Claudin-2 Expression Levels in Ulcerative Colitis: Development and Validation of an In-Situ Hybridisation Assay for Therapeutic Studies

Ulcerative colitis is a chronic inflammatory disease affecting the colon and is characterized by epithelial damage and barrier dysfunction. Upregulation of the tight junction protein claudin-2 by cytokines is hypothesized to contribute to the dysregulation of the epithelial barrier. New therapeutic agents which block the action of cytokines are being investigated in patients with ulcerative colitis. In order to understand the potential of these therapies, it is important to have reliable assays that can assess downstream endpoints that reflect drug mechanism of action. The aim of the current study was therefore to establish & validate an assay to reproducibly assess the expression and distribution of claudin-2 in human colon biopsy samples. Initially, the potential to measure claudin-2 protein by immunohistochemistry (IHC) was investigated. To identify suitable reagents to develop an IHC assay, preestablished criteria were used to screen five commercial antibodies by Western blotting, immunofluorescence and immunohistochemistry on claudin-2 positive and negative cells and healthy and ulcerative colitis colon tissue. Despite some of these antibodies specifically detecting claudin-2 using some of these techniques, none of the antibodies showed the expected specific staining pattern in formalin fixed human colon samples. As an alternative method to detect claudin-2 expression and distribution in formalin fixed biopsy sections, an in situ hybridization assay was developed. This assay underwent a novel tiered approach of validation to establish that it was fit-for-purpose, and suitable for clinical deployment. In addition, to understand the possible relationship of claudin-2 in the context of disease severity, expression was compared to the Geboes score. Overall, the microscopical Geboes score correlated with the claudin-2 biomarker score for samples that retained crypt morphology; samples with the highest Geboes score were not specifically distinguished, probably due to crypt destruction. In summary, we have applied a strategy for identifying target-specific antibodies in formalin fixed biopsy samples and highlighted that (published) antibodies may not correctly identify the intended antigen in tissues fixed using this method. Furthermore, we have developed and, for the first time, validated an in situ hybridization assay for detection of claudin-2 mRNA, suitable for use as a supportative method in clinical trials. Using our validated assay, we have demonstrated that increased claudin-2 expression correlates with the severity of ulcerative colitis, where crypt destruction is not seen.