No abstract

Spermatocytes, the most sensitive male germ cells to heatinduced apoptosis, do not respond to hyperthermia by inducing heat shock proteins (HSPs), including HSP70i, which has been previously shown to confer resistance to apoptosis in somatic cells. To dissect the mechanism of heat-induced apoptosis and to determine if we could protect spermatocytes by expressing HSP70i, we engineered transgenic mice that express in spermatocytes constitutively active heat shock transcription factor (HSF)1. Such HSF1 expression did not lead to transcription of inducible Hsp70 genes, but instead induced caspase-dependent apoptosis that mimicked heat shock-induced death of spermatogenic cells. Both mitochondria-dependent and death receptor-dependent pathways appear to be involved in such HSF1-induced apoptosis: the levels of Bcl-2 family proteins became increased, p53 protein accumulated and expression levels of caspase-8 and deathreceptor-interacting proteins (including Fas-associated death domain protein and TNF receptor associated death domain protein) became elevated. Surprisingly, the constitutive spermatocyte-specific expression of HSP70i in doubletransgenic males did not protect against such HSF1-induced apoptosis.

The human HSPA2 gene, which belongs to the HSP70 family of heat shock genes, is a counterpart of rodent testis-specific HspA2 gene. Rodent genes are expressed mainly in pachytene spermatocytes, while transcripts of human HSPA2 gene have been detected in various normal somatic tissues, albeit translation of the messenger RNA into corresponding protein has not been yet unambiguously demonstrated, except for several cancer cell lines. The aim of our work, a first step in search for HspA2 function in cancer cells, was to establish its intracellular localization at physiological temperature and during heat shock. First, we used qRT-PCR and a highly specific antibody to select cell lines with the highest expression of the HspA2 protein, which turned out to be A549 and NCI-H1299 lines originating from non-small cell lung carcinoma (NSCLC). Significant expression of the HspA2 was also detected by immunohistochemistry in primary NSCLC specimens. Intracellular localization of the HspA2 was studied using both the specific anti-HspA2 polyclonal antibody and transfection of cells with fusion proteins HspA2-EGFP and mRFP-HspA2. We found that, at physiological temperature, the HspA2 was localized primarily in cytoplasm whereas, during heat shock, localization shifted to nucleus and nucleoli. Moreover, we demonstrate that in heat-shocked cells HspA2 accumulated in centrosomes. Our results suggest that the HspA2, like Hsp70 protein, can be involved in protecting nucleoli and centrosomes integrity in cancer cells subjected to heat shock and, possibly, other cellular stressors.

Expression of constitutively active heat shock transcription factor 1 (HSF1) in mouse spermatocytes induces apoptosis and leads to male infertility. We report here that prior to the onset of massive apoptosis caused by expression of active HSF1 in spermatocytes a marked reduction in spermatocyte-specific Hsp70.2 mRNA and protein levels occurs. In addition, HSP70.2 protein relocalizes from a predominant cytoplasmic to a nuclear position in developing spermatocytes that express active HSF1. Later in the developmental stages, cells undergoing HSF1-induced apoptosis essentially lack the HSP70.2 protein. The down-regulation of Hsp70.2 gene expression by HSF1 is paradoxical because HSF1 is the prototypical activator of HSP genes. Furthermore, HSF1-mediated repression neither involved a heat shock element (HSE)-like sequence adjacent to the Hsp70.2 gene nor were Hsp70.2 promoter sequences associated directly with HSF1. Interestingly, other spermatocyte- and spermatid-specific transcripts are also down-regulated in testes of transgenic mice expressing active HSF1, suggesting involvement of a putative HSF1-dependent block of development of spermatogenic cells. Importantly however, transcription of the Hsp70.2 gene is down-regulated in testes of wild-type mice subjected to a hyperthermia that induces transient activation of HSF1, indicating that the spermatocyte-specific activity of HSF1 might misdirect a network of transcription factors required for proper regulation of Hsp70.2.

To determine whether DNA transfer to mouse testes by in vivo electroporation could be useful method for studying regulatory elements of genes specifically active in spermatocytes first we compared the expression pattern of a construct containing the EGFP reporter gene ligated to a fragment of the heat shock testis-specific hst70 gene promoter, both in testis of transgenic mice and in testis electroporated in vivo. While in transgenic mice the EGFP was expressed in all seminiferous tubules in a cell- and stage-specific manner, in the testes electroporated in vivo only small fraction of cells expressed this marker protein. In order to make a quantitative comparison between the specificity of these two experimental systems we used several vectors containing the CAT gene ligated to fragments of the hst70 gene 5' upstream of DNA sequences which either promoted or did not activate expression of the reporter gene in the testes of transgenic mice. Also, as a reference opposite to spermatogenic cells we examined the expression pattern of the same set of vectors in the rat hepatoma FTO 2B cells. Although electroporated testes retain some spermatocyte-specific features such as the ability to repress promoters which do not contain regulatory elements responsible for testis-specific transcription, several important drawbacks of the method are evident. They include basal activity of constructs which are not transcribed in testes of transgenic mice and low overall transfection efficiency. This may hamper studies in which subtle changes in the expression pattern are under investigation. However, the in vivo electroporation of the testis can be useful for preliminary screening of constructs aimed to study in transgenic mice.