Immunohistochemical staining for tilapia and human insulin demonstrates that a tilapia transgenic for humanized insulin is a mosaic

Abstract

In a previous paper we described the production of transgenic tilapia grown from fertilized eggs microinjected with a tilapia insulin transgene ‘‘humanized’’ by site directed mutagenesis; founder male NT 56 was able to pass this to his F1 offspring, and we were able to measure high levels of circulating humanized insulin in many of these offspring (Pohajdak et al. 2004) and subsequently F2 offspring (unpublished observation). In NT 56’s offspring, the size, shape, and distributions of cells staining for tilapia insulin and human insulin appeared to be identical, suggesting that they are the same cells expressing both peptides—as would be expected. Curiously, we were never able to demonstrate humanized insulin in the serum of NT56, suggesting that production of the transgene was silenced or highly inhibited, possibly due to mosaic integration in the genome. After siring many offspring, NT56 died and was necropsied. As shown in Fig. 1, immunoperoxidase staining of NT 56’s islet tissue for (a) tilapia insulin, using a monoclonal antibody specific for tilapia insulin (Snowdon 2003) and for (b) human insulin (Pohajdak et al. 2004) showed that, unlike his offspring, there were two distinct populations of insulin-positive b-cells. The b-cells that stained for tilapia insulin possessed abundant cytoplasm and were arranged in repetitive units of small tight clusters of multiple b-cells, surrounded by concentric layers of nonb-cells; this pattern is typical for wild-type tilapia (Yang et al. 1999) and for teleost fish in general. In stark contrast, the ‘‘b-cells’’ staining for human insulin were smaller, much less numerous, tended to occur as single cells rather than discrete clusters, and appeared excessively densely granulated (i.e., ‘‘constipated’’ with insulin). Studies with mammalian islets have shown that disruption of normal b-cell interrelationships (i.e., islet architecture), inhibits insulin secretion, so the abnormal architecture of the cells staining for human insulin may account for absence of measurable humanized insulin in NT56’s serum. J. R. Wright Jr (&) Department of Pathology & Laboratory Medicine (Calgary Laboratory Services) and the Julia McFarlane Diabetes Research Centre, Faculty of Medicine, University of Calgary, Room C410, Diagnostic & Scientific Centre, 9, 3535 Research Road NW, Calgary, AB, Canada T2L 2K8 e-mail: jim.wright@cls.ab.ca