The goal of this project is to apply and develop molecular biological methods that can ensure that vaccines and cell substrates are free from viral, and particularly retroviral, contamination. The recent publications of three sensitive methods for the detection of reverse transcriptase (RT) that are at least a million-fold more sensitive than conventional RT assays and the finding using one of these assays that certain vaccines had detectable levels of RT activity necessitated that CBER undertake research to address this issue. The three assays depend on the same principle: an RNA of known sequence is used as a template with an oligodeoxynucleotide primer in an RT reaction; because the sequence of the RNA is known, the cDNA product can be amplified by the polymerase chain reaction (PCR); the PCR product can be detected by a number of methods. The three assays use different RNA templates: brome mosaic virus (BMV) genome, MS2 genome, and an in vitro synthesized product from encephalomyocarditis virus (EMCV). Although the assays were variously termed PERT (for product-enhanced RT) or Amp-RT, we have chosen to call them by the generic name of PBRT for PCR-based RT. We set up the three assays at CBER in order to compare their sensitivities, specificities, and reproducibilities. We found that the MS2 assay (PERT) was about 10 to 50 times more sensitive than the BMV assay, which was about 10 to 100 fold more sensitive than the Amp-RT assay. By increasing the template-primer levels of the BMV assay to those of the PERT assay, the two assays became equally sensitive. In addition, we designed primers to be used with a third commercially-available RNA, TMV, and the assay with this template was similarly sensitive. The inherent background signal of the assay was effectively eliminated by incorporating an RNase digestion prior to the PCR step and using a thermostable polymerase lacking RT activity and. Because of the high sensitivity of the PBRT assay, cell extracts produce a positive signal as do many DNA-dependent DNA polymerases. Reasoning that genuine RT~s would have properties that would enable the enzyme to copy RNA, we designed primers that were around regions of secondary structure and that required that a longer cDNA had to be made. With this modified assay, the signal produced by the cell extract was diminished several logs without significant loss in sensitivity. Importantly, retroviruses from all known classes scored positive. Recently, we have adapted the PBRT assay for use with the real time quantitative system, the TaqMan system, for use with the Perkin-Elmer 7700 system. This modified PBRT assay, the TM-PERT assay, is linear over at least 6 orders magnitude and is as sensitive as the original PBRT assays. We are currently trying to modify this assay to make it selective for retroviruses.
