DEC-7/SUSD2, a sushi domain-containing protein, regulates an ultradian behavior mediated by intestinal epithelial Ca²⁺ oscillations in Caenorhabditis elegans

Abstract

In Caenorhabditis elegans, rhythmic posterior body wall muscle contractions mediate the highly regular defecation cycle. These contractions are regulated by inositol-1,4,5-trisphosphate (InsP3) receptor-dependent Ca²⁺ oscillations in intestinal epithelial cells. Here, we find that mutations in dec-7, which encodes the nematode ortholog of the human Sushi domain-containing 2 protein (SUSD2), lead to an increase in InsP3 receptor-dependent rhythmic posterior body wall muscle contractions. DEC-7 is highly expressed in the intestinal epithelia and localizes to the cell-cell junction. The increase in rhythmic activity caused by the loss of dec-7 is dependent on the innexin gap junction protein INX-16. Moreover, DEC-7 is required for the clustering of INX-16 to the cell-cell junction of the intestinal epithelia. We hypothesize that DEC-7/SUSD2 regulates INX-16 activity to mediate the rhythmic frequency of the defecation motor program. Thus, our data indicate a critical role of a phylogenetically conserved cell-cell junction protein in mediating an ultradian rhythm in the intestinal epithelia of C. elegans.

NEW & NOTEWORTHY The conserved complement group protein DEC-7/SUSD2 acts at the apical cell-cell junction of C. elegans intestinal epithelia to mediate gap junction protein organization and function to facilitate a Ca2+ wave-regulated ultradian behavior.

INTRODUCTION

Biological rhythms are widespread behaviors that are observed from single-cell organisms to humans and are thought to be a critical evolutionary adaptation. A primary example of this adaptation, which is observed in most organisms, is the 24-h circadian rhythm that has emerged as an adaptation to the Earth’s rotation. Importantly, biological rhythms are diverse and range from subsecond to years. Examples include repetitive firing of neurons (subsecond), breathing (seconds), heart contractions (seconds), intestinal peristalsis (minutes to hours), hormone release (minutes to hours), sleep periods (hours), ovulation cycles (weeks), and cicada eclosion rhythm (years). Since these rhythms are intrinsic to all living organisms, to understand organismal behavior or physiological processes, it is fundamental to understand the mechanisms underlying rhythmic behaviors. Unlike rhythms with a longer period, such as the widely studied circadian rhythms, which are dependent on changes in gene regulation and expression, ultradian rhythms (rhythm with a cycle less than a day) will likely depend on changes in cellular physiology and signaling pathways. To gain a mechanistic understanding of ultradian rhythms, there are many advantages to studying rhythmic behaviors in a model system that is amenable to a genetic approach. Indeed, the Caenorhabditis elegans defecation motor program (DMP) has proven to be a simple ultradian rhythmic behavior that is amenable to genetic and physiological analyses.

This rhythmic motor program consists of three distinct motor sequences that occur every ∼45 s. At the behavioral level, the DMP initiates with a posterior body wall muscle contraction (pBoc), followed by an anterior motor contraction (aBoc), and culminates with enteric muscle contraction that allows for the expulsion of waste material (Exp). Through several genetic and physiological studies, the DMP has been found to be regulated by a rhythmic endogenous posterior to anterior intestinal Ca²⁺ wave that is mediated by the inositol 1,4,5-trisphosphate (InsP3) receptor (1–6). In agreement with the InsP3 receptor functioning as the central timekeeper, a weak reduction of function mutation in the gene encoding the sole InsP3 receptor in C. elegans, itr-1, results in an extended DMP period, and, correspondingly, overexpression of itr-1 results in a shorter DMP period and, lastly, loss of itr-1 abolishes the DMP behavior (1). Several other phylogenetically conserved gene products have been identified that can modulate the time-keeping rhythm of the intestinal epithelial InsP3 receptor. For example, mutations in the gene products encoding the sole C. elegans phospholipase C (PLC) gamma, plc-3, and a TRPM (transient receptor potential melastatin) ion channel, gon-2, result in an extended DMP period (2, 7). PLC-3/PLCγ mediates the hydrolysis of phosphatidylinositol 4,5-bisphosphate to generate InsP3 and mediates Ca²⁺ entry via GON-2 TRPM channels to regulate ITR-1 activity (2, 5, 6). However, unlike loss of itr-1, which abolishes the Ca²⁺ wave and the DMP, loss of plc-3 and gon-2 result in only an asynchronous and extended DMP interval and Ca²⁺ wave frequency, indicating the central role of ITR-1 as the DMP timekeeper and the modulatory role of PLC-3/PLCγ and the GON-2 TRPM channel.

Here, to further investigate the mechanisms regulating the rhythmic DMP in C. elegans, we have found evidence that VAV-1, a Rho family guanine nucleotide exchange factor, acts in the same pathway as PLC-3/PLCγ to modulate InsP3 receptor activity and, through an RNA inference (RNAi) screen, we have identified a phylogenetically conserved complement control protein, orthologous to human SUSD2 (sushi domain-containing 2), that acts to negatively regulate the DMP rhythmic frequency. Loss of function mutations in the gene encoding the C. elegans SUSD2, dec-7, or RNAi depletion of dec-7 result in a shortened DMP cycle and suppresses the extended DMP cycle observed in gon-2, plc-3, and vav-1 mutants. SUSD2 is expressed in the intestinal epithelia cells and is localized to cell-cell junctions. We have found that the shortened DMP cycle frequency observed in dec-7 mutants is dependent on the cell-cell junction localized gap junction protein INX-16/innexin and that dec-7 function is required for the proper clustering of INX-16 to the cell-cell junction in intestinal epithelia. These studies provide evidence that DEC-7/SUSD2 has a novel role in mediating DMP rhythmic frequency.

MATERIALS AND METHODS

Nematode Culture and Genetics

Nematodes were cultured at 20°C using standard methods on nematode growth medium (NGM) agar plates seeded with OP50 E. coli bacteria (8, 9). Wild-type worms were the Bristol N2 strain and where indicated the VC2010 strain. The following alleles were used in this study: dec-7(sa296) III, dpy-17(e164) dec-7(qm166) III, dec-7(gk183060) III, dec-7(gk183066) III, dec-7(gk339739) III, dec-7(gk434141) III, dec-7(gk656760) III, dec-7(gk733664) III, dec-7(gk183067gk183068) III, egl-8(sa47) V, gon-2(ok465) I, inx-16(tm1589) I, plc-3(tm1340) II, and vav-1(ak41) X. The following transgenic strains were used in this study: mcIs46 [dlg-1::RFP + unc-119(+)], oxIs143 [inx-16::GFP, lin-15(+)] V, takEx240 (K03H1.5p::GFP), takEx395 (K03H1.5p::K03H1.5::GFP), takEx516 (K03H1.5p::K03H1.5::GFP-NIDO/IPT deletion), takEx517 (K03H1.5p::K03H1.5::GFP-NIDO/IPT deletion), takEx519 (K03H1.5p::K03H1.5::GFP-NIDO/IPT deletion), takEx521 (K03H1.5p::K03H1.5::GFP-AMOP deletion), takEx524 (K03H1.5p::K03H1.5::GFP-AMOP deletion), takEx525 (K03H1.5p::K03H1.5::GFP-AMOP deletion), takEx495 (K03H1.5p::K03H1.5::GFP-vWD deletion), takEx505 (K03H1.5p::K03H1.5::GFP-vWD deletion), takEx509 (K03H1.5p::K03H1.5::GFP-vWD deletion), takEx497 (K03H1.5p::K03H1.5::GFP-CCP/SUSHI deletion), takEx500 (K03H1.5p::K03H1.5::GFP-CCP/SUSHI deletion), takEx501 (K03H1.5p::K03H1.5::GFP-CCP/SUSHI deletion), takEx506 (K03H1.5p::K03H1.5::GFP-TM/CYTO deletion), takEx508 (K03H1.5p::K03H1.5::GFP-TM/CYTO deletion), and takEx510 (K03H1.5p::K03H1.5::GFP-TM/CYTO deletion).

DNA Constructs and Transgenesis

K03H1.5 promoter and coding sequence were isolated from fosmid WRM0629cB06 (C. elegans Vancouver fosmid library, Reverse Genetics Core Facility). For the dec-7p::GFP transcriptional reporter construct, primers were engineered with PstI and BamHI restriction enzyme sites and amplified using Q5 High-Fidelity PCR Kit (NEB). The resulting amplicon was ligated into pPD95.75 (Fire Lab Vector Kit) using Quick Ligation Kit (NEB). After ligation, DNA was subsequently transformed into MAX Efficiency DH5 alpha competent cells (Invitrogen).

The K03H1.5::GFP translational fusion was created using primers optimized for Gibson Assembly (NEB). In short, primers were designed to amplify vector (pPD95.75) and inserts with a nonpriming overlap region homologous to the 5′ terminal sequence of the adjacent fragment. Vector was amplified with primers that lacked overlaps. K03H1.5 was amplified from the WRM0629cB06 fosmid with a forward primer that overlaps the 3′ end of the vector, and the reverse overlapped with the 5′ end of the vector. These overlapped regions, which are located on the 5′ end of the primers, allow contiguous fragments to effectively ligate in a single reaction. The resulting reactions were then transformed into MAX Efficiency DH5 alpha competent cells (Invitrogen).

DEC-7::GFP deletion reporter constructs were made using QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent). Two complementary oligonucleotides were designed to omit the domain sequence desired. After amplification, the parental template was digested with DpnI and transformed into MAX Efficiency DH5 alpha competent cells (Invitrogen).

Before injection into animals, all constructs were sequenced to verify no spurious DNA mutations were acquired during cloning. Using standard procedures (10), DNA constructs were either directly introduced into wild-type worms (the transcriptional and full-length translational reporters were injected into wild-type animals) or dec-7(sa296) animals [the deletion constructs were all injected into dec-7(sa296) mutants] using microinjection with a ttx-3p::GFP coinjection marker that fluorescently marks the bilaterally symmetrical AIY neurons (gift from O. Hobert, Columbia University).

RNA Interference

C. elegans strains were fed E. coli [HT115(DE3)] bacteria producing dsRNA to knock down targeted gene expression (11). For the RNAi screen, genes on chromosome III represented from four 384-well plates were selected from the Ahringer RNAi library (Source BioScience - plates III-2, III-3, III-4, and III-5). The dsRNA bacteria used for additional RNAi experiments [e.g., egl-8, gar-2 (as a control), itr-1, and plc-3] were also obtained from the Ahringer RNAi library (Source BioScience).

Behavioral Analysis

For all non-RNAi genetic analyses, P0 young adults were grown on NGM supplemented with OP50 bacteria, and F1 young adults were assayed on fresh NGM plates supplemented with OP50.

For all genetic analyses using RNAi, NGM containing 100 µg/mL ampicillin and 1 mM IPTG (isopropylthio-β-galactoside) were seeded with HT115(DE3) RNAi bacteria targeting genes of interest. P0 young adults were grown on RNAi feeding plates, and F1 young adults were assayed on fresh NGM plates supplemented with OP50.

For behavioral assays, animals were grown at 20°C (unless otherwise noted) and assayed as young adult animals at a room air temperature of 22–24°C.

DMP frequency was determined by observing pBoc intervals using a Zeiss SteREO Discovery.V8 stereo dissecting microscope. As previously defined (12), the timing of 5–10 consecutive pBoc cycles was determined in each animal.

Dextran Feeding

Strains were grown on NGM plates for 2–3 days at 20°C. Worms were washed off plates with M9 buffer, pelleted, and subsequently transferred to a light-blocking microcentrifuge tube. Dextran-tetramethylrhodomine (TRITC), 40,000 MW (ThermoFisher #D1842) labeling of animal intestines was carried out by incubating animals in 100 µL of dextran-TRITC (5 mg/mL stock) and 10 µL of concentrated OP50 on bench top for 30 min at room temperature with brief gentle mixing every 5 min. After incubation, 500 µL of 50 mM NaN₃ was added, mixed, and incubated for 1 min. Tubes were centrifuged, supernatant was removed, and pellet was washed with M9. After 1 min of centrifugation and subsequent supernatant removal, the remaining 100 µL was pipetted onto a fresh NGM plate seeded with OP50. Young adult animals were imaged immediately.

Microscopy

Microscopy analyses were conducted on young adult staged animals. mcIs46 (DLG-1::RFP)-expressing young adult animals were immobilized on a 2% agarose pad using 50 mM NaN₃ and imaged using an Olympus Fluoview 1200 confocal laser scanning microscope with a ×60 oil immersion objective lens and z-motor step of 0.2 µm (50–125 z-sections were taken for each image). Images were compiled and analyzed using Olympus Fluoview software.

Similarly, oxIs146 (INX-16::GFP)-expressing age-matched animals were immobilized on a 2% agarose pad using 50 mM NaN₃ and imaged using a Nikon A1R HD25 confocal laser scanning microscope with a ×60 oil immersion objective lens and z-motor step of 0.125 µm (75–100 z-sections were taken for each image). Images were compiled and analyzed using Nikon Elements software. Post hoc analyses of images were carried out using ImageJ (Fiji). For determining INX-16::GFP clustering, z-projections were made using maximum intensity projection, and the extent of INX-16::GFP clustering was calculated using the straight-line measurement tool in ImageJ (Fiji). Three measurements were made for each animal and experiments and analyses were done blind to genotype.

All other transgenic and dextran-labeled, age-matched animals were immobilized on a 2% agarose pad using 50 mM NaN₃ and imaged under a ×63 oil immersion objective lens on a Zeiss AxioObserver inverted microscope equipped with an Andor Clara CCD camera. Images were compiled using Metamorph software.

Dark-field microscopy to determine lipid stores was carried out essentially as previously described (13). In brief, using identical light exposure, camera, and microscope settings, age-matched young adult animals were imaged on a Nikon SMZ18 microscope equipped with a dark field illuminator and a pco.edge CMOS camera. Nikon Elements software was used for image collection. Post hoc analyses of images were carried out using ImageJ (Fiji).

Statistical Analysis

pBoc cycle length data are presented as mean ± SEM. Gaussian distribution was determined using D’Agostino-Pearson and Shapiro–Wilk normality tests. If the data followed a normal distribution (Gaussian), a parametric test was employed, and if not, a nonparametric test was used. Statistical significance was determined using Student’s two-tailed t test or Mann–Whitney U test. When comparing three or more groups, statistical significance was determined by one-way ANOVA with appropriate multiple comparisons post hoc test (Dunnett’s or Kruskal–Wallis). P values of ≤ 0.05 were taken to indicate statistical significance. The rhythmicity of the pBoc cycle is quantified as coefficient of variance, which is the standard deviation expressed as a percentage of the sample mean. GraphPad Prism software (Version 9 for macOS) was used for statistical analysis and graphical representation.

RESULTS

Loss of DEC-7/SUSD2 Suppresses the Long Arrhythmic Defecation Cycles Observed in vav-1, plc-3/PLCγ, and gon-2/TRPM Mutants

In addition to the loss of plc-3 and gon-2 function, loss of function mutations in the gene product encoding VAV-1, a phylogenetically conserved Rho family guanine nucleotide exchange factor, have also been implicated in regulating the cycle timing of the DMP (4, 14). Indeed, vav-1 mutants show extended and irregular intestinal epithelial Ca²⁺ waves and DMP frequency comparable to plc-3 and gon-2 mutants (2, 4, 5, 7). Due to this similar DMP phenotype observed in vav-1, plc-3, and gon-2 mutants, we sought to determine whether vav-1 is acting in the same genetic pathway as plc-3 and gon-2 to regulate DMP rhythm. Thus, we analyzed the DMP frequency (pBoc to pBoc interval) in vav-1 loss of function mutants treated with plc-3(RNAi). We found that plc-3 RNAi-depleted animals show extended and arrhythmic (as indicated by a high coefficient of variance) DMP cycles as previously observed (Fig. 1, A and B) (2). RNAi depletion of plc-3 in loss of function vav-1 mutants did not further exacerbate the defecation period length or arrhythmicity defects unlike egl-8 mutants treated with plc-3(RNAi) (Fig. 1, A and C). egl-8 encodes PLCβ that has been shown to act in a parallel pathway to plc-3/PLCγ in mediating InsP3 receptor activity in the intestinal epithelia (2). Indeed, loss of both plc-3/PLCγ and egl-8/PLCβ abolishes the DMP (Fig. 1C), which phenocopies the loss of the Ins3P receptor (Fig. 1J). These data suggest that vav-1 and plc-3 act in a common genetic pathway to mediate Ins3P receptor timekeeper activity.

Figure 1.

Loss of dec-7/K03H1.5 function suppresses the extended DMP cycle observed in vav-1, plc-3, and gon-2 mutants. A: plc-3 RNAi depletion and loss of vav-1 extends DMP periodicity. The extended DMP cycle of vav-1 mutants is not further extended by plc-3 RNAi depletion. Number of animals examined is indicated in bar. B: DMP coefficient of variance is increased in vav-1, plc-3(RNAi), and plc-3(RNAi); vav-1 animals compared with wild-type control animals. Number of animals examined is the same as indicated in A. C: plc-3 RNAi depletion in egl-8 mutants abolished DMP behavior. For all genotypes n = 15 animals examined. D: dec-7 RNAi depletion suppresses the extended DMP period observed in vav-1 mutants. Number of animals examined is indicated in bar. E: DMP coefficient of variance is increased in vav-1 mutants compared with wild-type control animals and is not suppressed by dec-7 RNAi depletion. Number of animals examined is the same as indicated in D. F: dec-7 RNAi depletion suppresses the extended DMP period observed in plc-3 mutants. Number of animals examined is indicated in bar. G: DMP coefficient of variance is increased in plc-3 mutants compared with wild-type control animals and is not suppressed by dec-7 RNAi depletion. Number of animals examined is the same as indicated in F. H: dec-7 RNAi depletion suppresses the extended DMP period observed in gon-2 mutants. Number of animals examined is indicated in bar. I: DMP coefficient of variance is increased in gon-2 mutants compared with wild-type control animals and is not suppressed by dec-7 RNAi depletion. Number of animals examined is the same as indicated in H. J: loss of dec-7 function does not restore DMP behavior in itr-1 RNAi-depleted animals. For all genotypes n = 12 animals examined. Data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Kruskal–Wallis multiple comparison test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, *P < 0.05, **P < 0.01, ****P < 0.0001. DMP, defecation motor program; pBoc, posterior body wall muscle contraction.

To help further understand the mechanisms and identify additional gene products involved in the regulation of epithelial Ca²⁺ oscillations that control DMP frequency in C. elegans, we conducted an RNAi screen in the vav-1 mutant background to identify gene products that could improve the DMP periodicity defect observed in vav-1 mutants. From this RNAi suppressor screen, we found that RNAi depletion of the gene encoded by the K03H1.5 open reading frame (from here on referred to as dec-7, see next section titled Loss of dec-7 Increases the Frequency of the Rhythmic DMP Cycle for details) suppressed the long DMP cycle period of vav-1 mutants (Fig. 1D). Although the extended period of the DMP cycle of vav-1 mutants is suppressed when dec-7 is depleted, the DMP frequency remained arrhythmic (Fig. 1E). As we found evidence that vav-1 acts in a common genetic pathway with plc-3 (Fig. 1A), we investigated whether knockdown of dec-7 also suppressed the defecation timing defects observed in plc-3 mutants. Indeed, similar to vav-1, dec-7 depletion also suppressed the extended DMP period defect observed in plc-3 mutants (Fig. 1F). However, similar to vav-1 mutants, dec-7 depletion did not improve the DMP arrhythmicity observed in plc-3 mutants (Fig. 1G). PLC-3/PLCγ is proposed to mediate the defecation cycle by modulating the activity of the GON-2 TRPM channel. Similar to plc-3 and vav-1 mutants, mutations in gon-2 disrupt the defecation cycling timing and epithelial Ca²⁺ oscillations (5). Comparable to the analyses with vav-1 and plc-3 mutants, we found that the extended timing but not the arrhythmic frequency of the DMP observed in gon-2 mutants was suppressed by RNAi depletion of dec-7 (Fig. 1, H and I). As InsP3 receptor is the central mediator of the Ca²⁺ oscillations that regulate DMP frequency (1), we tested whether dec-7 loss of function mutants (see below) could suppress the defecation defect observed in InsP3 receptor-depleted animals. Depletion of the InsP3 receptor using itr-1(RNAi) results in the absence of the DMP, consistent with its essential role in mediating the Ca²⁺ oscillation driving the DMP cycle (1). Unlike the suppression of the long defecation cycles observed in gon-2, plc-3, and vav-1 mutants, we found that dec-7 mutations could not restore defecation behavior in itr-1(RNAi)-depleted animals. Indeed, unlike untreated wild-type or dec-7 mutant animals, wild-type and dec-7 mutants treated with itr-1(RNAi) failed to produce observable pBoc behavior (Fig. 1F). These data indicate that dec-7 acts upstream of InsP3 receptor and either parallel or downstream of gon-2, plc-3, and vav-1 to mediate the DMP timing.

Loss of dec-7 Increases the Frequency of the Rhythmic DMP Cycle

As the loss of dec-7 suppresses the extended DMP rhythm in vav-1, plc-3, and gon-2, we next examined the defecation cycle in dec-7(RNAi)-depleted animals and dec-7 loss of function mutants. We found that wild-type animals treated with dec-7(RNAi) resulted in animals with significantly shorter DMP period (Dec-s) than is observed in control RNAi-treated wild-type animals (Fig. 2A). Indeed, control RNAi animals have a mean defecation rate of ∼47 s, whereas dec-7(RNAi) animals have a mean defecation rate of ∼33 s (Fig. 2A). Although the DMP cycle frequency was faster by dec-7 depletion, the DMP rhythmicity as indicated by the DMP coefficient of variance remained intact (Fig. 2B). Next, we searched the Million Mutation Project strain collection to determine if any dec-7 mutants were identified. The Million Mutation Project is a collection of C. elegans strains harboring randomly induced mutations whose genomes are fully sequenced (15). From this collection, we have determined the DMP frequency of seven independently isolated strains; six of these strains have premature stop codon mutations and one strain has two missense mutations in the dec-7 gene product (Fig. 3A). All seven of these strains show a short DMP cycle length (Dec-s) compared with wild-type animals and the Million Mutation Project parental strain (VC2010) (Fig. 2C). In addition, like dec-7 RNAi depletion, all of the Million Mutation Project dec-7 mutants displayed a highly regular defecation cycle (Fig. 2D). Lastly, we found that dec-7/K03H1.5 gene product maps to a region of chromosome III that contains a previously uncloned locus that is implicated in DMP frequency. Indeed, the previously identified mutants dec-7(sa296) and dec-7(qm166) were found to have a Dec-s phenotype (12, 16). DNA sequencing of both dec-7(sa296) and dec-7(qm166) mutants revealed G to A nucleotide transitions that cause a C1144Y amino acid change in the K03H1.5 gene product in dec-7(sa296) mutants and a G791E amino acid change in the K03H1.5 gene product in dec-7(qm166) mutants. Consistent with previous studies, we found that both of these mutants have a Dec-s phenotype similar to RNAi knockdown and the Million Mutation Project dec-7 mutant strains and do not show DMP arrhythmicity (Fig. 2, E–H). Moreover, since the dec-7(sa296) mutant was originally described to have a temperature dependence on DMP frequency (12), we examined this mutant as well as dec-7 RNAi-depleted animals and dec-7(qm166) mutants at 20°C and 25°C. Similar to what was observed for dec-7(qm166) (16), we found that dec-7(sa296), dec-7(qm166), and dec-7 RNAi-depleted animals all had a significantly shorter DMP period that was not impacted by temperature (Fig. 3). Taken together, these data indicate that dec-7 encodes a critical regulator of DMP period frequency.

Figure 2.

Loss of dec-7/K03H1.5 function shortens DMP periodicity. A: dec-7 RNAi depletion shortens DMP periodicity. Number of animals examined is indicated in bar. B: DMP coefficient of variance of dec-7 RNAi depletion is indistinguishable from wild-type control animals. Number of animals examined is indicated in bar. C: Million Mutation Project dec-7 mutants have a shortened DMP periodicity. Number of animals examined is indicated in bar. D: DMP coefficient of variance of Million Mutation Project dec-7 mutants is indistinguishable from wild-type control animals. Number of animals examined is the same as indicated in C. E: dec-7(sa296) mutants have a shortened DMP period. Number of animals examined is indicated in bar. F: DMP coefficient of variance of dec-7(sa296) mutants is indistinguishable from wild-type control animals. Number of animals examined is indicated in bar. G: dec-7(qm166) mutants have a shortened DMP period. Number of animals examined is indicated in bar. H: DMP coefficient of variance of dec-7(qm166) mutants is indistinguishable from wild-type control animals. Number of animals examined is indicated in bar. Data are presented as means ± SE. Statistical significance was determined by an unpaired t test (A, B, E, F, G, and H) or one-way ANOVA with Dunnett’s multiple comparison test (C and D). Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, ****P < 0.0001. DMP, defecation motor program.

Figure 3.

The reduced DMP periodicity of dec-7 mutants or RNAi depletion is not impacted by temperature. Quantification of DMP period. dec-7 RNAi depletion shortens DMP periodicity in animals raised at 20°C (A) and 25°C (C). Number of animals examined is indicated in bar. B and D: quantification of DMP coefficient of variance. Number of animals examined is indicated in A and C, respectively. Data are presented as means ± SE. Statistical significance was determined by Mann–Whitney test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, ****P < 0.0001. Quantification of DMP period or animals raised at 20°C (E) and 25°C (G). dec-7 mutants show short DMP period compared with wild-type animals at both 20°C (E) and 25°C (G). Number of animals examined is indicated in bar. F and H: quantification of DMP coefficient of variance. Number of animals examined is indicated in E and F, respectively. Data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Kruskal–Wallis multiple comparison test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, *P < 0.05, ****P < 0.0001. DMP, defecation motor program.

dec-7 Encodes a SUSD2 Ortholog That Localizes to the Apical Junction of Intestinal Epithelia

The dec-7/K03H1.5 open reading frame encodes a putative large 1,385 amino acid phylogenetically conserved complement control protein that is orthologous to the mammalian sushi domain-containing 2 protein (SUSD2) and Drosophila Mesh. The primary structure of DEC-7 contains a NIDO (Nidogen) domain, an IPT (Ig-like) domain, an AMOP (adhesion-associated domain in MUC4 and other proteins) domain, a von Willebrand factor, type D (vWD) domain, a Sushi/CCP domain, and a characteristic transmembrane domain near the C-terminus (Fig. 4A). Interestingly, several dec-7 missense mutations that cause a Dec-s phenotype fall in the coding region that encodes several of the phylogenetically conserved domains of DEC-7. Indeed, the dec-7(qm166) mutation alters the AMOP domain (G791E) and dec-7(gk183067gk183068) mutations affect the vWD domain (S923R and R936K) (Fig. 4A).

Figure 4.

DEC-7 is phylogenetically conserved and localizes to the apical junction of intestinal epithelia. A: schematic representation of human Sushi domain-containing 2 (SUSD2), Drosophila Mesh, and C. elegans DEC-7. The asterisks indicate location of premature stop codons associated with the Million Mutation Project dec-7 mutants. *1 dec-7(gk183060), *2 dec-7(gk434141), *3 dec-7(gk339739), *4 dec-7(gk183066), *5 dec-7(gk733664), and *6 dec-7(gk656760). B: representative images of a live animal expressing a dec-7p::GFP transcriptional reporter in the intestinal epithelia. White arrowhead indicates coinjection marker and arrows highlight the intestinal epithelia. Scale bar equals 100 µm. C: schematic representation of intestinal epithelia. The yellow boxes indicate the apical junction region shown in D and subsequent representative apical junction organization micrographs. D: representative images of live animals expressing a functional DEC-7::GFP fusion protein and DLG-1::RFP showing the colocalization of these proteins at the intestinal epithelia cell-cell apical junction. Scale bar equals 10 µm. AMOP, adhesion-associated domain in MUC4 and other proteins domain; IPT, Ig-like domain; NIDO, Nidogen domain; SB, somatomedin B domain; Sushi, Sushi/complement control protein domain; vWD, von Willebrand factor, type D domain.

To determine where dec-7 is expressed, we generated transgenic animals expressing a dec-7p::GFP transcriptional fusion and found that dec-7 is highly expressed in the intestinal epithelia (Fig. 4B). Next, to establish the subcellular distribution of dec-7 gene product, we generated animals expressing a functional GFP-tagged DEC-7 construct. Analysis of these animals demonstrated that DEC-7::GFP is localized to the cell-cell junction of the intestinal epithelia. In C. elegans, as in other organisms, the epithelial cell-cell junction provides a permeability barrier, mediates strong adhesive linkages between neighboring epithelial cells, and establishes and maintains cellular polarity (Fig. 4C). The C. elegans cell-cell junction, known as the apical junction, is a single discernable electron-dense structure that is composed of classical cadherin-catenin adhesion complexes, the MAGUK (membrane-associated guanylate kinase) family protein DLG-1/Disks large and several other phylogenetically conserved proteins (17–19). DLG-1 is required for apical polarity and formation of the C. elegans apical junction. We found that coexpression of DEC-7::GFP with a functional RFP-tagged DLG-1 (DLG-1::RFP) (20) demonstrated that DEC-7::GFP colocalizes with DLG-1::RFP in the same region within the apical junction of the intestinal epithelia (Fig. 4D). These colocalization imaging analyses are consistent with a previous report that discovered that endogenous DLG-1 copurifies with the dec-7/K03H1.5 gene product (21).

As the conserved domains of DEC-7 are implicated in adhesion molecule protein-protein interactions and we found that mutations within two of the conserved domains of DEC-7 disrupt DMP frequency, we set out to determine the functional domains of DEC-7 that are required for its localization and activity in maintaining normal DMP frequency. To accomplish this, we generated a panel of deletion constructs where individual domains were removed from the functional DEC-7::GFP fusion protein construct that rescues the dec-7 loss of function Dec-s phenotype (Fig. 5A). We characterized three independently isolated transgenic lines for each deletion construct. Unlike full-length DEC-7::GFP, we found that each deletion construct failed to rescue the Dec-s phenotype observed in dec-7 mutants (Fig. 5, A and C). Moreover, we found that the DEC-7 deletion variants also failed to localize to the apical junction of the intestinal epithelia and were instead found in a diffuse unorganized pattern (Fig. 5, B and C). These data indicate that the domain composition of DEC-7 is important for its subcellular localization and function in mediating DMP periodicity.

Figure 5.

Functional analysis of the conserved domains of DEC-7. A: deletion of individual domains of DEC-7, unlike full length DEC-7 [dec-7(-); Rescue], failed to rescue the Dec-s phenotype associated with dec-7(sa296) mutants. Data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Dunnett’s multiple comparison test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, ****P < 0.0001. For all genotypes, n is ≥ 16. B: representative images of DEC-7::GFP localization. Deletion of individual domains of DEC-7, unlike full length DEC-7, failed to localize to the apical junction of the intestinal epithelia. Arrowheads indicate DEC-7 apical junction localization. Scale bar equals 10 µm. C: table summarizing the DEC-7 domain function in the rescue of the dec-7(sa296) Dec-s phenotype and subcellular localization of DEC-7::GFP constructs in dec-7(sa296) mutants. “+” indicates rescue/apical junction localization and “−” indicates failure to rescue/failed apical localization.

Intestinal Integrity is Not Affected by Loss of dec-7

Previous work carried out in Drosophila has shown that the Drosophila ortholog of DEC-7, Mesh, localizes to the intestinal septate junctions and is critical for septate junction integrity and barrier function (22–24). Therefore, to investigate the integrity of the analogous apical junction in intestinal epithelia of C. elegans, we analyzed the localization and organization of DLG-1::RFP, a critical component of apical junctions (25, 26). DLG-1::RFP in dec-7 loss-of-function animals is organized in a linear pattern within the apical junction of the intestinal epithelia and is indistinguishable from wild-type animals (Fig. 6, A and B). Moreover, we analyzed intestinal integrity by feeding dec-7 animals TRITC-labeled dextran-40 and examined epithelial cohesion. Consistent with the DLG-1::RFP analyses, dextran-40 accumulation in dec-7 loss-of-function animals was indistinguishable from wild-type animals. Dextran-40 was evenly distributed throughout the lumen of the intestine and no ruptures were evident in dec-7 mutants (Fig. 6, C and D). Typically, animals with mutations that disrupt intestinal function are metabolically challenged and show developmental delay, low brood size, and lowered fat content (27). Although dec-7 mutants appear superficially wild type, to determine whether loss of dec-7 causes metabolic defects, we investigated brood size, developmental growth rate, and lipid content. Consistent with normal intestinal morphology and development, dec-7 mutants have a similar brood size, developmental growth rate, and lipid accumulation to age-matched wild-type animals (Fig. 6, E–H). Therefore, unlike the Drosophila Mesh protein, these data indicate a nonessential role of DEC-7 in intestinal integrity.

Figure 6.

Intestinal integrity is not affected by loss of dec-7. Representative images of DLG-1::RFP organization at the apical junction in wild-type (A) and dec-7 RNAi-depleted animals (B). Scale bar equals 10 µm. C and D: representative brightfield and fluorescence (TRITC-Dextran-40) images showing normal intestinal structure and integrity in dec-7(sa296) mutants. Scale bar equals 10 µm. E: quantification of progeny number from wild-type and dec-7(sa296) mutant animals. Data are presented as means ± SE. Statistical significance was investigated by an unpaired t test. ^nsP > 0.05. n ≥ 11. F: quantification of developmental stages hours after egg laying. The four larval stages are indicated by L1, L2, L3, and L4, and Y. Adult indicates young adult stage. G: representative dark-field images showing relative lipid content by light scattering density. H: quantification of light scattering densities in wild-type and dec-7(sa296) mutant animals. Data are presented as means ± SE. Statistical significance was determined by an unpaired t test. ^nsP > 0.05. n ≥ 24.

The Dec-s Phenotype Associated with Loss of dec-7 Function is Suppressed by Mutations in egl-8/PLCβ and inx-16/Innexin

To gain insight into the role DEC-7 has in the apical junction of intestinal epithelia to regulate DMP frequency, we investigated the impact that loss of DEC-7 has on two apically localized gene products that influence DMP Ca²⁺ wave propagation. Mutations in these two gene products, egl-8/PLCβ and inx-16, which encodes an innexin gap junction protein, have been shown to initiate ectopic intestinal epithelial Ca²⁺ waves that disrupts DMP rhythm (2, 4, 28). Although examination of DMP frequency in egl-8/PLCβ and inx-16 mutants showed a mean DMP frequency like wild-type animals, we found, as previously observed, that both egl-8/PLCβ and inx-16 mutations cause a significant defect in the DMP rhythm as indicated by the high coefficients of variance (Fig. 7, A–D). Analysis of DMP frequency in egl-8/PLCβ and inx-16 mutants with dec-7 depleted revealed that both egl-8/PLCβ and inx-16 suppress the Dec-s phenotype observed in dec-7-depleted animals. These data indicate that the dec-7 loss of function Dec-s phenotype is dependent on egl-8/PLCβ and inx-16. Since both INX-16 and DEC-7 are transmembrane proteins localized to the apical junction of intestinal epithelia, we examined the localization of INX-16 using a functional INX-16::GFP fusion protein (28). Unlike the normal DLG-1::RFP localization at the apical junction (Fig. 6, A and B), analysis of dec-7 loss-of-function mutant or dec-7 RNAi-depleted animals showed an expansion of INX-16 distribution at the apical junction domain of the intestinal epithelia (Fig. 7, E, F, and G), indicating that INX-16 clustering at the apical junction requires DEC-7 function. Lastly, since both egl-8/PLCβ and inx-16 are localized to the apical junction of the intestinal epithelia and mutations in egl-8/PLCβ and inx-16 show ectopic intestinal epithelial Ca²⁺ waves and suppress the Dec-s phenotype associated with loss of dec-7, we investigated whether egl-8/PLCβ and inx-16 act in the same genetic pathway to mediate DMP frequency. To perform this, we examined the DMP frequency in inx-16 mutants with egl-8/PLCβ RNAi depleted. Although inx-16 mutants and egl-8/PLCβ-depleted animals showed a mean DMP frequency similar to wild-type animals, inx-16 mutants depleted for egl-8/PLCβ showed an extended and more arrhythmic DMP cycle than compared with the single mutant or egl-8/PLCβ RNAi-depleted animals (Fig. 7, G and H). These data indicate that egl-8/PLCβ and inx-16 act in separate genetic pathways. However, both loss of egl-8/PLCβ and inx-16 suppress the Dec-s phenotype associated with loss of dec-7. These data along with the role of GON-2/TRPM, PLC-3/PLCγ, and VAV-1 suggest a complex genetic network involved in mediating the DMP periodicity. The ability of loss of egl-8/PLCβ and inx-16 to suppress the Dec-s phenotype of dec-7(RNAi) animals suggests dec-7 acts upstream or parallel to both egl-8/PLCβ and inx-16 to mediate DMP frequency. However, the increased arrhythmic DMP period found in inx-16 with egl-8/PLCβ depleted indicates that these gene products act in separate genetic pathways to mediate DMP timing. Lastly, we found that the localization of DEC-7 was not impacted by loss of egl-8/PLCβ or inx-16 (Fig. 7J).

Figure 7.

dec-7(−) Dec-s phenotype is suppressed by the loss of the INX-16 innexin and EGL-8/PLCβ. A: the Dec-s phenotype associated with dec-7 RNAi depletion is suppressed by the loss of egl-8/PLCβ. Number of animals examined is indicated in bar. B: dec-7 RNAi depletion does not impact the arrhythmic DMP (increased coefficient of variance) associated with loss of egl-8/PLCβ. Number of animals examined is the same as indicated in A. C: the Dec-s phenotype associated with dec-7 RNAi depletion is suppressed by the loss of inx-16. Number of animals examined is indicated in bar. D: dec-7 RNAi depletion does not impact the arrhythmic DMP (increased coefficient of variance) associated with loss of inx-16. Number of animals examined is the same as indicated in C. A–D: Data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Kruskal–Wallis multiple comparison test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, *P < 0.05, **P < 0.01, ****P < 0.0001. E: representative images of INX-16::GFP localization to the apical junction in wild type, dec-7(sa296), dec-7(gk183067gk183068), dec-7(gk776634), and wild-type control and dec-7 RNAi-depleted animals. Arrowheads indicate width of apical junction INX-16::GFP localization and arrows indicate disorganized localization of INX-16::GFP. Scale bar equals 5 µm. F and G: quantification of INX-16::GFP clustering at the apical junction. Data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Kruskal–Wallis multiple comparison test (F) and Mann–Whitney test (G). ****P < 0.0001. n = 42 for wild type, n = 39 for dec-7(sa296), n = 26 for dec-7(gk183067gk183068), n = 27 for dec-7(gk776634), and n = 36 WT control and n = 54 dec-7(RNAi)-depleted animals. H: quantification of DMP frequency. Mean DMP frequency is extended in inx-16 mutant animals with egl-8/PLCβ depleted. Number of animals examined is indicated in bar. I: quantification of DMP rhythm. Depletion of egl-8/PLCβ in inx-16 mutants enhances DMP arrhythmicity. Number of animals examined is the same as indicated in H. H and I: data are presented as means ± SE. Statistical significance was determined by one-way ANOVA with Kruskal–Wallis multiple comparison test. Comparisons are made to wild type unless otherwise indicated. ^nsP > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. J: representative images showing normal localization of DEC-7::GFP in egl-8/PLCβ and inx-16 mutants. Scale bar equals 5 µm. For all data shown in A–D, H, and I, n = number of animals examined. For all data shown in F and G, n = number of measurements taken (3 per animals described in methods). DMP, defecation motor program.

DISCUSSION

In this study, we have found that dec-7 encodes a phylogenetically conserved complement control protein that is orthologous to human SUSD2 and Drosophila Mesh. Like its orthologs, DEC-7 is a type I transmembrane protein that has several conserved extracellular domains that are frequently found in cell-cell adhesion molecules. Consistently, DEC-7 colocalizes with DLG-1 at the apical junction of intestinal epithelia. However, unlike Drosophila Mesh (22–24), DEC-7 does not have an essential role in maintaining intestinal integrity. Instead, we find that DEC-7 acts as a negative regulator of DMP rhythmic frequency. Indeed, using both RNAi and mutant analysis, we found loss of DEC-7 results in more frequent DMP rhythmic activity. Importantly, this Dec-s mutant phenotype requires the InsP3 receptor, an innexin gap junction protein (INX-16), and EGL-8/PLCβ, suggesting that DEC-7 may act upstream to modulate their activity or in a parallel pathway to mediate DMP rhythm.

The C. elegans DMP is an ultradian rhythm that is regulated by InsP3 receptor-dependent Ca²⁺ signaling, which acts as the master oscillator that mediates DMP cycle length (1). Although the InsP3 receptor is vital for the DMP, several other gene products influence DMP timing. These include two PLCs (PLC-3/PLCγ and EGL-8/PLCβ), VAV-1, and GON-2-containing TRPM channels (2, 4–7, 14). Although evidence indicates that PLC-3/PLCγ acts to regulate GON-2/TRPM activity, which in turn regulates InsP3 receptor activity (5), the role EGL-8/β has in mediating DMP activity is not clear. Here, we provide evidence that the phylogenetically conserved Rho guanine nucleotide exchange factor, VAV-1, acts in the same genetic pathway as PLC-3/PLCγ and GON-2/TRPM to regulate DMP rhythm. The role VAV-1 has in this pathway is not clear. However, it has been shown that dysregulation of RHO-1 disrupts DMP activity (29), suggesting that the guanine nucleotide exchange activity of VAV-1 may be important for its role in regulating DMP frequency. Moreover, in mammalian cells, VAV proteins have been shown to be important for promoting PLCγ activity, in part, by acting as a scaffolding protein (30–32). In agreement with VAV-1 acting in a common genetic pathway with plc-3/PLCγ and gon-2/TRPM, we found that loss of dec-7 suppresses the DMP defects associated with gon-2, plc-3, and vav-1 null mutants. These data indicate that dec-7 acts either downstream of these gene products or in a parallel pathway to mediate DMP frequency.

Using genetic analysis to investigate the role of DEC-7 in DMP regulation, we found that the dec-7(-) Dec-s phenotype, in addition to being dependent on InsP3 receptor function, is dependent on the innexin gap junction protein, INX-16, and EGL-8/PLCβ. These data suggest that DEC-7 may act upstream to regulate the activity of these gene products (Fig. 8). Despite the phenotypic similarities between inx-16 and egl-8 mutants, e.g., they have a similar arrhythmic DMP defect and loss of INX-16 or EGL-8 suppresses the dec-7(-) Dec-s phenotype, we found evidence that these genes act in parallel genetic pathways to regulate DMP activity. Indeed, when egl-8 is RNAi-depleted in the inx-16 null mutants, the DMP defect observed is exacerbated compared with the single loss of egl-8 or inx-16.

Figure 8.

Schematic model of DEC-7/SUSD2 function in the apical junction of intestinal epithelia of C. elegans. DEC-7 regulates DMP frequency by mediating the action of EGL-8/PLCβ and the activity and localization of INX-16 gap junction complexes. DMP, defecation motor program; SUSD2, Sushi domain-containing 2.

To further explore the interaction between dec-7 and inx-16, we examined the localization of INX-16::GFP to determine if the loss of dec-7 disrupts its localization. From our microscopy analyses, we observed an expansion of INX-16 clustering at the apical junction. While loss of dec-7 did not impact the localization of DLG-1, a critical component and regulator of the apical junction, or overall intestinal morphology or function, the mislocalization of INX-16 indicates that DEC-7 has a functional role in apical junction organization. Since loss-of-function mutations in inx-16 suppress the Dec-s phenotype associated with loss of dec-7, the mislocalization of INX-16 in dec-7 mutants may be the underlying mechanism causing the deregulation of DMP timing. However, to understand the role INX-16 mislocalization has on DMP behavior will require further studies. Interestingly, the human ortholog of DEC-7, SUSD2, has been found to be important for the establishment of synaptic density in hippocampal neuronal cultures, suggesting an important role of SUSD2 in synaptic organization (33). Consistently, the complement control-like (Sushi domain) protein, LEV-9, has been shown to mediate synaptic clustering in GABAergic neurons of C. elegans (34), indicating the role this protein family has in subcellular organization.

In sum, we have identified a novel role of a phylogenetically conserved complement control protein, DEC-7, in mediating an InsP3 receptor-dependent Ca²⁺ signaling ultradian rhythm in C. elegans. DEC-7 is highly expressed in the intestinal epithelia and localizes to the apical junction. Through genetic analysis, we have found evidence that dec-7 acts upstream of the Ins3P receptor oscillator, EGL-8/PLCβ, and an innexin gap junction protein, INX-16, to modulate their activity to regulate DMP frequency.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by funds from National Science Foundation (1337280) and NIH (GM088213, GM145364, and AG064175).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.T.L. and K.R.N. conceived and designed research; J.T.L., J.B., and K.R.N. performed experiments; J.T.L., J.B., and K.R.N. analyzed data; J.T.L. and K.R.N. interpreted results of experiments; J.T.L. and K.R.N. prepared figures; J.T.L. and K.R.N. drafted manuscript; J.T.L. and K.R.N. edited and revised manuscript; J.T.L., J.B., and K.R.N. approved final version of manuscript.