Functional characterization of Caenorhabditis elegans cbs-2 gene during meiosis

Cystathionine β-synthase (CBS) is a eukaryotic enzyme that maintains the cellular homocysteine homeostasis and catalyzes the conversion of homocysteine to L-cystathionine and Hydrogen sulfide, via the trans-sulfuration pathway. In Caenorhabditis elegans, two cbs genes are present: cbs-1 functions similarly as to human CBS, and cbs-2, whose roles are instead unknown. In the present study we performed a phenotypic characterization of the cbs-2 mutant. The null cbs-2 mutant is viable, fertile and shows the wild-type complement of six bivalents in most oocyte nuclei, which is indicative of a correct formation of crossover recombination. In absence of synaptonemal complex formation (syp-2 mutant), loss of cbs-2 leads to chromosome fragmentation, suggesting that cbs-2 is essential during inter-sister repair. Interestingly, although proficient in the activation of the DNA damage checkpoint after exposure to genotoxic stress, the cbs-2 mutant is defective in DNA damage-induced apoptosis in meiotic germ cells. These results suggest possible functions for CBS-2 in meiosis, distinct from a role in the trans-sulfuration pathway. We propose that the C. elegans CBS-2 protein is required for both inter-sister repair and execution of DNA damage-induced apoptosis.

By the end of pachytene, about half of the oocytes in C. elegans gonad undergo physiological cell death 23 . Additional exogenous or endogenous DNA damage trigger the DNA-damage checkpoint leading to increased germline apoptosis 20,21,[24][25][26] .
The presence of DNA damage produces cell cycle arrest at G2 phase in the premeiotic proliferative region of gonad, inducing the activation of CEP-1 protein 27,28 . CEP-1 in turn induces EGL-1 expression, which disrupts the association between CED-4 (Apaf1) and CED-9 protein (Bcl2). This disruption induces the accumulation of CED-4 at the nuclear membrane 29,30 and its re-localization to the nucleus periphery. The CED-4 accumulation activates the caspase CED-3 [31][32][33] . Apoptosis culminates in the engulfment and degradation of the apoptotic cell 34 .
In this paper, we investigate the role of cbs-2 gene in C. elegans. We first determined the phenotypic effects of cbs-2 depletion, revealing cbs-2 function in meiosis and in DNA damage apoptosis.

Results
C. elegans cbs-2 mutant is hypersensitive to DNA damage. The F54A3.4 (cbs-2) gene is the homolog of human CBS (UniProt entry P355320). The cbs-2 gene contains seven exons (Fig. 1a) and the predicted encoded protein exhibits 55% sequence identity with human CBS 9 . Its catalytic core contains the pyridoxal phosphate binding domain (PLP) followed by CBS core domain 35 . Protein domain prediction obtained by using InterPro 36 , revealed that CeCBS-2 contains two PLP domains and lacks the CBS core at C-terminal (Fig. 1b). Moreover, protein sequences alignment highlighted a greater homology between the C-terminal PLP domain of CeCBS-2 and HsCBS PLP domain ( Fig. 1b and S1).
To determine whether cbs-2 gene is transcribed, we analyzed, by quantitative RT-PCR, the absolute levels of cbs-2 transcript and of pmp-3, used as "housekeeping gene", in wild-type worms. The average of the CT (cycle threshold) obtained from three biological replicates were 27.40 +/−0.27 and 23.96 +/−0.17 , respectively. Since CTs ≤ 29 were strong positive reactions and are indicative of abundant target nucleic acid in the sample, we moved onto analyzing the cbs-2 mutant.
The cbs-2(ok666) allele presents a deletion of 692 bp in the third exon (Fig. 1a). The putative amino acid sequence of ΔCBS-2 obtained by ExPASy translate tool shows the presence of an early stop codon ( Fig. 1b and S2).
Viability assessment of cbs-2 mutants revealed increased embryonic lethality as well as occasional postembryonic developmental defects in a frequency that was statistically different from wild-type, and that might be a consequence of defects in the repair of DNA damage (Table 1, Fig. 1c and Table S1).
We investigated whether CBS-2 could be involved in DNA damage repair by analyzing the sensitivity of cbs-2 depleted worms to cis-platin (CDDP), a DNA damage inducing agent that causes ICLs.
Upon exposure to CDDP we observed increased embryonic lethality in cbs-2 mutant compared to wild-type. After 24-48 h of CDDP treatment, the percentage of dead embryos in cbs-2 mutant was five times higher than what we observed in wild-type (33.6% and 7.85% respectively), and 60% in cbs-2 mutant and 26.6% in wild-type after 48-72 h of CDDP treatment (Fig. 1d).
Taken together, these results support the hypothesis that the observed phenotypes in absence of cbs-2 gene are due to defects in the DNA damage repair. cbs-2 mutant is defective in inter-sister DNA repair during meiosis I. Taking advantage of the C. elegans germline, we investigated the proficiency of cbs-2 mutants in repairing endogenous DNA damage, by analyzing key meiotic steps: repair of recombination intermediates and consequential cross-over formation. Chromosome pairing is coupled with the assembly of the SC and is followed by the formation of multiple physiological DSBs induced by the SPO-11 protein, which trigger meiotic recombination. RAD-51 recombinase binds single-stranded regions adjacent to resected DSBs forming distinct foci, which follow precise kinetics of loading onto and release from the chromatin during meiotic prophase I. Hence, analysis of RAD-51 foci formation and disappearance is widely used as a cytological marker for progression and resolution of recombination intermediates 18,20,21,37 . Mutants affecting homologous recombination repair of meiotic DSBs show altered levels and/or distribution of RAD-51 foci and irregular chromatin aggregates at diakinesis stage.
The distribution of RAD-51 foci in cbs-2 mutant appeared similar to wild-type, however the frequency of RAD-51 foci appeared significantly different compared to wild-type strain. Indeed, the number of RAD-51 foci in cbs-2 significantly decreased in transition zone, middle and late pachytene whilst significantly increased in diplotene (Fig. 2a, Table S2).
At diakinesis stage, each condensed chromosome pair forms a compact body, which can be easily visualized by DNA staining (DAPI); diakinetic nuclei normally contain six DAPI-stained bodies. In cbs-2 mutant most of the nuclei showed six DAPI-stained bodies corresponding to the expected six bivalents, although a low percentage of oocyte (2%) showed poorly condensed chromatin (Fig. 2b,c). HR in C. elegans leads to the formation of one inter-homolog CO event per each chromosome. The rest of the SPO-11 induced DSBs are repaired by non-CO mechanisms, using sister chromatid-mediated recombination, a process dependent on brc-1, ortholog of the mammalian BRCA-1 gene 20,38 .
The absence of the SC, such as in syp-2 mutant, leads to a strong reduction of crossover recombination. Furthermore, in absence of SC, RAD-51 outbreaks disappear in late pachytene, impling that all DSBs are repaired, probably using the sister chromatid as a template 13,14,18 . In addition, in the syp-2 mutant 12 achiasmatic chromosomes (univalent) are present in the diakinesis nuclei 39 , confirming that DSBs are repaired eventually. When inter-sister repair is inhibited, in absence of the SC, the meiotic DSB repair is impaired leading to both Scientific Reports | (2020) 10:20913 | https://doi.org/10.1038/s41598-020-78006-w www.nature.com/scientificreports/ chromosomal fragmentation in diakinesis nuclei and persistence of RAD-51 foci at diplotene stage 13,15,40,41 . To determine whether cbs-2 mutant is competent in repairing meiotic DSBs via inter-sister repair, we quantified DAPI bodies at diakinesis stage in cbs-2; syp-2 double mutants as well as the number of RAD-51 foci, comparted to syp-2 mutant. As shown in Fig. 2b, in cbs-2; syp-2 double mutants 40% of the diakinesis nuclei displayed ≥ 13 DAPI bodies (Fig. 2b,c), a phenotype reminiscent of that observed in the brc-1; syp-2 double mutant (Fig. 2b,c).
This data suggests that CBS-2 may play a role during inter-sister repair. Unexpectedly, RAD-51 staining was dramatically reduced in the diplotene stage in cbs-2; syp-2 compared to syp-2 mutant (Fig. 2a, Table S2). Never the less, diplotene RAD-51 reduction resulted to be similar to that observed in brc-1; syp-2 double mutant diplotene 20,42 ( Figure S3).  23 . The physiological germline apoptosis can be scored in vivo using SYTO-12 and an average of about 3 apoptotic nuclei in wild-type are observed 43 . In addition, accumulation of DNA damage can occur in the germline, either upon failure in faithfully executing meiotic recombination or in response to genotoxic stress, which activates the DNA damage checkpoint, resulting in cell cycle arrest at the G2 phase in order to allow DNA repair. When DNA damage levels are excessive, cells might undergo apoptosis and an increase in the SYTO-12-stained nuclei can be observed. In fact, mutants with defective DNA repair in C. elegans, as brc-1 and fcd-2 mutants, trigger the DNA-damage checkpoint leading to an increase in the apoptotic levels in the germline 20,21,24,25,44 . We investigated the apoptotic levels in the hermaphrodite germline of cbs-2 mutant and unexpectedly, even if cbs-2 mutant is defective in inter-sister repair, the apoptotic levels were not significantly different from wild-type animals (Fig. 3a). We therefore tested whether exogenous DNA damage, induced by CDDP treatment, was able to enhance apoptosis in cbs-2 mutant. Although the apoptotic levels were significantly increased in wild-type after 48 h of CDDP treatment, no enhancement was observed in treated cbs-2 mutant (Fig. 3a). We therefore confirmed our data inducing DSBs with another source of exogenous damage, γ-rays (irradiating worms at 120 Gy), and also in this case we did not observed any increase in the apoptosis levels (Fig. 3a).
To further investigate cbs-2 role in DNA damage apoptosis, we tested whether endogenous DNA damage was proficient in inducing apoptotic cell death. We assayed the apoptotic levels of cbs-2 mutant in a fcd-2 mutated background. fcd-2 mutant is defective in DNA repair and shows higher levels of germline apoptosis due to the activation of the pachytene DNA damage checkpoint 21 . The depletion of cbs-2 reduced the apoptotic level of fcd-2 mutant to wild-type levels (Fig. 3a). Thus, our data support a role for CBS-2 in eliciting the DNA damage apoptosis after endogenous and exogenous DNA damage.
Activation of the DNA damage checkpoint in the mitotic and meiotic region of the germline is intact in the cbs-2 mutant. Since we demonstrated cbs-2 involvement in eliciting DNA damage apoptosis, we next tested whether the DNA damage checkpoint is correctly activated in cbs-2 mutant. The nuclei residing in the tip of the C. elegans germline are engaged in mitotic proliferation and respond to DNA damage triggering a temporary block of cell cycle progression to allow DNA repair. As a consequence of cell cycle arrest, these nuclei appear enlarged. To establish whether cbs-2 mutant is able to activate cell cycle arrest, we measured the diameter of the nuclei in mitotic zone before and after hydroxyurea (HU) treatment, compared to wild-type as positive control and to mrt-2 as negative control 45 . As shown in Fig. 3c, HU treatment induced similar increase in the average diameter of mitotic nuclei in cbs-2 mutant and wild-type animals. Instead, in mrt-2 mutant the average diameter of the mitotic nuclei remained constant before and after HU treatment, as expected upon defective arrest of cell cycle. These data suggest that cbs-2 is not implicated in cell cycle arrest activation (Fig. 3b,c). cbs-2 is not required for the expression and localization of the proapoptotic factors CEP-1 and EGL-1 after DNA damage. We tested whether cbs-2 mutated worms were competent in the expression and localization of the proapoptotic factors CEP-1 and EGL-1 after treatment with DNA damage agent (Fig. 4a).
cbs-2 mutant showed wild-type CEP-1 loading in the mitotic and in the late pachytene nuclei 24 h after CDDP treatment (Fig. 4b). To assess the egl-1 gene expression we analyzed, by quantitative RT-PCR, the relative egl-1 transcript levels before and after DNA damage in cbs-2 mutant and in wild-type. CDDP treatment induced similar increase in the expression levels of egl-1 transcript. After CDDP treatment in both wild-type and cbs-2 mutant, egl-1 transcript levels increased four times (Fig. 4c). Thus, in absence of cbs-2 gene, CEP-1 is competent to induce egl-1 expression after DNA damage.
We can conclude that cbs-2 is not involved in triggering the DNA damage checkpoint in meiotic and mitotic region of the gonad but it may have a role in the execution of the DNA damage apoptosis.

CBS-2 is not required for proper localization of the apoptotic effector CED-4 after DNA damage.
Since we demonstrated that the DNA damage checkpoint was correctly activated in cbs-2 mutant after www.nature.com/scientificreports/ CDDP treatment, we then wondered whether the failure in DNA damage apoptosis execution might depend on altered localization of the apoptotic factors acting downstream EGL-1. We then assessed CED-4 localization, before and after DNA damage, using a transgenic strain expressing CED-4::GFP under the ced-4 promoter 29 .
As shown in Fig. 4d, GFP expression was detected in cbs-2 mutant and wild-type untreated worms, moreover CDDP treatment induced an increase in GFP signal in the mitotic region and late pachytene regions in both strains (Fig. 4d), suggesting that the apoptotic response defects observed in cbs-2 depleted worms is not due to the downregulation of CED-4.

Discussion
Human CBS is known to catalyze the production of hydrogen sulfide and cystathione from Homocysteine. In C. elegans two cbs genes are present, cbs-1 and cbs-2. Previous studies demonstrated that, unlike cbs-1, cbs-2 is not implicated in hydrogen sulfide metabolism 9 . Using a protein domain prediction software we have shown that CeCBS-2 PLP domain have a high degree of conservation with HsCBS PLP domain and further, we found that cbs-2 is regularly transcribed under physiological conditions of growth in wild type animals. In this paper we determine the phenotypes caused by the absence of cbs-2 gene in C. elegans. The hypersensitivity to DNA damage, the increase in embryonic lethality and developmental defects, as well as the occasional presence of condensed chromatin in oocyte nuclei of cbs-2 mutated worms (Figs. 1c,d, 2b,c) suggest an inefficient repair of recombination intermediates in meiosis. We demonstrate for the first time that  www.nature.com/scientificreports/ stained body higher than 13, instead of 12 achiasmatic chromosomes (univalents) (Fig. 2b,c). Furthermore, we have not observed RAD-51 foci persistence at diplotene stage in cbs-2;syp-2 typical of the defective meiotic DSB repair (Fig. 2a, Table S2). This phenotype is similar to that observed for brc-1 mutant (Fig. S3). It had been demonstrated that BRC-1 is essential for both inter-sister repair and recruitment/stabilization of the RAD-51 factor to DNA damage in absence of SC 20,42 . We speculate that CBS-2, as BRC-1, may play a role in recruitment and/or stabilization of RAD-51 recombinase to DNA damage during meiosis I. Our findings are in line with a previous observation regarding the specific germline expression of cbs-2 46 . These results led us to speculate that in C. elegans the two CBS proteins, CBS-1 and CBS-2, act in different tissues. In fact, previous findings showed that CBS enzymatic function seems to be unaltered in cbs-2 mutant 9 suggesting a main role of CBS-1 in CBS enzymatic activity in the whole organism and a different specific function of CBS-2 in the germline.
In C. elegans, mutants with genetic defects in DNA repair display activation of the DNA damage checkpoint, which ultimately results in an increase of the apoptotic level in the germline 20,21 .
The physiological apoptotic levels in the germline of cbs-2 mutant are similar to wild-type strain. Our results show that CBS-2 is required for the execution of germ-line apoptosis in response to DNA damage originated by defective meiotic repair, as well as by DNA damage agents (CDDP and IR) (Fig. 3a).
We demonstrate that the DNA damage checkpoints activation function properly in cbs-2 mutated worms, as in fact, the HU treatment produces cell cycle arrest in the premeiotic proliferative region of gonad (Fig. 3b,c). Germ cells in absence of cbs-2 gene showed a normal localization of CEP-1 as well as a correct increased in the expression levels of egl-1 transcript after treatment with DNA damage agents (Fig. 4b,c). Moreover, germ cells in cbs-2 mutant showed increased CED-4 localization after DNA damage, similar to wild type (Fig. 4d).
These data suggest that the failure in the activation of DNA damage apoptosis in cbs-2 mutant does not depend on defective apoptotic machinery.
Other meiotic genes, as msh-4, msh-5 ,zhp-3 and rad-51 (isoform A) whose actions are essential for DNA damage-induced apoptosis during C. elegans gametogenesis, appear to act downstream of the pro-apoptotic signal 37,43 . Further studies of these processes are needed to identify the exact step in which CBS-2 regulate the DNA damage apoptosis.
Studying the roles and the functions of genes involved in the regulation of the DNA damage apoptosis and in DNA repair is essential to highlight new processes that can be targeted by novel therapeutic approaches in the treatment of diseases caused by incorrect DNA repair and inappropriate apoptosis regulation.

Experimental procedures
Strains. Nematodes have been cultured at 20 °C on NGM plates with Escherichia coli OP50 as food source according to standard methods 47 .
Bioinformatics. Alignment of amino acid sequences of C. elegans CBS-2 (WormBase entry CE47819) and human CBS (UniProt entry P35520) was performed using Cobalt Constraint-based Multiple Alignment online software with default parameters 48 . Predictions of protein domains both C.elegans CBS-2 (WormBase entry CE47819) and human CBS (UniProt entry P35520) were obtained using InterPro online software 36 . The putative amino acid sequence of C. elegans ΔCBS-2 (ok666) was obtained using ExPASy translate tool 49 . Screening of laying worms. Worms were isolated and cloned during the L4 larval stage on Petri plates and left at 20 °C to lay eggs for 3 days. Worms were transferred every 12 h onto fresh plates until the deposition of non-fertilized oocytes. Each plate was monitored every for 24/72 h to analyze the following parameters among the progeny: embryonic lethality, presence of males, aberrant phenotypes and larval arrests 21,37 . The percentage of embryonic lethality is calculated as the ratio of unviable eggs over laid eggs. The percentage of males, aberrant phenotypes and larval arrests is calculated as the ratio of males/aberrant/larval phenotype over the hatched eggs. DNA damage sensitivity. CDDP. Young adult worms were transferred onto fresh seeded plates containing 180 μM cis-diamminedichloroplatinum(II) (CDDP, Sigma). Worms were picked after 24 h for cytological procedures or transferred onto fresh plates every 12 h to score embryonic lethality (unhatched eggs) and organisms with evident postembryonic developmental defects under a dissecting microscope.
HU. Young adult worms were transferred onto freshly seeded plates containing 40 mM Hydroxyurea (HU, Sigma) and picked after 24 h for cytological procedures.
IR. Young adult worms were transferred on freshly seeded plates and exposed to ionizing irradiation from a 137 Cs source for a time corresponding to 120 Gy. Animals were scored 24 h after treatment. www.nature.com/scientificreports/ quality was checked by 2% agarose gel electrophoresis. cDNA was prepared from 2 µg of total RNA by using a high-capacity cDNA Reverse Transcription kit (Applied Biosystem, Foster City, CA). For real-time quantitative PCR, 5 µl of the 1:10 dilution of cDNA was used in each reaction; the Power SYBR Green Master Mix (Applied Biosystems) was used 37 . RT-PCR results were recorded as relative gene expression changes after normalizing for pmp-3 gene expression and computed using the comparative CT method (2-ΔΔCT) as previously described 50 . The 2-ΔΔCT value was > 1 for gene more highly expressed and < 1 for genes lowly expressed. RT-PCR data are the mean ± SD of three biological replicates. Primers used for gene expression analysis listed in Table S3.
Quantitative analysis of germline apoptosis. Adult nematodes were suspended in M9 solution and stained by incubation with 33 µM SYTO-12 (Molecular probes) for 1 h and 30 min at room temperature in the dark. The worms were then transferred to seeded plates to allow stained bacteria to be purged from the gut. After 30 min, the animals were mounted on 2% agarose pads in 2 mM levamisole 21 . The estimation of apoptotic levels for each genotype was calculated as the average number of apoptotic nuclei per gonadal arm 22 .
Analysis of DAPI-stained germlines. Adult nematodes were suspended in M9 solution on glass slides, permeabilized and fixed with 10 µl of 100% Et-OH. Finally, slides were mounted in 10 µl of DAPI (2 ng/µl) diluted in M9. Numbers of scored nuclei are indicated in the legends of charts.
Image collection and processing. Collection of images was performed using a Leica DM6 fluorescence microscope, Hamamatsu camera under the control of Leica LAS AF 6000 software. Images were processed using Leica LAS AF 6000 software and Image J/Photoshop programs. Quantitative analysis of RAD-51 foci and DAPIstained bodies along the germline were performed on z series. Optical sections were collected at 0.18 µm and 0.50 µm increments respectively. The quantitative analysis of RAD-51 foci was performed by dividing the germline into seven 36 × 36 μm zones (mitotic tip, mitotic zone, transition zone, early pachytene, middle pachytene, late pachytene diplotene stage), beginning from the distal tip 18 .
Statistical tools. Statistical analyses for independent samples were computed through Mann-Whitney t-test. The level comparison of aberrant phenotypes and larval arrests of the different genotypes were computed through χ 2 test. The statistical analyses of diameter of mitotic nuclei before and after treatment were computed through Mann One-way ANOVA Kruskal Wallis.