Validation of nonradioactive in situ hybridization as a quantitative approach of messenger ribonucleic acid variations: a comparison with northern blot.

Abstract

The recent improvement of in situ hybridization (ISH) procedures, the increased sensitivity of immunohistochemical detection systems, and the development of assisted image analysis now enable the quantification of specific messenger RNAs (mRNAs) detected by nonisotopic probes on histological sections. However, the reliability and accuracy of this type of mRNA quantification are still to be determined. To this end, we compared in an experimental model of rat malnutrition the densitometric analysis of proinsulin mRNA detected by nonradioactive ISH with those obtained from radioactive Northern blot hybridization (NBH). Proinsulin gene expression was quantified by ISH and by NBH in the pancreatic islets of normally fed rats, rats fasted for 3 days, and rats refed for 8, 24, and 48 h after fasting. Starvation decreased the pancreatic proinsulin mRNA signal by 34% and 38%, as evaluated by ISH and NBH, respectively. Also, with both methods, mRNA levels returned to normal after refeeding. Taken together, the results derived from nonradioactive quantitative ISH were closely correlated to those obtained by quantitative NBH (r = 0.975, p < 0.005). It is thus possible to evaluate variations of mRNA content accurately by quantitative ISH as is currently done by NBH, but with the invaluable advantage of integrating the data with a morphological analysis.