Multiple ShKT domain-containing MUL-1 proteins act as redox-responsive modulators of oxidative stress signaling in C. elegans

Abstract

Organismal survival depends on coordinated responses to oxidative stress and DNA damage. Using Caenorhabditis elegans, we investigate mul-1, a robust transcriptional target of ionizing radiation and reactive oxygen species. Although annotated as a mucin, MUcin-Like 1 protein (MUL-1) is a small Stichodactyla helianthus K⁺ channel toxin (ShKT) domain-containing protein belonging to an invertebrate expanded family of cysteine-rich proteins. mul-1 is selectively induced by oxidative stress, including ionizing radiation, hydrogen peroxide, Pseudomonas aeruginosa infection, or loss of the peroxiredoxin PeRoxireDoXin-2 (PRDX-2), via the p38 Mitogen-Activated Protein Kinase-cAMP-dependent Transcription Factor (MAPK-ATF-7) pathway in intestinal cells. Loss of mul-1 and its paralogs increases reactive oxygen species accumulation, oxidative stress sensitivity, and C. Elegans P53-like protein (CEP-1)/p53-dependent germ cell apoptosis. Combined deletion of mul-1 paralogs causes constitutive apoptosis, reduced fecundity, and compensatory activation of abnormal DAuer Formation (DAF)-16/FoxO and SKiNhead-1 (SKN-1)/Nuclear factor erythroid 2-related factor 2 (Nrf2) stress response pathways. Together with genetic analysis of SYstemic Stress signaling Mediator 1 (SYSM-1), these findings suggest MUL-1-like ShKT proteins buffer oxidative stress.

Graphical abstract

INTRODUCTION

Organismal survival depends on the activation of coordinated stress response pathways. Ionizing radiation (IR) and reactive oxygen species (ROS) are among the most potent inducers of cellular stress, triggering DNA damage and oxidative insults. The nematode Caenorhabditis elegans provides a genetically tractable model to dissect the regulation and functional impact of these pathways at the organismal level. Although IR can directly induce DNA strand breaks, most DNA damage associated with IR exposure arises indirectly through ROS generation. ROS include superoxide (O₂^-), hydrogen peroxide (H₂O₂), and hydroxyl radicals (∙OH), produced when radiation interacts with cellular water and organic molecules (Roots and Okada, 1975, Ward, 1994). In addition to oxidizing bases, generating abasic sites, and causing DNA strand breaks, ROS also inflict cellular damage by oxidizing metabolites, lipids, and proteins (Dalle-Donne et al., 2006, Yohe and Davies, 2014). ROS are also generated endogenously, for instance, through mitochondrial electron leakage and Nicotinamide Adenine Dinucleotide Phosphate (NADPH) oxidase activity (Forman et al., 2010). Among ROS, H₂O₂ is a precursor to highly reactive ∙OH generated via Fenton chemistry (Forman and Zhang, 2021; Koppenol, 1993), but can also act as a signaling molecule (Forman et al., 2010, Miranda-Vizuete and Veal, 2017). For instance, in C. elegans, H₂O₂ regulates FMRF-Like Peptide 1 (FLP-1) neuropeptide release from AIY interneurons during diet-induced stress response in the gut (Jia and Sieburth, 2021), and this response is potentiated by the H₂O₂-dependent release of the FMRF-Like Peptide 2 (FLP-2) peptide from the intestine (Jia et al., 2024). Elevated ROS in AWC neurons causes Neuropeptide-Like Protein 1 (NLP-1) peptide secretion, which induces the mitochondrial unfolded protein response in the gut and reduces its digestive capacity (Liu et al., 2024). Cellular detoxification of H₂O₂ is primarily mediated by antioxidant enzymes such as superoxide dismutases, peroxiredoxins, and glutathione (GSH) peroxidases, which rely on conserved cysteine residues or thiol-containing cofactors for redox cycling (Aranda-Rivera et al., 2022, Juan et al., 2021).

Transcriptomic analyses following IR in C. elegans revealed no induction of canonical DNA repair genes (Greiss et al., 2008). Among the DNA damage response genes, only the proapoptotic Bcl2 Homology 3 (BH3)-domain only genes, egl-1 and ced-13, both C. Elegans P53-like protein (CEP-1)/p53 targets and required for DNA damage-induced germ cell apoptosis, were upregulated (Greiss et al., 2008, Schumacher et al., 2005). In contrast, a broad CEP-1/p53-independent transcriptional activation of oxidative stress-related and innate immunity-associated genes was observed, many of which are nematode-specific (Greiss et al., 2008). Notably, mul-1 emerged as the most robustly IR-induced transcript, and its induction requires the conserved p38 Mitogen-Activated Protein Kinase (MAPK) pathway (Greiss et al., 2008, Kimura et al., 2012). Recently, a MUL-1 high-copy transgene was shown to be expressed in the gut, and mul-1 deletion was associated with reduced sensitivity to Pseudomonas aeruginosa infection, possibly by limiting bacterial association with gut epithelium (Hoffman et al., 2020). However, while annotated as a mucin-like gene, MUL-1 lacks some hallmark features of vertebrate mucins, which are typically thousands of amino acids long, highly enriched in serine and threonine, and heavily glycosylated to form gel-like protective barriers in gut epithelia (Johansson et al., 2013). Instead, MUL-1 is a small 259-amino-acid protein composed mainly of five ∼36- to 42-amino-acid Stichodactyla helianthus K+ channel toxin (ShKT) domains (InterPro Entry IPR003582: ShKT domain). Only a 42-amino-acid unstructured region between the two C-terminal ShKT domains is highly enriched in serine/threonine residues. ShKT domains were initially characterized as potent toxins derived from sea anemones that inhibit mammalian potassium channels (Castañeda et al., 1995, Gerdol et al., 2019 Cotton et al., 1997; Shafee et al., 2019; Tudor et al., 1998). Except for the human metalloprotease Matrix MetalloProteinase 23 (MMP23), which contains a single ShKT module, this motif is otherwise absent from vertebrate proteomes (Rangaraju et al., 2010). Structurally, ShKT domains are defined by six conserved cysteine residues that form three disulfide bonds, stabilizing a compact two-α-helix fold commonly used to engage and modulate potassium channels (Castañeda et al., 1995, Shafee et al., 2019, Tudor et al., 1998). Given this distinctive organization, we posit that MUL-1 may perform roles unrelated to, or in addition to, those of traditional mucins.

The identification of mul-1 as an IR-responsive gene is reminiscent of sysm-1, a small protein also induced by IR and composed of two ShKT domains (Soltanmohammadi et al., 2022). Like mul-1, sysm-1 induction depends on the p38 MAPK pathway (Soltanmohammadi et al., 2022). Functional studies have shown that SYstemic Stress signaling Mediator 1 (SYSM-1) is secreted from the intestine and is required for germ cell apoptosis following IR, acting in parallel to the C. elegans CEP-1/p53 pathway. Notably, the induction of the two proapoptotic BH3 domain-only genes, egl-1 and ced-13 remains intact in sysm-1 mutants, suggesting that SYSM-1 conveys stress signals across tissues, independent of CEP-1 transcriptional activity (Soltanmohammadi et al., 2022).

Here, we employed a mul-1 transcriptional reporter as an inroad to dissect regulatory circuits involved in the oxidative stress response. mul-1 is induced by IR, H₂O₂, Pseudomonas infection, and loss of the peroxiredoxin PeRoxireDoXin-2 (PRDX-2). Peroxiredoxins are abundant cysteine-based peroxide reductases that detoxify H₂O₂ through the oxidation of conserved N-terminal cysteines to sulfenic acid, followed by disulfide bond formation with a receiving cysteine (Rhee, 2016). We found that mul-1 expression is induced in the intestine and depends on p38 signaling and its downstream transcription factor, cAMP-dependent Transcription Factor (ATF-7). mul-1 mutants are hypersensitive to oxidative stress and exhibit increased p53-dependent germ cell apoptosis upon IR. MUL-1 belongs to a family of proteins expanded in invertebrates, and we included the three most closely related paralogs, as well as related sysm-1, in our analysis. MUL-1 family quadruple mutants display a further increase in radiation-induced apoptosis, and excessive apoptosis occurs even in the absence of IR. Furthermore, both mul-1 and the quadruple mutant bypass the apoptosis defect of sysm-1. In compound mul-1 paralog mutants and prdx-2 single mutants, compensatory abnormal DAuer Formation (DAF)-16-dependent SOD-3 and SKiNhead-1 (SKN-1)-dependent GST-4 induction occurs even in the absence of exogenous stress. We argue that MUL-1-like proteins are part of a regulatory circuit that has a key role in the organismal responses to oxidative stress. We hypothesize that MUL-1-like genes may act via their ShKT domains as scavengers or rheostats of oxidative damage.

MATERIALS AND METHODS

Experimental Design

The aim of this study was to determine if and how MUL-1 and its ShKT domain paralogs regulate organismal responses to oxidative stress and DNA damage in C. elegans. We used genetically defined wild-type and mutant strains to compare responses to IR, chemical genotoxins, oxidative stress, osmotic stress, heat shock, starvation, and Pseudomonas aeruginosa infection. Stress-induced signaling and ROS levels were monitored using single-copy fluorescent transcriptional and translational reporters at endogenous loci, CellROX Green staining, and quantitative microscopy. Strains were generated by Clustered Regularly Interspaced Palindromic Repeats-CRISPR-associated protein 9 (CRISPR-Cas9) genome editing and genetic crosses. Alleles were validated by Polymerase Chain Reaction (PCR) and sequencing. All assays were performed with age-synchronized populations. Phenotypic outcomes were quantified using standardized assays. All experiments included at least three independent biological replicates. Sample sizes varied depending on the assay and are indicated in the corresponding figure legends. For most microscopy-based assays, 20 to 30 animals per condition were analyzed, whereas lifespan and progeny assays were performed using assay-appropriate cohort sizes. The number of animals (n) refers to individual worms scored per condition, unless otherwise specified.

Strain Maintenance and Genetics

Caenorhabditis elegans strains were maintained using standard procedures as originally described by Brenner (Brenner, 1974). Animals were cultured on Nematode Growth Medium (NGM)-lite agar plates seeded with Escherichia coli OP50 and maintained at 20 °C under standard laboratory conditions unless otherwise indicated. For specific assays, animals were propagated at 25 °C (eg, P. aeruginosa PA14 survival assays) or at 15 °C for the maintenance of selected strains. Strains are listed under Suppl Table 2.

Age-synchronized populations were generated either by alkaline hypochlorite treatment of gravid adults followed by overnight L1 arrest in M9 buffer, or by filtration-based synchronization methods, as indicated. Transgenic and CRISPR-Cas9-edited strains were generated by standard microinjection protocols (Dokshin et al., 2018, Ghanta and Mello, 2020, Wang et al., 2018) or obtained from the Caenorhabditis Genetics Center or SunyBiotech. Compound mutant strains were generated through standard genetic crosses and verified by PCR genotyping and/or DNA sequencing. All strains used in this study are listed in Suppl. Table 2, and the corresponding reagents can be found in Suppl. Table 3.

Genotoxic Stress Analysis

For IR experiments, age-synchronized animals were exposed to X-rays using an RS2000 X-ray irradiator (Rad Source Technologies) at the indicates doses. Following irradiation, animals were returned to OP50-seeded NGM plates and allowed to recover under standard conditions.

For chemical genotoxic stress assays, freshly prepared aliquots of cisplatin dissolved in saline and methyl methanesulfonate (MMS)diluted in water were used according to established protocols (Volkova et al., 2020). All solutions were protected from light until use. Groups of 20 to 30 age-synchronized animals were transferred into 2 mL of S-basal buffer supplemented with 5 μl of concentrated OP50 bacterial suspension as a food source. For MMS treatment, animals were exposed to 0.8 mM MMS for 16 h at 20 °C, as previously reported. For cisplatin treatment, animals were incubated with 10 μM cisplatin for 16 h at 20 °C. Samples were incubated under gentle agitation throughout the treatment period. After exposure, animals were washed thoroughly to remove residual genotoxins, transferred to OP50-seeded NGM plates, and allowed to recover.

For UV irradiation, age-synchronized animals were placed on unseeded NGM plates without lids and exposed to 200 mJ/cm² UV light using a CL-1000 UV cross-linker (UVP), as previously described (Yue et al., 2024). Animals were transferred immediately to OP50-seeded NGM plates following exposure and imaged 24 h post treatment.

Western Blot Analysis of MUL-1::3xHA

Synchronized C. elegans populations were collected and washed three times with M9 buffer, flash-frozen in liquid nitrogen, and stored at −80 °C. Worm pellets were thawed on ice and mixed 1:1 with a 2x Laemmli buffer containing 5% β-mercaptoethanol, boiled at 95 °C for 5 min, and briefly centrifuged. Proteins were resolved on hand-cast 12% SDS-PAGE gels in Tris-glycine-SDS buffer (Jeong et al., 2018) and transferred to a PVDF membrane using a semidry system (15 V, ∼0.8 mA cm⁻², 15 min). Membranes were blocked in 5% nonfat milk in PBST for 1 h, incubated with mouse anti-HA (Sigma, clone 16B12; 1:1000) overnight at 4 °C, washed, and probed with HorseRadish Peroxidase (HRP)-conjugated anti-mouse secondary antibody (1:5000) for 1 h. Signals were detected using Pierce Enhanced ChemiLuminescence (ECL) Plus substrate and imaged on a Bio-Rad ChemiDoc system.

Microscopy and Image Acquisition

Images were acquired on a Zeiss Axio Imager microscope equipped with an Axiocam 503 mono camera and controlled by ZEN software. Z-stacks were collected at 1-μm intervals. For each experiment, exposure times, illumination intensity, and acquisition settings were kept constant across all genotypes and conditions. Detailed acquisition parameters for each experiment are provided in the corresponding Methods section.

Quantification of Intestinal Nuclei Fluorescence

Fluorescence intensity from the transcriptional reporter and from all mutant strains generated in this reporter background was quantified by performing line-scan measurements across the first pair of anterior intestinal nuclei in mul-1(syb3342) IV animals. A transverse (10 pixels wide) line was manually positioned through the nuclei, spanning 30 μm in L1 to L3 larvae or 40 μm in L4 and adult animals. Line placement was optimized to minimize background contributions from adjacent intestinal cells and out-of-focus planes. For each nucleus, the maximum fluorescence intensity peak along the line profile was extracted and used for quantitative analysis.

For translational reporters and CellROX Green staining, fluorescence intensity was quantified by measuring the corrected total cell fluorescence. Images were acquired as z-stacks, and a single optical section representing a comparable focal plane was selected for each animal. Whole-animal regions of interest were manually delineated for individual animals, and corrected total cell fluorescence values were calculated as the integrated fluorescence intensity after background subtraction and used for quantitative analysis.

Stress Response Experiments

To assess stress responses, age-synchronized C. elegans at the indicated developmental stages were subjected to defined stress conditions, following established protocols.

For starvation stress, L1 larvae were transferred to unseeded NGM plates and incubated at 20 °C for 6 h, as previously described for starvation-induced stress responses (Baugh, 2013). For osmotic stress, synchronized L1 animals were placed on OP50-seeded NGM plates supplemented with 250 mM NaCl and incubated for 24 h at 20 °C, following standard hyperosmotic stress assays (Urso et al., 2020). For heat-shock treatment, synchronized L4 animals were incubated on OP50-seeded NGM plates at 35 °C for 1 h, followed by a 1 h recovery period at 20 °C, as previously described (Golden et al., 2020, Lithgow et al., 1995). For oxidative stress, synchronized L1 animals were exposed to 10 mM H₂O₂ in liquid culture. Briefly, animals suspended in M9 buffer were treated by adding 10 μl of a 5x H₂O₂ stock solution to 40 μl of the animal suspension and incubated for 1 h at 20 °C under gentle agitation (Offenburger and Gartner, 2018). After treatment, animals were washed four times with 1 mL of M9 buffer to remove residual H₂O₂, transferred to OP50-seeded NGM plates and incubated at 20 °C. Animals were imaged 24 h post treatment.

Generation of a Stable Integrated Multicopy mul-1::eGFP Line

A multicopy mul-1::eGFP reporter line was generated by first establishing an extrachromosomal array in the temperature-sensitive lin-15(n765) background using a PCR-amplified mul-1P::linker::eGFP fragment co-injected with the rescue plasmid pL15EK. Transgenic F1 animals showing robust intestinal Green Fluorescent Protein expression and phenotypic rescue of the lin-15 defect were selected. To obtain a stable genomic insertion, animals carrying the array were exposed to 50 Gy of IR, and subsequent generations were screened for lines that maintained uniform GFP expression in the absence of selection, indicative of successful array integration.

Detection of of Reactive Oxygen Species

A 5x CellROX Green solution was prepared by diluting the 2.5 mM stock solution in M9 buffer and protected from light until use, as previously described for ROS detection in C. elegans (Min et al., 2021). For each reaction, synchronized L1 animals were collected in M9, and 160 μl of the suspension was mixed with 40 μl of the 5x CellROX Green solution to obtain a final reaction volume of 200 μl. Samples were incubated for 2 h at 20 °C in the dark under gentle agitation to ensure uniform staining. Following incubation, animals were pelleted by centrifugation at 1000 rpm for 1 min, washed three times with fresh M9 buffer to remove residual dye, mounted on 2% agarose pads, and imaged.

Sensitivity to Stress

Sensitivity to oxidative stress was assessed using H₂O₂ treatment in liquid culture following established protocols in C. elegans (Offenburger and Gartner, 2018). A H₂O₂ stock solution was freshly prepared from 30% (w/v 9.8 M) H₂O₂ and diluted with water to generate a 5x stock solution of 50 mM. For treatment, the 5x stock was added to synchronized L1 animals suspended in M9 buffer to obtain the desired final concentration. Samples were incubated for 1 h at 20 °C under gentle agitation. After treatment, animals were washed three times with 1 mL of M9 buffer to remove residual H₂O₂ and transferred to OP50-seeded NGM plates for recovery. Animals were plated in triplicate, and developmental progression was scored 48 h post treatment.

Sensitivity to IR was assessed by exposing age-synchronized animals to X-ray irradiation, followed by analysis of post IR developmental progression (Ermolaeva et al., 2013). Briefly, synchronized animals were irradiated with a single dose of 100 Gy, transferred to OP50-seeded NGM plates for recovery under standard conditions, and plated in triplicate. Sensitivity to IR was quantified by scoring developmental stage 48 h post IR, using vulval morphology and overall body size as staging criteria.

Lifespan, Reproductive Fitness, and Apoptosis Assays

Age-synchronized L4 animals from different genetic backgrounds were transferred to OP50-seeded NGM plates and maintained under standard conditions. For lifespan analysis, groups of 20 animals were plated on each NGM plate and transferred to fresh plates daily until egg-laying ceased. Viability was assessed daily by gently prodding the head or tail with a platinum wire; animals unresponsive to stimulation were scored as dead. Animals that escaped, ruptured, or died due to internal hatching (“bagging”) were censored from the analysis, following standard lifespan assay criteria (Kenyon et al., 1993).

For reproductive fitness assays, individual animals were placed on 35 mm OP50-seeded NGM plates and transferred to fresh plates daily until egg-laying ceased. After 24 h, unhatched embryos were scored as dead embryos, and after 48 h, live larvae were counted to determine brood size. Total progeny counts were analyzed and compared across genotypes, as previously described (Andux and Ellis, 2008).

For apoptosis assays, animals were collected at 24 h after the L4 stage, immobilized with 1 mM levamisole, and mounted on 2% agarose pads. Germ cell corpses in the gonad arms were visualized and quantified using fluorescence microscopy with the ced-1::gfp(bcIs39) reporter strain, as previously described (Zhou et al., 2001). To assess DNA damage-induced apoptosis, L4 animals were exposed to IR, and apoptotic germ cells were quantified at 24 h post IR treatment (Gartner et al., 2000).

Pseudomonas Survival Assays

Survival assays on P. aeruginosa PA14 were performed as previously described (Kwon et al., 2024) with minor modifications. Briefly, PA14 was grown overnight in LB broth at 37 °C, seeded evenly across the entire surface of NGM plates, and incubated at 37 °C for 24 h followed by an additional 24 h at 25 °C prior to use. Age-synchronized L4 animals were transferred to PA14-seeded plates supplemented with 50 μM 5-fluoro-2’-deoxyuridine to prevent progeny production and maintained at 25 °C. Survival was monitored every 12 h. Animals that ruptured internally (“bagging”), crawled off the agar, or exhibited vulval bursting were censored from the analysis. At least 60 animals per genotype were scored per assay, and three independent biological replicates were performed.

Statistical Analysis

Statistical analyses were performed using GraphPad Prism. For experiments involving two independent variables, data were analyzed by ordinary two-way ANalysis Of VAriance (ANOVA). When significant effects were detected, multiple comparisons were performed using Šidák’s or Tukey’s post hoc tests, as indicated in the figure legends. Comparisons between two independent groups were performed using an unpaired two-tailed Student’s t-test. Survival curves were compared using the log-rank (Mantel-Cox) test and the Gehan-Breslow-Wilcoxon test. All tests were two-sided. Data are presented as mean ± Standard Error of the Mean (SEM) unless stated otherwise. A p value <0.05 was considered statistically significant.

RESULTS

Transcriptional Regulation of mul-1

To assess if and where mul-1 (F49F1.6) is induced upon IR, we developed a transcriptional reporter strain, mul-1(syb3342), in which the coding sequence of mul-1 is replaced with mCherry fused to histone H2B for fluorescent detection and nuclear targeting (Fig. 1A). mul-1 induction was typically measured 6 h after irradiation (Fig. 1B). Under control conditions, the basal expression of mul-1 is predominantly localized to the nuclei of gut cells, with the strongest expression observed in the two anterior-most gut nuclei (Fig. 1C), and expression gradually increasing during larval development, reaching a maximum in L4 larvae and adults (Fig. 1C, E-F, black lines). To examine the induction of mul-1 under DNA-damaging conditions, we exposed mul-1(syb3342) animals at all larval stages to 100 Gy of IR and analyzed transcriptional activation 6 h post treatment (Fig. 1C). IR exposure results in a strong induction of mul-1 across all gut cells at every developmental stage, most notably in the anterior two nuclei (Fig. 1D, E-F, red lines). The induction is dose- and time-dependent, becoming detectable after 2 h and peaking at 6 h (Figs. S1 and S2). For Western blotting, we generated a knock-in strain with a C-terminal 3xHA tag at the endogenous mul-1 locus. In untreated controls, MUL-1::3xHA protein was undetectable by immunoblotting, consistent with very low basal expression. However, 6 h after IR treatment, a specific ∼35-kDa band corresponding to the predicted molecular weight appears (Fig. S3).

Fig. 1.

Ionizing radiation (IR) induces transcriptional activation of mul-1 in the intestine of C. elegans. (A) Schematic representation of the mul-1(syb3342) IV transcriptional reporter, in which the mul-1 coding sequence is replaced by mCherry::H2B under the endogenous mul-1 promoter. (B) Experimental design. Animals were exposed to ionizing radiation (IR), and reporter expression was analyzed 6 h post IR. (C) Under control conditions, mul-1 expression in mul-1(syb3342) IV animals is detected predominantly in intestinal nuclei, with the strongest signal in the two anterior-most nuclei and a gradual increase during larval development. (D) Six hours after exposure to 100 Gy IR, mul-1 expression is robustly induced in intestinal nuclei across all larval stages, initiating in the anterior cells and subsequently expanding throughout the gut. (E) Representative fluorescence intensity profiles measured along a transverse line across the first pair of anterior intestinal nuclei in mul-1(syb3342) IV animals. (F) Quantification of nuclear fluorescence intensity derived from the maximum intensity peaks corresponding to each nucleus, following optimized line placement to minimize background contributions from adjacent intestinal cells in different focal planes. Scale bars, 20 μm. Statistical analysis was performed using two-way ANOVA with Šidák’s multiple comparisons test. Quantification includes animals from at least three independent experiments (n = 20–30 animals per developmental stage and condition).

To determine if the induction of mul-1 is specific to IR-induced DNA damage, we tested other genotoxic agents. Neither cisplatin, a DNA cross-linking agent, nor MMS, an alkylating agent, induced mul-1 expression (Fig. 2A-D, U). In contrast, some induction was observed following UV treatment (Fig. 2E, U). UV exposure not only generates cyclobutane pyrimidine dimers and 6-4 photoproducts but also produces ROS through photochemical reactions (Yoshiyama et al., 2023), suggesting that mul-1 induction might be linked to oxidative stress rather than DNA damage itself. Additionally, no induction was observed after starvation, heat shock, or osmotic stress (Fig. 2F-I, V). However, a strong induction occurred after exposure to H₂O₂, a potent ROS generator that produces ∙OH and superoxide anions (O₂^-) (Kumsta et al., 2011) (Fig. 2J, V). ROS are also produced during normal metabolism, and the two 2-Cys peroxiredoxins, PRDX-2 and PRDX-3, serve as key antioxidants, with PRDX-2 playing a vital role in detoxifying H₂O₂. Loss of PRDX-2 results in increased sensitivity to H₂O₂, a shortened lifespan, and developmental abnormalities (Kumsta et al., 2011, Olahova et al., 2008). To investigate whether impaired H₂O₂ detoxification can trigger mul-1 induction, we used CRISPR/Cas9 to introduce the prdx-2(gk169) mutation into mul-1(syb3342) animals. Under normal conditions, basal mul-1 expression in prdx-2(gk169); mul-1(syb3342) animals was similar to that of controls during early larval stages (Fig. 2K, W). However, as development progressed, mul-1 transcription first appeared in anterior gut cells. It gradually expanded along the intestine, eventually resulting in widespread, strong expression in adult animals (Fig. 2L-O, W). Our results suggest that both exogenous and endogenous ROS induce mul-1 expression.

Fig. 2.

mul-1 is selectively induced by IR and oxidative stress and requires the p38 MAPK-ATF-7 signaling following ionizing radiation (IR). (A-E) Representative DIC/mCherry merged images of mul-1(syb3342) IV animals under control (A) conditions or following exposure to IR (B), cisplatin (C), methyl methanesulfonate (MMS) (D), or UV irradiation (E). Robust induction of mul-1 expression in intestinal nuclei is observed specifically after IR (B), whereas other DNA-damaging agents elicit little or no reporter activation (C-E). Animals were treated with the respective agents at the L4 stage and imaged 6 h after IR (B), 24 h after cisplatin treatment (C), 16 h after MMS treatment (D), and 24 h after UV treatment (E) (see Materials and Methods). (F-I) Representative images of mul-1(syb3342) IV animals subjected to starvation, heat shock, or osmotic stress show no detectable reporter activation in intestinal nuclei. (J) Treatment with H₂O₂ induces robust mul-1 expression in intestinal nuclei. (F-J) All animals were treated at the L1 stage and imaged after 24 h; note that starved and H₂O₂-treated worms are developmentally arrested. (K-L) Basal mul-1 expression in prdx-2(gk169) II; mul-1(syb3342) IV animals is comparable to controls during early larval stages (L1–L2). (M-O) As development progresses, mul-1 activation in prdx-2 mutants initiates in anterior intestinal nuclei and gradually extends toward posterior regions of the gut. (P) IR-induced mul-1 reporter expression in mul-1(syb3342) IV animals. (Q-R) Loss of the core p38 MAPK components sek-1 and pmk-1 abolishes mul-1 induction in response to IR, indicating an essential role for this signaling pathway. (S-T) Genetic analysis of downstream transcription factors shows that atf-7, but not skn-1, is required for IR-dependent mul-1 activation. (U) Quantification of nuclear mCherry fluorescence intensity in intestinal cells showing a selective activation of the mul-1 transcriptional reporter upon IR treatment. (V) Quantification of fluorescence intensity in anterior intestinal nuclei confirms selective mul-1 induction by oxidative stress. (W) Comparative quantification under control conditions reveals elevated basal mul-1 activation in prdx-2 mutants. (X) Quantification of fluorescence intensity in anterior intestinal nuclei across genotypes and conditions. Scale bars, 20 µm. Data are shown as mean ± SEM and include animals from at least three independent experiments (n = 20–30 animals per condition or genotype). Statistical analysis was performed using two-way ANOVA followed by Tukey’s multiple comparisons test.

The p38 MAPK Pathway and Its Downstream Effector ATF-7 Regulate mul-1

Next, we examined the role of key stress response pathways in regulating mul-1. Previous RNAi studies and quantitative PCR showed that both the p38/PMK-1 and insulin/IGF-1 signaling pathways are necessary for mul-1 induction after IR treatment (Kimura et al., 2012). Using our mul-1 transcriptional reporter, we systematically dissected the contribution of the p38/PMK-1 pathway and its downstream effectors, including SKN-1 and ATF-7, as well as the upstream regulator stress-activated protein kinase-1 (SEK-1), in response to IR-induced DNA damage. In C. elegans, the TIR-1-NSY-1-SEK-1-PMK-1 signaling cascade mediates the innate immune response in the gut (Inoue et al., 2005). SEK-1, a mitogen-activated protein kinase kinase, acts upstream of p38/PMK-1, modulating its activity through phosphorylation in response to various stress stimuli, including infection, oxidative stress, or environmental insults (Kim et al., 2002). We found that IR-induced mul-1 upregulation is compromised in sek-1(km4) and pmk-1(km25) mutants (Fig. 2P-R, X). We also found that ATF-7, but not the SKN-1 downstream effector, is required for mul-1 induction (Fig. 2S-T, X). ATF-7 and SKN-1 have distinct roles downstream of p38/PMK-1. SKN-1 mediates responses to oxidative stress by regulating classical phase II detoxification genes, whereas ATF-7 governs immune responses, such as resistance to P. aeruginosa (Foster et al., 2020, Zhu et al., 2022). Although mul-1 induction was strongly reduced in atf-7 mutants, a residual response remained detectable, suggesting that additional factors may contribute to mul-1 activation in parallel with ATF-7 (Fig. 2S, T, X). Mutations in the daf-2 insulin receptor and the daf-16 transcription factor showed no effect on mul-1 induction (Fig. S4).

To visualize MUL-1 protein, we created a translational reporter, mul-1::linker::eGFP(gt3545). Detecting MUL-1 was challenging due to autofluorescence from gut granules, which interfered with signal clarity. To overcome this, we introduced the glo-1(zu391) mutation, which disrupts gut granule formation, thus reducing autofluorescence without affecting intestinal function (Hermann et al., 2005) (Fig. S5A). Under control conditions, MUL-1 expression was not detectable in the mul-1(gt3545); glo-1(zu391) reporter strain (Fig. S5B, F, G-H). After IR and H₂O₂ treatment, MUL-1 expression was induced, resulting in a low but visible diffuse cytoplasmic signal in intestinal cells, along with distinct cytoplasmic puncta (Fig. S5C, D, F). MUL-1 expression was increased in prdx-2(gk169); mul-1(gt3545); glo-1(zu391) animals, indicating that endogenous oxidative stress promotes MUL-1 induction (Fig. S5I-J) in line with the transcriptional induction (Fig. 2K-O, W). An intense cytoplasmic signal in the gut can be observed in a high-copy MUL-1::eGFP transgene (Fig. S5E) (Hoffman et al., 2020).

MUL-1 Mitigates Oxidative Stress and Modulates DNA Damage-Induced Germ Cell Apoptosis

We could not identify an overt phenotype associated with the mul-1 reporter line lacking the open reading frame under basal conditions, based on progeny numbers, embryonic viability, or lifespan (Fig. 3A-C). Additionally, after exposure to IR, we did not observe any deviation from wild-type in developmental progression from the L1 stage or in progeny survival at the L4 stage (see below, Fig. 4H). To further explore the role of mul-1 in oxidative stress management, we used the CellROX Green assay. This fluorescent probe detects multiple ROS, including H₂O₂, superoxide, and ∙OH (Palacin-Martinez et al., 2024). Under control conditions, both wild-type and mul-1(syb3342) mutants showed minimal green fluorescence, indicating low basal ROS levels (Fig. 3D, E, G). Following IR exposure, wild-type animals showed a moderate increase in ROS levels (Fig. 3D, F), whereas mul-1(syb3342) mutants exhibited a stronger signal, especially in intestinal cells (Fig. 3D, H). We hypothesize that the increased signal might be due to reduced buffering of oxidative stress in mul-1 animals (see below). We next directly tested sensitivity to H₂O₂ exposure and first identified the most suitable concentration, finding that L1 worms treated with 1 mM H₂O₂ developed normally, while those treated with 5 mM H₂O₂ were uniformly arrested at the L1 stage; treatment with 2.5 mM resulted in an intermediate response (Fig. S6A). When assessing sensitivity to 2.5 mM H₂O₂, we observed that developmental progression was moderately delayed in mul-1(syb3342), comparable to daf-16 and pmk-1 mutants, which served as positive controls (Fig. 3I).

Fig. 3.

MUL-1 buffers oxidative stress, promotes developmental progression under oxidative stress, and restrains ionizing radiation (IR)-induced germline apoptosis. (A) Lifespan analysis under control conditions reveals no significant difference between wild-type and mul-1 mutants (n = 10 animals per genotype). (B) Brood size analysis of wild-type and mul-1(syb3342) IV animals under control conditions shows no significant difference in total progeny (n = 10–20 animals per genotype). (C) Embryonic viability, assessed by the fraction of hatched embryos, is comparable between wild-type and mul-1 mutants (n = 10–20 animals per genotype). (D) Quantification of CellROX Green fluorescence intensity in wild-type and mul-1(syb3342) IV animals under control conditions and following IR reveals exaggerated reactive oxygen species (ROS) accumulation in mul-1 mutants after IR (n = 20–30 animals per genotype and condition). (E-F) Representative CellROX Green images of wild-type animals show low basal ROS levels under control conditions and a moderate increase following IR. (G-H) In contrast, mul-1(syb3342) IV animals display low basal CellROX Green signal under control conditions, but accumulate excessive ROS after IR, with strong fluorescence particularly evident in intestinal cells. (I) Developmental stages distribution 48 h after exposure to 2.5 mM H₂O₂ at the L1 stage (n = 60 animals per genotype). Developmental stages were scored based on vulval morphology and overall body size. mul-1 (syb3342) IV mutants display a moderate delay in developmental progression compared with wild-type animals, comparable to that observed in daf-16 and pmk-1 mutants, which were included as positive controls for oxidative stress sensitivity. (J) Quantification of germ cell corpses 24 h after exposure to IR reveals a hyperinduction of apoptosis in mul-1 mutants, which is suppressed in cep-1; mul-1 double mutants (n = 20–30 animals per genotype and condition). This phenotype was independently confirmed using the mul-1(STOP-IN casette, 43-nucleotide-long knock-in cassette) null gt3459 allele. Germ cell corpses were scored using the ced-1::gfp reporter. L1 animals were analyzed unless otherwise indicated. Scale bars, 20 µm. Data are shown as mean ± SEM from at least three independent experiments. Lifespan (A) was analyzed using log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests. Brood size and embryonic viability (B-C) were analyzed using unpaired two-tailed Student’s t-tests. CellROX Green fluorescence (D) and germ cell apoptosis (J) were analyzed by two-way ANOVA followed by Tukey’s or Šidák’s multiple comparisons tests.

Fig. 4.

A ShKT-containing MUL-1 paralog cluster acts redundantly to maintain germline homeostasis and enable developmental recovery after ionizing radiation (IR). (A) Schematic representation of the domain architecture of MUL-1 and its closest paralogs, all encoding ShKT domain-containing proteins. (B) Pairwise sequence identity matrix comparing MUL-1 and related paralogs. (C) Schematic representation of the F49F1 genomic locus on chromosome IV showing the organization of mul-1 and its paralogs. The syb8776 allele corresponds to a deletion of an approximately 7.9 kb region encompassing all four genes. (D) Lifespan analysis of F49F1 quadruple mutant shows no significant difference compared with wild-type under control conditions (n = 10–20 animals per genotype). (E) Progeny production is significantly reduced in the F49F1 quadruple mutant (n = 10–20 animals per genotype). (F) Embryonic viability, assessed as the percentage of hatched embryos, is comparable between wild-type and F49F1 quadruple mutants (n = 10–20 animals per genotype). (G) Germ cell apoptosis is elevated in the F49F1 quadruple mutants under both control and IR conditions (n = 20–30 animals per condition). (H) Developmental stage distribution of animals exposed to IR at the L1 stage and scored after 48 h of recovery (n = 60 animals per genotype). Data are shown as mean ± SEM. Lifespan analyses (D) were performed using log-rank (Mantel-Cox) and Gehan-Breslow-Wilcoxon tests. Progeny production (E) was analyzed using an unpaired two-tailed Student’s t-test. Embryonic survival (F) was calculated for each individual mother as the percentage of embryos that hatched relative to the total number of embryos laid and analyzed using a two-tailed Mann-Whitney test (p = 0.0294). Germ cell apoptosis (G) was analyzed by two-way ANOVA followed by Tukey’s multiple comparisons test.

Next, we analyzed germ cell apoptosis using a widely used reporter where the Cell Death abnormality protein 1 (CED-1) apoptotic corpse receptor is tagged with GFP (Zhou et al., 2001). Under basal conditions, mul-1(syb3342) animals exhibited normal levels of apoptosis (Fig. 3J, lanes 1 and 5). In contrast, following IR, apoptosis was hyperinduced in mul-1(syb3342), a finding confirmed when using a second mul-1 allele (Fig. 3J, lanes 2, 6, 8). Excessive IR-dependent apoptosis was suppressed in cep-1(lg12501), which is defective for the nematode p53-like transcription factor (Fig. 3J, lanes 4 and 10). Complementation analyses using mul-1::eGFP and mul-1::3xHA alleles under the same conditions confirmed that both tagged MUL-1 proteins are functional, as they suppress the extra apoptosis phenotype of mul-1 to wild-type apoptosis levels (Fig. S6B).

MUL-1 Belongs to a Conserved ShKT-Containing Gene Cluster That Modulates Oxidative Stress and Germline Homeostasis

We aimed to investigate whether mul-1 might act redundantly. Performing Basic Local Alignment Search Tool (BLAST) searches for MUL-1 paralogs and scanning through the WormBase, we indeed found multiple MUL-1 paralogs. mul-1 is part of a four-gene cluster on chromosome IV that includes three additional closely related genes (drd-50, F49F1.5, and F49F1.7) (Fig. 4A, B, Suppl. Table 1). MUL-1 paralogs contain multiple ShKT domains with six conserved cysteines, forming three disulfide bonds that stabilize a compact double α-helix structure (Fig. 4A). This clustering suggests potential co-regulation or shared functions. Consistent with this, transcriptional analysis revealed a strong induction of mul-1 upon IR, whereas its paralogs display only modest changes, indicating differential regulation within the cluster (Fig. S7A). We refer to the syb8776 mutation when taking out all four paralogs as the quadruple mutant (Fig. 4C). To test for redundancy, we started by analyzing the quadruple mutant and focused on developmental progression and lifespan, a key measure of organismal resilience affected by genes involved in stress response, genomic stability, and cellular homeostasis (Lopez-Otin et al., 2023). The quadruple mutant showed no change in lifespan compared with wild-type (Fig. 4D). However, the quadruple mutant had a significant 70% reduction in progeny (Fig. 4E). In contrast, embryonic survival was complete in the quadruple mutant and nearly complete in wild-type, with 100% and 99.69% hatching, respectively (Fig. 4F), indicating that while progeny production of the quadruple mutant is reduced, embryonic viability of the remaining embryos is not affected.

Importantly, expression of the multicopy mul-1(gtIs3000) transgene restored brood size in the quadruple mutant without affecting embryonic viability (Fig. S6B-C). We then tested if reduced progeny in the quadruple mutant was correlated with excessive germ cell apoptosis and found that it was the case, both with and without IR treatment (Fig. 4G). Excessive apoptosis was CEP-1/p53-dependent under both conditions. Given the redundancy of MUL-1-like proteins, we tested whether larval development is delayed upon treating L1-stage animals with IR, and found that this is true for the quadruple mutant (Fig. 4H). In contrast, single deletions of mul-1 or pmk-1 do not impair developmental recovery after IR, whereas the combined loss of mul-1 paralogs slows development.

MUL-1 Paralogs Protect From Pseudomonas aeruginosa Infection

We hypothesized that MUL-1 and its closely related paralogs may protect against bacterial infection, commonly associated with oxidative stress. Therefore, we tested susceptibility to P. aeruginosa PA14 infection, a model commonly used in vertebrates and C. elegans (Tan et al., 1999) and known to induce oxidative stress in the nematode (Zhang et al., 2025). We observed that mul-1(syb3342) animals died slightly earlier than wild-type, with the quadruple mutant being the most sensitive, comparable to the pmk-1-positive control (Fig. 5A). mul-1 expression was robustly induced upon exposure to PA14, as indicated by increased reporter fluorescence intensity relative to OP50-fed controls (Fig. 5B-D).

Fig. 5.

Redundant MUL-1 paralogs contribute to resistance against P. aeruginosa infection and restrain basal reactive oxygen species (ROS) accumulation. (A) Survival analysis of wild-type animals, mul-1(syb3342) IV, pmk-1, and F49F1 mutants following exposure to P. aeruginosa PA14. While mul-1 single mutants do not display increased sensitivity, the quadruple mutant exhibits reduced survival comparable to the pmk-1-positive control (n = 60 animals per genotype). (B-C) Representative images showing reporter expression in adult animals exposed to E. coli OP50 (B) or P. aeruginosa PA14 (C). (D) Quantification of fluorescence intensity reveals robust induction of mul-1 reporter expression upon PA14 exposure compared with OP50-fed controls. (E-F) Representative images of wild-type (E) and F49F1(syb8776) IV animals (F) stained with CellROX Green under control conditions. (G) Quantification of fluorescence intensity reveals increased basal ROS levels in the quadruple mutant compared with wild-type animals (n = 20 animals per genotype). Data are shown as mean ± SEM from three independent experiments. Fluorescence intensity (D and G) was analyzed using unpaired two-tailed Student’s t-tests with Welch’s correction.

Loss of MUL-1 Paralogs Activates Oxidative Stress Response Pathways via daf-16/FoxO and skn-1/Nrf2

If MUL-1-like proteins act by scavenging or sensing oxidative stress, their absence might lead to increased endogenous ROS and possibly activate compensatory stress response pathways. We thus tested if ROS is induced in the quadruple mutants even in the absence of IR, and found that this is the case using the CellROX Green assay (Fig. 5E-G).

To examine if this leads to the activation of compensatory pathways, we investigated oxidative stress response pathways regulated by daf-16/FoxO and skn-1/Nuclear factor erythroid 2-related factor 2 (Nrf2) (Doonan et al., 2008, Inoue et al., 2005, Leiers et al., 2003). We generated translational reporters for sod-3 and gst-4 to assess pathway activation by fusing eGFP to sod-3 (sod-3(gt3598)) and mCherry to gst-4 (gst-4(gt3596)). We combined both reporters with the glo-1(zu391) mutation to reduce autofluorescence from intestinal gut granules (Hermann et al., 2005). sod-3 encodes a mitochondrial manganese superoxide dismutase, which neutralizes superoxide radicals and is often linked to increased stress resistance and longevity (Doonan et al., 2008). GST-4 is a phase II detoxification enzyme, GSH S-transferase, that exhibits peroxidase activity and catalyzes the conjugation of GSH to electrophilic compounds, promoting detoxification and excretion (Hurst et al., 1998, Inoue et al., 2005, Leiers et al., 2003). Analysis of GST-4::mCherry and SOD-3::eGFP expression in wild-type L1larvae confirmed that GST-4 is primarily expressed in the intestine, with additional localization in head hypodermal cells (Fig. 6A). In contrast, SOD-3::eGFP fluorescence was localized to the pharynx, especially in the anterior bulb, with a faint but detectable signal around the terminal bulb of the pharynx. No sod-3 expression was observed in the intestine under normal conditions, nor in the hypodermis, body wall muscles, neurons, or tail (Fig. 6A). Deletion of mul-1 alone did not significantly change GST-4 expression but increased SOD-3 expression in the pharynx (Fig. 6B). Remarkably, the F49F1(syb8776) quadruple deletion caused a strong induction of GST-4 throughout the body, especially in the anterior gut, with widespread upregulation of SOD-3 (Fig. 6C). Consistent with these findings, prdx-2 mutants, which are known to accumulate endogenous ROS and which we show to induce mul-1 (Fig. 2K-O and S5 I, J), also showed strong GST-4 induction and pharyngeal SOD-3 expression (Fig. 6D). As expected, GST-4 induction in prdx-2 mutants was skn-1-dependent (Fig. 6E), while SOD-3 induction required daf-16 (Fig. 6F). Overall, these results suggest that deleting mul-1 and its paralogs increases oxidative stress and the expression of key genes involved in oxidative stress response.

Fig. 6.

Compensatory induction of GST-4::mCherry and SOD-3::eGFP. (A) Expression pattern of the oxidative stress reporters gst-4::mCherry and sod-3::eGFP in wild-type L1 larvae carrying the glo-1(zu391) X mutation. gst-4::mCherry is predominantly expressed in the intestine, whereas sod-3::eGFP localizes mainly to the pharynx. (B) Deletion of mul-1 alone does not alter gst-4 or sod-3 expression. (C) The quadruple mutant shows strong induction of gst-4 throughout the body, particularly in the anterior intestine, together with widespread upregulation of sod-3. (D) prdx-2(gk169) II mutants exhibit robust induction of gst-4 and increased sod-3 expression, consistent with elevated endogenous oxidative stress. (E) Induction of gst-4 in prdx-2 mutants requires skn-1. (F) Induction of sod-3 in prdx-2 mutants is abolished in the absence of daf-16. (G-H) Quantification of gst-4::mCherry (G) and sod-3::eGFP (H) fluorescence intensity across the indicated phenotypes. Representative images are shown. Quantification of gst-4::mCherry and sod-3::eGFP fluorescence was performed on 20 to 30 animals per genotype per experiment. Data are shown as mean ± SEM from at least three independent experiments. Statistical analysis was performed using one-way ANOVA followed by Tukey’s multiple comparisons test.

Genetic Interaction With Further MUL-1 Paralogs

Our results are consistent with MUL-1 proteins acting redundantly to protect animals from oxidative stress. Given that the nematode genome encodes multiple additional MUL-1 paralogs, we wanted to test this notion more generally and examined SYSM-1, a more distantly related MUL-1 paralog. SYSM-1 is an apoptosis effector, whose deficiency leads to decreased, rather than increased, DNA damage-induced germ cell apoptosis. Additionally, SYSM-1 acts cell-nonautonomously, being secreted from the gut and functioning independently of p53 (Soltanmohammadi et al., 2022). We confirmed that sysm-1 mutants are defective for DNA damage-induced apoptosis (Soltanmohammadi et al., 2022) (Fig. S8A). However, analysis of sysm-1; mul-1 double mutants, as well as quadruple mutants in conjunction with sysm-1, revealed that the excessive apoptosis phenotype observed in mul-1 single mutants as well as in the quadruple mutant, where excessive apoptosis occurs even without IR, is not suppressed by sysm-1. In other words, the apoptosis defect associated with sysm-1 is bypassed by mul-1 and its paralogs (Fig. S8A).

We next generated and analyzed syb8669, a deletion of F46B3.1, which is the most closely related MUL-1 paralog located outside the MUL-1 paralog cluster. Double-mutant analysis of syb8669 reveals complex genetic interactions with the quadruple mutant; the quintuple mutant suppresses some phenotypes while enhancing others. syb8669 suppressed the reduced fecundity of the quadruple mutant (Fig. S8B). Conversely, the reduced H₂O₂ sensitivity of the quadruple mutant was further suppressed to an extent such that quintuple mutants are partially resistant compared with wild-type (Fig. S8C). Likewise, the hypersensitivity toward Pseudomonas PA14 of the quadruple mutant is suppressed in the quintuple mutant (Fig. S8D). In contrast, quintuple mutants are hypersensitive to IR (Fig. S8E). All in all, the experiments including sysm-1 and syb8669 point toward a more complex picture where MUL-1 paralogs can have opposing functions, in line with the hypothesis that MUL-1-like proteins, besides being scavengers, might act as rheostats for ROS.

DISCUSSION

We initiated our study by focusing on MUL-1 and later included the four closest paralogs in our analyses. MUL-1 and its close paralogs are unstructured, except for possessing 3 to 5 ShKT domains. These domains are cysteine-rich motifs initially described in sea anemone toxins and are widely found in invertebrate proteins, although their function in nematodes remains largely unexplored (Rangaraju et al., 2010, Sachkova et al., 2020). We postulate that nematode multi-ShKT domain proteins may act as scavengers or rheostats of oxidative stress, owing to their potential to scavenge ROS via disulfide formation facilitated by the six cysteines in each ShKT domain (Fig. 7, see below).

Fig. 7.

Ionizing radiation (IR) increases reactive oxygen species (ROS), engaging the stress-activated protein kinase-1 (SEK-1)/PMK-1 p38 MAPK pathway and its downstream transcription factor ATF-7 to induce MUL-1 and other ShKT domain proteins. Under physiological conditions, PRDX-2 limits basal ROS levels. Upon stress, ShKT proteins function as redox-responsive modulators that limit the magnitude of antioxidant gene activation. In parallel, elevated ROS activate SKN-1/Nrf2 to promote GST-4 expression and DAF-16/FoxO to drive SOD-3 expression. In the absence of ShKT proteins, derepression of GST-4 and SOD-3 enhances antioxidant capacity, thereby supporting genome stability and normal development while mitigating p53-/CEP-1-dependent germ cell apoptosis.

We certainly do not rule out that MUL-1 and its paralogs are mucin-like proteins (Hoffman et al., 2020). MUL-1 encodes a 42-amino-acid domain highly enriched in serine/threonine residues, which is akin to mammalian mucins that are highly enriched for serine/threonine throughout most of their length. The parasitic nematode Toxocara canis encodes four secreted proteins, each with an N-terminal signal peptide for secretion and an 83- to 97-amino-acid S-/T-enriched mucin domain, N-terminal to two ShKT domains (Loukas et al., 2000). For MUL-1, prominent enrichment of serine and threonine residues, which occurs along the entire length of mammalian mucins comprising several thousand amino acids (Johansson et al., 2013), occurs only in a 42-amino-acid unstructured region between the 4th and 5th ShKT domains. Irrespective, our combined genetic analysis indicates that MUL-1 and related paralogs are involved in a regulatory circuit associated with oxidative stress, as discussed below.

We discovered that the transcriptional induction of MUL-1 by IR, which generates ROS, is restricted to the gut, with the most pronounced effects observed in the anterior cells. Notably, MUL-1 induction is not triggered by DNA-damaging agents such as the methylating agent MMS or the DNA cross-linking agent cisplatin, nor by osmotic stress or starvation, but rather by oxidative stress, as demonstrated by direct exposure to H₂O₂ or increased endogenous H₂O₂ levels in peroxidase-deficient prdx-2 mutants. mul-1 induction is mediated by the p38 MAPK signaling pathway, consistent with previous RNAi-based studies (Kimura et al., 2012), and requires the transcription factor ATF-7, but not SKN-1. While SKN-1 is widely recognized as a master regulator of oxidative stress responses by inducing phase II detoxification genes (Blackwell et al., 2015, Foster et al., 2020), our findings underscore a previously underappreciated role for ATF-7 in orchestrating transcriptional responses to ROS accumulation. At first glance, gut expression seems intriguing; however, it aligns with bacterial-nematode infections, in which bacterial pathogens and C. elegans produce ROS upon pathogen exposure (Chavez and Mohri-Shiomi, 2007, Hoeven et al., 2011, Jansen et al., 2002, Miranda-Vizuete and Veal, 2017). Indeed, Rhizobium infection and the associated oxidative stress led to defective genome integrity during larval gut development, resulting in excessive DNA bridges and karyokinesis defects in gut nuclei, with the phenotype most prominent in the anterior-most gut cells (Kniazeva and Ruvkun, 2019). We did not find decreased susceptibility to P. aeruginosa infection in the mul-1 single mutant, as previously reported (Hoffman et al., 2020), but did find increased sensitivity in the quadruple mutant. This is due to us using 5-fluoro-2’-deoxyuridine to prevent germ cell proliferation, animals otherwise producing embryos that hatch inside their parents (bagging phenotype), leading to lethality (Kwon et al., 2024).

Functionally, in our study, mul-1 mutants exhibited a small increase of ROS after IR, a modest delay in development under oxidative stress conditions, and elevated CEP-1-dependent germ cell apoptosis. mul-1 is part of a cluster of three additional paralogs, each encoding three to five ShKT domains. Deleting these three paralogs, along with mul-1 (quadruple mutant), reduced progeny numbers and exhibited CEP-1/p53-dependent germline apoptosis even without irradiation. Also, genetic interactions with SYSM-1, a small, 99-amino-acid, unstructured protein that carries two ShKT domains and is induced by IR (Soltanmohammadi et al., 2022), yielded surprising results. Like MUL-1, SYSM-1 is a transcriptional target of p38 signaling and the downstream ATF-7 transcription factor (Soltanmohammadi et al., 2022). In contrast to MUL-1, which reduces CEP-1 p53-induced germ cell apoptosis, SYSM-1 is essential for DNA damage-induced apoptosis. We find that radiation-induced apoptosis of sysm-1 mutants is bypassed by mul-1 single and quadruple mutants. SYSM-1 was suggested to be secreted from the gut to facilitate DNA damage-induced apoptosis in the germline (Soltanmohammadi et al., 2022).

We show that in the absence of the four MUL-1 paralogs, endogenous ROS accumulates, and this aligns with our observation that DAF-16-dependent SOD-3::eGFP and SKN-1-dependent GST-4::mCherry are induced in the mul-1 quadruple paralog mutant. Together, these data support a circular model of redox signaling, in which MUL-1 and MUL-1-like ShKT domain proteins may function as scavengers or rheostats of ROS (Fig. 7). Thus, the loss of the MUL-1 cluster may lead to increased oxidative stress and the compensatory activation of stress pathways, conferring increased survival under oxidizing conditions, but is insufficient to protect against apoptosis induction, sensitivity to IR, and Pseudomonas infection. A localized balance between the expression of various MUL-1 paralogs and the differential activation of compensatory pathways might determine the activity of different stress response pathways. At present, we do not know how signals associated with mul-1 single and compound mutants are transmitted across worm tissues, especially to the germline, where CEP-1-dependent apoptosis is induced. Signaling, and this hypothesis remains to be tested, might be conferred via direct translocation of closely related MUL-1 paralogs, as shown for SYSM-1 (Soltanmohammadi et al., 2022). Alternatively, intercellular signaling could be directly mediated by H₂O₂ diffusion across plasma membranes mediated by aquaporins (Sies and Jones, 2020).

ShKT domains were initially characterized as potent toxins derived from sea anemones that inhibit mammalian potassium channels (Castañeda et al., 1995; Cotton et al., 1997; Shafee et al., 2019; Tudor et al., 1998). We postulate that ShKT domains may be involved in redox reactions. If so, ShKT domains are used in redox regulation, and their cysteines could be oxidized, akin to the 3-amino-acid GSH peptide to its oxidized dimeric (GSSG) form. Each ShKT domain contains six cysteine residues, potentially allowing for extensive disulfide bond formation and redox reactivity. Although such a system may appear inefficient, especially if reductive recycling does not occur, it could serve as a buffering mechanism during acute oxidative stress. In this context, the expansion of proteins primarily composed of ShKT domains in nematodes and other invertebrates might reflect an evolutionary strategy to cope with transient yet potentially lethal oxidative insults. Akin, peroxiredoxin ShKT domains might act as direct H₂O₂ scavengers or enable thiol oxidation by relaying H₂O₂-derived oxidation equivalents to other proteins (Stöcker et al., 2018).

The hypothesis that ShKT domains might be linked to redox reactions is supported by invertebrate redox-active proteins, such as peroxidases and tyrosinases that carry ShKT domains (Rangaraju et al., 2010). For instance, C. elegans MLT-7 and SKPO-1, -2, and -3 peroxidases have acquired an N-terminal ShKT domain and are closely related to the human peroxidasin PXDN, which lacks a ShKT domain (Thein et al., 2009, Tiller and Garsin, 2014). These proteins have been shown to cross-link collagen, regulate endothelial basement membrane structure, and protect against E. faecalis infection (Thein et al., 2009, Tiller and Garsin, 2014). Peroxidase reactions use H₂O₂ to catalyze the oxidation of various substrates, and C. elegans peroxidases modify cuticle collagen structure and permeability (Edens et al., 2001, Myllyharju and Kivirikko, 2004, Thein et al., 2009). Beyond peroxidases, ShKT domains are also present in several C. elegans tyrosinases-like proteins (TYR-1 through TYR-6), which belong to the type-3 copper enzyme family and are annotated to contain tyrosinases copper-binding domains together with an N-terminal ShKT module. Although the specific biochemical activities of TYR proteins in C. elegans remain untested and are inferred primarily from homology, mammalian tyrosinases are well-established type-3 copper oxidoreductases that function through catalytic redox cycling (Pretzler and Rompel, 2024).

Overall, our combined results indicate that MUL-1-like proteins may act as buffers or rheostats for oxidative stress. It remains to be directly tested if and when ShKT domains are oxidized and if this involves disulfide bond formation. Certainly, it is possible, and this remains to be tested, that MUL-1-like proteins have a role in connecting neuronal circuits and gut behavior, where H₂O₂ has a role in signaling (Jia and Sieburth, 2021, Jia et al., 2024, Liu et al., 2024). Irrespective, the expansion of ShKT domains in nematodes and other invertebrates may facilitate rapid evolutionary adaptation to the various challenges posed by oxidative stress. The expansion of MUL-1 paralogs may also have facilitated different MUL-1 paralogs having overlapping and opposing functions.

Limitations of the Study

Overall, our combined results indicate that MUL-1 is part of a regulon induced by oxidative stress via p38 MAP kinase signaling, and that MUL-1 and its paralogs may act as buffers or rheostats for oxidative stress. It remains to be directly tested if and when ShKT domains are oxidized and if this involves disulfide bond formation. MUL-1 paralog SYSM-1 was previously shown to be secreted from the gut and taken up in the germline. Analyzing high-copy MUL-1, we do not see any evidence for germline localization, but acknowledge that this might be due to the limited sensitivity of GFP. We recognize that we have not investigated whether MUL-1 and its paralogs act cell-nonautonomously, which will be an interesting future question. Also, our analysis largely depends on the analysis of the mul-1 single mutant and the quadruple mutant, where all 4 paralogs of the locus are deleted. It will be interesting to investigate how all single-, double-, and triple-mutant combinations behave relating to apoptosis induction, H₂O₂ resistance, the accumulation of ROS, as well as the activation of compensatory GST-4 and SOD-3. Finally, we acknowledge that we do not provide direct evidence that the heightened sensitivity of the quadruple mutant to Pseudomonas infection PA14 is due to excessive oxidative stress.

CRediT Authorship Contribution Statement

Emilio Carranza-García: Writing – review & editing, Writing – original draft, Visualization, Validation, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, and Conceptualization. Abe Gayle Santos: Investigation. Kyoung-Hye Yoon: Investigation, Funding acquisition, and Formal analysis. Anton Gartner: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, and Conceptualization.

Conflict of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

Supplementary material