhabditis elegans , with its invariant cell lineage and simple nervous system of 302 neurons, is an ideal organism to explore the earliest stages of neural development. We used a comparative transcriptome approach to examine the role of cnd-1 /NeuroD1 in C. elegans nervous system development and function. This basic helix-loop-helix transcription factor is deeply conserved across phyla and plays a crucial role in cell fate specification in both the vertebrate nervous system and pancreas. We find that cnd-1 controls expression of ceh-5 , a Vax2-like homeobox class transcription factor, in the RME head motorneurons and PVQ tail interneurons. In addition, NeuroD1 is expressed abundantly in the brain after terminal fate specification, which suggests a secondary role in nervous system homeostasis and/or neural maturation and survival ( 2016 ), a comprehensive list of NeuroD1 targets has not been compiled and many questions on its role in neural development remain unanswered. The nematode Caenorhabditis elegans , with its invariant cell linage and well-defined nervous system, is an excellent model to study cell lineage determination and terminal fate specification ( Sulston and Horvitz 1977 ; Sulston et al. 1983 ). Second, does a single transcription factor control the fate of a single neuron, or is terminal fate specified in a combinatorial manner, with multiple transcription factors controlling different aspects of the final cell fate? Extensive work has identified a battery of transcription factors known as “terminal selectors”, which are required for terminal fate specification in C. elegans neurons ( Hobert 2016 and references therein). These transcription factors generally act in a combinatorial fashion to specify cell fates, although individual transcription factors may specify the fate of multiple cells that are unrelated by cell lineage, type, or circuit. elegans bHLH transcription factor cnd-1 is orthologous to the human NeuroD1 gene and is one of the earliest proneural genes to be activated during C. elegans embryonic development ( Hallam et al. 2000 ). However, the only reported defects seen in cnd-1 loss-of-function mutants are a relatively mild back-coiler phenotype caused by misspecification of 2-3 dorsal D (DD) motorneurons required for inhibitory GABAergic neuromuscular innervation, in addition to axon guidance and synapse remodeling defects in the remaining D neurons ( Hallam et al. 2000 ). To gain a better understanding of CND-1 ’s role during C. elegans neural development, we performed an RNA-seq assay comparing embryonic wild type and cnd-1 ( ju29 ) mutant transcriptomes. We find that CND-1 positively regulates the expression of homeobox transcription factor ceh-5 /Vax2 in the head RME and tail PVQ neurons. We also confirm that CND-1 is required for the generation of cnd-1 expressing cells during ventral nerve cord fate specification. Finally, we show that cnd-1 functions in parallel with the Hox gene ceh-13 /labial to specify a subset of embryonic DD class ventral nerve cord motorneuron fates. Materials and Methods Strains and maintenance C. Extra-chromosomal arrays used in this study were leEx2489 [ceh-5p :: GFP + unc-119 (+)] and dbEx724 [flp-6p :: tax-2 (cDNA) :: SL2 :: GFP + lin-15 (+)] . The cnd-1 ( gk718 ) allele was identified by the C. elegans deletion mutant consortium (2012). All mutants were outcrossed at least twice prior to analysis. Table S1 shows details of strains generated during the course of this study including strain numbers and sources. Figure 6S and 6T summarize average DD neuron and ventral cord motorneuron counts for the above assays. Overall, the difference in ventral cord neuron count parallels the loss of DD neurons observed in cnd-1 and ceh-13 mutant backgrounds and argues against a change of DD neuron fate to another neuronal cell type. Figure 6 cnd-1 and ceh-13 control the birth of DD motorneurons but have no obvious role in DA or DB motorneuron birth.

In the nervous system, the specific identity of a neuron is established and maintained by terminal selector transcription factors that directly activate large batteries of terminal differentiation genes and positively regulate their own expression via feedback loops. However, how this is achieved in a reliable manner despite noise in gene expression, genetic variability or environmental perturbations remains poorly understood. We addressed this question using the AIY cholinergic interneurons of C. elegans, whose specification and differentiation network is well characterized. Via a genetic screen, we found that a loss of function of PRC1 chromatin factors induces a stochastic loss of AIY differentiated state in a small proportion of the population. PRC1 factors act directly in the AIY neuron and independently of PRC2 factors. By quantifying mRNA and protein levels of terminal selector transcription factors in single neurons, using smFISH and CRISPR tagging, we observed that, in PRC1 mutants, terminal selector expression is still initiated during embryonic development but the level is reduced, and expression is subsequently lost in a stochastic manner during maintenance phase in part of the population. We also observed variability in the level of expression of terminal selectors in wild type animals and, using correlation analysis, established that this noise comes from both intrinsic and extrinsic sources. Finally, we found that PRC1 factors increase the resistance of AIY neuron fate to environmental stress, and also secure the terminal differentiation of other neuron types. We propose that PRC1 factors contribute to the consistency of neuronal cell fate specification and maintenance by protecting neurons against noise and perturbations in their differentiation program. In the nervous system, the specific identity of a neuron is established and maintained by terminal selector transcription factors that directly activate large batteries of terminal differentiation genes and positively regulate their own expression via feedback loops. However, how this is achieved in a reliable manner despite noise in gene expression, genetic variability or environmental perturbations remains poorly understood. We addressed this question using the AIY cholinergic interneurons of C . elegans , whose specification and differentiation network is well characterized. These observations raise an intriguing question: how can the identity of a neuron be specified and maintained in a reliable manner despite variability in gene expression? In addition, neuronal fate has to be robust against internal (genetic) or external (environmental) perturbations. C . elegans is a good system to study noise and precision during nervous system development. C . elegans has a fixed number of neurons (302 in the adult hermaphrodite) that are generated via an invariant cell lineage, allowing the monitoring of neuronal development with a high level of precision [ 7 – 10 ]. Thus, in PRC1 mutants, while TTX-3 and CEH-10 expression is completely absent in some AIY neurons, their levels seem mostly unaffected in the remaining AIY neurons. This may be explained by the positive feedback loop that maintains TTX-3 and CEH-10 expression as such network motifs can lead to bistable ON-OFF states [ 5 ]. However, in PRC1 mutants, we noticed some very rare cases (less than 0.5%) of neurons with very low levels of TTX-3 or CEH-10 proteins (see for example Fig 4C , red arrow). They may represent r = Spearman’s correlation. PRC1 is also implicated in the differentiation of other neuronal types We showed that the PRC1 complex affects the differentiation program of the AIY cholinergic interneuron. As PRC1 factors are also expressed in other neurons, we next asked whether the PRC1 complex also affects the identity of other neuronal types. We first looked at the effect on dopaminergic neuron fate. C . elegans larvae have 3 pairs of dopaminergic neurons in the head: CEPD, CEPV and ADE ( Fig 9A ). (D) Percentage of late larvae (L4) that display for each class of touch receptor neurons a loss of mec-4p :: gfp ( zdIs5 ) expression in at least one neuron (error bars show standard error of proportion, n = number of animals analyzed, * p<0.05, ** p<0.01 Fisher’s exact test). We next characterized the effect of PRC1 on the touch receptor neurons. Late larvae of C . elegans have 6 touch receptor neurons: 2 ALM and 1 AVM in the anterior region; 1 PVM in the mid region; 2 PLM is the posterior region ( Fig 9C ). In a loss of function of the PRC1 gene spat-3 , we observed no effect on the expression of touch receptor neuron marker mec-4 at late larval stage ( Fig 9D ). This suggests that PRC1 also affects the consistency of the differentiation program of the ALM and PVM touch receptor neurons. Taken together, these data suggest that the PRC1 complex affects the consistency of the differentiation of several neuronal types in C . elegans , and that the sensitivity of neurons to a loss of PRC1 varies depending on neuronal types. Discussion In the nervous system, neurons need to acquire and maintain their differentiated type identity in an accurate manner. However, the mechanisms that ensure the precision of neuronal cell fate acquisition during development and maintenance throughout the life of the animal remain poorly characterized. The rare neurons that we observed with intermediate levels may represent neurons that transition from the ON state to the OFF state. We observed that PRC1 factors act cell autonomously in the AIY neuron and that their activity seems continuously required, as restoration of PRC1 activity during the maintenance phase partially rescues the loss of AIY fate. How are PRC1 factors acting on ttx-3 and ceh-10 transcription? While PRC1 factors have been initially characterized as repressors of gene transcription, it was more recently shown that they can also act in a positive manner on the transcription of their direct targets [ 19 – 21 ].