A STANDARDIZED APPROACH TO PCR-BASED SEMIQUANTITATION OF MULTIPLE CYTOKINE GENE TRANSCRIPTS FROM SMALL-CELL SAMPLES

A simple, rapid, reproducible, and nonradioisotopic method for semiqua ntitative analysis of cytokine mRNAs based on polymerase chain reactio n (PCR) is described. RNA isolated from 2.5 million cells has proven s ufficient to perform semiquantitative analysis of mRNA for 10 differen t cytokines. By this approach accurate assessment of mRNA levels for m ultiple cytokines can be made from as little as 2 ml of blood or about 3 mg of biopsy material. Total cellular RNA is quantitatively recover ed by guanidinium isothiocyanate-acid-phenol extraction of a constant number of cells. Further quantitation of RNA is unnecessary. Highly re producible PCR product formation occurs after specific amplification o f aliquots of reverse transcribed test RNA. The photographic image of the ethidium bromide-stained gel accurately reflects the amount of PCR product loaded, both densitometrically and visually. PCR product gene ration is not affected by the presence of carrier RNA. Thus quantitati ve recovery of total RNA is possible even from a very small number of cells. Similarly, presence of a large excess of nonspecific RNA from n onexpressing cells does not affect amplification of the specific mRNA under study. A linear relationship between mRNA frequency and PCR prod uct formation is observed over a 256- to 512-fold range. The actual mR NA concentration for each cytokine varies depending on the relative ab undance of mRNA for that cytokine relative to total RNA. By performing two amplification cycles (28 and 35) on undiluted and 10-fold diluted RNA samples, the range of detection linearity is extended over a 5000 -fold difference in input RNA levels. This allows for nearly exponenti al amplification of a wide variety of cytokine messages, that vary sub stantially in relative abundance, under standardized experimental cond itions. The quasiduplicate amplification of each sample also provides an internal check for experimental error. It is possible to reproducib ly assign 7 grades to the amplified product of each cytokine simply by visual examination of the bands on an ethidium bromide-stained gel. A s little as 4-fold differences in mRNA levels can be detected for seve ral cytokines. This simple semiquantitative approach should help to st andardize, and thus widen, the research application of this technique to the study of complex patterns of cytokine networking in various imm unopathologic conditions.