Abl kinase inhibits the engulfment of apoptotic [corrected] cells in Caenorhabditis elegans

Abstract

The engulfment of apoptotic cells is required for normal metazoan development and tissue remodeling. In Caenorhabditis elegans, two parallel and partially redundant conserved pathways act in cell-corpse engulfment. One pathway includes the adaptor protein CED-2 CrkII and the small GTPase CED-10 Rac, and acts to rearrange the cytoskeleton of the engulfing cell. The other pathway includes the receptor tyrosine kinase CED-1 and might recruit membranes to extend the surface of the engulfing cell. Although many components required for engulfment have been identified, little is known about inhibition of engulfment. The tyrosine kinase Abl regulates the actin cytoskeleton in mammals and Drosophila in multiple ways. For example, Abl inhibits cell migration via phosphorylation of CrkII. We tested whether ABL-1, the C. elegans ortholog of Abl, inhibits the CED-2 CrkII-dependent engulfment of apoptotic cells. Our genetic studies indicate that ABL-1 inhibits apoptotic cell engulfment, but not through CED-2 CrkII, and instead acts in parallel to the two known engulfment pathways. The CED-10 Rac pathway is also required for proper migration of the distal tip cells (DTCs) during the development of the C. elegans gonad. The loss of ABL-1 function partially restores normal DTC migration in the CED-10 Rac pathway mutants. We found that ABI-1 the C. elegans homolog of mammalian Abi (Abl interactor) proteins, is required for engulfment of apoptotic cells and proper DTC migration. Like Abl, Abi proteins are cytoskeletal regulators. ABI-1 acts in parallel to the two known engulfment pathways, likely downstream of ABL-1. ABL-1 and ABI-1 interact physically in vitro. We propose that ABL-1 opposes the engulfment of apoptotic cells by inhibiting ABI-1 via a pathway that is distinct from the two known engulfment pathways.

Figure 1. Molecular Pathways Required for the Engulfment of Apoptotic Cells

Proteins of the CED-10 Rac pathway are labeled in yellow. Proteins of the CED-1 pathway are labeled in green. C. elegans protein names are written above and their mammalian homologs are below. The dashed arrow from CED-6 to CED-10 indicates that CED-1, CED-6, and CED-7 might also signal through CED-10 [19]. The CED-1 pathway is required for engulfment only, whereas the CED-10 Rac pathway is required for both engulfment and DTC migration. PSR-1 might act upstream of CED-2 [21].

Figure 2. abl-1 Mutation Does Not Affect the Timing or Morphology of Cell Corpses

(A) The number and time of appearance of apoptotic cell corpses that occurred from 200–340 min after the first embryonic cell cleavage was recorded at 3-min intervals in wild-type and abl-1(n1963) animals using time-lapse DIC microscopy (see Materials and Methods). Mean numbers of corpses at each time point were calculated from three embryos for both wild-type and abl-1(n1963) animals. The curves are similar (p = 0.49). (B) The duration of cell-corpse appearance is similar in wild-type and abl-1(n1963) embryos. The percentage of cell corpses that lasted for a given period were recorded. The duration of appearance of all cell corpses recorded from three wild-type (n = 162 cell corpses) and three abl-1(n1963) (n = 171) embryos was analyzed. The curves are similar (p = 0.97). (C) The morphology of cell corpses in wild-type and abl-1(n1963) embryos are similar. Arrowheads, apoptotic corpses. Embryos were at a similar stage of development, approximately 300 min after the first cell corpse appeared.

Figure 3. abl-1 Suppresses the Cell-Killing Effect of Engulfment Pathway Genes

Animals doubly mutant for ced-3 and an engulfment gene with or without abl-1(n1963) were scored for the presence of extra cells in two tissues. (A) abl-1 suppresses the engulfment gene cell-killing effect in the pharynx. Extra cell nuclei in the pharynges of animals in the early third larval stage (early L3) were counted using DIC microscopy. All animals harbored the ced-3(n2427) mutation. Means and standard deviations are shown. Error bars, standard deviation. At least ten animals were scored for each genotype. *, p < 0.005; **, p < 0.001. (B) abl-1 suppresses the engulfment gene cell-killing effect in the ventral nerve cord. Extra GFP⁺ cells were counted in the ventral nerve cords of late fourth larval stage (L4) animals. All animals carried the ced-3(n2424) mutation and the insertion nIs96[lin-11::gfp], which labels VC neurons (P3–8.aap) and VC neuron-like cells (P1.aap, P2.aap, and P9–12.aap) that are normally fated to die. P1.aap is variably labeled by nIs96[lin-11::gfp] and was not scored. 100 animals of each genotype were scored. The number of extra VC-like cells were compared between strains containing ced-3(n2424) and a mutation in an engulfment gene with or without abl-1(n1963). p-Values for differences between strains were as follows: for ced-10(n1993)-containing strains, p = 2.2 × 10⁻¹⁶; for ced-1(n2091)-containing strains, p = 4.4 × 10⁻⁹; for ced-12(tp2)-containing strains, p = 3.3 × 10⁻⁸. For all other strains, p > 0.7.

Figure 4. abl-1 Mutation Suppresses the DTC Migration Defects of all CED-10 Rac Pathway Gene Mutations

The gonads of animals mutant for an engulfment gene with or without abl-1 mutation were observed and scored for morphology using DIC microscopy. Scoring was as described in Materials and Methods. Percentages of abnormal gonad arms are shown. At least 50 gonad arms were scored for the wild-type and abl-1(n1963) mutants. More than 100 gonad arms were scored for all other genotypes. All mutant abl-1 strains used the abl-1(n1963) allele except for ced-10(n3417), which used abl-1(ok171). Statistical analysis used Fisher's exact test, *, p < 1x10⁻⁵; **, p < 0.005.

Figure 5. abi-1 Is a C. elegans Abi Gene

(A) Alignment of three Homo sapiens Abi proteins (Abi-1, Abi-2, and Abi-3) with C. elegans ABI-1. The dashed line indicates the Wave-binding domain. The dotted line shows the proline-rich region. The continuous line shows the SH3 domain. The boxed residues in ABI-1 indicate the sequence removed by the tm494 deletion. Dark gray indicates identities, light gray similarities. (B) Similarity and identity indices between ABI-1 and each of the three human Abi proteins.

Figure 6. Loss of abi-1 Function Enhances the DTC Migration Defects of Engulfment Pathway Genes

The gonads of animals mutant for an engulfment gene with or without abl-1(n1963) were treated with abi-1 RNAi or a control RNAi, observed, and scored for morphology using DIC microscopy. Scoring was as described in Materials and Methods. Percentages of abnormal gonad arms are shown. At least 50 gonad arms were scored for the wild-type and abl-1(n1963) mutants. More than 100 gonad arms were scored for all other genotypes. Statistical analysis used Fisher's exact test, *, p < 0.04; **, p < 1 × 10⁻⁴.

Figure 7. ABL-1 and ABI-1 Interact In Vitro

GST, GST-ABI-1, and GST-ABIΔ were expressed, bound to glutathione beads, and incubated with in vitro translated Luciferase and portions of ABL-1: ABL-1(112–611) and ABL-1(606–1,224); numbers in parentheses represent amino acid positions. Beads were washed, diluted in sample buffer, and separated on SDS-PAGE. The gel was stained with Coomassie Blue, dried, and exposed. Left panel, autoradiograph. Input lanes contain 2% of each in vitro translation reaction. 25% of each in vitro translation reaction was used per binding experiment. GST, GST-ABI, and GST-ABIΔ lanes contain 25% of each binding experiment. Size markers in kilodaltons are shown on the left. Right panel, Coomassie Blue-stained gel showing relative amounts of GST, GST-ABI, and GST-ABIΔ used in each binding experiment. There is a nonspecific band that comigrates with GST-ABIΔ.

Figure 8. ABL-1 and ABI-1 Likely Function in Parallel to the CED-10 Rac and CED-1 Engulfment Pathways

We suggest that ABL-1 inhibits ABI-1, which acts to promote the engulfment of apoptotic cells. ABI-1 might signal either independently of the CED-10 Rac pathway (Arrow 1) or through CED-10 Rac, in parallel to the CED-10 GEF CED-5/CED-12 (Arrow 2). This model also applies to the roles of ABL-1 and ABI-1 in the regulation of DTC migration, but we do not show DTC migration in the figure, because the CED-1 pathway does not act in this process. Since ABL-1 and ABI-1 act in DTC migration, ABL-1/ABI-1 cannot act solely through the CED-1 pathway since the proteins of the CED-1 pathway have no role in DTC migration.