Multiplex allele-specific PCR with optimized locked nucleic acid primers.

The ease and advantages of allelespecific PCR (AS-PCR) for SNP genotyping, along with the difficulty of assay specificity with certain mismatched DNA primers, have been described in prior studies (1–3). More recently, the use of real-time PCR detection formats has been described for one or both alleles of AS-PCR (4–6). These detection enhancement methods do not overcome the inherent difficulties with mismatch specificity and may also be limited by the choice of dyes or instrumentation for larger multiplex AS-PCR protocols. Researchers have attempted to utilize multiplex PCR in allele-specific applications (7,8). These approaches add a level of complexity to the design and optimization of DNA primers, negating some of the ease and speed of the original AS-PCR technique. Several molecular applications of locked nucleic acid (LNA) (9,10) have been described in review articles (11,12). These LNA applications include antisense and therapeutic techniques (13,14), SNP genotyping by fluorescence polarization (15), and hybridization-based probe detection (16). Our prior work showed improved ASPCR specificity with 3′ LNA primers at the polymorphic position in SNP genotyping with human and plasmid DNA targets with all possible mismatches (17). LNA is believed to exert its effects of increased hybrid stability (higher Tm) and improved specificity in AS-PCR through the interaction of its rigid conformational structure with complementary strands or enzymes, resulting in slower rates of dissociation (18). In this report, SNPs within the human cystic fibrosis gene (CFTR) were analyzed in individual PCRs with 3′ LNA and DNA primers under various reaction conditions. Once optimization was completed, multiplex PCR amplification reactions were investigated with individual and pooled templates and detected by gel analysis. LNA primers demonstrated higher specificity than DNA primers in multiplex AS-PCR with human SNP targets. This strong allelic discrimination with LNA primers was evident over a broad range of parameters, demonstrating a simplified approach to highly specific multiplex AS-PCR. All heterozygous human DNA samples containing mutant CFTR alleles were purchased from the Coriell Cell Repository (Camden, NJ, USA). Wildtype human genomic DNA was purchased from Promega (Madison, WI, USA) and verified for each lot. All PCR primers were produced by Proligo (Boulder, CO, USA) and are listed in Table 1 with either 3′ LNA or native DNA residues. PCR variables for Figures 1 and 2 are listed in Table 2. These include concentrations of MgCl2, primer, AmpliTaq Gold® DNA polymerase (Applied Biosystems, Foster City, CA, USA), and template DNA. Constant PCR parameters were 1× Gold reaction buffer (Applied Biosystems), dNTP concentration (200 μM), reaction volume (25 μL), and thermal cycling using a DNA Engine (MJ Research, Waltham, MA, USA): 95°C for 7 min, followed by 30 cycles of 94°C for 30 s, 61°C for 30 s, and 72°C for 30 s. AS-PCR optimization was done with 3′ LNA and DNA primers in more than 20 target SNPs in three human genes (CFTR, adenosine deaminase, and T cell receptor Vβ). LNA-containing primers required no difference in preparation or significant changes in reaction and thermal cycling parameters compared with DNA primers. Figure 1 is a comparison of LNA primers with DNA primers for detection of a cystic fibrosis G551D mutation (G→A) in wild-type and heterozygous human genomic DNA samples under five PCR conditions (labeled A–E) listed in Table 2. Wild-type DNA template was used to assess whether mutant AS-PCR primers produced false-positive products in the absence of a mutant allele. In each panel, LNA mutant primers AS-PCR primer sequences (with LNA bases in capital letters) are listed along