2012/05
Evaluation of Secreted Frizzled-Related Pro- Tein 2 Expression in Canine Thyroid Tumors Using Rt-Pcr and Immunohistochemistry

Deitz K. and Metivier K.

Conference Proceedings, (2012). American College of Veterinary Internal Medicine, New Orleans:

We have previously reported up-regulation of secreted frizzled- related protein 2 (SFRP2) gene expression in canine thyroid tumors using microarray and SFRP2 protein expression and cel- lular localization were characterized with immunohistochemistry (IHC). The purpose of this study was to validate the gene expres- sion of SFRP2 using RT-PCR. Additional samples were also evaluated via IHC and analyzed statistically with the samples from the initial study to increase statistical power. Tissue samples were obtained from dogs undergoing neck mass removal with informed client consent. Only dogs with thyroid carcinomas con- firmed by histopathology were included in the study. Control samples of histologically normal thyroid were obtained from dogs lacking thyroid disease that were euthanized for unrelated reasons. Thyroid tumors from the dogs were categorized based on the degree of invasion: tumor cells confined to the capsule were classified as non-invasive, and those with capsular or vascu- lar invasion were classified as invasive. There were 9 samples in each of the 3 groups (control, non-invasive, and invasive). For RT-PCR, a 1-2cm3 section of tissue was placed in RNA later® and samples were assessed for RNA quality by agarose gel elec- trophoresis. Only samples with no evidence of degradation were used for further analysis. Beta actin was used as the internal ref- erence gene. The mean expression of SFRP2 mRNA in invasive carcinoma was 288 times higher versus control samples. Mean SFRP2 mRNA expression in non-invasive versus control samples was 50 times higher. Samples for IHC were placed in formalin and assessed for SFRP2 antigen using goat anti-SFRP2 antibody. There were a total of 35 tumor samples (including 20 from the previous study) and 10 controls (including 4 from the previous study). In all sam- ples, SFRP2 antigen was present in the cytoplasm of neoplastic thyroid follicular epithelial cells. Tumor samples had significantly higher distribution (p=0.000) and intensity scores (p=0.0021) versus control tissues, which lacked immunoreactivity for SFRP2. This study verifies our previous microarray and IHC results using RT-PCR to assess expression of SFRP2 in thyroid tumors. SFRP-2 is up-regulated in these tumors at all levels and likely plays a role in pathogenesis that requires further elucidation