Zebrafish HSF4: a novel protein that shares features of both HSF1 and HSF4 of mammals

Heat-shock proteins (hsps) have important roles in the development of the eye lens. We previously demonstrated that knockdown of hsp70 gene expression using morpholino antisense technology resulted in an altered lens phenotype in zebrafish embryos. A less severe phenotype was seen with knockdown of heat-shock factor 1 (HSF1), suggesting that, while it likely plays a role in hsp70 regulation during lens formation, other regulatory factors are also involved. Heat-shock factor 4 plays an important role in mammalian lens development, and an expressed sequence tag encoding zebrafish HSF4 has been identified. The deduced amino acid sequence shares structural similarities with mammalian HSF4 including the lack of an HR-C domain. However, the HR-C domain is absent due to a severe C-terminal truncation within zebrafish HSF4 (zHSF4) relative to the mammalian protein. Surprisingly, the amino acid composition of the zHSF4 DNA binding domain shares a greater degree of identity with HSF1 proteins than it does with mammalian HSF4 proteins. Consistent with this, the binding affinity of in vitro synthesized zHSF4 for discontinuous heat-shock response element sequences is more limited, similar to what has been previously observed for HSF1 proteins. Hsf4 mRNA is expressed in zebrafish adult eye tissue but is only observed in developing embryonic tissue at 60 h post-fertilization or later. This, together with the lack of an observable phenotype following morpholino-based antisense knockdown of hsf4, suggests that zHSF4 is unlikely to play a role in regulating early embryonic lens development.     

Expression levels of heat shock factors are not functionally coupled to the rate of expression of heat shock genes

The expression patterns of two mammalian heat shock factors (HSFs) were analysed in cell systems known to reflect an altered heat shock response. For being able to discriminate between the two closely related factors HSF 1 and HSF 2, specific cDNA sequences were cloned and used to generate antisense RNAs as hybridization probes. In general, in various cell lines expression of the two heat shock factors was clearly different. These expression patterns of the HSF genes were not influenced by retinoic acid-induced differentiation of human NT2 and mouse F9 teratocarcinoma cells. Generally, HSF 2 expression was extremely low, whereas the significantly higher expression of HSF 1 revealed cell specific differences. The highest expression rates of both HSFs were observed in 293 cells. To examine whether these high levels are involved in the constitutive expression of heat shock genes in these cells, we analysed the binding pattern of 293 cell proteins to the heat shock elements (HSEs). As with other cells, HSE-binding activity in 293 cells was only observed after heat shock treatment. This points to an HSE-independent way for high level expression of heat shock genes in these cells.