Amplification of differentially displayed PCR products isolated from untreated denaturing polyacrylamide gels.

Analysis of differentially expressed genes has been facilitated by introduction of two related techniques: differential display (3,4) and RNA arbitrarily primed polymerase chain reaction (RAP-PCR) (10). These powerful techniques have been used to identify developmentally, environmentally and hormonally regulated genes in animals and plants (5,7–9). In each method, two or more RNAs are used as templates to generate cDNA. Subsequently, the cDNA fragments are amplified by PCR using an arbitrary primer in the presence of a radiolabeled dNTP. After separation by denaturing polyacrylamide gel electrophoresis, the gels are fixed and dried. Differentially amplified cDNAs are identified by autoradiography. To produce sufficient DNA for further analysis, the differentially expressed DNA must be eluted from the gel and re-amplified in a second, highstringency PCR. This last step has proven to be problematic for researchers, with poor yields of re-amplified product being common. Two solutions to the problem have been purification of the eluted DNA by ethanol precipitation or addition of a third high-stringency PCR. We present an alternative solution, which consistently gives high yields of DNA in the second round of PCR and requires fewer manipulations of the gel. RAP-PCR was done according to the manufacturer’s protocol (Stratagene, La Jolla, CA, USA) using either [32P]dCTP or [33P]dCTP (2–5 μCi/25 μL reaction) as the label (2). Following RAP-PCR, the amplified cDNAs were separated by electrophoresis on a 0.4-mm-thick 4% polyacrylamide gel containing 7 M urea. Because no suitable gel drying apparatus was available, a procedure for direct autoradiography of DNA sequencing gels was adapted to identify differentially expressed cDNAs. Before use, the glass plates of the electrophoresis unit (Sequi-Gen Nucleic Acid Sequencing Gel; Bio-Rad, Hercules, CA, USA) were treated to facilitate gel handling. To increase adherence of the gel to the glass, the outer plate was treated with 0.5% 3-(trimethoxysilyl)propyl methacrylate and 0.3% acetic acid in 95% ethanol. The other plate was coated with Rain-X (siloxane-based glass/ windshield treatment; Unelko, Scottsdale, AZ, USA) to increase separation of the gel from the plate (1). After electrophoresis, the Rain-X-treated plate was removed. The gel, on the outer electrophoresis plate, was covered with plastic wrap and autoradiographed for 24 h (33P-labeled DNAs) or 14 h (32Plabeled DNAs) at -70°C. Bands representing selectively expressed cDNA were excised from the gel (12–16 μL gel/band). DNA was eluted from the polyacrylamide into 50 μL of TE buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA) for 1 h at 70°C, followed by 8–20-h incubation at 25°C. DNA was either used directly for re-amplification under high stringency conditions, or it was purified from the eluate using Wizard PCR Prep according to the manufacturer’s instructions (Promega, Madison, WI, USA) before use (2). DNA concentrations were determined fluorometrically using the TKO-100 Mini Fluorometer (Hoefer Pharmacia Biotech, San Francisco, CA, USA). When DNA eluted from a differentially expressed band was used directly for re-amplification, yields of re-amplified cDNA were extremely low (Table 1). More often, no DNA at all was detected by agarose electrophoresis. Because the gel was not fixed or dried before excision of the DNA-containing band, we suspected that some component of the gel might inhibit PCR. To remove possible contamination, we purified the eluted DNA before PCR reamplification using the Wizard PCR Prep. As a result of this purification of DNA, yields of amplified products increased more than 10-fold (Table 1). For more than thirty re-amplification reactions, the yield ranged from 200 to 750 ng DNA per reaction. To determine whether some component of the gel eluate was inhibiting PCR, the effect of a “blank” gel eluate and urea on amplification of a plasmid cDNA was tested. A high yield (850 ng) of amplified DNA was found for the control reaction (Figure 1, lanes 3 and 4). In contrast, the presence of the gel eluate reduced the yield to less than