Identification of genetic suppressors for a BSCL2 lipodystrophy pathogenic variant in Caenorhabditis elegans Open Access

Lipid droplets (LDs) are cellular energy reservoirs to maintain fat homeostasis and balance, ensuring proper lipid metabolism and healthy physiology (Fei et al., 2008). Failure to maintain the size and morphology of LDs can lead to many diseases and disorders, including obesity, cancer, lipodystrophy and neurodegenerative disorders. (Onal et al., 2017; Pressly et al., 2022; Teixeira et al., 2021; Lim et al., 2021; Patni and Garg, 2015; Zadoorian et al., 2023) Seipin is one of the primary LD scaffold proteins to govern LD size, and may also regulate lipid biosynthesis and sorting to promote LD synthesis at the endoplasmic reticulum (ER). (Szymanski et al., 2007; Su et al., 2019; Wang et al., 2016). Clinical studies have identified multiple seipin pathogenic variants in patients with Berardinelli–Seip congenital lipodystrophy 2 (BSCL2), Silver syndrome and teratozoospermia syndrome (Magre et al., 2001; Agarwal and Garg, 2004; Windpassinger et al., 2004; Jiang et al., 2014). Therefore, understanding the molecular and cellular mechanisms governing LD formation and the nature of LD scaffold proteins, such as seipin, are critical to developing therapeutic strategies for LD-associated diseases.

Proper lipid metabolism and transfer are critical to oogenesis, fertilization, embryogenesis and other physiological processes, such as lifespan and locomotion, in Caenorhabditis elegans. Two lipid reservoirs, yolk granules and LDs, were identified in C. elegans oocytes and embryos as regulating lipid transfer and metabolism. However, the molecular and cellular mechanisms of how these two lipid structures contribute to healthy physiology remain unexplored. Our recent study established a BSCL2 (seipin) disease model for understanding seipin-associated disorders in vivo (Bai et al., 2020a). Three seip-1 null alleles and an orthologous BSCL2 pathogenic variant allele, seip-1(A185P), caused severe embryonic lethality and abnormally enlarged LDs in C. elegans oocytes and embryos (Bai et al., 2020a). The embryonic lethality in the seipin mutants was correlatively linked to the impaired permeability barrier of the eggshell, a lipid-enriched extracellular matrix layer surrounding the embryo. Lipidomic studies showed that the C20 polyunsaturated fatty acid content was significantly reduced in the seipin mutants. The dietary supplementation of two types of linoleic acids restored embryonic lethality and extracellular eggshell formation in the seipin mutants. Intriguingly, the supplementation of the linoleic acids failed to rescue the abnormal/enlarged LD size and instead led to excessively enlarged LDs in oocytes and embryos (Bai et al., 2020a), suggesting that proper eggshell and LD formation were independently regulated by seipin. However, the exact contributions of seipin during these two cellular events remain obscure.

The nematode model C. elegans has become an emerging system for understanding lipid and lipid metabolism diseases (Shi et al., 2016; Yue et al., 2021). At least ten lipodystrophy-associated genes have been identified, including phosphate acetyltransferase (AGPAT2), phosphoinositide-dependent serine-threonine protein kinase (AKT2), seipin (BSCL2), caveolin 1 (CAV1) and perilipin 1 (PLIN1) (Kuivenhoven and Hegele, 2014). The conserved orthologs of all these lipodystrophy-associated genes are found in the C. elegans genome. Disturbing expression of the C. elegans orthologous genes led to various phenotypes, including embryonic lethality, sterility, lifespan alterations and locomotion defects (Bai et al., 2020a; Hertweck et al., 2004; Scheel et al., 1999; Chughtai et al., 2015). These easily scorable phenotypes allowed geneticists to conduct forward genetic screens to identify previously unidentified genetic modifiers for LD biogenesis and lipid signaling pathways (Wu et al., 2018; Ruiz et al., 2018; Li et al., 2016; Zhu et al., 2022). The development of rapid mapping strategies using molecular inversion probes (MIP-MAP; Mok et al., 2017) combined with high-throughput whole-genome sequencing provide a fast and cost-effective method for pinpointing candidate mutations in a mutagenized genetic model.

To explore more specific roles that seipin plays during embryonic and LD biosynthesis, we focused on identifying previously unknown genetic determinants and pathways. Taking advantage of the embryonic lethality in the seip-1(A185P) mutant as a readout, we conducted an unbiased forward genetic screen to identify genetic modifiers that restore embryonic viability in the seip-1(A185P) mutant. Using MIP-MAP genomic mapping and the whole-genome sequencing technique, we identified two independent missense alleles of a target suppressor gene R05D3.2, which we renamed lmbr-1. We then generated those two putative suppressor alleles of lmbr-1, lmbr-1(S647F) and lmbr-1(P314L), in the wild-type background. The lmbr-1 missense mutants were then crossed back into the seip-1(A185P) mutant to assess their suppression. The homozygous lmbr-1(S647F) and lmbr-1(P314L) alleles significantly suppressed embryonic lethality and eggshell formation defects caused by seip-1(A185P). lmbr-1(S647F) did not alleviate the enlarged LD size caused by seip-1(A185P), whereas the other allele, lmbr-1(P314L), partially suppressed enlarged LD size in the seip-1(A185P) background. In summary, we describe the results of the mutagenesis screen designed to identify regulators that genetically interact with a seipin BSCL2 pathogenic variant. The discovery of new suppressor candidates will shed light on the molecular mechanisms contributing to seipin-associated lipodystrophy and related physiological disorders.

Previously, we established a BSCL2 disease model by generating an orthologous pathogenic variant, seip-1(A185P), in C. elegans (Bai et al., 2020a). This corresponds to the orthologous alanine 212 in the human BSCL2 gene. Using this disease model, we could functionally characterize the seip-1(A185P) allele in vivo. seip-1(A185P) caused penetrant embryonic lethality, defective eggshell formation and abnormally enlarged LDs (Bai et al., 2020a). To determine whether this missense mutation disrupts the temporal and spatial expression pattern of SEIP-1, we generated the seip-1(A185P) allele by CRISPR/Cas9 genome editing into an endogenously tagged red fluorescent reporter strain, seip-1::mScarlet (Fig. 1A), which we designated seip-1(A185P)::mScarlet.

Embryonic localization of SEIP-1(A185P)::mScarlet was similar to the wild-type reporter, with enrichment around the nuclear envelope (Fig. 1B,C). This enrichment corresponds to the ER, as SEIP-1::mScarlet was shown previously to co-localize with the ER marker SP12 (Magre et al., 2001). However, the cytosolic distribution was more diffuse and less punctate in the seip-1(A185P)::mScarlet mutant compared with the wild type, suggesting a defect in ER structure or the association of seipin with the ER. The most apparent difference was the presence of multiple nuclei in the seip-1(A185P)::mScarlet strain (Fig. 1C, green arrows). This defect has not been reported previously and likely contributes to the embryonic lethality observed in the seip-1(A185P) strain. seip-1::mScarlet has previously been shown to express in additional C. elegans tissues, including the pharynx and epidermis, where the LDs are enriched (Bai et al., 2020a; Cao et al., 2019); similar localization patterns were found in the seip-1(A185P)::mScarlet mutant (Fig. S1). Imaging of the adult germline (see following) also revealed that seip-1(A185P)::mScarlet distribution was normal (Fig. 1G,H). Taken together, results from the fluorescent reporter genes suggest that the seip-1(A185P) mutation may compromise protein function instead of altering the trafficking and cellular localization of SEIP-1 in vivo.

We performed a variety of functional assays to test this hypothesis. We used DAPI, a small molecule that can penetrate into embryos with an impaired eggshell permeability barrier, to assess eggshell integrity in the seip-1(A185P)::mScarlet mutant. In the seip-1::mScarlet control strain, no DAPI staining was detected in the cytosol, demonstrating that knock-in of mScarlet at the endogenous loci of seip-1 did not disrupt eggshell integrity (DAPI staining found in 0 out of 95 embryos) (Fig. 1D). In contrast, zygotic chromatin labeled by DAPI was frequently observed in the fertilized embryos of the seip-1(A185P)::mScarlet mutant (67 out of 68 embryos), which is indicative of impaired eggshell formation (Fig. 1E); we also observed multiple chromatin bodies (Fig. 1E, yellow arrows) associated with the multi-nucleate phenotype, which is a previously unreported observation in the seip-1 mutants. Next, we assessed embryonic viability, as previous work indicated that the wild-type seip-1::mScarlet reporter did not affect survival (Bai et al., 2020a); in contrast, we found that seip-1(A185P)::mScarlet exhibited reduced brood size and embryonic lethality comparable with the seip-1(A185P) allele (Fig. 1F,F′). Finally, we examined LD size by staining animals with a lipophilic fluorescent probe BODIPY 493/503 in both seip-1::mScarlet control (Fig. 1G) and the seip-1(A185P)::mScarlet mutant (Fig. 1H). Enlarged LDs were observed in the oocytes on the proximal side of the gonad, which is adjacent to the spermatheca of the seip-1(A185P)::mScarlet animals (Fig. 1H). The defective phenotypes in the seip-1(A185P)::mScarlet mutant are consistent with other seip-1 mutants characterized in the previous study (Bai et al., 2020a). We conclude that the primary impact of the A185P mutation is on protein function rather than localization, although a modest effect on the latter cannot be ruled out.

To identify novel genetic modifiers that regulate seipin function, we conducted an ethyl methane sulfonate (EMS)-based forward genetic screen with the seip-1(A185P) mutant. The seip-1(A185P) allele caused a significantly reduced brood size, which includes both unhatched/dead embryos and hatched larvae, and penetrant embryonic lethality (Fig. 2A-D) (Bai et al., 2020a). During the course of screen optimization, we determined that these defects were enhanced at a higher temperature (25°C) and that the embryonic lethality in seip-1(A185P) worsened when the animals were maintained at 25°C for one to two generations. Therefore, we performed the mutagenesis screen under this optimized temperature to isolate and recover suppressor lines of seip-1(A185P) that restored embryonic viability and produced viable larvae (Fig. 2B). The screen yielded a total of five independent suppressor lines from ∼64,000 haploid C. elegans genomes (∼32,000 mutagenized F1s). To generate the homozygous suppressor lines, embryonic lethality in each line was assessed for at least two generations, testing more than 20 animals per generation. Only the candidate lines displaying consistently elevated viability rates in each tested animal were collected and identified as the homozygous suppressor lines. Sanger sequencing confirmed the seip-1(A185P) allele in each suppressor line. The embryonic viability rate in each suppressor line was significantly increased when compared with seip-1(A185P) allele alone (Fig. 2D). Yet, the brood size in each suppressor line was reduced (Fig. 2C). Overall, our findings suggest that homozygous suppressor variants might alleviate the embryonic lethality caused by the seip-1(A185P) mutant. However, these variants could potentially have additional impacts on reproduction, such as causing a reduction in brood size. It is also plausible that other essential mutations in the suppressor background could contribute to the observed sub-fertility.

The previous study indicated that the embryonic lethality of the seip-1(A185P) mutant was correlatively linked with impaired eggshell (Bai et al., 2020a). To test whether the eggshell formation deficiency was restored in each suppressor line, we imaged the DAPI penetration in the early embryos. In the wild type, only the first polar body, which is located outside the permeability barrier (one layer of the eggshell) that blocks small-molecule penetration, was stained by DAPI (Fig. 3A,C-E). In the seip-1(A185P) mutant, DAPI could penetrate the eggshell and stain the zygotic chromatin, indicating the defective permeability barrier (Fig. 3B,F-H). Using the DAPI imaging protocol, we tested over a hundred early embryos of each suppressor line. We found that DAPI penetration in each tested suppressor line was significantly reduced compared with that in the seip-1(A185P) mutant (Fig. 3I). These observations suggested that the putative modifiers in the suppressor lines restore the proper formation of the permeability barrier. Not surprisingly, the suppressor lines seip-1(A185P); sup1 and seip-1(A185P); sup3, which displayed the least DAPI penetration, correlated with the highest levels of embryonic viability (Figs 2D and 3I), supporting our proposed hypothesis that the embryonic lethality in the seip-1(A185P) mutant background is linked with the proper eggshell formation.

The canonical function of seipin is to maintain LD size and biosynthesis. We found that irregularly enlarged LDs were observed in the oocytes and embryos of the seip-1(A185P) mutant compared with wild type (Bai et al., 2020a) (Fig. 4A,B,H). To test whether the abnormal LD size was alleviated in the suppressor lines, we imaged the oocytes in which LDs were enriched by staining with a lipophilic BODIPY neutral fluorophore that binds to neutral lipids (Fig. 4A-G). The parameter of LD size was defined and quantified by measuring the number of LDs with diameters larger than 1.5 μm in the −1 to −3 oocytes, as described in the previous study (Bai et al., 2020a). In wild type, the average LD size was less than 0.5 μm (Fig. 4A,H), but more than 20 enlarged LDs (>1.5 μm) were readily stained in the −1 to −3 oocytes in the seip-1(A185P) mutant (Fig. 4B,H). The enlarged LD size (>1.5 μm) was only alleviated in two of five suppressor lines: seip-1(A185P); sup4 and seip-1(A185P); sup5 (Fig. 4F-G,H). These two suppressor lines displayed the most rescued LD phenotypes. Furthermore, they had the highest levels of DAPI penetration and embryonic lethality (Figs 2D and 3I), suggesting that genetic modifiers in each suppressor line may function somewhat independently in regulating eggshell formation, embryonic viability and the maintenance of proper LD size in the oocyte. These observations are consistent with our previous observation that dietary supplementation of polyunsaturated fatty acids significantly restored embryonic lethality and defective eggshell formation but enhanced the abnormal enlarged LD size (Magre et al., 2001). This finding was also observed in a study indicating that the changes in phospholipid synthesis could suppress embryonic lethality but not affect LD size in a seip-1 null mutant (Zhu et al., 2022). In summary, our data suggested that maintaining proper LD size primarily contributes to maintaining lipid balance and metabolism homeostasis, but may only play a limited role in orchestrating eggshell formation.

Using a fast and high-throughput genomic mapping strategy that involves molecular inversion probes (MIP-MAP; Mok et al., 2017), we were able to identify the genetic modifiers in the suppressor line seip-1; sup1 and seip-1; sup4 (Fig. S2). We used a superficial wild-type strain, VC20019, with sufficient genomic diversity to provide∼1 Mb mapping resolution. To maintain the seip-1(A185P) allele during the mapping process, we generated seip-1(A185P) in the VC20019 background by CRISPR-mediated genome editing. After mating and propagation to select for animals that contained the suppressor, we constructed MIP-MAP and whole-genome libraries for next-generation sequencing. For two of the suppressor lines, seip-1; sup1 and seip-1; sup4, we mapped the suppressors to overlapping genomic regions on chromosome III (sup1, 5.3-9.2 Mb, Fig. S2B; sup-4, 5.5-9.2 Mb, Fig. S2E). The mapping plots of the remaining suppressors either contained multiple gaps and peaks (seip-1; sup3) (Fig. S2D), indicative of multiple loci under selection, or no gaps (seip-1; sup2 and seip-1; sup5) (Fig. S2C,F), consistent with dominant suppressors (Fig. S2). Therefore, we focused on suppressor lines seip-1; sup1 and seip-1; sup4. After restricting the list of mutations to homozygous, protein-coding variants in the mapped regions and comparing the lists, we identified a single candidate modifier gene, R05D3.2, that contained different missense mutations in the two suppressor lines: R05D3.2(S647F) in seip-1; sup1 and R05D3.2(P314L) in seip-1; sup4 (Fig. 5A). R05D3.2 is an ortholog of human limb development membrane protein 1 (LMBR1) and we renamed the C. elegans R05D3.2 to lmbr-1.

To validate lmbr-1 as the suppressor candidate, we regenerated the two putative suppressor alleles of the lmbr-1 gene, lmbr-1(P314L) and lmbr-1(S647F), in the wild-type background by CRISPR/Cas9-mediated genome editing (Fig. 5A). Both lmbr-1(P314L) and lmbr-1(S647F) mutations alone caused reduced brood size compared with wild type (Fig. 5B); however, there were no embryonic lethality or eggshell formation defects observed in the lmbr-1 missense mutants (Fig. 5C,D). To further characterize the role of lmbr-1 in C. elegans, we also generated a null mutation, lmbr-1Δ, which bears a full-length deletion of the lmbr-1 gene. Compared with the P314L and S647F alleles, the homozygous lmbr-1Δ mutant caused smaller brood size and ∼25% embryonic lethality (Fig. 5B,C); this result indicates that lmbr-1 is critical to C. elegans reproduction and embryogenesis, and suggests that the missense mutations are not loss-of-function alleles. Additionally, we did not observe any eggshell defects in the lmbr-1Δ mutant (Fig. 5D), suggesting that the embryonic lethality caused by lmbr-1Δ is not associated with the integrity of eggshell. To further investigate the expression pattern of lmbr-1 in C. elegans, we successfully knocked in a fluorescent reporter gfp at the N terminus loci of the lmbr-1 gene. However, we could not detect any visible fluorescence signal of GFP::LMBR-1, likely due to the low expression level of lmbr-1 in C. elegans.

We then crossed the newly generated lmbr-1(P314L) and lmbr-1(S647F) alleles into the seip-1(A185P) background to test their suppression in the context of embryonic lethality, eggshell formation and enlarged LD size. Both lmbr-1(P314L) and lmbr-1(S647F) significantly suppressed embryonic lethality and the defective eggshell formation but did not affect the brood size compared with the effects of seip-1(A185P) alone (Fig. 5B-D). Additionally, all lmbr-1 mutants, including lmbr-1(P314L), lmbr-1(S647F) and lmbr-1Δ alone did not affect LD size in the −1 to −3 oocytes compared with wild type (Fig. 5E,I-L). Intriguingly, the lmbr-1(P314L) allele partially suppressed the irregularly enlarged LD size in the −1 to −3 oocytes of the seip-1(A185P) mutant (Fig. 5F,G,L). This finding is consistent with the alleviated enlarged LDs phenotype in the original suppressor line seip-1(A185P); sup4, which contains the lmbr-1(P314L) allele. The other allele lmbr-1(S647F), like the seip-1(A185P); sup1 strain from which it was identified, did not affect the enlarged LDs compared with the seip-1(A185P) mutant (Fig. 5F,H,L). Overall, the CRISPR-edited lmbr-1 alleles displayed identical suppression of embryonic lethality, defective eggshell formation and enlarged LD phenotypes that were observed in the original suppressor lines. Additionally, different lmbr-1 suppressor alleles, such as P314L and S647F, may contribute independently to regulating embryonic lethality/defective eggshell integrity and enlarged LD size.

The primary objective of the forward genetic screen in the orthologous lipodystrophy seip-1(A185P) mutant background was to uncover previously unreported genetic determinants and regulators of seipin. We sought to identify and characterize the genetic modifiers capable of restoring cellular and developmental deficiencies in the seip-1(A185P) mutant background. In the course of this study, we successfully isolated five distinct suppressor lines. Furthermore, we refined the MIP-MAP strategy by introducing the pathogenic variant into the mapping strain using CRISPR/Cas9 editing. This strategic innovation significantly streamlined the mapping process and reduced both labor and time required for identifying potential modifier candidates. During the trial screen, we identified two suppressor alleles, including lmbr-1(P314L) and lmbr-1(S647F), within the newly implicated genetic determinant, lmbr-1. These alleles demonstrated a remarkable ability to suppress embryonic lethality and defective eggshell formation resulting from the seip-1(A185P) mutation.

The C. elegans gene R05D3.2 is the only ortholog of human LMBR1 and LMBR1-like membrane protein (LMBR1L). Both LMBR1 and LMBR1L exhibit high conservation, sharing 60% identity in their amino acid sequence. A distinguishing feature found among all LMBR1 proteins is the presence of multiple transmembrane segments. These LMBR1 proteins are predicted to express at the ER membrane, similar to seipin expression (Choi et al., 2019). Notably, the LMBR1L protein is also known as lipocalin 1-interacting membrane receptor (LIMR) (Hesselink and Findlay, 2013), which interacts with lipocalin 1, an extracellular protein responsible for binding hydrophobic ligands, fatty acids and phospholipids, thereby facilitating their internalization and subsequent degradation (Tsukamoto et al., 2009; Glasgow, 2021; Glasgow and Abduragimov, 2021).

Therefore, the lmbr-1 suppressor mutations are plausible candidates to influence lipid transfer or lipid metabolism, potentially acting to restore the disturbed lipid balance within the seip-1(A185P) mutant. Consequently, these mutations may contribute to the suppression of embryonic lethality and other defective phenotypes. Further investigations will be directed toward a better understanding of the underlying molecular mechanism driving this suppression. In particular, we aim to characterize the intricate interplay between lipid metabolism and the functional coordination between seipin and lmbr-1.

In the scope of this study, we also characterized the functional contribution of lmbr-1 by generating multiple mutants, including two missense suppressor alleles and a lmbr-1 null mutant bearing a full-length deletion. All tested lmbr-1 mutants displayed a reduction in brood size, and the null mutant exhibited about 30% of embryonic lethality. Interestingly, previous research has shed light on the role of LMBR1L as a pivotal player in cell signaling regulation. Specifically, it is known to modulate Wnt/catenin and BMP signaling pathways in contexts of lymphocyte development and Drosophila oogenesis (Cao et al., 2019; Glasgow and Abduragimov, 2021). Collectively, these findings underscore that lmbr-1 may serve diverse functional roles that significantly contribute to the intricate dynamics of C. elegans reproduction and embryonic development. The observations also serve as an explanation for the reduced brood size in the original seip-1 suppressor lines.

Finally, the CRISPR-generated lmbr-1 mutations showed a relatively low level of suppression of DAPI penetration compared with the EMS-derived strains, suggesting that additional modifiers likely await discovery. However, we could not exclude the possibility that other cellular signaling pathways may be involved in regulating lipid droplet, embryogenesis and eggshell formation in the seipin mutants. lmbr-1 suppressor alleles may also be involved in other signaling pathways to coordinate the suppression of the defects.

In this study, we observed the emergence of a multi-nucleate and defective nuclear envelope formation, which appears to play a pivotal role in the observed lethality of the seip-1(A185P) mutant. Furthermore, similar nuclear envelope phenotypes were also associated with lipodystrophies, such as partial lipodystrophy of the Dunnigan type (FPLD2) (Guenantin et al., 2014). Thus, cellular-level similarities could be relevant to human diseases, even though C. elegans lacks adipose tissues, which show significant phenotypes in human lipodystrophy conditions.

Our study also introduces an accessible in vivo system for investigating seipin in the C. elegans germline tissue, which is enriched in LDs and necessitates the transfer of lipids and fatty acids between somatic tissue during ovulation and fertilization. The mechanisms involving lipid transportation and metabolic regulation of these structures are likely conserved between humans and C. elegans. Moreover, the observed embryonic lethality in seip-1(A185P) mutants has allowed us to screen genetic antagonists specific to seip-1 patient-specific alleles in vivo. For future studies, we intend to identify the currently unknown genetic modifiers in the other three suppressor lines using the alternative mapping crosses more suitable for polygenic or dominant mutations. Our suitable paradigm of embryonic lethality in seipin mutants should allow us to continue forward screens to identify molecular pathways or determinants to suppress or reverse the defects. The findings in our genetic study may provide insights for future therapeutic targets. A successful reproductive cycle is precisely coordinated by lipid or modified lipid signaling to fertilization, oocyte growth, meiotic progression and gonadal muscle contraction in C. elegans. The overarching theme of our study is to delineate the signaling mechanism of seipin in coordinating these physiological processes. Understanding these signaling pathways involving seipin should advance our understanding of lipid biology in other organisms, including humans.

The strains used in this study were maintained on MYOB plates as previously described at 20°C or 25°C when screening the suppressors (Burns et al., 2006). Detailed information of the strains is as follows: N2 Bristol (wild-type); AG429, seip-1(av169[A185P]) V. CRISPR/Cas9 Edit; AG444, seip-1(av169[seip-1::mScarlet]) V. CRISPR/Cas9 Edit; AG666, seip-1(A185P) V; MIP-MAP. CRISPR/Cas9 Edit; AG685, seip-1(av304[seip-1(A185P)::mScarlet]) V. CRISPR/Cas9 Edit; AG670, seip-1(av169[A185P]) V; sup1; AG671, seip-1(av169[A185P]) V; sup2; AG686, seip-1(av169[A185P]) V; sup3; AG687, seip-1(av169[A185P]) V; sup4; AG688, seip-1(av169[A185P]) V; sup5; AG743, lmbr-1(av288[P314L]) III; AG746, lmbr-1(av291[S647F]) III; AG750, lmbr-1(av293[lmbr-1Δ] III; AG751, seip-1(av294) V; AG755, seip-1(av294) V; lmbr-1(av288[P314L]) III; and AG756, seip-1(av294) V; lmbr-1(av291[S647F]) III.

seip-1(A185P) early L4 hermaphrodites were washed three times in M9 buffer and soaked in 48 mM ethyl methane sulfonate (EMS) solution for 4 h at room temperature. The EMS-treated animals were washed three times in M9 buffer and were then transferred to a fresh 100 mm MYOB plate with OP50 on one side. The animals were allowed to recover for up to 4 h before being picked to 100 mm MYOB plates with fresh OP50. Only the recovered animals that were able to crawl across the plates to the OP50 food were transferred to the fresh plates. A total of 15 MYOB plates with ten to 15 mid-L4 (P0s) on each were incubated at 20°C. Gravid F1 adult progeny (∼32,000) were collected for synchronizing the F2 population using hypochlorite treatment. The F2 embryos were shaken in a glass flask with M9 buffer overnight at 20°C, the hatched larvae were shifted to 25°C and their F3 progeny were screened for viable larvae. A total of five suppressor lines were isolated from the screen. The males of seip-1(A185P^MM) were mated with the homozygous hermaphrodites of each suppressor line (Fig. S2A). We carefully pooled F2 progeny from the cross and allowed the F2 population to expand for over ten generations for MIP-MAP analysis, which is sufficient to distribute the MIP-MAP molecular probes into the suppressor line background and provides a high molecular resolution for identifying the mutation regions (Fig. S2A). Theoretically, we should be able to narrow down the target regions bearing the putative modifiers by reading the frequency of MIP-MAP probes. As the flanking regions of the modifiers originated from a wild-type background, the occurrence frequency of the MIP-MAP probes at the nearby loci would drop to nearly zero (Mok et al., 2017).

For imaging SEIP-1::mScarlet and SEIP-1(A185P)::mScarlet expression, animals were immobilized on 7% agar pads with anesthetic (0.1% tricaine and 0.01% tetramisole in M9 buffer). Differential interference contrast (DIC) and mScarlet image acquisition were conducted using a Nikon 60×1.2 NA water objective with 1 μm z-step size; 20-25 z-planes were captured. The imaging was performed on a scanning-disk confocal system, including a Yokogawa CSU-X1 confocal scanner unit, a Nikon 60×1.2 NA water objective, and a Photometrics Prime 95B EMCCD camera. The images were analyzed using NIS imaging software (Nikon) and ImageJ/FIJI Bio-formats plugin (National Institutes of Health) (Linkert et al., 2010; Schindelin et al., 2012).

Gravid hermaphrodites were subject to three washes in M9 buffer and subsequently dissected using 23G×3/4″ needles. Embryos at various developmental stages were then transferred into a handing drop chamber that was pre-filled with blastomere culture medium (BCM), as described in the previous study (Bai and Bembenek, 2017). The BCM was prepared freshly, containing 10 μg/ml DAPI (BD Biosciences, BD Pharmingen DAPI solution, #564907) for staining. Before imaging, the hanging drop chamber was sealed with molten Vaseline. Image acquisition was performed using a Nikon 60×1.2 NA water objective with a z-step size of 1 μm.

The lipophilic molecule BODIPY 493/503 (Invitrogen, D3922) was dissolved in 100% DMSO to 1 mg/ml. The working solution of BODIPY was diluted by M9 buffer to 6.7 μg/ml BODIPY (the final concentration of DMSO was 0.8%). The tested animals were picked and incubated in 100 μl of 6.7 μg/ml BODIPY for 30 min and soaked in M9 buffer for 5-10 min until the animals were hydrated and started moving. The recovered animals were then anesthetized with 0.1% tricaine and 0.01% tetramisole in M9 buffer for 15 to 30 min before being transferred to 7% agarose pads for imaging. The diameter of LD was quantified in each oocyte by ImageJ/FIJI.

The sequence information of CRISPR design is listed in Table S1. We followed the optimized CRISPR/Cas9 editing protocol that was used in our previous studies (Bai et al., 2020a). The gene-specific guide RNAs were designed with the help of a guide RNA design checker from Integrated DNA Technologies (https://www.idtdna.com/) and were ordered as 4 nmol products from Horizon Discovery (https://horizondiscovery.com/en/dharmacon), along with tracRNA. Repair template design followed the standard protocols (Paix et al., 2015). Young gravid animals (∼20) were injected with the prepared CRISPR/Cas9 injection mix, as described in the literature (Paix et al., 2015; Bai et al., 2020b). The Cas9 protein was purchased from PNA Bio (CP01). All homozygous animals edited by CRISPR/Cas9 were validated by Sanger sequencing. The detailed sequence information for the repair template and guide RNAs are listed in Table S1.

Candidate mutations (defined as novel, homozygous and nonsynonymous) were identified by whole-genome sequencing (WGS) as described previously (Smith, 2022). Briefly, sequencing libraries were constructed using the Invitrogen Pure Link Genomic DNA Mini Kit (K1820-01) with genomic DNA from homozygous suppressor-bearing strains. The WGS libraries were pooled and sequenced on a HiSeq 3000 instrument (Illumina) to at least 20-fold coverage. Variants were identified with a pipeline of BBMap (https://sourceforge.net/projects/bbmap/), SAMtools (Li et al., 2009), FreeBayes (Garrison and Marth, 2012 preprint) and ANNOVAR (Wang et al., 2010). Mapping loci for suppressors were identified using molecular inversion probes (MIPs) to single-nucleotide polymorphisms (SNPs), as described previously (Mok et al., 2017). Briefly, suppressor-bearing strains were mated to SNP mapping strain VC20019 (Thompson et al., 2013), engineered via CRISPR to contain the seip-1(A185P) mutation. F1 cross-progeny were allowed to self-fertilize, and a minimum of 50 homozygous F2 progeny were pooled for the construction of MIP libraries. SNP allele frequencies were determined using a custom script and plotted with R (http://www.R-project.org) to delimit the mapping interval.

The sample size for each experimental group or condition is labeled in the figures and figure legends. Statistical methods and sample numbers are detailed in the corresponding figure legends. Statistical significance for other assays was determined using an unpaired two-tailed t-test, a one-way ANOVA with Tukey's post hoc test or a χ² test. ns, not significant; *P<0.05; **P<0.005. Both the Shapiro–Wilk and Kolmogorov–Smirnov normality test indicated that all data follow normal distributions.

This article is part of the collection ‘Translating Multiscale Research in Rare Disease’, which was launched in a dedicated Special Issue edited by Monica Justice, Monkol Lek, Karen Liu and Kate Rauen. See related articles in this collection at https://journals.biologists.com/dmm/collection/39/Rare-Disease.

We thank past and present members of A.G's lab at the National Institutes of Health (NIH), National Institute of Diabetes and Digestive and Kidney Diseases for insightful discussions during the screening and preparation of the manuscript. We thank all members of the National Institutes of Health (NIH) Worm Club and the Baltimore Worm Club, Dr Chao-Wen Wang (SINICA, Taiwan) and Dr Will Printz (UT Southwestern Medical Center) for providing critical comments on the manuscript and on our seipin project. The N2 strain was provided by the Caenorhabditis Genetics Center, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440). This work is dedicated in memory of Dr Andy Golden, whose invaluable support greatly contributed to the success of this project. He will be deeply missed.

This article is part of the Special Issue ‘Translating Multiscale Research in Rare Disease’, guest edited by Monica Justice, Monkol Lek, Karen Liu and Kate Rauen. See related articles at https://journals.biologists.com/dmm/collection/39/Rare-Disease.