After proteotoxic stress, interphase (differentiated) cells may either initiate apoptosis (if this program is available in the cells) or they may undergo necrosis. Hsp, by their chaperone activity, may inhibit or repair the damage and thus take away the trigger for activating cell death. Similarly, storage of poly-Q-diseased proteins may prevent (mini)aggregates from disturbing cellular functions; thus cells maintain their biochemical activity and stay alive (Arrasate et al. 2004). In how far Hsp play a role in this storage mechanism remains to be elucidated, but, for example, aggresome formation clearly requires an intact cytoskeleton (Johnston et al. 2002) and thus stabilization of the cytoskeletal by, for instance,
Hsp27 may certainly be important for storage under conditions of heat stress. For dividing cells, life or death ultimately depends on the ability to move through the cell cycle and divide. If checkpoints are not functional and/or apoptosis is not initiated, cells may attempt to progress through S-phase, but nuclear protein damage may obstruct DNA synthesis, leading to replication errors and lethal chromatid aberrations (Dewey et al. 1990) and subsequent death by apoptosis or necrosis. When cells enter mitosis with unresolved protein damage (or when heated in mitosis), cells may undergo a mitotic catastrophe or aberrant divisions (Hut et al. 2005; Nakahata et al. 2002) that will mostly result in secondary apoptosis or necrosis of the daughter cells (Fig. 3). There are clear indications that the accumulation of heat-unfolded proteins at the centrosome are linked to these aberrant mitoses (Hut et al. 2005; Nakahata et al. 2002), implying that storage of heat damage at centrosomes must be restored before proper cell divisions can occur.
Besides being involved in heat damage repair, Hsp70 (as well as Hsp27 and Hsp90) have been implicated in the process of apoptosis execution. Hsp70 was suggested to inhibit JNK activation after heat shock (Gabai et al. 1997), but these effects may not be uncoupled from the Hsp70 chaperone activity either upstream or at the level of JNK (Mosser et al. 2000). However, based on in
Fig. 3 Hypothetical model for causes and expression of heat-induced cell lethality and the protective effects of Hsp. See text for further explanation
Fig. 3 Hypothetical model for causes and expression of heat-induced cell lethality and the protective effects of Hsp. See text for further explanation vitro observations, Hsp70 was also suggested to be able to interfere with the formation of the apoptosome (Beere et al. 2000; Saleh et al. 2000), one of the cell death execution caspase complexes that is induced after the leakage of cytochrome-c from the mitochondria. This would suggest that Hsp70 could act downstream in the apoptosis pathway and thus could be cytoprotective, irrespective of the type of cell death trigger. This has not yet been confirmed in living cells and recent in vitro data suggest that the presumed Hsp70 effects on apoptosome formation may have been due to salt artifacts (Steel et al. 2004). Moreover, if true, these or other presumed specific interferences of Hsp70 with the apoptosis pathway should also have an impact when nonproteotoxic stresses such as FAS-ligand or radiation are used.
When investigating FAS-mediated apoptosis, Jaattela's group had already repeatedly shown that Hsp70 did not prevent apoptosome formation in cells, and in this case Hsp70 acts even further downstream in the apoptotic cascade (Jaattela et al. 1998; Nylandsted et al. 2004). In fact, their studies are among the few that do indicate that Hsp70 has anti-cell death functions that may be separated from its chaperone activity acting upstream in the apoptotic pathway to prevent proteotoxic damage. Nevertheless, here also a link with protein quality control is suggested by their findings that Hsp70-mediated protection against FAS-induced cell death seems to occur at the level of lysosomes (Nylandsted et al. 2004). Also, it was recently shown that human Fas associated factor 1 (hFAFl) inhibits Hsp70 chaperone activity in mammalian cells (Kim et al. 2004), suggesting that stimulating the FAS pathway may inhibit physiological Hsc70/Hsp70 functions (including those involved in lysosomal control) and thus explain why overexpression of Hsp70 is protective in the case of FAS-induced apoptosis.
Maybe the most compelling data to show that Hsp70 (and perhaps other Hsp) may have no general inhibitory function in the pathway of apoptosis execution comes from studies on the interaction between hyperthermia and radiation, longbefore specific studies on apoptosis became popular. These data showed that the temporal and physiological upregulation of the cohort of all heat-inducible proteins sufficient to provide cells with a thermotolerant state does not alter the cellular radiosensitivity (Dikomey and Jung 1992; Hartson-Eaton et al. 1984; Haveman et al. 1987; Jorritsma et al. 1986; Kampinga et al. 1997; Mivechi and Li 1987; Raaphorst and Azzam 1983, and many more). Also, when Hsp70 or Hsp27 are transiently upregulated using inducible systems, no effect of radiosensitivity was found (unpublished data). Inversely, RNAi-mediated transient downregulation of Hsp70 was shown not to affect radiation-induced apoptosis (Ekedahl et al. 2003). Finally, although some stable cell lines overexpressing individual Hsp may become radiation resistant (Gehrmann et al. 2005; Lee et al. 2001), in other cell systems, Hsp overexpression does not protect (Fortin et al. 2000; Kampinga et al. 1997; Stege et al. 1995) or even sensitizes for radiation-induced cell death (Liu et al. 2003). As mentioned before, clonal overexpression of Hsp may require adaptation of cellular gene and protein expression profiles. Depending on cell type-specific adaptations, indirect effects rather than functions of Hsp per se may be causally related to cellular phenotypes such as an altered radiation sensitivity.
The presumed function ofHsp70 in apoptosis has often been used to explain why it is repeatedly found to be associated with tumorigenesis. For example, elevated Hsp70 levels have been found in almost all oncogene-transformed cell lines and tumor cells (Li et al. 1995). Hsp70 is highly expressed in many tumors (Jaattela, 1999) and the oncogenic potential of Hsp70 has been demonstrated (Jaattela 1995; Volloch and Sherman 1999). If not through direct interference with the apoptotic pathway, how can one explain these associations? First, cellular transformation may require the Hsp70 machine, as suggested by findings that several of the transforming viral T antigens contain regions of significant homology with the conserved J domain of the DnaJ co-chaperones (Kelley and Georgopoulos 1997). Alternatively, Hsp70 can regulate the stability and function of the tumor suppressor p53 (King et al. 2001; Zylicz et al. 2001). A third possibility, attractive but even more speculative, is that Hsp70 may act as a buffer for the negative consequences of aberrant protein expression due to the genomic instability in tumor cells. As such, its elevated expression is required for tumor cell survival, without Hsp70 being in the transformation process as such.
Was this article helpful?