Discussion

The evaluation of the effectiveness of antiviral therapy requires an accurate and quantitative measurement of HIV-1 from clinical samples.Basic information on sample handling，processing and storage are required for a credible interpretation of the quantitative measurements.PCR is a powerful tool for assessing the presence of small amounts of nucleic acids.The application of this technique has yielded a great deal of information on the pathogenesis of HIV and the efficacy of anti-HIV drugs（Ou et al.，1988，1990；Kellogg et al.，1990；Holodniy et al.，1991；Schnittman et al.，1991；Michaelet al.，1992；Semple et al.，1993）.However，the use of PCR for routine quantitative analysis has been questioned due to the exponential nature of the amplification process and the effects of small quantities of PCR inhibitors that may be present in clinical samples（Gilliland et al.，1990；Stieger et al.，1991；Piatak，et al.，1993a，b；Siebert et al.，1993）.A minor change in the amplification efficiency and kinetics will result in a large change in the magnitude of a PCR signal.In addition，the attempted measurement of high copy numbers of targets that yield signals in the PCR plateau range fails to quantitate nucleic acids.

A simple PCR protocol was evaluated for quantitation and relied upon carefull aboratory practice to control the variability of the amplification process.This approach is based upon：① using and optimizing an RNAzol extraction procedure in the presence of added carrier tRNA which immediately denatures proteins，including any RNases to ensure high yield of RNA；② a full-length synthetic HIV-1 RNA is used as a quantitative standard and amplified concurrently with a series of patient samples under the same conditions；③ the HIV-1 RNA extract from a patient specimen is diluted so that quantitation occurs within a valid range of PCR amplification，which is between 50 and 1，000 copies for the assay described here.This assay，however，is able to quantitatively detect 10 copies of HIV-1 RNA（see Figure 1）and 1 copy of HIV-1 DNA（data not shown）.Although the level of reproducibility of simple PCR protocols described in the literature spans a large range from two-to ten-fold，as described by Ferre（1992），the variability was controlled to within two-fold under cautious laboratory practice.

The application of the RT-PCR assay described now shows that acidic RNAzol exclusively isolates HIV RNA from plasma preparations in the presence of expected DNA contamination from lymphocytes.Using this reagent，we observed that intact HIV is stable in plasma at room temperature for at least 24 h.Apparently，the envelope and coat proteins of the virus protect the RNA genome from RNases in plasma.In contrast，we have observed that purified HIV-1 RNA is extremely sensitive to plasma RNases.One thousand copies of purified HIV-1 RNA are completely degraded in a 1-min exposure to the plasma（data not shown）.Intact HIV is also resistant to the multiple freezing and thawing treatments.This observation is in agreement with another report（Shizuko et al.，1990）.The stability of HIV-1 RNA as intact virus in plasma and the stability of isolated HIV-1 RNA in RNAzol，as measured by RT-PCR of a short segment of the RNA genome，indicate that sample handling and storage are not an impediment for long-term clinical studies that are required for evaluating therapeutic vaccines and anti-HIV compounds.

In this study，the distribution of HIV RNA was also evaluated within platelet-free plasma，isolated platelets and peripheral blood lymphocytes from 3 patients with relatively high plasma loads of HIV.Ho et al.（1989）previously reported that the HIV-1 in plasma is platelet associated in many HIV-infected patients，possibly on the outside of the cell and perhaps bound to the Fc receptor.The 3 patients evaluated in this study had plasma viral loads ranging from 32，000 copies/ml to 110，000 copies/ml.We observedthat the plateletassociated HIV fraction（PP）amounts to 500—1，000 molecular copies/ml of plasma，corresponding to only 0.5%—3.1%of the total number of copies of HIV-1 in the plasma.Since the typical human platelet count is 108/ml，the results indicate that the number of platelets containing a single virus is in the order of 1 cell per 105 platelets.These results imply that RT-PCR analysis of PRP preparations measures predominately free HIV viral particles，as opposed to cell associated viral particles.These results are in agreement with the observations of Piatak et al.（1993c），but are not supported by the observations of Lee et al.（1993）.Although our results indicate that as mall fraction of the viral RNA is platelet associated，a different fraction may be determined with culture measurements.Since the laboratory-grown virus has a 2，000∶1 ratio of molecular titer（as measured by RT-PCR）to infectious titer，it is possible that the population of infectious virus has a higher affinity for the platelets than the non-infectious，packaged viral particles.Such a differential distribution of virus will result in a higher proportion of the HIV being associated with the platelets when measured by infectious culture assays.

The peripheral blood cells from the 3 patients displayed a range of HIV RNA from 120 copies per 106 PBLs to 7，000 copies per 106 PBLs.These values are in agreement with other observations（Michael et al.，1992）.On a per milliliter of blood basis，thisrange of PBLassociated virus corresponds to a wide fraction of the total viral load.The HIV-1 RNA copies in the PBLs was found to range from 0.3%of the plasma load to 21%of the plasma load.The implication of such a change of distribution of HIV-1 between the plasma and cellular fractions in the pathogenesis of AIDS is certainly worthy of further investigation.Recently，Saksela et al.（1994）correlated disease progression with an increase of HIV RNA in PBLs.(https://www.daowen.com)

While our efforts have concentrated on a simple PCR procedure for quantitation of HIV，several groups have developed competitive PCR procedures for quantitation of HIV DNA and RNA.These protocols are based on quantitation of the relative amounts of PCR products from the target sequence and a competitor of known copy number in a co-amplification system.The attractive feature of the competitive approach is to eliminate the variability of both RT and PCR.However，some drawbacks of this method may limit its application（Terre，1993）.A plasmid with a deletion must be constructed and cloned for each target.The competitor RNA，which is designed to have a different size and partial sequence homology relative to the true target genome，may have a secondary structure that differs from that of the true target RNA.Therefore，the efficiency and kinetics of amplification of the target and the competitor must be established for each system，particularly in the PCR plateau range.Furthermore，since a competitor of at least 5 different concentrations must be introduced to each specimen to be analyzed，the work load is greatly magnified and may limit the application of the competitive approach for routine use in the clinical laboratory.

With the assay described here，0.1 ml of plasma is routinely used in a simple protocol that reproducibly differentiates two-fold differences in HIV copy number within the range of 50—1，000 copies input into the assay.The competitive approach of Piatak et al.（1993a，b）describes a sensitivity of 200 HIV-1 RNA copies from 0.5 and 2.8 ml of plasma.

The issue of whether the HIV DNA copy number in the PBLs and the HIV RNA copy number in these same cells track each other during progression of the disease has not yet been resolved.This，in part，has been limited by the inability to simultaneously isolate both RNA and DNA in a quantitative manner.Our preliminary observations with basic RNazol indicate that quantitative recovery of both HIV RNA and HIV DNA can be performed simultaneously under conditions where the RNA remains intact.Application of this method will permit a single extraction process to quantitatively yield both RNA and DNA that can be assessed by quantitative PCR.