Influence of the decrease of intracellular antigenic content on morphometric analysis: effect of the type and...

The aim of the study is to determine the effect degranulation of B cells on their immunohistochemical detection and to evaluate whether this effect depends on the technique of immunological detection. The biological model is the pancreatic insulin-containing B cell. To decrease the insulin content of the pancreas, insulin release was stimulated by five intraperitoneal injections of glibenclamide (2 mg/kg). Specimens of the pancreas were taken for insulin extraction and quantitation by radioimmunoassay (RIA), ultrastructural analyses and immunocytochemistry. The sections were treated either by polyclonal or monoclonal anti-insulin serum at various concentrations and peroxidase-antiperoxidase (PAP) technique eventually followed by silver amplification. The B-cell insular fraction or relative B-cell area (RBA) was measured with an automatic image analyser. The reference value for the relative B-cell area was established in control rats and corresponds to the ratio of the insular area occupied by immunostained B cells to the islet area. Its value was confirmed by calculation of the area occupied by non-B cells. RIA indicates a decrease of 83% of the insulin content in treated rats while the number of B granules decreases by 72% at the ultrastructural level. In usual conditions (polyclonal serum, 1/1500) the degranulation leads to a 16% underestimation of the B-cell insular fraction. This underestimation increases when monoclonal antibodies are used and further increases when higher dilutions are tested. The silver amplification does not prevent this underestimation and, in this particular model, exclusively acts by increasing the contrast. The only means of restoring the correct level of detection is to use the serum at a higher concentration.(ABSTRACT TRUNCATED AT 250 WORDS)