Heat shock proteins (HSP) are highly conserved across eukaryotic and prokaryotic species. These proteins play a role in response to cellular stressors, protecting cells from damage and facilitating recovery. In tumor cells, HSPs can have cytoprotective effects and interfere with apoptotic cascades. This study was performed to assess the prognostic and predictive values of the gene expression of HSP family members in canine osteosarcoma (OS) and their potential for targeted therapy. Gene expressions for HSP were assessed using quantitative PCR (qPCR) on 58 snap-frozen primary canine OS tumors and related to clinic-pathological parameters. A significant increased expression of HSP60 was found in relation to shorter overall survival and an osteoblastic phenotype. Therefore, the function of HSP60 was investigated in more detail. Immunohistochemical analysis revealed heterogeneous staining for HSP60 in tumors. The highest immunoreactivity was found in tumors of short surviving dogs. Next HSP expression was shown in a variety of canine and human OS cell lines by qPCR and Western blot. In two highly metastatic cell lines HSP60 expression was silenced using siRNA resulting in decreased cell proliferation and induction of apoptosis in both cell lines. It is concluded that overexpression of HSP60 is associated with a poor prognosis of OS and should be evaluated as a new target for therapy.

Background: Quantitative PCR (qPCR) is a common method for quantifying mRNA expression. Given the heterogeneity present in tumor tissues, it is crucial to normalize target mRNA expression data using appropriate reference genes that are stably expressed under a variety of pathological and experimental conditions. No studies have validated specific reference genes in canine osteosarcoma (OS). Previous gene expression studies involving canine OS have used one or two reference genes to normalize gene expression. This study aimed to validate a panel of reference genes commonly used for normalization of canine OS gene expression data using the geNorm algorithm. qPCR analysis of nine canine reference genes was performed on 40 snap-frozen primary OS tumors and seven cell lines. Results: Tumors with a variety of clinical and pathological characteristics were selected. Gene expression stability and the optimal number of reference genes for gene expression normalization were calculated. RPS5 and HNRNPH were highly stable among OS cell lines, while RPS5 and RPS19 were the best combination for primary tumors. Pairwise variation analysis recommended four and two reference genes for optimal normalization of the expression data of canine OS tumors and cell lines, respectively.