Various body fluids contain agents that inhibit the PCR. Such fluids include urine, (Lz) serum, (3-5) feces, (6'7) amniotic fluid, (8) and cerebrospinal fluid. (8) When detection or quantification of microbial genomes from clinical specimens is desired, the presence of PCR inhibitory factors potentially leads to false-negative results or underestimations, respectively.We have been engaged in PCR detection and quantification of the EpsteinBarr virus (EBV) genome in oral fluids. Preliminary attempts to demonstrate viral DNA in whole mouth fluid (WMF) and parotid saliva (PS) from study subjects resulted in poor positivity yields. This was despite having established PCR conditions for the sensitive detection of EBV derived from cell culture and of plasmids that contain subgenomic EBV inserts. These results were surprising in view of previous reports that the virus could be isolated from saliva at high rates by the lymphocyte transformation assay. (9'1~ We therefore investigated the possibility that saliva might also contain PCR inhibitory factors. We demonstrate here that saliva potently inhibits PCR and describe a simple procedure to eliminate this effect.
MATERIALS AND METHODSWMF was collected from 10 healthy human subjects after mechanical stimulation by the chewing of sterile rubber bands. PS was obtained by applying a Lashley cup to Stensens' duct; stimulation was achieved by placing citric acid crystals on the tongue at -30-sec intervals. WMF collection preceded PS collection. WMF was clarified, passed through 5.0-1~m Acrodisc filters, aliquoted, and stored at -70~ until use. PS samples were passed through 0.45-1~m filters, similiarly aliquoted, and frozen.Plasmid pBR322 containing the nonreiterated BamHI K fragment of the B95-8 strain of EBV (11) was used as the principal source of DNA template for PCR. Oligonucleotide PCR primers and probe used, respectively, for the amplification and detection of the BamH! K region of EBV were synthesized in an Applied Biosystems 380B synthesizer; the sequences of these oligomers have been reported previously. (~2) PCR was performed in a Perkin-Elmer Cetus PCR DNA thermal cycler. Thirty-five cycles were performed per run. Cycling conditions were as follows: denaturation at 94~ for 1 min; annealing at 50~ for 40 sec; and extension at 72~ for 1 min.WMF and PS samples, individually spiked with plasmid DNA, underwent each of the following DNA extraction procedures: Geneclean II (Stratech Scientific, Luton, UK); Magic DNA Clean-up System (Promega, Madison, WI); Sephadex G-50 chromatography; phenol-chloroform; guanidinium thiocycanate/silica with or without proteinase K pretreatment; microwaving; boiling for 5 min; and Chelex-100. The Geneclean and Magiprep methods were performed according to manufacturers' instructions. G-50 chromatography was performed as described by de Franchis et al. (13) Extraction with phenol-chloroform was performed as follows: Each 100-1~1 WMF or PS sample was vortexed with an equal volume of 25:24:1 phenol/chloroform/ isoamyl alcohol and spun briefly; t...