Detection of HIV Provirus by in Situ Polymerase Chain Reaction

Publisher Website
Google Scholar
Abstract
Attempts to quantify the results of in situ polymerase chain reaction (PCR) experiments as reported by Bagasra et al. (May 21 issue)1 run the risk of serious misinterpretation. We have studied the use of this technique for specific amplification not only of foreign (viral) DNA but also of unique rearrangements of endogenous DNA. In studies with artificial mixtures of cells containing known numbers of positive and negative cells, we evaluated the effects of different fixatives and other pretreatments and found that when we used single primer pairs and the same fixatives and number of PCR cycles as employed by Bagasra et al., artifacts were a substantial problem.2 3 4 One important factor is the diffusion of PCR products from fixed cells onto bystander cells. The demonstration (by agarose-gel electrophoresis and Southern blot assay) of specific PCR products in the supernatant during cellular PCR experiments and the observation of hybridization signals as rims around cells known to be negative for the particular genomic sequences under study confirm this phenomenon.4 In the article of Bagasra et al., Panel C in Figure 1 and Panel A in Figure 2 show clear examples of such rims, which probably represent diffusion artifacts. Of similar concern are the faintly staining but apparently diffusely positive cells, as seen in the lower right quadrant of Panel B in Figure 2. It is unclear how the authors distinguished between positive and negative cells. Misinterpretation is clearly possible.