ALG-4 is the gene responsible for upregulating the mRNA of the Fas-ligand. The
dominant-negative clone that was isolated diminished the transcriptional activation
of the Fas-ligand promoter. This effect seems to be accomplished by inhibiting (or
activating, in the case of wild-type ALG-4) the inducible transcription factor NF-κB,
which is a known transcriptional regulator of the Fas ligand promoter (Lacana and
D’Adamio, 1999). The molecular cause of this effect is currently unknown.
Selection strategies can also be applied to dominant activities in apoptosis. In view
of the observation that most apoptosis repressors are activated upon overexpression,
several laboratories have performed experiments solely to isolate inhibitors of
apoptosis. To this end, cDNA expression libraries were transfected into cells, a
proapoptotic stimulus was applied, and the cells surviving this signal were selected.
After several rounds of plasmid extraction from surviving cells and amplification in
bacteria, genes could be isolated that repress TNF-mediated cell death. Two of these
genes turned out to be the known apoptosis inhibitors Bcl-2 and Bcl-X (Jäättelä et
al., 1995). In a similar experiment, researchers tried to clone genes that impaired
amyloid-β-induced apoptosis and discovered a secreted polypeptide, called
‘humanin’. Interestingly, this gene seems to be specific in its cell-death repression for
β-amyloid or other Alzheimer’s disease genes (such as presenilins), since cell death
caused by etoposide, huntingtin, and SOD mutants were not affected by humanin
(Hashimoto et al., 2001). A functional explanation for this effect is still lacking, but
its investigation is certainly underway.
Other investigators discovered TIAF1 in a functional expression assay. TIAF is an
apparently nuclear protein that features a Wilms’ tumor protein family domain. Its
expression inhibits TGF-β-induced apoptosis but also cell death caused by the TNF
receptor and the overexpression of a variety of proapoptotic molecules. This is
evidently achieved by downregulating IκB, the specific inhibitor of NF-κB, an effect
that would result in the constitutive activation of its antiapoptotic effects (Chang et
al., 1998).
Gerry Nolan and coworkers used another selection strategy. Aiming at
pharmacologically relevant gene targets in apoptosis, they did not concentrate on
cDNAs that express endogenous proteins; instead, they used a retroviral gene library
encoding an artificial permutation of random short peptides (Xu et al., 2001). A
peptide that inhibited taxol-induced cell death was isolated and shown to upregulate
the multiple-drug resistance transporter (ABCB1). This artificial construct was used
to isolate interacting proteins that might explain the peptide’s antiapoptotic activity.
Two subunits of the proteasome were determined, suggesting that the observed effect
is regulated by this protein degradation complex.
In summary, a number of different stimuli have been applied to apoptosis
induction to select genes or genetic elements (both antisense and fragments of
cDNAs) that repress apoptosis. The information that could be inferred from these
results is at the moment too complex to draw a coherent picture. Many approaches
have yielded only individual genes that still require integration into a larger network
of interacting proteins for apoptosis signaling. However, some systems, such as the
one established by Adi Kimchi, which, it should be noted, is focused on only one
206 GENETICS OF APOPTOSIS