GOP-1 promotes apoptotic cell degradation by activating the small GTPase Rab2 in C. elegans

Abstract

Apoptotic cells generated by programmed cell death are engulfed by phagocytes and enclosed within plasma membrane-derived phagosomes. Maturation of phagosomes involves a series of membrane-remodeling events that are governed by the sequential actions of Rab GTPases and lead to formation of phagolysosomes, where cell corpses are degraded. Here we identified gop-1 as a novel regulator of apoptotic cell clearance in Caenorhabditis elegans Loss of gop-1 affects phagosome maturation through the RAB-5-positive stage, causing defects in phagosome acidification and phagolysosome formation, phenotypes identical to and unaffected by loss of unc-108, the C. elegans Rab2 GOP-1 transiently associates with cell corpse-containing phagosomes, and loss of its function abrogates phagosomal association of UNC-108. GOP-1 interacts with GDP-bound and nucleotide-free UNC-108/Rab2, disrupts GDI-UNC-108 complexes, and promotes activation and membrane recruitment of UNC-108/Rab2 in vitro. Loss of gop-1 also abolishes association of UNC-108 with endosomes, causing defects in endosome and dense core vesicle maturation. Thus, GOP-1 is an activator of UNC-108/Rab2 in multiple processes.

Figure 1.

gop-1 is required for cell corpse removal. (A) DIC and fluorescent images of wild-type and gop-1(tm5384) embryos expressing Annexin V::GFP. Cell corpses were identified by their raised button–like morphology and surface-exposed phosphatidylserine indicated by Annexin V::GFP labeling (arrowheads). The percentage of cell corpses labeled by Annexin V::GFP was quantified and is shown as mean ± SD at right. Unpaired t test was performed to compare data derived from wild type with that of gop-1(tm5384). No significant difference was observed. Bar, 2.5 µm. (B, C, G, and H) Time-course analysis of cell corpse appearance during embryonic (B and G) and germline (C and H) development was performed in the indicated strains. The y axis indicates the number of cell corpses. At least 15 animals were scored in each strain at each stage, and data are shown as mean ± SD. Data derived from different genetic backgrounds at multiple developmental stages were compared by two-way ANOVA followed by Bonferroni posttest. *, P < 0.05; **, P < 0.001; N.S., not significant. (D) Embryonic cell deaths occurring 200–380 min after the first cleavage were followed in wild type (WT; n = 3) and gop-1(tm5384) mutants (n = 3). The mean number of total cell deaths (± SD) is shown in parentheses. The y axis indicates the number of cell deaths at each time point. (E and F) Cell corpse duration in soma (E) and germ line (F) was analyzed by 4D microscopy in WT and gop-1(tm5384) mutants. 33 somatic cell corpses or 25 germ cell corpses were examined in each strain. The mean duration time of cell corpses (± SD) is shown in parentheses.

Figure 2.

gop-1and unc-108 act in the same pathway to regulate phagosome maturation. (A and B) Percentage of somatic (A) and germ cell corpses (B) labeled by various phagosomal markers at the 1.5-fold (A) or adult (48 h after L4/adult molt; B) stage was quantified in the indicated strains. (C–P) DIC and fluorescent images of the gonads stained by Lysosensor green (C–H′) or expressing LAAT-1::GFP (J–O′) in the indicated strains. White arrowheads, cell corpses labeled by Lysosensor green or LAAT-1::GFP; yellow arrowheads and boxes, unlabeled corpses. Quantification is shown in I and P. In A, B, I, and P, at least 15 animals were scored in each strain, and data are shown as mean ± SD. In A and B, data derived from different genetic backgrounds were compared by two-way ANOVA followed by Bonferroni posttest. *, P < 0.05; **, P < 0.001. All other points had P > 0.05. In I and P, one-way ANOVA with Tukey’s posttest was performed to compare mutant datasets with wild type or datasets linked by lines. **, P < 0.0001. Bars, 5 µm.

Figure 3.

Loss of gop-1 and unc-108 affects the dynamics of RAB-5 and PtdIns3P on phagosomes. Time-lapse images of somatic cell corpses in wild type (WT), gop-1(tm5384), unc-108(n3263), tbc-2(qx20), and unc-108;tbc-2 embryos expressing GFP::RAB-5 (A) or YFP::2xFYVE (B). 0 min, time point before RAB-5 or FYVE was first detected around the cell corpse; arrowheads, time point when GFP::RAB-5 or YFP::2xFYVE disappears from phagosomes. Bars, 2.5 µm. Quantifications are shown in C and D. Data are shown as mean ± SD. One-way ANOVA with Tukey’s posttest was performed to compare mutant datasets with wild type. **, P < 0.0001.

Figure 4.

Loss of gop-1 and unc-108 affects phagosomal association of RAB-7 and phagolysosome formation. Time-lapse images of somatic cell corpses in wild type (WT), unc-108(n3263), and gop-1(tm5384) embryos expressing GFP::RAB-7 (A) or coexpressing mCherry::MTM-1(C378S) and LAAT-1::GFP (B–F). 0 min, time point before RAB-7 or MTM-1 was first detected around the cell corpse; yellow arrowheads, MTM-1 release from phagosomes; white arrowheads, shrinkage of RAB-7– or LAAT-1–positive phagosomes. The percentage of phagosomes with the representative pattern is shown at left (A) or top (B–F). Quantification of RAB-7 duration on phagosomes (A) or the time between MTM-1 release and LAAT-1 recruitment is shown beneath the images (mean ± SD). n, number of phagosomes quantified. Bars, 2.5 µm.

Figure 5.

Recruitment of GOP-1 to phagosomes requires RAB-5 but not RAB-7 function. (A–C″) Confocal fluorescence images of the sheath cell in wild type coexpressing GFP::GOP-1 and Cherry::UNC-108 (A–A″) or Cherry::FAPP1-PH (B–B″, trans-Golgi marker) or coexpressing GFP::UNC-108 and Cherry::FAPP1-PH (C–C″). White arrowheads, vesicles containing overlapping GFP and Cherry fluorescence; pink arrowheads, GFP- and Cherry-positive vesicles that are in close proximity; yellow arrowheads, GFP-positive and Cherry-negative puncta. Quantification is shown at right. At least 10 animals were scored in each strain. Data are shown as mean ± SD. (D–I′) DIC and fluorescence images of embryos expressing GFP::GOP-1 in the indicated strains. White arrowheads, cell corpses labeled by GFP::GOP-1; yellow arrowheads, unlabeled ones. (J and K) Quantification of somatic (J) and germ cell (K) corpses labeled by GFP::GOP-1 in the indicated strains is shown. At least 15 animals were scored in each strain. Data are shown as mean ± SD. Recruitment of GOP-1 in unc-108(n3263) and tbc-2(qx20) was compared with that in wild type, whereas data derived from rab-5 RNAi or rab-7 RNAi were compared with the control group by unpaired t tests. **, P < 0.0001. All other points had P > 0.05. Bars, 5 µm.

Figure 6.

Loss of gop-1 affects phagosomal recruitment of UNC-108. (A and B) Time-lapse images of somatic cell corpses in wild type (WT; A) and gop-1(tm5384) (B) expressing mCherry::MTM-1(C378S) and GFP::UNC-108. 0 min, time point when MTM-1(C378S) was first detected around the cell corpse. The time period until the cell corpse was surrounded by MTM-1(C378S) was defined as the pseudopod extension stage. Quantification of pseudopod extension and the phagosomal duration of MTM-1 and UNC-108 is shown at bottom (mean ± SD). n, number of phagosomes quantified. Yellow arrowheads, MTM-1 release from phagosomes; white arrowhead, recruitment of GFP::UNC-108 to the phagosomal membrane. (C and D) Time-lapse images of germ cell corpses in wild type (WT; C) and gop-1(tm5384) (D) expressing GFP::UNC-108 and mCherry::H2B (Histone 2B). 0 min, time point when phagosomal degradation of the apoptotic cell, indicated by diffuse H2B (white arrowheads), was observed. In wild type, UNC-108 was enriched on the phagosomal membrane 36 min before diffuse H2B was observed, whereas it was absent from phagosomes in gop-1(tm5384). Percentage of phagosomes with the representative pattern is shown in parentheses. n, number of phagosomes that were followed and quantified. In A–D, pink arrowheads indicate the first time point when the intensity of GFP::UNC-108 on the phagosomal surface (white dashed line) was measured. Bars, 5 µm. (E) Cartoon showing a cross section of a pair of gonadal sheath cells expressing GFP::UNC-108. Two apoptotic germ cells engulfed by sheath cells are shown. Red arrow, appearance of GFP::UNC-108 on the phagosomal surface (dark green); black arrows, diffuse GFP::UNC-108 in the sheath cell cytosol (light green) separated by germ cell nuclei. (F and G) Total fluorescence intensity of GFP::UNC-108 on the phagosomal surface in embryos (F) or germ line (G) at different time points as shown in A–D was measured, and the normalized fluorescence intensity is shown. Enrichment of GFP::UNC-108 on the phagosomal surface was seen in wild type but not gop-1(tm5384) mutants.

Figure 7.

Loss of gop-1 affects endosome and dense core vesicle maturation. (A–J) DIC and fluorescent images of coelomocytes expressing RME-8::GFP (A–D′) or LMP-1::GFP (E–H′) in the indicated strains. RME-8::GFP and LMP-1::GFP associate with endosomes and lysosomes, respectively, in wild type (white arrowheads), but label enlarged vacuoles (yellow arrowheads) in gop-1(tm5384), unc-108(n3263), and unc-108;gop-1 worms. Quantification is shown in I and J. At least 30 animals were scored in each strain. (K–O) Confocal fluorescence images of the dorsal nerve cord in the indicated strains expressing NLP-21::VENUS (K–N). Quantification is shown in O. At least 30 animals were scored in each strain. Data are shown as mean ± SD. One-way ANOVA with Tukey’s posttest was performed to compare mutant datasets with wild type or datasets linked by lines. **, P < 0.0001; N.S., no significance. (P–T) Confocal fluorescence images of the neuronal cell body in the ventral nerve cord expressing NLP-21::VENUS in the indicated strains. White arrowheads, small vesicles (<0.4 µm²); yellow arrowheads, large ones (>0.8 µm²). The surface area of NLP-21–positive vesicles in neuronal bodies was quantified in the indicated strains, and the size distribution of vesicles is shown in T. Purple lines indicate the mean surface area. At least 15 animals were scored in each strain. One-way ANOVA with Tukey’s posttest was performed to compare mutant datasets with wild type or datasets linked by lines. **, P < 0.0001; N.S., no significance. Bars: (A–H′ and K–N) 5 µm; (P–S) 2 µm.

Figure 8.

Loss of gop-1 disrupts endosomal recruitment of UNC-108. (A and B) Time-lapse images of coelomocytes in wild type (WT; A) and gop-1(tm5384) (B) expressing GFP::UNC-108 and NUC-1::Cherry. 0 min, time point before GFP::UNC-108 was initially detected on endosomes. White arrowheads, recruitment of GFP::UNC-108 to the endosome membrane; blue arrowheads, appearance of NUC-1::Cherry; yellow arrowheads, absence of UNC-108 from endosomes in gop-1(tm5384). The duration of UNC-108 on endosomes was quantified and is shown beneath the images (mean ± SD). n, number of endosomes quantified. (C and D) DIC and fluorescent images of coelomocytes in wild type (WT; C and C′) and gop-1(tm5384) (D and D′) expressing GOP-1::GFP. White arrowheads, GOP-1::GFP-positive vesicles. (E) Percentage of animals that contain GOP-1::GFP-positive vacuoles in coelomocytes was quantified and is shown. n, number of worms that were quantified. Bars, 5 µm.

Figure 9.

GOP-1 interacts with UNC-108 and disrupts the UNC-108–GDI-1 complex. (A and B) The interaction between GOP-1 and UNC-108 was examined by yeast two-hybrid analyses (Y2H; A) and GST pull-down assays (B). GOP-1 specifically interacts with the GDP-bound and nucleotide-free forms of UNC-108 in Y2H (A), whereas significantly more GOP-1-HIS is pulled down by GST-UNC-108(S20N) (GDP-bound) and GST-UNC-108(N119I) (nucleotide-free) than by GST-UNC108(WT) and GST-UNC-108(Q65L) (GTP-bound; B). At least four independent experiments were performed, and data are shown as mean ± SD. One-way ANOVA with Tukey’s posttest was performed to compare mutant datasets with wild type. *, P < 0.05; N.S., no significance. (C) Cherry::UNC-108 is coprecipitated with GFP::GOP-1 by GBP pull-down in lysates prepared from adult worms coexpressing both fusion proteins. (D–D″′) DIC and fluorescent images of the wild-type germ line coexpressing GFP::GOP-1 and Cherry::UNC-108. GOP-1 and UNC-108 colocalized on the phagosome (arrowheads). n, number of worms quantified. Bars, 5 µm. (E) GOP-1 disrupts the UNC-108–GDI-1 complex. Schematic illustration of the GDI displacement assay is shown at left. Addition of GST-GOP-1 but not GST led to association of UNC-108 but not GDI-1 with GST-beads. At least three independent experiments were performed, and a representative result is shown.

Figure 10.

GOP-1 promotes loading of GTP onto and membrane recruitment of UNC-108. GOP-1 promotes association of UNC-108 with liposomes. Binding of UNC-108–GDI-1 to PC + PtdIns3P (8%) liposomes was detected by Western blot. UNC-108 but not GDI-1 was recovered in the top fraction with GOP-1 in a liposome flotation assay in the absence or presence of GTPγS. Quantification is shown at right. At least three independent experiments were performed. Data are shown as mean ± SD. Unpaired t test was performed to compare datasets linked by lines. *, P < 0.05. (B) GOP-1 promotes loading of GTP onto UNC-108. Schematic illustration of the GTP incorporation assay is shown at left. At least three independent experiments were performed, and a representative result is shown. (C–H) Fluorescent images of the gonads (C and D) or coelomocytes (F and G) in gop-1(tm5384) expressing GFP::UNC-108 (C and F) or GFP::UNC-108(K120E) (D and G). K120E (white arrowheads) but not wild-type UNC-108 (yellow arrowheads) was seen on phagosomes (C and D) and endosomes (F and G). Quantification is shown in E and H. In E, at least 15 animals were scored, and data are shown as mean ± SD. Unpaired t test was performed to compare data derived from wild type and gop-1(tm5384). **, P < 0.0001; N.S., no significant difference. In H, the percentage of animals containing GFP-positive vacuoles in coelomocytes was quantified. n, number of animals followed and quantified. Bars, 5 µm.