MEC-10 and MEC-19 Reduce the Neurotoxicity of the MEC-4(d) DEG/ENaC Channel in Caenorhabditis elegans

The Caenorhabditis elegans DEG/ENaC proteins MEC-4 and MEC-10 transduce gentle touch in the six touch receptor neurons . Gain-of-function mutations of mec-4 and mec-4(d) result in a hyperactive channel and neurodegeneration in vivo. Loss of MEC-6, a putative DEG/ENaC-specific chaperone, and of the similar protein POML-1 suppresses the neurodegeneration caused by a mec-4(d) mutation. We find that mutation of two genes, mec-10 and a new gene mec-19 (previously named C49G9.1), prevents this action of POML-1, allowing the touch receptor neurons to die in poml-1mec-4(d) animals. The proteins encoded by these genes normally inhibit mec-4(d) neurotoxicity through different mechanisms. MEC-10, a subunit of the mechanosensory transduction channel with MEC-4, inhibits MEC-4(d) activity without affecting MEC-4 expression. In contrast, MEC-19, a membrane protein specific to nematodes, inhibits MEC-4(d) activity and reduces MEC-4 surface expression.

Accumulation of high levels of constitutively-open ENaC channels or hyperactivation of gated DEG/ENaC channels can be very detrimental. For example, the excessive accumulation of ENaC channels in the kidney leads to increased sodium reabsorption and hypertension in Liddle syndrome in humans (Shimkets et al. 1994;Hansson et al. 1995a,b;Goulet et al. 1998). The hyperactivation of ASIC1 channels by ischemia and stroke-induced local acidosis causes massive neuronal death in mouse brains (Xiong et al. 2004). Gain-of-function mutations affecting Caenorhabditis elegans (C. elegans) DEG/ENaC proteins produce hyperactive channels that cause neuronal lysis and degeneration (Shreffler et al. 1995;Driscoll and Chalfie 1991;Chalfie and Wolinsky 1990) or hypercontraction of muscle (Park and Horvitz 1986;Liu et al. 1996). Studying the molecular mechanisms that regulate hyperactive DEG/ ENaCs can better our understanding of both channel hyperactivationinduced toxicity and normal channel physiology.
were confirmed as alleles of mec-10 by sequencing mec-10 DNA amplified from mutant worms by PCR. We assayed for gentle touch sensitivity in blind tests as described (Chalfie and Sulston 1981). We quantified the response by counting the number of responses to a total of 10 touches delivered alternately near the head and tail in 30 young adult animals (Hobert et al. 1999). We performed in vivo electrophysiology as described previously (O'Hagan et al. 2005).

Microscopy and immunofluorescence
Fluorescence and immunofluorescence were observed with a Zeiss Axio Observer Z1 inverted microscope equipped with 63· and 100·, NA 1.40 oil immersion objectives and a Photometrics CoolSnap HQ 2 camera (Photometrics, Tucson, AZ). Confocal images were acquired using Confocal ZEISS LSM700 equipped with a 63· NA 1.40 oil immersion objective. Live animals were anesthetized using 0.1 mM 2, 3-butanedione monoxime in 10 mM HEPES, pH 7.4.
MEC-4::TagRFP or immunofluorescence intensity in the cell body was determined by measuring the mean intensity of the entire cell body (20230 mm 2 ) and subtracting the mean intensity of nearby background of the same size using Image J (rsbweb.nih.gov/ij/). The intensity of the MEC-4::TagRFP puncta in TRN neurites was measured using the Puncta Analysis Toolkit beta developed by Dr. Mei Zhen (Samuel Lunenfeld Research Institute, Toronto, Canada). Puncta were examined over a region equivalent to approximate ten cell body lengths (~50 mm) starting near the cell bodies. The intensity of MEC-4 immunofluorescence in the TRN neurite was determined by measuring the mean intensity of 30250 mm lengths of the PLM neurite between cell bodies of PLM and PVM using Image J. We performed single-molecule fluorescence in situ hybridization as described previously .
Oocyte experiments cRNA expression and electrophysiology in Xenopus laevis oocytes followed the procedures and used the plasmids described in Goodman et al. (2002) except for the experiments with CaV2.1, which followed Fan et al. (2012). mec-19 cDNA of 390 bp was obtained by reversetranscription PCR from cDNA library (generated by reverse-transcription using wild-type mRNA) and was cloned in pGEM-HE (Liman et al. 1992). A total of 10 ng cRNA of mec-4(d), mec-2, and mec-10; 1 ng mec-6; and 1 ng cRNA of mec-19 were injected to oocytes unless noted (oocytes were a gift of Dr. Jian Yang and were obtained from frogs from Xenopus I, Dexter, MI, or Nasco, Fort Atkinson, WI). Oocytes were maintained as described previously (Árnadóttir et al. 2011). Membrane current was measured 426 d after RNA injection using a two-electrode voltage clamp as described previously (Goodman et al. 2002).
Immunoprecipitation of C-terminally HA-tagged MEC-19 and N-terminally Myc-tagged MEC-4(d) were performed 526 d after cRNA injection as described previously (Goodman et al. 2002) by using a rabbit polyclonal antibody against the HA tag (sc-805; Santa Cruz Biotechnology, Dallas, TX) and Protein A/G PLUS-Agarose (Santa Cruz Biotechnology). Protein was detected by using mouse monoclonal n Table 2 poml-1 suppression of mec-4(d) requires mec-10 and mec-19 A . T the 3rd nucleotide, intron 6 R 2 8 u894 G . A splicing junction, exon 9 -intron 9 Semi-D 2 2 u895 G . A splicing junction, exon 14 -intron 14 a DNA from u897 animals could not be amplified using primers that were 120 bp upstream of the start ATG and 80 bp downstream of the stop codon. n = 50 animals.
antibodies against the Myc (9E10; Sigma-Aldrich, St. Louis, MO) and the HA (sc-7392; Santa Cruz Biotechnology) tags and horseradish peroxidase2conjugated secondary antibodies (Jackson Immuno-Research Laboratories, West Grove, PA). Approximately three oocytes equivalents were loaded for the immunoprecipitation, and total lysate of one oocyte were loaded for the input. The specificity of the immunoprecipitation was confirmed in three ways. First, EGFP::HA, a negative control generated by the injection of 1 ng of the encoding cRNA, did not immunoprecipitate Myc::MEC-4(d).
Second, MEC-19::HA did not immunoprecipitate Myc::EGFP when 1 ng cRNA of constructs encoding each were coinjected. Third, the oocyte membrane protein b-integrin was not detected in the immunocomplexes by a monoclonal antibody against it (8C8; Developmental Studies Hybridoma Bank, University of Iowa, IA). Imaging and stoichiometry analysis of protein complexes on oocyte membranes using total internal reflection fluorescence microscopy were performed 122 d after cRNA injection as described previously Isacoff 2008, 2007;Abuin et al. 2011). The constructs of N and C-terminally EGFP-tagged MEC-4 have been described in Chen et al. (2015).

Statistics
Statistical analysis was performed using the Student's t-test, one-way analysis of variance (ANOVA), one sample t-test or the Mann2Whitney U-test using GraphPad Prism 5 software (http://www.graphpad. com/scientific-software/prism/) unless otherwise noted. The Student's t-test was used for most of the experiments, with the Welch's correction when data being compared did not have equal variances. The Mann2Whitney U-test was used to analyze the number of MEC-4 spots on the surface of Xenopus oocytes. P-values were adjusted with a Bonferroni correction when multiple comparisons were performed, and the raw P-values were also provided. The one sample t-test was used to analyze the western blots of MEC-4 expression in total lysates of Xenopus oocytes. One-way ANOVA was used to compare the number of mRNA molecules in wild type and two mec-19 mutants. In all figures, Ã , ÃÃ , ÃÃÃ indicate Bonferroni-corrected P-values of , 0.05, , 0.01, and , 0.001, respectively; ns, not significant.

Data and reagent availability
All strains used and/or generated in this study are available upon request. Strains are given in Table 1 and Table 2.

RESULTS
Loss of mec-10 or mec-19 enhances TRN cell death in poml-1 mec-4(d) animals Loss of poml-1 (e.g., with the ok2266 mutation) lowers MEC-4 protein levels and suppresses mec-4(d)2induced TRN degeneration (90% of the TRNs live; Chen et al. 2016). To identify genes whose products normally reduce MEC-4 activity and hence increase the TRN cell death when mutated, we screened for mutations that increased TRN cell death in poml-1(ok2266) mec-4(d) animals. The starting strain (TU3871) also contained mec-3p::tagrfp to label the TRNs and the FLP neurons and mec-17p::gfp to label the TRNs. Mutations that allowed TRN deaths would lack the TRN label but not the FLP label.
Seventeen such mutations were found among F2 progeny representing 20,000 haploid genomes after EMS mutagenesis [ Table 2; one mutation was a mec-3 non-coding mutation, which gave the phenotype by causing mec-3 expression in the FLP neurons, but not in the TRNs]. Fifteen of the mutations were X-linked and failed to complement each other. All 15 strains had mec-10 mutations; these mutations included nonsense alleles, missense alleles, a deletion allele, and several splice junction alleles. Several of these mec-10 mutations acted semidominantly. The mec-10(ok1104) allele, which is considered to be a loss-of-function deletion (Árnadóttir et al. 2011), also enhanced the TRN cell death in poml-1(ok2266) mec-4 (d) animals semidominantly ( Figure 1A). Addition of the wild-type gene rescued the effects of the mec-10 mutations ( Figure 1A). The inhibitory effect of MEC-10 on MEC-4(d)2induced TRN neurodegeneration is consistent with our previous finding that MEC-10 decreased MEC-4(d) activity in Xenopus oocytes (Goodman et al. 2002). Thus, both the in vivo and in vitro data suggest that MEC-10(+) inhibits MEC-4(d) channel activity.
The remaining mutation deleted a 288-bp sequence containing 19 bp upstream of start codon, the first exon and part of the first intron from C49G9.1. This mutation enhanced the mec-4(d) phenotype recessively ( Figure 1A). The effect on mec-4(d) degeneration was caused by this mutation, because it could be rescued by the wild-type gene ( Figure  1A). Given that a larger deletion allele (ok2504) gave a similar phenotype, both mutations are likely to be null alleles ( Figure 1A). Because of its effect on touch-sensitivity in a sensitized background (see MEC-19 reduces MEC-4 expression in the TRNs), we have renamed the gene mec-19.
We also tested the effect of the mec-10 and mec-19 mutations on the suppression of mec-4(d) by crt-1 and mec-6 mutations, which are known to suppress mec-4(d) deaths (Chalfie and Wolinsky 1990;Xu et al. 2001). (Both CRT-1 and MEC-6 act as endoplasmic reticulum chaperones for the production of MEC-4; Chen et al. 2016). Loss of mec-10 and mec-19 enhanced cell death in crt-1; mec-4(d) animals, but to a lesser extent ( Figure 1A) than they did in the poml-1 animals. In contrast, neither mec-10 nor mec-19 mutations promoted mec-4(d) degeneration when mec-6 gene was absent ( Figure 1A), probably due to a broader role of mec-6 in mec-4(d) function.
We next tested the effect of mec-10 or mec-19 mutations on touch sensitivity with or without the poml-1 mutation. The mec-10 null allele ok1104 caused a modest loss of the touch sensitivity (as previously seen Figure 2 The amino acid sequence of MEC-19 and its homologs in other nematode species. The predicted transmembrane (TM) region is in the black box. Sequence alignment was performed using ClustalW2 (http://www.ebi.ac.uk/ Tools/msa/clustalw2/). The sequences deleted in mec-19(u898) and mec-19(ok2504) are highlighted in red and blue, respectively.
by Árnadóttir et al. 2011), which was further reduced by poml-1 null mutations (ok2266 and u882; Figure 1B). The mec-10 poml-1 double mutation had a stronger effect on anterior touch sensitivity than posterior touch sensitivity ( Figure 1B). These data suggest that MEC-10 and POML-1 act additively in touch sensitivity but against each other with regard to MEC-4(d) channel activity. In contrast to mec-10, loss of mec-19 did not detectably change touch sensitivity either with or without a poml-1 mutation ( Figure 1B).

MEC-19 reduces MEC-4 expression in the TRNs
mec-19 encodes a novel membrane protein of 129 amino acids with one predicted transmembrane domain near its N-terminus (Figure 2). We identified similar proteins in other nematodes but not in other organisms (Figure 2). The gene is expressed in the TRNs, FLP neurons, and PVD neurons . A MEC-19:: GFP translational fusion was found throughout the TRN neurite and also on the plasma membrane and spots within the TRN cell body (Figure 3, A and B); its expression overlapped only partially with MEC-4 ( Figure 3A) and MEC-2  in the proximal neurite and cell body. In the cell body, MEC-19 spots also were found to partially overlap with the Golgi marker AMAN-2::TagRFP ( Figure 3B).
Thus, MEC-19 affects the amount of MEC-4 in the TRN neurite. The increase in cell death in mec-19; poml-1 mec-4(d) animals was likely due, at least in part, to elevated levels of surface . In contrast, mec-10 did not appear to affect MEC-4 protein levels and presumably enhanced mec-4(d) cell deaths through a different mechanism.
Consistent with the increased amount of MEC-4 in mec-19 TRN neurites, mec-19 loss increased the touch sensitivity of mec-4 ts animals (Gu et al. 1996) at various temperatures ( Figure  4A). However, loss of mec-19 did not detectably affect touch sensitivity in wild-type or poml-1 mutants ( Figure 4A and Figure 1B) and had only modest effects on the response of the mechanoreceptor current to different pressures, the peak amplitude at saturating stimuli, and the kinetics of the mechanoreceptor current (Figure 4, B and C).

MEC-19 reduces MEC-4 surface expression and activity in Xenopus oocytes
We next tested the effect of MEC-19 on MEC-4(d) currents in Xenopus oocytes. MEC-19 dramatically reduced the amiloride-sensitive current of MEC-4(d) coexpressed with MEC-6, POML-1, MEC-2, or MEC-10 by approximately 70-80% ( Figure 5A). [MEC-19 alone produced an amiloride-resistant current when expressed at a greater concentration in oocytes: I (at 285 mV) = 22.5 6 0.4 mA (mean 6 SEM) for 2.5 ng cRNA vs. I = 20.2 6 0.2 mA (n = 4) for 1 ng cRNA for oocytes 5 d after injection.] Thus, both in vivo and in vitro experiments suggest that wild-type MEC-19 inhibits MEC-4(d) channel activity. Part or all of this inhibition likely resulted from the loss of surface MEC-4 in oocytes, which was seen with total internal reflection fluorescence microscopy ( Figure 5, B and C). MEC-19 reduced MEC-4 surface expression with or without MEC-10 ( Figure 5, B and C; MEC-10 did not affect MEC-4 surface expression). Even in the presence of MEC-6, MEC-19 still reduced MEC-4 surface expression by nearly 50% (Figure 5B). The reduced MEC-4 surface expression in the presence of MEC-19 was not due to generally poor surface expression, because MEC-19 was well expressed on the surface of oocytes ( Figure 5B). The reduced MEC-4 surface expression also was not due to a reduction in total MEC-4 protein level in oocytes (relative amount was 1 without MEC-19 vs. 0.99 6 0.02 with MEC-19, mean 6 SEM, n = 3 independent experiments, not significant by one sample t-test). The action of MEC-19 on MEC-4(d) could be due to its physical interaction with it, since C-terminally HA-tagged MEC-19 coimmunoprecipitated with N-terminally Myc-tagged MEC-4(d) in oocytes ( Figure 6A). Figure 4 The effect of mec-19 mutations on touch sensitivity and on the mechanoreceptor current (MRC) in vivo. (A) mec-19(u898) and mec-19(ok2504) increase touch sensitivity of mec-4ts(u45) animals (mean 6 SEM, n = 30). Difference of touch responses between mec-4ts and mec-19(u898); mec-4ts or mec-19(ok2504); mec-4ts at 21°, 22°, 23°, and 24°; all had Bonferronicorrected P , 0.001 (raw P , 0.0001) by the Student's t-test, whereas the difference at 20°and 25°was not significant by the Student's t-test. Touch response between mec-19(u898); mec-4ts and mec-19(ok2504); mec-4ts was not significantly different from 20°to 25°b y the Student's t-test. (B) mec-19(u898) did not produce significant changes in the current vs. pressure (I vs. P) relation of MRCs. The peak amplitude of MRCs recorded from PLM (at -74 mV) at the onset of a mechanical stimulus was normalized to the maximum MRC current. Wild type is represented by the gray curve and white symbols. Each symbol (rectangle or circle) represents a recording from a different PLM cell. mec-19 or mec-19; poml-1 is represented by the black curve and black symbols. Wild type: P 1/2 = 4.5 6 0.7 nN/mm 2 , P slope = 3.1 6 0.7, N = 3 (Chen et al. 2016). mec-19: P 1/2 = 7.3 6 0.9 nN/mm 2 , P slope = 3.0 6 0.6, N = 2. mec-19; poml-1: P 1/2 = 7.0 6 1.2 nN/mm 2 , P slope = 5.0 6 1.0, N = 2. Data are represented as mean 6 SD. N indicates the number of cells tested. (C) mec-19 mutation had little effect on the average peak MRC amplitude, latency, activation (t1), and adaptation (t2) calculated from MRC response at the onset and offset of mechanical stimuli (mean 6 SEM). The data of wild type are from Chen et al. 2016. Ã P , 0.05, compared to the wild-type and mec-19; poml-1 double mutants, one-way analysis of variance with Tukey post hoc.
MEC-19 affected at least one other membrane channel, since it largely reduced the current from the human P/Q-type calcium channel CaV2.1 in frog oocytes (the maximal current of CaV2.1 was 26.3 6 1.1 mA without MEC-19 vs. 20.7 6 0.2 mA with MEC-19, mean 6 SEM, n = 5, P , 0.01, Student's t-test). MEC-19, however, did not affect channel proteins generally, since the surface expression of the BEST1 chloride channel (Sun et al. 2002) was unchanged in oocytes (the number of EGFP::BEST1 fluorescent spots on the surface was 99 6 21 without MEC-19 and 162 6 26 with MEC-19, mean 6 SEM, n = 15 patches from 7-8 cells, not significant by Student's t-test).
Because the expression of MEC-19 overlapped with that of MEC-4 and MEC-2 in the TRNs and coimmunoprecipitated with MEC-4(d) in oocytes, we asked whether it was part of the MEC-4/MEC-10 channel. We tagged MEC-19 with EGFP/mCherry at its C termini and expressed them in oocytes. The tagged protein retained its normal function because it acted like the untagged protein in rescuing the mec-19 enhancement of TRN cell death in poml-1 mec-4(d) animals (surviving TRNs, ALM 94 6 2%, PLM 92 6 3%, mean 6 SEM, n = 40 from five stable lines), and reduced the MEC-4(d) current amplitude when coexpressed with MEC-6 in oocytes [I Amil (at 285 mV) = 20.17 6 0.07 mA, mean 6 SEM, n = 4]. The stoichiometry of MEC-19 could not be determined because the molecules moved on the surface of oocytes even in the presence of MEC-4, and they did not colocalize with MEC-4 (Supporting Information, File S1). In addition, MEC-19 did not change the stoichiometry of the MEC-4 trimer (Chen et al. 2015) on the oocyte surface ( Figure 6B), an indication that this protein is not incorporated into the MEC-4 channel complex.

DISCUSSION
The poml-1 mec-4(d) double mutant provides a sensitized background in which to screen for genes that normally inhibit mec-4(d) degeneration. Using this double mutant, we identified two inhibitors, MEC-10 and MEC-19, that function downstream of POML-1. The average mutation rate in C. elegans for EMS mutageneses is approximately 1 in 2000 haploid genomes (Brenner 1974;Greenwald and Horvitz 1980). By examining the animals representing 20,000 haploid genomes, we are, thus, likely to have saturated for genes whose loss causes TRN degeneration in the poml-1 mec-4(d) background. The number of mec-10 alleles (15) supports this conclusion. The mec-10 alleles we found had a variety of defects, including missense, nonsense, and deletion mutations. In contrast, our previous screens for and MEC-10 (mean 6 SEM, n = 9-12 patches from 7-10 cells). 2.5 ng cRNA for EGFP::MEC-4 and mCherry::MEC-10, 1 ng cRNA for MEC-19 were injected to oocytes. Ã P , 0.05 by Mann2Whitney U-test with the Bonferroni correction (raw P = 0.009).
touch insensitive mutants only resulted in mec-10 missense mutations (Huang and Chalfie 1994). In fact animals lacking MEC-10 retain considerable touch sensitivity, a result that suggested that MEC-10 was partially redundant for touch sensitivity (Árnadóttir et al. 2011). The present screen, however, revealed a role for MEC-10 in the control of the MEC-4 channel.
The role for MEC-10 remains, however, elusive, because MEC-10 seems to have opposite effects on MEC-4 and MEC-4(d) channels. MEC-10 is needed for the optimal activity of the MEC-4 mechanotransduction channel, because its loss in vivo decreases the mechanoreceptor current amplitude by 25% and modestly decreases touch sensitivity (Árnadóttir et al. 2011). In contrast, MEC-10 inhibits MEC-4(d) both in vivo and in vitro: MEC-10 loss increases mec-4(d) toxicity in poml-1 mutants, and MEC-10 decreases the macroscopic MEC-4(d) current amplitude carried by either Na + or Ca 2+ in Xenopus oocytes (Goodman et al. 2002;Bianchi et al. 2004). These differences may result because the MEC-4 and MEC-4(d) channels function differently. Specifically, the wild-type MEC-4 channel may need MEC-10 to allow it to be maximally gated, whereas the MEC-4(d) channel, which is constitutively open, allows more current when MEC-10 is absent. Because MEC-10 does not affect MEC-4(d) surface expression (Árnadóttir et al. 2011), single-channel conductance, or open probability (Brown et al. 2008) in oocytes, it may act by inactivating some MEC-4(d) channels, making them unable to be opened.
In contrast to yielding many independent mec-10 mutants, our screen gave a single mec-19 strain, albeit one that contained an early deletion within the gene. The small size of the gene (MEC-19 has only 129 amino acids) is a likely explanation for the dearth of alleles identified in our screen. (The single non-null allele of mec-3 we identified is a non-coding mutation that affects the expression pattern of the gene; such mutations are expected to be rare.) Whereas MEC-10 modulates channel function, MEC-19 affects channel surface expression and counters the action of POML-1. POML-1 acts as an endoplasmic reticulum-resident chaperone for MEC-4 production and folding (Chen et al. 2016). In contrast, MEC-19, which is localized to the plasma membrane and, perhaps, the Golgi, reduces MEC-4 surface expression. MEC-19 is not part of MEC-4 channel complex, although it may transiently interact with MEC-4. Thus, the loss of mec-19 activity causes TRN degeneration in poml-1 mec-4(d) animals likely by increasing the number of MEC-4(d)-containing channels on the surface of the TRNs. The mechanism of MEC-19 action on the MEC-4 channel remains to be studied, in part, at least because MEC-19 is a novel protein we could find only in Caenorhabditis species. Given the localization of MEC-19 on the plasma membrane and its negative effect on MEC-4 surface expression, one possible hypothesis is that it may regulate the removal of the transduction channel from the plasma membrane. Alternatively, MEC-19 could inhibit the insertion of channel into the membrane. Although MEC-19 has not been found in other species, a similar mechanism may exist for other membrane proteins.
Our screen identified two genes that generated mec-4(d) deaths in the poml-1 background, and the protein products of these genes normally restrict the action of MEC-4(d). By screening F2 progeny from P0 animals, we biased the screen for mutations with very strong effects. Weaker suppression of poml-1 or enhancement of mec-4(d) might be revealed by testing specific candidates, such as the genes that are expressed in the TRNs, but whose loss does not produce touch insensitivity . Testing the effect of RNAi for these genes on TRN cell death in poml-1 mec-4(d) animals may identify more components that restrict mec-4(d) toxicity.

ACKNOWLEDGMENTS
We thank Jian Yang and Qiao Feng for providing Xenopus laevis oocytes and cRNA for CaV2.1; Oliver Hobert, Alexander Boyanov, and Gregory Minevich for whole-genome sequencing; and members of our laboratory for discussion. This work was supported by grants GM30997 to M.C. and NS35549 to E.Y.I. from the National Institutes of Health. R.O was supported by NJCSCR Postdoctoral Fellowship 10-2951-SCR-E-0.