ABSTRACT
Heat shock protein 90 (Hsp90) is a molecular chaperone required for folding, maturation and activity of a select set of proteins important for signal transduction and development. The last decade witnessed this chaperone come into the limelight after the observation that Hsp90 inhibition in developing fruit fl ies results in a large phenotypic variation (Rutherford and Lindquist 1998). This has been interpreted to mean that genetic variation existing within a population does not translate into phenotypic variation under normal circumstances due to ‘buffering’ activity of Hsp90. A recent study has questioned this interpretation and it is proposed that Hsp90 may in fact actively suppress the generation of genetic variation rather than or in addition to merely buffering the phenotypic consequences (Specchia et al. 2010). Specifi cally it was found that fruit fl ies with an impaired Hsp90 activity exhibit an enhanced mobilization of transposons in the germ-line leading to an increased mutation rate. When a phenotypically uniform laboratory population was exposed to sub-lethal doses of Hsp90 inhibitor during development, a variety of morphological abnormalities were expressed. This was found to be the case even when Hsp90 loss-of-function alleles were introduced in an otherwise wild-type genetic background. These initial observations made by Susan Lindquist and colleagues in the fruit fl y Drosophila melanogaster were later on extended to be valid in plants, fungi, slime moulds and fi sh (reviewed in Rutherford et al. 2007). The spectrum of morphological traits in fl ies was shown to be strain-background dependent. Moreover, continued phenotypic selection for an abnormal morphology resulted in increased frequency of that morphology over a few generations, and in expression of the trait even when Hsp90 function was restored (Sangster et al. 2004). This indicates an existence of a genetic component to the Hsp90-induced morphological alterations. Taken together, these observations led to the belief that under normal conditions Hsp90 minimizes phenotypic consequences of random mutations arising in a population, thus exhibiting a ‘canalized’ uniform phenotype despite genetic differences. This in turn means that genetic variation can accumulate over long periods within an interbreeding group of organisms with little deleterious fi tness consequences, if the environmental conditions remain constant. However, when such a genetically diverse population is exposed to stress akin to experimental inhibition of Hsp90, the genetic variation could then manifest itself as a phenotypically heterogeneous population. This can serve as a substrate upon which natural selection can act, resulting in propagation of those phenotypic variants which survive the stressful environment. Hsp90 was thus termed as a ‘capacitor’ for morphological evolution with an analogy to electrical capacitors that store charge only to release it when required. This line of thinking was expanded in its scope later by the observation that Hsp90 inhibition causes phenotypic variation even in highly inbred, genetically uniform population of fruit fl ies (Sollars et al. 2003). Thus Hsp90 is thought to buffer genetic as well as heritable epigenetic variation. The Hsp90 capacitor hypothesis brings to the fore and provides a mechanism for Waddington’s observations of genetic assimilation (Waddington 1959). Taking as an example of evolution of callosities in the ostrich footpads, Waddington proposed that cryptic variation, initially unmasked by an environmental agent (e.g. stress), could provide a driving force for new morphological traits. Hsp90 may have been an important mechanism for genetic assimilation during evolution. Central to the Hsp90 capacitor hypothesis is an ill-supported assumption that phenotypic variation observed upon Hsp90 inhibition is due to pre-existing genetic and/or epigenetic variation. A recent study (Specchia et al. 2010) shows that this assumption need not be true. Working with the fruit fly.