SCAV-3 affects apoptotic cell degradation in Caenorhabditis elegans

As the ﬁnal step in apoptosis, apoptotic cells (ACs) are swiftly removed by specialized phagocytes, such as macrophages, or nonprofessional phagocytes, such as epidermal cells. Genetic studies of model organisms such as Caenorhabditis elegans have helped to elucidate the mechanisms of AC clearance and the underlying causes of disorders related to the dysregulation of these pathways. C. elegans possesses six class B scavenger receptor homologs, but whether they affect apoptosis is unknown. Here, we show that only the loss of function of scav-3 , the C. elegans homolog of human lysosomal integral membrane protein-2, resulted in a considerable accumulation of cell corpses, which was caused by a failure in degradation rather than engulfment. SCAV-3 was found to be widely distributed and localized in lysosomes to maintain the integrity of the lysosomal membrane. Further study revealed that loss of scav-3

As the final step in apoptosis, apoptotic cells (ACs) are swiftly removed by specialized phagocytes, such as macrophages, or nonprofessional phagocytes, such as epidermal cells. Genetic studies of model organisms such as Caenorhabditis elegans have helped to elucidate the mechanisms of AC clearance and the underlying causes of disorders related to the dysregulation of these pathways. C. elegans possesses six class B scavenger receptor homologs, but whether they affect apoptosis is unknown. Here, we show that only the loss of function of scav-3, the C. elegans homolog of human lysosomal integral membrane protein-2, resulted in a considerable accumulation of cell corpses, which was caused by a failure in degradation rather than engulfment. SCAV-3 was found to be widely distributed and localized in lysosomes to maintain the integrity of the lysosomal membrane. Further study revealed that loss of scav-3 had no effect on phagosome maturation or the recruitment of lysosomes to phagosomes carrying cell corpses. Moreover, we discovered that the hydrolytic enzymes contained in the lysosomes were reduced in phagosomes in scav-3 mutants. Thus, hydrolases may leak from the damaged lysosome during phagolysosome formation due to the loss of scav-3 function, which reduces lysosome digestion activity and thus directly contributes to the elimination of ACs.
As the final step in apoptosis, apoptotic cells (ACs) are swiftly removed by specialized phagocytes such as macrophages or nonprofessional phagocytes such as epidermal cells [1]. The clearance of ACs (efferocytosis) is essential for maintaining tissue homeostasis [2]. In contrast, defective efferocytosis results in the release of harmful contents from ACs, which leads to chronic severe autoimmune diseases and neurodegenerative diseases [3]. Genetic studies of model organisms such as Caenorhabditis elegans have helped to understand AC clearance mechanisms and the underlying causes of disorders related to the dysregulation of these pathways. In C. elegans, two parallel redundant regulatory pathways control the process of engulfment of ACs. In the ced-1/6/7 pathway [4][5][6], the phagocytic receptor CED-1 recognizes 'eat-me' signal phosphatidylserine (PS) [7][8][9]. In the ced-2/5/12 pathway [10][11][12][13], several receptors including the phosphatidylserine receptor PSR-1 [14], MOM-5, and integrin INA-1 [15] are proposed to recognize and bind to the PS.
It is reasonable to propose that the engulfment of ACs in C. elegans is mediated by additional receptorbased recognition systems. CD36 and SR-BI, two B-class scavenger receptors, have played crucial roles in AC clearance in mammals, but the precise molecular mechanism is unclear [16][17][18]. SCAV-1-6 comprises the six class B scavenger receptor homologs found in C. elegans (SRB). However, whether SCAV-1-6 affect apoptosis in C. elegans is unknown.
In the present study, we found that only loss of function of SCAV-3 causes the accumulation of ACs in C. elegans. We show that scav-3 acts specifically in phagocytes for cell corpse clearance. In addition, we found that the persistent cell corpses in scav-3 mutants were internalized but failed to be degraded promptly. Consistent with previous studies, we found that SCAV-3 is widely distributed and localized on lysosomes to maintain the integrity of the lysosomal membrane, and loss of scav-3 causes damage to lysosome membranes. Our further analysis found that loss of scav-3 did not affect the phagosome maturation process or the recruitment of lysosomes to the cell corpsecontaining phagosome. Furthermore, we discovered that lysosomal hydrolytic enzymes could enter phagosomes, but were reduced in phagosomes in scav-3 mutants. We propose that hydrolases may leak from damaged lysosomes during the formation of phagolysosomes in scav-3 mutants, leading to the accumulation of ACs. Our findings provide direct evidence for the significance of lysosomal membrane integrity in the removal of ACs during programmed cell death.

Plasmid construction
To construct the plasmid for RNAi, the genomic sequence of the scav-1 or scav-6 was amplified by PCR and inserted between the BglII and HindIII sites of the vector pPD129.36, and genomic sequence of scav-4 was amplified by PCR and inserted between the XhoI and HindIII sites of the vector pPD129.36. To make the construct of P scav-3 scav-3::gfp, a 5.3-kb fragment containing the genomic sequence of the gene and 2.1 kb of its promoter region was inserted between the HindIII and BamHI sites of the vector pPD95.77. To make P ced-1 scav-3 and P egl-1 scav-3 constructs, the full-length scav-3 cDNA was cloned into the P ced-1 vector via its SalI and BamHI sites or into the P egl-1 vector through its BamHI and XmaI sites.

RNAi experiments
RNAi experiments were performed by using bacterial feeding assays as described previously [20]. In most cases, L4stage animals were transferred to plates seeded with bacteria expressing either control double-stranded RNA (dsRNA; L4440 empty vector; Control RNAi) or dsRNA of genes of interest. Germ and embryo cell corpses were observed at different time points.

Quantification of cell corpses
Cell corpses were scored by using Nomarski optics. For embryonic cell corpses, only those in the head regions of embryos at the comma, 1.5-fold, twofold, threefold, and fourfold stages were scored. At least 15 embryos at each embryonic stage were scored, and the average numbers of cell corpses were shown. Germ cell corpses in the germline meiotic region of one gonad arm were scored 12, 24, 36, 48, and 60 h after the L4 larval stage. Data derived from different genetic backgrounds were compared using unpaired t-tests.

Acridine orange staining of cell corpses
To stain the germ cell corpses with Acridine orange (AO), adult worms were incubated in M9 medium containing AO (100 mgÁmL À1 ) and OP50 bacteria in the dark for 2 h at room temperature. Worms were then transferred to nematode growth medium plates for recovery for 1 h to decrease background gut fluorescence. The stained worms were examined under fluorescence microscopy to score the AOpositive germ cell corpses from > 150 germ cell corpses of each examined strain.

Four-dimensional analysis of cell corpse duration
Time-lapse imaging was performed under 1009 objectives at 20°C. To record the duration of embryonic cell corpses, early embryos (2-cell stage) were put in egg salt buffer (118 mM NaCl and 48 mM KCl) and mounted on slides with agar pads. The slides were sealed with beeswax and petroleum jelly (1 : 1). Images in 20 Z-serial sections (1 lm/section) were captured every minute for 400 min by using a Zeiss Axioimager M2 (Zeiss, Imager M2, Oberkochen, Germany) coupled with an AxioCam monochrome digital camera.
We found that only scav-3(ok1286) mutants displayed a significant increase in ACs at various stages of embryonic development ( Fig. 2A-F). We also observed that germ cell corpses accumulated significantly in an age-dependent manner in scav-3(ok1286) animals ( Fig. 2G-L). In addition, we found that scav-3 RNAi-treated worms had a significantly higher apoptotic phenotype in embryos and germ cells when compared to controls (Fig. S1A,B). The scav-3(ok1286) contains a 1146-bp deletion and a 2-bp insertion that removes most of the second and third exons and generates an early stop codon, resulting in a truncated protein containing only the first 118 amino acids, representing a null mutation of the scav-3 gene ( Fig. 1) [21]. While the scav-3 mutant was able to produce healthy fertilized oocytes, we were unable to detect any defects in the mutant's growth while keeping it in culture. These results demonstrate that loss of scav-3 function caused apoptotic program abnormalities.

SCAV-3 affects cell corpse clearance in C. elegans
The increase of both embryonic and germ cell corpses in scav-3(ok1286) could be caused by either excessive apoptosis or cell corpse clearance deficiencies. To distinguish between these two possibilities, we first used the promoter of ced-1, which is only expressed in engulfing cells, to drive the expression of scav-3 (P ced-1 scav-3) and discovered that it fully rescued the defects  in cell corpse clearance in scav-3(ok1286) embryos (Table 1) as well as in adult germ lines (Table 2). In comparison, we also used the promoter of egl-1 [22][23][24], which was exclusively expressed in dying cells, to drive the expression of scav-3 (P egl-1 scav-3) but found it did not rescue the cell corpse clearance defects in scav-3(ok1286) embryos (Table 1). We next carried out time-lapse recording analysis to measure the duration of persistence of both embryonic (Fig. 3A,B) and germ cell corpses. In wild-type animals, the average duration of embryonic cell corpses was 25 AE 3 min, whereas 60% of cell corpses persisted more than 30 min and some exceeded 100 min in scav-3(ok1286) animals (Fig. 3C). In germ lines, the average duration of germ cell corpses was 51.7 AE 5.8 min, whereas we seldom observed germ cell corpses persisting 60 min in wild-type animals. In scav-3(ok1286) mutant, the average persistence time of germ cell corpses significantly increased, with ≥ 90% of examined cell corpses persisted over 100 min (Fig. 3E). Moreover, there were no significant differences between wild-type and scav-3 (ok1286) in the number of cell death events from the first cell division to 400 min during embryogenesis ( Fig. 3D). We then monitored the uptake and degradation of ssGFP by coelomocytes in a time course manner. At 20 h after heat shock, ssGFP in the body cavity and coelomocytes had vanished in the wild-type. However, ssGFP was observed both in the body cavity and in coelomocytes even 36 h after heat shock in scav-3 mutant animals, indicating that endosomal/lysosomal protein degradation was strongly impeded in the mutant coelomocytes (Fig. S2A,B). Together, these findings demonstrate that SCAV-3 activity in engulfing cells is essential for cell corpse clearance, and the increase of cell corpses in scav-3(ok1286) animals is caused by defects in the removal of cell corpses.

SCAV-3 is widely distributed and localizes to lysosomes
The open reading frame Y49E10.20 found on linkage group III and encoding the 534 amino acid SCAV-3 protein in C. elegans. Previous studies have found that SCAV-3 can maintain the integrity of the lysosomal membrane and that aberrant SCAV-3 function can affect the life span of C. elegans [17]. We generated a Table 1. Expression of scav-3 cDNA in engulfing but not dying cells rescued a scav-3 phenotype in embryo of scav-3(ok1286).

The persistent cell corpses in scav-3(ok1286) are internalized but not digested
To investigate how scav-3 influences the clearance of cell corpses, we first examined whether scav-3 affected the recognition of ACs by phagocytes. In scav-3 (ok1286) mutant animals, the majority of the persisting germ cell corpses were surrounded by CED-1:: GFP generated by germ line sheath cells, indicating that scav-3 does not affect the recruitment of phagocytes around germ cell corpses (Fig. 5A,B). AO can be used to label the cell corpses internalized in live cells. We next used AO staining to determine whether the persisting cell corpses in scav-3(ok1286) animals were internalized. In this experiment, the mutant ced-1 (e1735) served as a negative control because majority of cell corpses were not phagocytosed in this mutant, whereas the mutant vps-18(tm1125) served as a positive control because most cell corpses were phagocytosed [27] but not degraded in this mutant. In the germ line of wild-type animals, cell corpses were rapidly engulfed, of which 60% were AO positive. In contrast, only a small number of germ cell corpses in ced-1(e1735) animals were labeled by AO, but majority of germ cell corpses in vps-18(tm1125) animals were stained by AO. We found that the gonad ACs of scav-3(ok1286) were rapidly engulfed by phagocytes, and the ratio of AO staining was higher than that of wild-type N2 (P < 0.05) (Fig. 5C,D). These results indicate that these persistent germ cell corpses in scav-3 (ok1286) were indeed internalized by the engulfing cells. To directly monitor the degradation process of cell corpses, we introduced in scav-3(ok1286) mutants a transgenic marker H2B::mCHERRY [28], which is specifically expressed in the germ line to label chromatin. Our results showed that wild-type and mutant animals had obvious 'button-like' ACs at 0 min, and the H2B::mCHERRY fusion protein surrounded the cells as the starting point for degradation. In the wild-type, the chromatin in early germ cell corpses was condensed and disappeared within 40 min (39 AE 1.9 min; n = 5). The chromatin in early germ cell corpses was also condensed in scav-3(ok1286) mutants, it became diffusive in the later stage, and the mCHERRY signal remained for 80 min (77 AE 5.8 min, n = 5) (Fig. 5E,F), showing that chromatin breakdown is greatly delayed. These results reveal that the accumulation of cell corpses produced by the scav-3 mutation is due to deficiencies in the digestion of cell corpses in engulfing cells.

SCAV-3 affects lysosome digestion activity
To clarify how SCAV-3 affects AC degradation, we analyzed the recruitment of maturation-required phagosomal components. We introduced integrated arrays expressing GFP-fluorescent protein-tagged phagosome markers (GFP::RAB-7) [20,29], a small GTPase required for late endosome-phagosome fusion, into scav-3(ok1286) mutants. We found that phagosomes in scav-3(ok1286) germlines are positive for GFP::RAB-7 at a similar level as in the wild-type, indicating that loss of scav-3 did not affect the phagosome maturation process (Fig. 6A,E). After phagosomes mature, they combine with lysosomes to form phagolysosomes. Then hydrolytic enzymes in lysosomes enter the phagosomes for degradation. Thus, SCAV-3 should function at the stage of cell corpses digestion in phagolysosomes. To examine whether the internalized germ cell corpses (phagosomes) fused with lysosomes in both wild-type and scav-3 mutants, we introduced integrated arrays expressing GFP or mCherry fluorescent protein-tagged lysosome markers into scav-3(ok1286) mutants, including LMP-1(LMP-1::mCHERRY and LMP-1::GFP), a lysosomal membrane protein [30]. We discovered that phagosomes in scav-3(ok1286) germ lines exhibit similar levels of LMP-1 positivity as those in wild-type (Fig. 6B,C,E). To examine whether hydrolytic enzymes in lysosomes entering the phagosomes were affected in scav-3 (ok1286), we introduced integrated arrays expressing mChOint fluorescent protein-tagged lysosome hydrolytic enzymes, CPL-1(CPL-1::mChOint), into scav-3 (ok1286) mutants [31]. We found that phagosome association with CPL-1::mChOint was significantly reduced in scav-3(ok1286) than in wild-type (Fig. 6D,  E), indicating that loss of scav-3 affected the hydrolytic enzyme CPL-1 in lysosomes entering the phagosomes. Thus, in scav-3(ok1286) mutants, although LMP-1positive organelles, likely aberrant lysosomes, could be recruited to the cell corpse-containing phagosome and were shown to be able to fuse further into a phagolysosome, hydrolytic enzymes leaked from the damaged lysosome. Thus, the digestive function of lysosomes was impaired, resulting in the accumulation of undigested cell corpses. These results indicate that the loss of function of SCAV-3 affects lysosome digestion activity during the AC clearance process, resulting in the accumulation of ACs.

Discussion
There are three scavenger receptor class B members in vertebrates: CD36, scavenger receptor SR-BI, and lysosomal integral membrane protein-2 (LIMP-2, also known as SCARB2 or CD36B like-2). CD36 and SR-BI have been implicated in the engulfment of ACs in mammals [16]. SCAV-1-6 consists of the six homologs of class B scavenger receptors in C. elegans. However, whether they affect apoptosis is not clear. Here, we found that only genetic inactivation of scav-3 gave rise to a substantial increase of both embryonic and germ cell corpses. The SCAV family appeared not to be functionally redundant during AC clearance, as only SCAV-3 affects the clearance of ACs. However, SCAV-3 and other phagocytic receptors may together maintain the phagocytosis of ACs. The scavenger receptors CD36 and SR-BI are highly expressed in macrophages and are able to bind anionic phospholipids, serine phospholipids, and ACs, which is one of the important conditions for macrophages to participate in AC clearance [17]. In this paper, we found that the lysosomal membrane protein SCAV-3 involved in the clearance process of ACs, unlike mammals, SCAV-3 seems not to be involved in the binding process, but to degrade ACs by maintaining the membrane integrity of lysosomes. Our further analysis revealed that defects in cell corpse clearance, not excessive apoptosis, were responsible for the increased cell corpses and that scav-3 functions in engulfing cells for cell corpse clearance. These findings established that SCAV-3 plays a critical role in the clearance of ACs.
SCAV-3 is the C. elegans homolog of human LIMP-2, which is a type III glycoprotein in the membranes of lysosomes and endosomes. LIMP-2 plays a vital role in the delivery of b-glucocerebrosidase from the endoplasmic reticulum to lysosomes and has been identified as the cellular receptor for multiple virus entry into host cells [32]. Recent studies found that SCAV-3 maintains lysosome integrity and dynamics in C. elegans, although lysosomal localization of bglucocerebrosidase is not affected by the loss of scav-3 [21]. Consistent with this study, we found that SCAV-3 is widely distributed and localized on lysosomes, and loss of scav-3 causes damage to lysosome membranes.
In C. elegans, two parallel and partially redundant regulatory pathways induce cytoskeletal rearrangement of engulfing cells through recognition and transduction of engulfment signals, leading to pseudopod encapsulation for the internalization of ACs and finally to sealing to form phagosomes. Then, phagosomes fuse with lysosomes to form phagolysosomes [33]. Finally, the cell corpses are digested in the phagolysosome by lysosomal acid hydrolases, such as CPL-1 and NUC-1 [25,33,34]. We found that the persistent cell corpses in scav-3 mutants were internalized but failed to be degraded promptly. Our further analysis found that loss of scav-3 did not affect the phagosome maturation process or the recruitment of lysosomes to the cell corpse-containing phagosome for further fusing into phagolysosome. In addition, we found lysosomal hydrolytic enzyme CPL-1 could enter phagosomes but were reduced in phagosomes in scav-3 mutants (Fig. 7).
Lysosomes are 'digestive factories' in cells that can degrade extra-and intracellular material delivered by endocytosis, phagocytosis, and autophagy. Consequently, the dysfunction of lysosomes contributes to the pathogenesis of many disorders, such as lysosomal Fig. 4. Damaged lysosomes accumulated in scav-3 mutants. (A-P) Expression and subcellular localization of SCAV-3::GFP driven by the scav-3 promoter. Images of DIC, SCAV-3::GFP, and CPL-1::mChOint (driven by the cpl-1 promoter) and merged SCAV-3::GFP with CPL-1:: mChOint are shown for a 1.5-fold embryo, an L1 larvae, the germ line, and the hypodermis. Bars, 10 lm. Arrowheads indicate some of the foci with colocalization of SCAV-3::GFP and CPL-1::mChOint. The fluorescent images of the hypodermis in wild-type (WT; Q-S) and scav-3 (ok1286) (T-V) adults expressing Gal3::GFP. Bars, 50 lm. (W) The average number of positives for GFP::Gal3 (damaged) was quantified in wild-type and scav-3(ok1286) expressing Gal3::GFP. At least 15 animals were scored in each strain (n = 3, mean AE SEM). (X) The fluorescent images of the hypodermis in wild-type and scav-3(ok1286) adults expressing Gal3::GFP and CPL-1::mChOint. Colocalization indicated by white arrows; Bars, 10 lm. Data were compared with the unpaired t-test. ***P < 0.001.   We analyzed ≥ 100 cell corpses from ≥ 20 animals for each marker, and data were derived from three biologically independent replicates. Comparisons were performed between N2 and scav-3(ok1286) mutants for each marker using an unpaired t-test. ***P < 0.001. storage diseases, neurodegenerative disorders, and cancer [35][36][37]. Lysosomes contain a variety of hydrolytic enzymes that act in coordination to degrade all types of macromolecules [36]. Thus, the integrity of the lysosomal membrane becomes crucial, and it serves as the mechanical barrier that encloses potent luminal hydrolases and protects other cellular constituents from unwanted degradation [38]. Lack of SCAV-3 function causes the integrity of the lysosomal membrane to be destroyed. Therefore, we propose that various hydrolases lysosomal hydrolases could still enter the phagosome during phagolysosome formation in scav-3 mutants, but that they are leaking from the damaged lysosome, thereby affecting lysosomal digestion activity and causing the accumulation of ACs.

Conclusion
In this study, we investigate whether six homologs of scavenger receptors of class B in C. elegans affect apoptosis. We found that only loss-of-function scav-3 resulted in a large accumulation of cell corpses due to a failure in degradation rather than engulfment. SCAV-3 is widely distributed and localized on lysosomes to maintain the integrity of the lysosomal membrane. Our further analysis indicates that in the process of phagolysosome formation in scav-3 mutants, hydrolases may be leaked from the damaged lysosome, leading to defects in the degradation of cell corpses.
and AM wrote the manuscript with feedback from all authors.