be involved in apoptosis. These results point to a much more complex regulation of
p53-induced apoptosis.
To sum up, the correlative strategy has already yielded a number of attractive genes
whose transcriptional expression correlates with and is sometimes even responsible
for the induction of apoptosis. The use of microarrays allows us to detect global
changes in the expression pattern. It is therefore likely that array techniques will speed
up the search for genes whose expression is changed during apoptosis induction.
Nevertheless, individual genes that are transcriptionally altered during cell death have
to be functionally verified for their role in apoptosis.
2.2
Selection strategiesApoptosis allows researchers to select cells that withstand a proapoptotic stimulus in
order to determine gene functions in cell death. As mammalian cells are permanently
diploid (or polyploid, like many tumor cells), recessive gene activities are difficult to
investigate. Transfection experiments to inactivate endogenous genes (and their
proteins) have been established to tackle this problem. They comprise both antisense
cDNAs against cellular genes and cDNA fragments encoding proteins that can act
in a dominant-negative manner. By isolating transfected plasmids from the cells that
survived a proapoptotic stimulus, one can determine genes that mediate a signal for
apoptosis (Figure 1b).
This approach was pioneered by Adi Kimchi and her group (Deiss and Kimchi,
1991). They transfected an antisense cDNA library into HeLa cells and isolated cells
that survived the exposure to interferon-γ (a strategy called ‘technical knockout’).
These workers expected to identify genes that mediate the signal for apoptosis and
to define rate-limiting steps in this process (Levy-Strumpf and Kimchi, 1998). Five
so-called DAP-genes (DAP: Death Associated Proteins) were determined that
conferred resistance to interferon-γ. They turned out to be completely divergent and,
partially, even localized to different compartments of the cell. Thioredoxin was the
first gene described (Deiss and Kimchi, 1991). It encodes a thiol protein that exerts
a potent reducing activity and might be able to keep in check the reactive oxygen
species that are frequently formed in apoptosis. It also promotes DNA binding of
transcription factors, such as NF-κB, AP-1, and p53, all of which have been
implicated in apoptosis induction. Moreover, thioredoxin has an inhibitory effect on
the TNF-induced activation of ASK1, possibly by direct binding to this MAPK kinase
kinase (Ichijo et al., 1997).
The other isolated genes from the screen include a calcium/calmodulin-regulated
kinase with a death domain (DAP-kinase), a nucleotide-binding protein (DAP-3), a
homolog of the translation initiation factor eIF4G (DAP-5), a small proline-rich
cytoplasmic protein (DAP-1), and the lysosomal protease cathepsin D (Levy-Strumpf
and Kimchi, 1998). The diversity of these genes makes it already evident how
disparate their contribution must be and consequently how the complex signal for
apoptosis is regulated. A surprising finding was that the lysosomal aspartyl protease,
CELL-CULTURE SYSTEMS IN APOPTOSIS 203