To better understand motor neuron specification, we have conducted an EMS-induced mutant screen in which we sought to identify mutants with defects in the expression of an
unc-53::gfp marker (hdIs1 transgene), which is normally expressed in the cholinergic DA and AS motor neurons (Wacker et al., 2003). As previously reported, this screen identified the zinc finger transcription factor
bnc-1 (Kerk et al., 2017). Another allele identified from this screen is
ot762. Unlike
bnc-1 mutant alleles,
ot762 mutant animals are Unc and Egl.
ot762 mutants show a decrease in the number of
unc-53::gfp expressing neurons in the L4, but not L1 stage, indicating a loss of
unc-53::gfp expression in the postembryonically generated AS neurons, but not the embryonically born DA neurons (Fig.1A,B). The lineal sister of the cholinergic AS neurons are the GABAergic VD motor neurons (Sulston, 1976). We find that an
unc-47::mCherry marker (otIs348; (Gendrel et al., 2016)) also fails to be expressed in postembryonically generated VD motor neurons of
ot762 mutants; in addition, there is also a reduction in the number of embryonically generated DD neurons (Fig.1A,B). Analysis of a panneuronal marker,
rab-3, indicates a reduced number of neurons in the ventral nerve cord (Fig.1C), supporting the possibility that AS, DD and/or VD neurons may not be generated.
Using a combined polymorphic mapping and whole genome sequencing pipeline (Minevich et al., 2012) we found that
ot762 animals harbor a missense mutation in the 5th zinc finger domain of the highly conserved zinc finger transcription factor REF-2, called Zic1/2/3 in vertebrates and Opa in flies (Alper and Kenyon, 2002; Aruga, 2004) (Fig.1D). The mutation results in the substitution of a highly conserved serine to a leucine within the last Zn finger domain of REF-2. This specific amino acid position is predicted to contact DNA (Benos et al., 2002). Previous analysis in C. elegans has shown that
ref-2 null mutants display an L1 larval lethal phenotype, possibly due to a disruption of the excretory system (Bertrand and Hobert, 2009; Bordet and Bertrand, 2018). However, while
ot762 animals are Unc (consistent with a function in motor neuron development), they do not die at L1. Therefore,
ot762 separates distinct functions of
ref-2 in different cell types.
ref-2 has also previously been shown to act upstream of the terminal selector
ttx-3 to control AIY interneuron specification (Bertrand and Hobert, 2009). We find that
ot762 animals display strong defects in
ttx-3::gfp expression (Fig.1D,E), albeit at a somewhat lesser penetrance than the
ot327 null allele (Bertrand and Hobert, 2009).
ref-2 is expressed in a number of neuroblasts, but not in the mature, adult nervous system (Bertrand and Hobert, 2009). Among the neuroblasts that express
ref-2 are the P neuroblasts (Alper and Kenyon, 2002; Bertrand and Hobert, 2009), which give rise to a number of motor neurons, including the AS and VD sister neurons (Sulston, 1976). Together with the
ref-2(
ot762) mutant phenotype that we describe here, as well as with P cell developmental defects described in
ref-2 mutants (Alper and Kenyon, 2002), this suggests that
ref-2 acts at some point in the P lineage to affect the differentiation of the AS and VD motor neurons. The phenotype that we observe in the AIY neurons of
ot762 is a reflection of the previously described function of
ref-2 in the neuroblast that generates the AIY neuron (Bertrand and Hobert, 2009). DD neuron differentiation defects observed in
ot762 mutants may be a reflection of
ref-2 function in embryonic neuroblasts akin to the function of
ref-2 in the AIY-generating neuroblast.