Results

3　Results

A simple RT-PCR assay was developed for use with plasma extracts to evaluate the stability and distribution of HIV in plasma preparations.The sensitivity of the RT-PCR assay as shown in Figure 1.While HIV-1 RNA copies as low as 10 copies were detected，the quantitative range that is normally used is 50—1，000 copies with the synthetic HIV-1 RNA standards.Although 2，000 copies can be measured with the assay，the standard curve becomes flat above this level of target molecules，indicating PCR plateau（data not shown）.The variability of the RT-PCR assay for HIV-1 quantitation was assessed by analyzing 15 replicates of 100，500 or 1，000 copies of the RNA standard.The intra-assay coefficient of variation was found to be 23，16，and 6%for 100，500，and 1，000 copies，respectively.The assay variation in a plasma background was determined with 10 replicates of 1，000 copies of HIV-1 RNA spiked into 100 μl of plasma from a seronegative individual.After RNAzol extraction and RT-PCR，the intra-assay coefficient of variation was found to be 22%.The inter-assay difference was found to be less than two-fold when the same samples were extracted，amplified and quantitated at different times.

The quantitative assay described here utilizes platelet rich plasma for the assessment of viral load.Guanidinium extraction is commonly used for extraction of RNA from plasma.However，PRP is contaminated with lymphocytes to a level of about 103—104per 100 μl，depending upon the preparative procedure.Since HIV DNA from infected cells would contribute to the RT-PCR signal if it were simultaneously extracted with the genomic RNA from HIV particles，we evaluated the level of DNA contamination thatwould be tolerated with the extraction procedure.Both RNAzol A and RNAzol B，two similar commercial products，were evaluated as potential extraction reagents.105 copies of HIV-1 synthetic cRNA and 105 copies of linearized pBKBHlOS were simultaneously spiked into RNAzol extracts of PBL preparations（106 cells）.RNA was extracted with acidic RNAzol（pH 4.0）as usual.As a control，both RNA and DNA were simultaneously extracted with RNAzol that had been adjusted to pH 8.5.Nucleic acids from 100 μl aliquots of both the aqueous phase and the interface of the phase separated RNAzol extracts were analyzed by reverse transcription and PCR amplification［see Figure 2（a）］.Under these conditions both RNA and DNA are amplified.The DNA contribution to the PCR signal was assessed with the same aliquots by amplifying without reverse transcriptase［see Figure 2（b）］.The autoradiograph of Figure 2 shows that the HIV-1 plasmid DNA could be detected in the aqueous phase when the pH was 8.5.HIV DNA could not be detected in the aqueous region when the extraction occurred under acidic conditions（pH 4.0）.HIV DNA was，however，measurable from material obtained at the interface.These results demonstrate that HIV RNA is selectively isolated from the aqueous phase when acidic RNAzol is used.This is the preferred preparative procedure for the quantitative analysis of the molecular copy number of virus in plasmaor serum samples.Small amounts of contaminating DNA in these samples will not contribute to the RT-PCR signal since they are left in the interface.With the RTPCR assay described here，2×104 HIV DNA copies in the presence of nucleic acids extracted from 2×105 PBLs did not contribute to the PCR signal.Extreme care should used to avoid the interface region during extraction，since the potential for contamination with HIV DNA sequences from originating from cellular material is high.

In order to study the stability of HIV in plasma and to determine the HIV-1 RNA molecular copies of cultured virus，plasma from a normal individual was prepared and spiked with 1 TCID50/ml or 100 TCID50/ml of HIV LAl.These samples were kept either at room temperature for 0，1，4，or 24 h，or were subjected to repeated freezing and thawing manipulations.The HIV-1 RNA from 1 TCID50/ml and 100 TCID50/ml samples were extracted with acidic RNAzol A，and diluted to 1∶2 and 1∶200，respectively，for the RTPCR assay.This dilution was required to bring the PCR signal within the range that yields quantitative data.The results are shown in Figure 3.One TCID50/milliliter of viral culture was estimated to be equivalent to 2，000 copies of the cRNA standard.HIV-1 RNA extracted from the plasma containing 100 TCID50/ml of HIV LAI at 1∶200 dilution resulted in a RTPCR signal similar to that from the extracted HIV-1 RNA from the plasma with 1 TCID50/ml of HIV LAl at a 1∶2 dilution.These results suggest that the extraction efficiency of HIV RNA is the same for a high-titer sample（200，000 molecular copies/ml）and for a lowtiter sample（2，000 molecular copies/ml）.Figure 3 also shows that intact HIV is stable in the plasma at room temperature for at least 24 h.Additionally，a single freeze/thaw had no effect on the RT-PCR signal，although 10 cycles of freeze/thaw shows a slight loss of signal.Similar results with HIV-1 seropositive individuals are presented as shown in Figure 4.The plasma from individuals（a）and（b）were frozen and thawed once（1FT）or 10 times（10FT）.The HIV-1 RT-PCR signal was not affected by the repeated treatment.In addition，we found that HIV-1 RNA was stable in frozen RNAzol at least for 6 months or when subjected to repeated freezing and thawing manipulations in RNAzol（data not shown）.

Next，the effects of plasma constituents on the yield of an RT-PCR assay signal were investigated，with samples containing identical viral copy numbers.Plasmas were prepared from 9 different seronegative individuals.These plasmas were each spiked to 0.5 and 1 TCID50/ml of HIV LAI.HIV-1 RNAs from each sample（100 μl）were extracted，amplified and quantitated.The results（see Figure 5）show that the plasmas from different individuals have a small effect upon the RT-PCR assay.Although patient 3 displayed abnormally high signals with both the low and high spiked samples，upon repeat（2×）analysis with this patient's plasma，the PCR signals were comparable to those of the other volunteers.By quantitation，the average of the 1 TCID50/ml samples was determined to be 1，800±730 copies.The 0.5 TCID50/ml fractions were found to contain 780±150 copies.Duplicate plasma samples from volunteer 9 were identically spiked with virus and quantitated in duplicate to assess the reproducibility in a single individual.As shown，the results are consistent.Although the various plasmas resulted in a spread in the intra-assay variation，in all cases，the two-fold differences in each individual plasma were discernible.The high/low molecular copy ratios for the group of samples was 2.30±0.96.These data show that within a single individual，two-fold differences in the RT-PCR signal are statistically significant（2.4-times the coefficient of variation for the average of the ratios of the molecular titers of each individual plasma）.

Our target sample for the analysis of HIV copy number is PRP.To determine the fraction of free HIV vs platelet-associated HIV in the PRP preparations，the PRP fraction was further separated into a PFP portion and a PP as described in Materials and Methods.HIV-1 RNA signals from same amount of the PRP and PFP，after RT-PCR，were compared to that from the PP.The quantitative results，summarized as shown in Table 1，show that the PRP and PFP have similar concentrations of HIV-1 RNA，indicating that plasma HIV-1 may not be strongly platelet-associated.Furthermore，the majority of HIV-1 RNA was found in the PFP，whereas only 0.5%—3.0%of HIV-1 RNA was found in the PP.Theviral load in PBLs from these same patients was also analyzed by RT-PCR，and compared to the HIV copy number in the plasma fractions.It was found that the number of HIV-1 RNA copies in the plasma fractions were very much higher than the number of HIV molecules associated with the PBLs in these patients.