The LIM and POU homeobox genes ttx-3 and unc-86 act as terminal selectors in distinct cholinergic and serotonergic neuron types

Abstract

Transcription factors that drive neuron type-specific terminal differentiation programs in the developing nervous system are often expressed in several distinct neuronal cell types, but to what extent they have similar or distinct activities in individual neuronal cell types is generally not well explored. We investigate this problem using, as a starting point, the C. elegans LIM homeodomain transcription factor ttx-3, which acts as a terminal selector to drive the terminal differentiation program of the cholinergic AIY interneuron class. Using a panel of different terminal differentiation markers, including neurotransmitter synthesizing enzymes, neurotransmitter receptors and neuropeptides, we show that ttx-3 also controls the terminal differentiation program of two additional, distinct neuron types, namely the cholinergic AIA interneurons and the serotonergic NSM neurons. We show that the type of differentiation program that is controlled by ttx-3 in different neuron types is specified by a distinct set of collaborating transcription factors. One of the collaborating transcription factors is the POU homeobox gene unc-86, which collaborates with ttx-3 to determine the identity of the serotonergic NSM neurons. unc-86 in turn operates independently of ttx-3 in the anterior ganglion where it collaborates with the ARID-type transcription factor cfi-1 to determine the cholinergic identity of the IL2 sensory and URA motor neurons. In conclusion, transcription factors operate as terminal selectors in distinct combinations in different neuron types, defining neuron type-specific identity features.

Keywords: Caenorhabditis elegans; Homeobox; Neuron differentiation.

Fig. 1.

Expression pattern of the C. elegans ttx-3 LIM homeobox gene. (A) ttx-3 expression constructs and summary of neuronal expression pattern. The promA::gfp and promB::gfp constructs were described previously (Altun-Gultekin et al., 2001; Wenick and Hobert, 2004) and are shown here for comparison only. (B) ttx-3 fosmid expression (wgIs68) in first larval stage animals and in adult animals. D-V, dorsoventral. White asterisks indicate gut autofluorescence. (C) The seventh intron of the ttx-3 locus contains cis-regulatory elements driving reporter gene expression in AIA and NSM neurons. These regulatory elements do not depend on ttx-3. Expression is shown in adult animals.

Fig. 2.

ttx-3 affects the terminal differentiation of AIA neurons. (A) Schematic representation of the AIA interneuron pair [reproduced with permission (Altun et al., 2002-2013)]. (B) The expression of terminal differentiation markers of AIA identity is affected in ttx-3 mutants. Reporter gene arrays were crossed into ttx-3(ot22) null mutants. Positions of AIA neurons are outlined (dashed circles). The fraction of animals that show the indicated phenotype is presented in the bar charts. Transgenic arrays are: otIs317 for mgl-1, otIs326 for ins-1, otIs379 for cho-1, otEx4687 for glr-2 and otEx5056 for flp-2 (see Materials and methods for more detail on the arrays; the Ex[gcy-28d::gfp] and Ex[scd-2::gfp] arrays were kindly provided by Takeshi Ishihara). Anterior is up in all panels.

Fig. 3.

Co-regulation of AIA-expressed genes by two cis-regulatory motifs. (A-C) Mutational dissection of the cis-regulatory elements of three AIA-expressed terminal identity markers. (D) Position weight matrix of the two motifs required for AIA expression, based on the motifs from ins-1, cho-1 and mgl-1 and orthologs in other nematode species. Perfect (filled box) and imperfect (stippled box) matches to the two cis-regulatory motifs [blue, G(A/G)ATC; green, TAATTNGA] in other AIA terminal identity markers are shown on the right. (E) TTX-3 binds to cho-1 and mgl-1 regulatory elements containing the HD (TAAT) motif. Deletion of the HD motif abolishes binding. EMSA was performed with 250 nM and 750 nM TTX-3.

Fig. 4.

The effect of unc-86 and ttx-3 on the serotonergic identity of NSM neurons. (A) The 5HT pathway including tetrahydrobiopterin biosynthesis genes (Deneris and Wyler, 2012). ‘?’ indicates that a unique homolog of SR could not be identified in the worm genome. (B) Schematic representation of the NSM interneuron pair [reproduced with permission (Altun et al., 2002-2013)]. (C) The expression of serotonergic identity features of NSM (dashed circles) is affected in unc-86(n846), ttx-3(ot22) or unc-86(n846); ttx-3(ot22) double-null mutants. Reporter gene arrays were crossed into the respective mutant backgrounds. Transgenic arrays are: zdIs13 for tph-1; otEx4781 for mod-5; otIs225 for cat-4; otEx5280 for ptps-1; otIs226 for bas-1; and otIs224 for cat-1 (see Materials and methods for more detail on the arrays). Images are only shown for mutant genotypes with effects on reporter expression. (D) Serotonin antibody staining. Thirty animals were scored for each genotype. In the double mutant, no animal showed staining in NSM (circled), whereas in the other genotypes all animals showed staining.

Fig. 5.

The effect of unc-86 and ttx-3 on other identity features of NSM neurons. The expression of other identity features of NSM is also affected in unc-86(n846), ttx-3(ot22) or unc-86(n846); ttx-3(ot22) double-null mutants. Reporter gene arrays were crossed into the respective mutant backgrounds. Transgenic arrays are: vsIs33 for dop-3; otIs317 for mgl-1; otEx5163 for nlp-3; otEx5364 for mgl-3; otEx5163 for nlp-13; otEx5055 for scd-2; and otEx5363 for flr-2 (see Materials and methods for more detail on the arrays). Micrographs are only shown for mutant genotypes with effects on reporter expression. Dashed circles indicate the position of NSM neurons. See also Table 1.

Fig. 6.

Cis-regulatory analysis of NSM identity specification. (A-D) Dissection of the cis-regulatory elements of four NSM-expressed serotonin pathway genes. All minimal cis-regulatory elements contain predicted POU sites. (E) EMSAs with UNC-86 protein on bas-1 and tph-1 regulatory elements. Mutated POU sites are those that also disrupt reporter gene activity when deleted from the gene contexts of bas-1 and tph-1 (A). EMSA was performed with 10 nM and 30 nM UNC-86. Arrowhead indicates UNC86-bound DNA probe.

Fig. 7.

unc-86 and ttx-3 affect NSM morphology. (A) unc-86(n846) mutant adults display shorter ventral neurites and fewer and shorter axon arbors. mod-5p::gfp (olaEx1446) is used to visualize NSM morphology. (B) UNC-40::GFP localization (transgene: olaEx1448) remains in a juvenile state in unc-86 mutant animals. White arrows indicate UNC-40::GFP puncta. (C) Quantification of the unc-86(n846) arborization phenotype. Displayed is the fraction of animals with clusters of arbors in the nerve ring region in wild-type and unc-86(n846) animals. The difference between wild type and unc-86 is significant (^*P<0.0001). (D) Quantification of the unc-86(n846) UNC-40::GFP localization phenotype. The fraction of animals with multiple, bright UNC-40::GFP puncta in the nerve ring region is displayed. The difference between wild type and unc-86 is significant (^*P=0.0017). Error bars indicate 95% confidence intervals. (E) unc-86(n846); ttx-3(ot22) double mutants display numerous NSM morphology defects, as visualized with flp-4::gfp (olaEx1485). In all images, anterior is to the left and ventral down. Asterisks indicate cell bodies (A,B) or additional cell-body-like swellings (E). Brackets denote the nerve ring terminal field where arbors form. White arrows indicate NSM neurites, red arrows and asterisks denote other non-NSM structures. Fisher’s t-test was used for statistical analysis. Scale bars: 5 μm.

Fig. 8.

unc-86 and cfi-1 control cholinergic IL2 neuron identity. (A) Animals are late L4 or young adults, with the exception of the lag-2::gfp transgenic animals which are dauers. The differential importance of cfi-1 in the dorsal IL2DL/R and ventral IL2VL/R neurons versus the lateral IL2L/R neurons mirrors morphological differences of the ventral versus lateral neurons, with the lateral neurons having a distinct spectrum of synaptic partners (White et al., 1986). See also supplementary material Fig. S4. IL2 schematic reproduced with permission (Altun et al., 2002-2013). (B) Summary of terminal selector combinatorial codes in head ganglia of C. elegans. Colors refer to neurotransmitter identities: green, serotonergic; red, cholinergic; yellow, glutamatergic. Support or blast cells are in gray.