cathepsin D, plays a role in apoptosis induction and even mediates Fas- and TNF-
induced apoptosis (Deiss et al., 1996). Cathepsin D was one of the first proteases,
besides the established cysteine proteases of the caspase family, found to contribute
to apoptosis induction (see Chapter 13). This was even more astonishing as lysosomal
proteases, unlike caspases, are known to have a wide range of substrates. How, then,
could it induce such a defined cellular outcome as apoptosis? Subsequent studies
revealed that cathepsin D accumulates in dying cells and might therefore change its
subcellular localization and hence its substrates, a fact that could explain the
involvement of this house-keeping gene in the induction of apoptosis (Deiss et al.,
1996).
The aforementioned DAP-kinase turned out to be especially interesting, as it is
the founding member of a small kinase family that regulates cell death (Kogel et al.,
2001). This is a rare example in which a kinase is involved in apoptosis induction;
usually, proapoptotic signals are mediated by direct protein-protein contact rather
than by phosphorylation (see Chapter 5). The DAP kinase has been shown to be a
tumor-suppressor gene (Inbal et al., 1997): it is downregulated in highly metastasis-
forming cells, and, upon reconstitution, it reduces the ability of cells to establish
metastasis. This effect could be mediated by a signaling pathway involving the tumor
suppressors p19ARF and p53 (Raveh et al., 2001). The tumor-suppressor activity of
the DAP kinase is a case in point of how apoptosis induction protects the organism
against malignant cells.
One gene from this screen, DAP-3, is a supposed nucleotide-binding protein, as
it harbors a ‘P-loop’ motif. This domain is necessary for its functioning since a point
mutation in this domain abrogated its apoptosis induction. While one report localized
this protein to mitochondria (Berger et al., 2000), another study has recently shown
that DAP-3 can bind to the TRAIL receptor on the plasma membrane (Miyazaki and
Reed, 2001). This is a TNFR family member that has attracted much attention, as
it induces apoptosis in a tumor-specific way (Walczak et al., 1999). DAP-3 also binds
to the adapter protein FADD and links it to the TRAIL receptor in a GTP-dependent
manner. Consequently, DAP-3 might be an assembling factor in the TRAIL receptor
complex.
DAP-5 was isolated as a cDNA fragment in a sense orientation that presumably
encodes a dominant-negative version of this gene. It is homologous to the translation
initiation factor eIF4GI but is missing its N-terminus, which otherwise associates
with the cap-binding protein eIF4A. Interestingly, eIF4GI is cleaved by caspase-3. It
is therefore possible that DAP-5, like the cleaved form of eIF4GI, works to initiate a
cap-independent translation during apoptosis induction. This special form of protein
synthesis can lead to the generation of both pro- and anti-apoptotic genes, such as
XIAP and Apaf-1, whose mRNAs have IRES (internal ribosome entry sites). These
results indicate another layer of cell-death regulation by the translation of specific
genes (Holcik et al., 2000).
An interesting aspect of this work was the finding that all genes except thioredoxin
can also elicit a proapoptotic signal upon overexpression. This and all other results
with DAP genes underscore the validity of this cloning approach. While an ordering
204 GENETICS OF APOPTOSIS