Combining the Technique of RNA Fingerprinting and Differential Display to Obtain Differentially Expressed mRNA Luda B. Diachenko, John Ledesma, Alex A. Chenchik, and Paul D. Siebert CLONTECH Laboratories, Inc., 1020 East Meadow Circle, Palo Alto, California 94303-4230 Received January 11, 1996 We have modified recently developed methods of RNA fingerprinting and differential display based on arbitrarily primed PCR which can be used for detection and cloning of differentially expressed mRNAs. Our protocol requires only a single cDNA synthesis for each different RNA sample, in contrast to the multiple cDNA reactions required for differential display method, followed by selective amplification of cDNA sequence fraction by arbitrary and oligo(dT) primers. We have shown that the longer primers (25–29-mers) allow the use of optimal dNTP concentration and higher stringency PCR. Further improvements include using TaqStart antibody for hot start PCR and thermostable enzyme mixes suitable for long-distance PCR. Long-distance PCR enables the method to display bands of up to 2 kb and should result in a higher fidelity of PCR products to the original RNA template. When these improvements are combined the resulting method is highly reproducible, and more than 85% of the differentially expressed bands can be confirmed by Northern blot analysis. © 1996 Academic Press, Inc. Two rather similar PCR-based techniques for the identification of differentially expressed mRNAs - RNA fingerprinting by arbitrarily primed PCR (RAP-PCR) (1) and differential display (DDRT-PCR) (2) - are becoming widely used. Both methods are simple and rapid and only require small amounts of total RNA. However, many laboratories have had difficulty obtaining reproduc- ible results with these methods. It has been shown that at least 40% of the differentially displayed bands are not reproducible between experiments even in well trained hands (3). Furthermore, the pattern of differential expression often cannot be reproduced on Northern blots and percent of these false positive can arise up to 90% (4). In our attempts to improve the reliability of the differential display method, we found that one of the key parameters is the length of the PCR primers. In the differential display method, arbitrary 10-mer oligonucleotides are used. However, the optimal primer length for Taq DNA polymerase is at least 13 nt (5). Furthermore, short primers necessitate using a very low annealing temperature (40°C) for all PCR cycles. RAP-PCR method used 18-mers primers. Longer primers make it possible to increase the annealing temperature of PCR reaction to 60°C following one to four initial cycles of annealing at a low temperature (40°C) (1,6). In order to achieve a high level of mismatched synthesis, differential display uses very low concentrations of dNTP (2–5 M). However, the optimum dNTP concentration for Taq polymerase is substantially higher (K m  13 M) (7), so there is a reduction in Taq DNA polymerase activity. This combination of unfavorable factors decrease reproducibility and increase false positives in the differential display method. The use of longer primers in the RAP-PCR method makes it possible to use standard dNTP concentrations (50 M) (1,6). The combination of high-stringency PCR conditions and a more optimal dNTP concentration should provide higher specificity relative to differential display. The increased reproducibility obtained by using longer primers in the differ- ential display approach has recently been described elsewhere (8,9). The RAP-PCR method makes it possible to use one oligo(dT) primer for cDNA synthesis (6). This is more convenient than performing the 3–12 cDNA synthesis reactions required for differ- ential display (2,8). When we tried to use a single oligo(dT)-primed cDNA with DDRT-PCR method, we had difficulty obtaining any bands. Unfortunately, a single RAP-PCR fingerprint typically contains only 10 to 30 bands per track BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 219, 824–828 (1996) ARTICLE NO. 0317 824 0006-291X/96 $18.00 Copyright © 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.