Figure 2

Outline of experimental strategy. A: Algorithm for combining limiting dilution of cells with RT-PCR. HIV-RNAs (in this example MsRNAs) of serial 5-fold dilutions of cells (left panel) are measured by RT-PCR (middle panel). Analysis of replicates of each dilution (right panel) reveals both the viral RNA content and the frequencies (estimated by 50% end-points) HIV-RNA⁺ cells. Applying criteria listed in table 1, in this case expression of either MsRNA-tatrev or MsRNA-nef (class-II_Medium) and expression of both MsRNA-tatrev and MsRNA-nef (class-II_High), cell classes differing in HIV-RNA content can be discerned. Specific HIV-RNA expression in each class of MsRNA+ cells can then be normalized by dividing MsRNA copies by the numbers of infected cells. In specimens positive for class-II_High cells which always contain class-II_Medium, the contribution of class-II_Medium needs to be considered (see formulas in panel B). Note that analysis of UsRNA contents in different cell-classes followed the same schemes. B: Analysis of specific MsRNA per-cell expression exemplified for patient 112. MsRNA expression (middle panels) was normalized to the number of HIV-RNA⁺ cells (bottom panels) resulting in MsRNA expression per cell (top panels). The left three panels comprise specimens positive for class-II_Medium expression only. In the right three panels indicating specimens positive for class-IIHigh MsRNA, the average contribution of class-II_Medium⁺ cells (light grey bars) was subtracted from MsRNA copy numbers before normalization to the number of class-II_High + cells. The dotted lines in the top panels show the geometric means (mean_gm) of all data-points. Note that RNA copies per sample and frequencies of MsRNA⁺ cells (middle and bottom panels) are depicted in a linear scale which may result in column heights hardly discernible from zero. Formulas at the bottom describe the calculations performed. Bars show PCR results of separate replicates of PBMC dilutions, horizontal axes in the diagrams have no dimension