Nicole Brass , Dirk Heckel

With the rapid advances in tumor genetics, an increasing number of amplifications have been identified in various human neoplasms (4). Examples of oncogenes amplified in tumors include N-myc and HER-2/neu, which have been proposed as prognostic markers in neuroblastoma and breast cancer, respectively (5,7). To date, most of the information on the amplification of genes has been gathered from conventional Southern blot analysis, but this approach requires relatively large amounts of genomic DNA (5–10 μg) and lacks sufficient sensitivity. Alternatively, comparative genomic hybridization has been used to detect and localize extended amplification domains in human cancers (3). This method, however, is not tailored to the analysis of single loci. Recently, quantitative polymerase chain reaction (PCR) application has gained increasing importance for the evaluation of oncogene amplification. Several modifications of this approach have been described, all of them bearing significant limitations. One of the widely used methods, called differential PCR, requires amplification of both a single-copy control gene and the gene of interest in the same vial (1). This in turn requires PCR conditions that are appropriate for both primer pairs in the same reaction. In addition, it remains unclear how an interaction of the two primers interferes with the amplification. To overcome these limitations, a method termed competitive PCR makes use of artificial internal standards that share the primer-binding sites with the gene of interest (6). These standards, which are termed competitors, have to be synthesized and quantified for each gene separately. According to most protocols, the determination of the gene amplification level also involves scintillation counting. We present a simple and rapid PCR scheme to assess gene amplification while avoiding time-intensive synthesis and use of competitors. The method, which we termed comparative PCR, requires PCR amplification of a given gene in tumor DNA and normal DNA (most conveniently peripheral blood DNA). Because gene amplification is a hallmark of tumor DNA and is rarely found in normal human cells, it is not necessary to use a sample of blood DNA from each patient (9). The signal intensity of the PCR products reflects the copy number of the gene in the tumor and peripheral blood DNA. By comparing the signals in tumor and normal DNA, the presence of gene amplification can be detected. A necessary prerequisite for direct comparison of the band intensities in a stained agarose gel is the calibration of the system with a primer pair specific for a control gene not amplified in tumor DNA. Furthermore, the PCR amplification is to be performed on at least three different amounts of tumor and blood DNA. The implications of this step are threefold: (i) the three independent amplifications facilitate the assessment of the amplification status; (ii) they provide an internal control for the reproducibility of the experiment; and (iii) they serve as control to keep the PCR amplification within the exponential phase. A schematic representation of the experimental approach is given in Figure 1. Details to this fast and