The 'pause' unpacked

Abstract

a mutation disrupts cellular function. Karras and colleagues found that some mutant FANCA proteins show greater physical association with HSP90 than others, and that, in general, these mutant proteins were associated with less-severe disruption of cellular function. Inhibiting HSP90 increased the susceptibility to mitomycin C of cells producing these mutant FANCA proteins, confirming a long-hypoth-esized mechanism of HSP90 action. Accord-ing to this hypothesis, the chaperone stabilizes otherwise defective proteins, allowing them to fold into a somewhat normal formation that ena-bles them to partially function (Fig. 1a). Notably, the effect of HSP90 inhibition on the sensitivity to mitomycin C could be mimicked by increased temperature — an environmental stressor that impairs HSP90 function and that might be expe-rienced if a patient has a fever. Thus, the ability of FANCA to interact with HSP90 can influ-ence the course of disease, and depends on both genetic and environmental context. In the second study, Hummel et al. 4 revealed a role for HSP90 in modulating the effects of endogenous retroviruses (ERVs) — DNA sequences derived from viruses that inserted copies of themselves into host DNA. ERVs can increase in number in a host genome either by reinfection or by replicating in cells that form sperm or eggs, and currently make up about 5–10% of the human and mouse genomes 9 . The presence or absence of some ERV insertions differs between people and between mice. Transcription of ERVs can stimulate the activity of adjacent genes. Hummel et al. profiled gene-expression patterns in three mouse cell types after HSP90 inhibition, which revealed that HSP90 counteracts this activating tendency. The researchers demonstrated that HSP90 interacts with the protein KAP1, which directs the deposition of repressive molecular modifications on ERV DNA to prevent tran-scription. HSP90 inhibition prevents KAP1-mediated repression of ERVs. The authors therefore propose that HSP90 activity enables the accumulation of different ERV insertions in different individuals. These insertion differ-ences would be inconsequential under normal conditions, but could lead to diverse outcomes in times of stress (Fig. 1b). If this holds true in humans, it would be predicted that clinical presentation of disease traits could be dramati-cally affected by ERV-insertion differences between individuals. These two studies add to a growing literature that establishes the abundance of 'cryptic' genetic variation, which has no effect under normal conditions, but lurks in populations until other mutations or environmental per-turbations reveal it 10