Programmed cell death and cell corpse removal are important in the development and homeostasis of tissues and organisms. While much is known about apoptotic corpse clearance, clearance of cells that die by non-apoptotic programs is not well understood. The death of the linker cell, a male-specific cell that leads the elongation of the developing gonad, is independent of the apoptotic caspase CED-3.
ced-3 mutations do not block linker cell death or engulfment, and the fraction of surviving cells in known linker cell death mutants is not enhanced by
ced-3 lesions. However, we found that in animals carrying mutations in
ced-3 or its regulator,
ced-4/Apaf1, engulfed linker cells are degraded inefficiently. Mutations in the apoptotic regulators
egl-1/BH3-only or
ced-9/Bcl2 do not affect linker cell clearance.
ced-3 is expressed in the linker cell as it dies but not during migration. While many linker cell-containing phagosomes are eventually degraded in
ced-3 mutants, clearance is delayed by hours or even days. From a genetic screen in which artificial insemination was used to propagate relevant mutants, we recovered several in which linker cell degradation is affected. We found that two small GTPases, RAB-35 and ARF-6, act antagonistically within the U.l/rp cells, which engulf the dying linker cell, to degrade the linker cell phagosome. Genetic interaction studies suggest that RAB-35 inhibits ARF-6, which, in turn, blocks linker cell degradation. We identified RME-4 and TBC-10 as the relevant guanine nucleotide exchange factor (GEF) and GTPase activating protein (GAP) for RAB-35, and EFA-6 and CNT-1 as the relevant GEF and GAP for ARF-6. Persistent linker cell-containing phagosomes in
rab-35 mutants appear to have degraded linker cell contents, and this degradation is delayed by mutations in
ced-3. Importantly, the frequency of persistent phagosomes is similar in
rab-35(0) and
rab-35(0);
ced-3(0) animals. Thus, our observations suggest that
ced-3 may function upstream of
rab-35, which may then promote linker cell phagosome maturation. We also show that
rab-35, but not
arf-6, is involved in phagosome maturation of apoptotic corpses in the embryo and germline, and is localized to extending pseudopods and to phagosomal surfaces. Rab35 has been implicated in phagosome maturation in mammals, although not in the context of cell death. Caspases have been suggested to promote degradation but not initiation of cell dismantling during spermatogenesis in Drosophila and during neurite pruning in vertebrates. We propose therefore that the cell degradation pathway we uncovered may be conserved.