Live-cell imaging in Caenorhabditis elegans reveals the distinct roles of dynamin self-assembly and guanosine triphosphate hydrolysis in the removal of apoptotic cells

Abstract

Dynamins are large GTPases that oligomerize along membranes. Dynamin's membrane fission activity is believed to underlie many of its physiological functions in membrane trafficking. Previously, we reported that DYN-1 (Caenorhabditis elegans dynamin) drove the engulfment and degradation of apoptotic cells through promoting the recruitment and fusion of intracellular vesicles to phagocytic cups and phagosomes, an activity distinct from dynamin's well-known membrane fission activity. Here, we have detected the oligomerization of DYN-1 in living C. elegans embryos and identified DYN-1 mutations that abolish DYN-1's oligomerization or GTPase activities. Specifically, abolishing self-assembly destroys DYN-1's association with the surfaces of extending pseudopods and maturing phagosomes, whereas inactivating guanosine triphosphate (GTP) binding blocks the dissociation of DYN-1 from these membranes. Abolishing the self-assembly or GTPase activities of DYN-1 leads to common as well as differential phagosomal maturation defects. Whereas both types of mutations cause delays in the transient enrichment of the RAB-5 GTPase to phagosomal surfaces, only the self-assembly mutation but not GTP binding mutation causes failure in recruiting the RAB-7 GTPase to phagosomal surfaces. We propose that during cell corpse removal, dynamin's self-assembly and GTP hydrolysis activities establish a precise dynamic control of DYN-1's transient association to its target membranes and that this control mechanism underlies the dynamic recruitment of downstream effectors to target membranes.

Figure 1.

The GTPase and the Middle domains of dynamins in different organisms are highly conserved. (A) Domain structure of DYN-1. The locations of mutations identified from dyn-1 mutant alleles are indicated. (B and C) GTPase (B) and Middle (C) domain sequence alignment. The residues identical or similar among at least three proteins are shaded in black or gray, respectively. The residues mutated in the dyn-1 mutant alleles are framed in open red boxes, with the corresponding allele numbers and the amino acid changes indicated above. (B) An open blue box indicates the position of K46A. α-Helices and β-strands identified in DynA are underlined with gray bars and arrows, respectively. G1–G4, highly conserved motifs required for GTP binding and hydrolysis. SWI and II, switch regions I and II. This figure was adapted from figure S2 of Yu et al. (2006). (C) An open blue box indicates the position of mutation R402A.

Figure 2.

The cell corpse removal activities of truncated and mutated DYN-1. The number of cell corpses in late fourfold stage dyn-1(n4039) homozygous embryos [progeny of dyn-1(n4039)/+ mothers] that expressed dyn-1 cDNAs bearing domain deletions or point mutations under P_ced-1 control are indicated by the diagrams. Data are presented as mean ± SD. For each sample, at least 10 embryos were scored. %Rescuing activity = 100% × (24.4 − mean no. of cell corpses in embryos expressing a particular transgene)/(24.4–0.3). Percentage of rescuing activity is defined as 0 if the number is negative.

Figure 3.

Basal and assembly-stimulated GTPase activities of wild-type and mutant DYN-1. (A) The [³⁵S]GTPγS binding activities of wild-type and mutant DYN-1 are shown in the bar graph (top). The numbers on top of each bar represent means. Error bars represent SD of three independent experiments. Bottom, Coomassie Blue staining pattern of 10% SDS-PAGE containing 1 μg of each protein sample used in the GTP binding assay. (B) Time courses demonstrating the rate of GTP hydrolysis of rat dynamin and C. elegans DYN-1 in the absence or presence of lipid. (C) The GTPase activity of wild-type and mutant DYN-1 in the presence or absence of lipid. GTPase activity was represented as the mean percentage of GTP hydrolyzed measured 30 min after the reactions were started. Error bars represent SD of four independent experiments.

Figure 4.

The I401F mutation abolishes DYN-1's ability to assemble along lipid surface. (A–C) Cryo-TEM images of wild-type and mutant DYN-1 incubated with PS liposomes. Wild-type DYN-1 and DYN-1(G40E) both bind to the liposomes and form a protein-coated lipid tubes with defined diameter of ∼50 nm. Arrows in A indicate the helical assembly of DYN-1 on lipid tubes. DYN-1(I401F) fails to assemble on the liposome surface (C). (D and E) Cryo-TEM images of DYN-1-lipid (D) and DYN-1(G40E)-lipid samples (E) 7 and 12 s after the addition of 1 mM GTP to the grid containing the dynamin-lipid tubes, respectively. Black arrowheads indicate bare liposomes without DYN-1 coating. White arrowheads in D indicate twisted, constricted dynamin-lipid complexes, which are intermediates of the DYN-1 dissociation process. (F) Cryo-EM image of liposomes (black arrowheads) in the absence of DYN-1 or GTP.

Figure 5.

The I401F mutation impairs DYN-1 oligomerization in C. elegans. (A) Diagram illustrating the principle of the BiFC assay. DYN-1 is fused with each one of the two fragments of Venus, VN173 (Venus 1-173 aa) or VC155 (Venus 155-238 aa) and expressed in the target cells. The appearance of a yellow fluorescence signal indicates the detectable interaction between DYN-1 monomers. (B) Fluorescence (a–f) and corresponding DIC (g–l) images of fourfold stage wild-type or dyn-1(en9) homozygous embryos expressing different BiFC constructs. Each fluorescence image represents two-dimensional projection of five consecutive 1-μm z-sections. Arrowheads indicate the fluorescence signals detected along the apical surface of the intestinal cells. Arrows indicate the fluorescent puncta observed throughout the worm body and considered nonspecific aggregation products. Bars, 10 μm. (C) Quantification of BiFC-positive intestinal tracks scored from embryos expressing different BiFC constructs. The percentage of embryos with fluorescent signal labeling the full-length (from the end of the second pharyngeal bulb to anus) or part of the intestinal track or with no signal along the entire intestinal track are represented in blue, yellow, or red, respectively. n, number of embryos scored. (D) Fluorescence (a and c) and corresponding DIC (b and d) images of L1 wild-type larvae expressing BiFC constructs. Arrowheads indicate the fluorescence signals detected along the apical surface of the intestinal cells. Arrows indicate the fluorescent puncta considered nonspecific aggregation products. Bars, 10 μm.

Figure 6.

The middle domain mutations suppress the dominant-negative effect of the GTPase mutations of DYN-1 in cell corpse removal. (A) DIC (a–c) and the corresponding fluorescence (e–g) images of wild-type fourfold stage embryos overexpressing dyn-1 cDNA bearing different mutations as C-terminal GFP fusions and under the control of heat-shock promoters. (d) DIC image of a wild-type embryo carrying no transgene. (e–g) images were captured with the same exposure time. All embryos have been subjected to heat-shock treatment. Arrows indicate cell corpses. Bars, 10 μm. (B) Number of cell corpses in wild-type fourfold stage embryos overexpressing dyn-1 cDNA bearing different mutations as C-terminal GFP fusions and under the control of heat-shock promoters. For each transgene (except K46A, I401F, and R402A), the results from two independent transgenic lines are presented. Data are presented as mean ± SD, and n is number of embryos scored. (C) Western blots of C. elegans protein extracts probed with polyclonal antibodies against DYN-1 (top) and β-tubulin (bottom). The predicted sizes of endogenous DYN-1 and DYN-1::GFP are 93 and 130 kDa, respectively. All samples have been treated with heat-shock and harvested 3 h later.

Figure 7.

DYN-1(I401F) fails to localize to target membranes, whereas DYN-1(G40E) fails to dissociate from membranes. (A–C) DIC and fluorescence images of wild-type or dyn-1 mutant embryos that express DYN-1::GFP, DYN-1(G40E)::GFP, or DYN-1(I401F)::GFP under P_ced-1. The genotype of each embryo is labeled. Arrows indicate GFP signals observed on the apical surface of the intestinal cells. Arrowheads indicate intestinal cells. Bars, 10 μm. The histogram beneath each GFP image depicts the GFP signal intensity observed within the framed region. (D) Fluorescence (a) and DIC (b) images of a wild-type embryo expressing P_ced-1 gfp::rab-7 at ∼330 min postfirst-embryonic division. Phagosomes C1, C2, and C3 are labeled with enriched GFP::RAB-7 (arrows). The boundary of three ventral hypodermal cells (identities labeled) that have engulfed apoptotic cells C1, C2, and C3 are traced with yellow lines. Yellow arrowheads indicate two hypodermal cells that fuse at the ventral midline of the embryo. This fusion event is used as a landmark of embryonic development with which the timing of engulfment is compared with in Figure 10. Anterior is to the top. Ventral faces readers. Bars, 10 μm. (E) Temporal order and duration of the enrichment of DYN-1, RAB-5, and RAB-7 on phagosomal surfaces. Data represent mean values obtained from time-lapse recording monitoring through the engulfment and degradation processes of multiple C1, C2, and C3 cell corpses using GFP fused reporters in wild type embryos. 0 min represents the time point when budding pseudopods are first detected. (F) Time-lapse images of wild-type embryos that express DYN-1::GFP (a–f), DYN-1(I401F)::GFP (g–l), or DYN-1(G40E) (m–t) under P_ced-1. Arrows indicate C3 phagosomes. 0 min is the time point when engulfment is just completed. Bars, 5 μm. (G) The frequencies of phagosomes C1, C2, and C3 labeled with enriched DYN-1::GFP reporters in a period >30 min, starting when engulfment is complete, and the average duration of GFP on phagosomal surfaces. In this series of time-lapse experiments, our recording terminated at ∼40 min after the completion of engulfment. n, number of C1, C2, and C3 phagosomes recorded.

Figure 8.

Both the G40E and I401F mutations severely impair the degradation of apoptotic cells. (A) dyn-I (en9) and dyn-I (en40) homozygous embryos carry persistent cell corpses. (a–c) DIC images of fourfold stage wild-type embryos and homozygous dyn-1(en9) (m⁻z⁻) and dyn-I (en40) (m⁻z⁻) embryos generated by homozygous mother rescued by wild-type dyn-I genomic fragment as extrachromosomal array. Arrowheads indicate cell corpses. Scale bars: 10 μm. (B) The number of cell corpses (mean ± sd) and percentage embryonic lethality observed from homozygous (m⁺z⁻) mutant embryos generated by heterozygous mothers and from (m⁻z⁻) mutant embryos generated by homozygous mother rescued by wild-type dyn-I genomic fragment as extrachromosomal array. (C) Time-lapse images of embryos expressing P_ced-1 ced-l C::gfp that demonstrate the duration of phagosomes in different genetic backgrounds. 0 min: the time point when engulfment is just complete and a phagosome (arrows) is first detectable as a dark hole inside the engulfing cell. Scale bars are 2 μm. (a–e) images of a C3 phagosome in a wild-type embryo. (f–j) images of Cl phagosome in dyn-l(en9) embryos. (k–o) images of Cl phagosome in dyn-l (en40) embryos. (D) Histogram distribution of phagosome durations in embryos determined using CED-IC::GFP as a reporter. Duration is measured from the formation of a nascent phagosome until its degradation. n, number of phagosomes C1, C2 and C3 measured using time-lapse recording.

Figure 9.

G40E and I401F mutation cause differential effects in recruiting tethering factors to phagosomes. (A) Bar graph summarizing our time-lapse recording results showing the correlation between the defects in RAB-7 recruitment and phagosome maturation. 0 min, the time point when engulfment is just completed. Gray bars depict the duration of phagosomes. The reduction of bar width corresponds to the time when a phagosome reduces in size. Gray bars with open right ends represent phagosomes that did not change size within the recording period of 120 min. The time duration that GFP::RAB-7 is enriched on the same phagosome is labeled by green bars that superimpose on part of the gray bars. A phagosome is considered RAB-7(+) when GFP::RAB-7 signal intensity on its surface reaches 1.2-fold of that detected in the host cell cytosol. n, total number of phagosomes C1, C2, and C3 scored. (B–D) Time-lapse images of one C3 (B) and multiple C2 (C–D) phagosomes in embryos expressing P_ced-1gfp::rab-7. 0 min, the time point when engulfment is just completed and the C2 phagosome (arrows) is first detectable as a dark hole inside its engulfing cell. Bars, 2 μm. (B) Images of a C3 phagosome in a wild-type embryo. (C) Time-lapse images of three different phagosomes (a–e, f–k, and l–q) in dyn-1(en9) embryos. (D) Time-lapse images of three different phagosomes (a–e, f–k, and l–q) in dyn-1(en40) embryos. (E) Time-lapse images of embryos expressing P_ced-1gfp::rab-5 for measuring the recruitment of RAB-5 to phagosomal surfaces. 0 min, the time point when engulfment is just completed and the C1 phagosome (arrows) is first detectable as a dark hole inside the engulfing cell. Bars, 2 μm. (a–s) Images of C1 phagosomes in wild-type (a–d), dyn-1(en9) (e–l), and dyn-1(en40) (m–s) embryos. (t) Histogram indicating the distribution of the time point when RAB-5 is first recruited to phagosomal surfaces measured using GFP::RAB-5 as a reporter, which is defined as the first time point when the GFP signal intensity on phagosomal surfaces is at least 1.2-fold as high as that in the cytosol of the cell that hosts the phagosome. 0 min is the time point when engulfment is just completed. n, number of phagosomes C1, C2, and C3 measured using time-lapse recording.

Figure 10.

Both G40E and I401F mutations impair the engulfment of apoptotic cells. (A) Time-lapse images depicting the engulfment process of C3 and C1 (arrows) in wild-type and dyn-1 mutant embryos. Embryos all express P_ced-1 gfp::rab-5. 0 min, time point when the two ventral hypodermal cells ABplaapppp and ABpraapppp contact each other at the embryonic midline. Bars, 2 μm. Extended pseudopods are marked by arrowheads. (a–e) Engulfment process of cell corpse C3 in a wild-type embryo. (f–o) Example of an abortive attempt of pseudopod extension (arrowhead) in a dyn-1(en9) embryo. (f–j) DIC images showing the position of the cell corpse C3. (k–o) GFP images correspond to f–j, respectively, which show normally initiated (n, arrowhead) yet failed pseudopod extension from the supposed-to-be engulfing cell ABplaapppp during the recording period. (p–v) Example of prolonged pseudopod extension around the C1 cell corpse in a dyn-1(en40) embryo. (w) A diagram depicting the physical contact between the two extending hypodermal cells at the ventral midline at the 0-min time point, which is the reference point for measuring the timing of the initiation of pseudopod extension. (B and C) The frequency of each type of engulfment defects observed from dyn-1(en9) or dyn-1(en40) embryos measured using GFP::RAB-5 (B) or CED-1C::GFP (C) as a reporter, respectively. The initiation time of pseudopod extension was measured using the time point that ABplaapppp and ABpraapppp made contact as the 0-min time point. The time between the initiation of pseudopod extension and the closure of a phagocytic cup is defined as the duration of pseudopod extension. All data were obtained from observations of the engulfment of C1, C2 and C3. n, total number of engulfment events scored.

Figure 11.

Models depicting the molecular mechanisms that regulate the function of DYN-1 during the removal of apoptotic cells. (A) Table comparing the biochemical properties and functional defects of the mutant DYN-1 defective in self-assembly or GTP binding. NA, not applicable. (B) Model proposing an autoregulatory loop that regulates the temporal association of DYN-1 to its target membrane. In response to an upstream recruitment signal, DYN-1 molecules undergo self-assembly, an event that enables DYN-1 to associate with target membrane. Meanwhile, self-assembly enhances the GTP hydrolysis activity of DYN-1, which results in the disassembly of DYN-1 oligomers and the consequential dissociation of DYN-1 from its target membrane. In the continuing presence of the upstream signal, DYN-1 monomers will again assemble along the membrane surface. As a consequence, the apparently opposite effects of self-assembly and GTP hydrolysis reach a balance and together, they maintain a steady-state level of DYN-1 on the target membrane. Once the upstream signal disappears, the high-level GTPase activity of DYN-1 oligomers allows rapid dissociation of DYN-1 from the target membrane.