ultrasensitive in-house reverse transcription-competitive pcr for quantitation of hiv-1 rna in...

Journal of Virological Methods 87 (2000) 91–97

Ultrasensitive in-house reverse transcription-competitivePCR for quantitation of HIV-1 RNA in plasma

Giulietta Venturi a, Rebecca Ferruzzi a, Laura Romano a, Marinunzia Catucci a,Pier E. Valensin a,b, Maurizio Zazzi a,b,*

a Sezione di Microbiologia, Dipartimento di Biologia Molecolare, Uni6ersita di Siena, Siena, Italyb Ser6izio di Microbiologia e Virologia, Azienda Ospedaliera Senese, Siena, Italy

Received 15 November 1999; received in revised form 14 February 2000; accepted 15 February 2000

Abstract

An ultrasensitive version of an ‘in-house’ reverse transcription-competitive polymerase chain reaction assaydescribed previously for quantitation of human immunodeficiency virus type 1 (HIV-1) RNA in plasma wasdeveloped. The increase in sensitivity from 400 to 50 HIV-1 RNA copies/ml was achieved by pelleting virus particlesfrom 1.8 ml plasma by centrifugation prior to RNA extraction, modifying competitor DNA structure and amounts,and redesigning primers. Quantitation of HIV-1 RNA in 130 samples tested previously by the standard assay showedthat the two procedures yield comparable results (mean absolute difference, 0.2690.20 log) and that the ultrasensi-tive version detects HIV-1 RNA below the threshold of sensitivity of the standard method. The ultrasensitive‘in-house assay’ and the reference QUANTIPLEX HIV-1 RNA 3.0 had the same sensitivity and gave equivalentresults (mean absolute difference, 0.1990.11 log), as shown by parallel blinded testing of 47 plasma samples.Titration experiments with reconstructed plasma samples allowed the determination of a dynamic range of 50–500 000 HIV-1 RNA copies/ml for the ‘in-house’ system. The interassay coefficient of variation for samples nominallycontaining 200, 4000 and 80 000 HIV-1 RNA copies/ml were 33.4, 22.9 and 38.2%, respectively. The performance,turnaround time, and cost-effectiveness of this system make it suitable for medium-scale clinical application. © 2000Elsevier Science B.V. All rights reserved.

Keywords: Human immunodeficiency virus; RNA; Quantitation; Plasma; Competitive PCR

www.elsevier.com/locate/jviromet

1. Introduction

Since levels of human immunodeficiency virustype 1 (HIV-1) RNA in plasma were shown to bepredictive of disease progression and useful formonitoring the efficacy of antiretroviral treatment(Hammer, 1996; Mellors et al., 1996), much efforthas been devoted to the development of reliable

* Corresponding author. Present address: Sezione di Micro-biologia, Dipartimento di Biologia Molecolare, Universita diSiena, Via Laterina 8, 53100 Siena, Italy. Tel.: +39-577-233850; fax: +39-577-233870.

E-mail address: [email protected] (M. Zazzi)

0166-0934/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.

PII: S 0166 -0934 (00 )00151 -8

G. Venturi et al. / Journal of Virological Methods 87 (2000) 91–9792

and sensitive techniques for measuring HIV-1RNA load. Three commercially available methodshave been licensed for quantitating HIV-1 RNA,based on competitive reverse transcription-poly-merase chain reaction (PCR) (AMPLICOR HIV-1 MONITOR; Roche), quantitative nucleic acidsequence-based amplification (NUCLISENSHIV-1 RNA QT; Organon Teknika), andbranched DNA technology (QUANTIPLEXHIV-1 RNA; Chiron). Most viral load measure-ments in infected patients have been carried outusing the versions of these reference systems hav-ing a threshold of sensitivity of about 500 HIV-1RNA copies per milliliter of plasma. Extensive usein clinical practice and comparative studies haveshown that the three commercial assays are allsuitable for routine monitoring of the course ofHIV-1 infection (Coste et al., 1996; Schuurman etal., 1996). However, the advent of highly activeantiretroviral therapy (Montaner et al., 1998) hasmade it possible to achieve suppression of virusreplication below 500 HIV-1 RNA copies/ml,providing a rationale for development of moresensitive HIV-1 RNA quantitative assays. Thus,upgraded versions of the commercial systems withtenfold-increased sensitivity have been releasedrecently and preliminary data have been obtainedsuggesting that individuals with HIV-1 RNA lev-els below 50 copies/ml have a more prolongedresponse to treatment than those with 50–500copies/ml (Powderly et al., 1999). Large-scale ap-plication of the ultrasensitive HIV-1 RNA quanti-tation systems is thus warranted. However,commercial assays are still expensive, posing amajor problem for expanding the use and fre-quency of viral load monitoring in different clini-cal settings. Cost-effective ‘in-house’ methodssuitable for routine medium-scale clinical applica-tion have been described previously (Trabaud etal., 1997; Lin et al., 1998; Zazzi et al., 1999). Thereverse transcription-competitive PCR (RT-cPCR) procedure that was used to quantitateHIV-1 RNA in more than 3000 samples has asensitivity threshold of 400 copies/ml (Zazzi et al.,1999). In this report, an upgraded version of theassay is described, which allows titration of HIV-1RNA with sensitivity increased at a thresholdcomparable with that of the ultrasensitive refer-ence systems.

2. Materials and methods

The standard ‘in-house’ RT-cPCR (Zazzi et al.,1999) involves: (i) plasma RNA extraction by spincolumns (QIAmp Viral RNA kit; Qiagen, Hilden,Germany), (ii) noncompetitive reverse transcrip-tion using Ready-to-Go You-Prime first-strandcDNA synthesis beads (Pharmacia, Uppsala, Swe-den), (iii) four-well amplification of an HIV-1 poltarget in the presence of increasing amounts ofcompetitor DNA using premade reaction mix-tures in a microtiter plate, (iv) agarose gel elec-trophoresis of the reaction products, and (v)densitometric analysis of the digitized image ofthe gel and calculation of the number of HIV-1RNA copies in the starting sample. In order toimprove the sensitivity of the standard assay, thefollowing modifications were made. First, virusparticles from 1.8 ml plasma were pelleted bycentrifugation before RNA extraction. Second,amounts of the competitor DNA have beenadapted to higher sensitivity while maintaining alarge dynamic range. Third, competitor DNA andprimers have been redesigned to ensure a lowerdifference in length between the wild-type (wt)and competitor target and maximum base pairingwith different HIV-1 subtypes (Korber et al.,1997).

A new deleted competitor HIV-1 gag DNAfragment was generated and titrated using thepreviously described strategy (Zazzi et al., 1999)(Fig. 1). The HIV-1 Z6 DNA (plasmidpSYC1857; Perkin-Elmer, Emeryville, CA) wasamplified with the antisense primer SK39 (Ou etal., 1988) and the sense primer T55B (5%-CACC-TAGAACTTTAAATGCATGGGTAAATGTT-TTCAGCATTATCAGAAG-3%). Primer T55Bhybridizes to position 1307–1327 of the HIV-1SF2 (GenBank accession number KO2007) mi-nus-strand DNA with its 3% terminal 21 bases andcontains a 28-base 5% tail identical to plus-strandposition 1239–1266. Thus, PCR amplificationwith primers P71 (5%-CTAGAACTTTAAATG-CATGGGT-3%, position 1242–1263) and P82 (5%-GTTCCTGCTATYTCACTTCCCCTTGGTT-3%,position 1485–1512) generates a 271-base pair(bp) or a 234-bp product when wt HIV-1 RNA-

G. Venturi et al. / Journal of Virological Methods 87 (2000) 91–97 93

derived cDNA or the T55B/SK39 competitorDNA, respectively, is used as the template.

In the ultrasensitive RT-cPCR, plasma RNA(1.8 ml) was centrifuged at 23 000×g for 1 h at4°C after the addition of 20 ml of a 0.25% (w/v)suspension of 0.21-mm diameter red polystyrenebeads (Bangs Laboratories, Fishers, IN). The bot-tom 140 ml above the pelleted beads were thenused for RNA extraction by the QIAmp ViralRNA kit according to the manufacturer’s instruc-tions. As for the standard assay, about 1000copies of tobacco mosaic virus (TMV) RNA(Boehringer Mannheim, Germany) per samplewere added to the virus lysis buffer as an internalcontrol to check that HIV-1 RNA-negative sam-ples were suitable for RT-PCR. The RT step wasperformed for 30 min at 37°C by reconstitutingthe Ready-to-Go You-Prime first-strand cDNAsynthesis beads with 20 ml pre-heated templateRNA and 16 ml diethylpyrocarbonate-treated wa-ter containing 10 pmol of the HIV-1-specific RTprimer P81 (5%-TAGTAGTTCCTGCTATGT-CACT-3%, position 1496–1517) and 3 pmol TMV-specific RT primer TMV1 (Zazzi et al., 1999).Then, four 8-ml aliquots of the RT mixture wereused as PCR templates in the presence of increas-ing amounts (8, 64, 512 and 4096 DNA copies) ofthe competitor T55B/SK39 DNA fragment. The

amplification reactions (50 ml) were run for 40cycles in a 96-well PCR plate as described (Zazziet al., 1999), except that 8 pmol of the primer pairP71/P82 were used and the extension time incre-ment per cycle was reduced from 3 to 2 s. Load-ing of PCR products on agarose gels,densitometric analysis of the wt and competitoramplification products, and data analysis were asfor the standard assay (Zazzi et al., 1999). Am-plification of TMV cDNA by the primer pairTMV1/2 (Zazzi et al., 1999) was carried out onthe residual 2 ml cDNA for samples with unde-tectable HIV-1 RNA as a control for RNA recov-ery and reverse transcription.

The specificity of the redesigned primers used inthe RT-cPCR assay was controlled with bloodDNA and RNA obtained from 50 HIV-1-unin-fected donors and three HIV-2-infected subjects.The standard and ultrasensitive systems werecompared by retesting a panel of 130 samplesobtained from HIV-1-infected subjects and shownto contain detectable (n=30) or undetectable (B400 copies/ml; n=100) HIV-1 RNA by the stan-dard assay. In order to compare the ‘in-house’assay with a reference ultrasensitive assay, a sec-ond panel of 47 plasma samples obtained fromHIV-1-infected subjects were processed in paralleland blindly tested by the ‘in-house’ RT-cPCR andby the QUANTIPLEX 3.0 branched-DNA assay(Chiron Corp., Emeryville, CA) according to themanufacturer’s instructions. In addition, the ti-trated QUANTIPLEX high-positive controlplasma was thawed, diluted accurately with HIV-1-negative plasma to reconstruct samples contain-ing 25–500 000 HIV-1 RNA genomes, and usedto determine the dynamic range and reproducibil-ity of the ‘in-house’ system.

3. Results

Failure to amplify any sequence in the humangenome and HIV-2 gag gene demonstrated thatthe new primers are HIV-1-specific. A panel of 30samples shown previously to contain detectablelevels of HIV-1 RNA (median, 26 318 copies/ml;range, 408–817 938 copies/ml) by the standardassay were tested by the ultrasensitive procedure

Fig. 1. Schematic diagram of (A) location of primers used forconstruction of the competitor DNA and for RT-cPCR, (B)competitor DNA, and (C) wild-type cDNA. The numbersabove the top line representing the HIV-1 genome refer tonucleotide position in the SF2 isolate. The dotted lines indicatecontinuation of sequence. Sense and antisense primers areshown as right and left arrows, respectively. The solid andopen boxes represent target sites for the primers P71 and P82,respectively, used in the competitive PCR.


Fig. 2. Scatter plot of HIV-1 RNA levels in 30 plasma samplesas measured by the standard and ultrasensitive ‘in-house’RT-cPCR method.

263) HIV-1 RNA copies/ml were measured in theremaining 35 samples. The viral load measuredwas below 400 copies/ml in 25 (71.4%) of thesesamples, 400–1000 copies/ml in seven (20.0%)samples and 1000–2443 copies/ml in three (8.6%)samples.

When the ultrasensitive ‘in-house’ RT-cPCRassay was compared blindly with the referenceQUANTIPLEX HIV-1 RNA 3.0 assay, resultswere in good agreement for the 25 samples inwhich HIV-1 RNA was detected by both tests(Fig. 3). There was a consistent tendency (22 of 25samples) for the values obtained by the ‘in-house’system to be slightly higher than those obtainedby the commercial system. The mean differencefor the whole panel was +0.31 log (2.04-fold), inagreement with the difference between the mea-surement unit originally adopted by the QUAN-TIPLEX assay (genome equivalents) and theRT-cPCR system (RNA copies). Indeed, whenHIV-1 RNA copy numbers were normalized togenome equivalents (i.e. divided by two) the meandifference was +0.01 log. The mean9S.D. abso-lute difference between paired values was 0.1990.11 log, and differences were within twofold andfourfold for 22 (88%) and 25 (100%) samples,respectively. Of the remaining 22 samples, 14 hadHIV-1 RNA undetectable by both of the twoassays, five had HIV-1 RNA detected by RT-cPCR only (mean9SD, 110931 copies/ml) andthree had HIV-1 RNA detected by QUAN-TIPLEX only (mean9SD, 76926 copies/ml).

Quantitation of HIV-1 RNA in reconstructedplasma samples nominally containing 25–500 000HIV-1 RNA genomes indicated that the ultrasen-sitive RT-cPCR allows quantitation of HIV-1RNA in plasma in a linear range of 50–500 000genome equivalents (Fig. 4). Finally, the repro-ducibility of the ultrasensitive RT-cPCR was as-sessed by quadruplicate testing of three plasmasamples reconstructed from the same QUAN-TIPLEX high-positive control plasma used for thetitration experiment. The mean CV for samplesnominally containing 200, 4000 and 80 000 HIV-1RNA copies/ml were 33.4, 22.9 and 38.2% withmeasured mean values of 235, 3395 and 85 517copies/ml, respectively.

Fig. 3. Scatter plot of HIV-1 RNA levels in 25 plasma samplesas measured by the Chiron QUANTIPLEX 3.0 assay and theultrasensitive ‘in-house’ RT-cPCR method. Values obtained bythe Chiron QUANTIPLEX 3.0 assay were normalized tocopies per milliliter (i.e. divided by two).

(Fig. 2). The mean9S.D. absolute difference be-tween the results obtained by the two methodswas 0.2690.20 log. All but one (96.6%) pairedvalues were within fourfold (0.6 log), and 18(62.1%) paired values were within twofold (0.3log). There was no tendency of each system toyield results higher than the other, as shown bythe nearly identical mean HIV-1 RNA loads (4.23versus 4.19 log for the standard and ultrasensitiveassay, respectively). The ultrasensitive assay wasthen used to quantitate HIV-1 RNA in a secondpanel of 100 plasma samples containing fewerthan 400 HIV-1 RNA copies/ml as determined bythe standard assay. While viremia was still unde-tectable in 65 of the samples, 36–2443 (median,


4. Discussion

As the number of potent antiretroviral treat-ment options increases, viral load measurementmust rely on effective and sensitive procedures.While the latest generation of commercial assaysfulfils these requirements, less expensive ‘in-house’systems with comparable performance may beconveniently used in medium-scale clinical set-tings. The procedure already described is an en-hanced-sensitivity version of the system we havebeen using in clinical practice to reliably monitorHIV-1 infection in several hundreds of patients(Zazzi et al., 1999). Similar to the Amplicor refer-ence system (Sun et al., 1998), the key to increasedsensitivity is the plasma centrifugation step allow-ing concentration of virus particles. It must benoted that this additional procedure is basically awalkaway step and actually reduces the hands-ontime in our system because the spin column isthen loaded only twice with a concentrated vol-ume instead of four times with the original plasmaas in the standard assay (Zazzi et al., 1999). Thered polystyrene beads used for visualization of theviral pellet are dispensed conveniently into emptymicrocentrifuge tubes before processing of blood.

There is a negligible added cost with respect to thestandard assay, and thus the ‘in-house’ ultrasensi-tive system retains the same productivity andcost-effectiveness as the former version. Retestingof 130 plasma samples that had been previouslystored in duplicate and titrated by the standardassay showed that results obtained by the stan-dard and ultrasensitive assay are directly com-parable, similar to what has been reported for thefirst- and second-generation AMPLICOR HIV-1MONITOR assays (Sun et al., 1998; Erali andHillyard, 1999). This is important, since mostpatients expected to be monitored with ultrasensi-tive assays have been previously evaluated withthe standard methods. The sensitivity and repro-ducibility of the ultrasensitive ‘in-house’ RT-cPCR were shown to be comparable with those ofthe reference ultrasensitive AMPLICOR assay(Sun et al., 1998; Erali and Hillyard, 1999), whichis also based on RT-PCR in the presence of acompetitor target. However, the dynamic range ofthe ‘in-house’ system is one log wider than that ofthe Roche system (Sun et al., 1998; Erali andHillyard, 1999), obviating the need to dilute andre-analyse plasma samples harboring high levelsof HIV-1 RNA. This is due to the possibility toread the results of competitive amplificationagainst four different amounts of the competitortarget in the ‘in-house’ assay as opposed to read-ing of serial dilutions of a single competitivereaction in the AMPLICOR assay (Sun et al.,1998).

An important issue in clinical practice is thepossibility of using HIV-1 RNA values obtainedwith different assays. Patients may be followed inmore than one health care unit during the courseof HIV-1 infection, or the same unit may adoptsubsequently a new diagnostic system. The choiceto compare our method with the ultrasensitiveQUANTIPLEX assay was based on the higherreproducibility of branched DNA-based proce-dures with respect to target amplification systems(Lin et al., 1998). Evaluation of both clinicalsamples and dilution series prepared from theQUANTIPLEX titrated controls showed an ex-cellent agreement between the two assays. Failureto detect HIV-1 RNA in low-positive samplesoccurred with comparable frequencies for the two

Fig. 4. Scatter plot of expected and measured HIV-1 RNAlevels in a plasma dilution series derived from the QUAN-TIPLEX high-positive control plasma. The titrated plasmawas diluted with HIV-1-negative plasma to reconstruct sam-ples nominally containing 500 000, 100 000, 20 000, 4000, 800,400, 200, 100, 50 and 25 HIV-1 RNA copies/ml. Resultsshown for samples with 25–800 HIV-1 RNA copies/ml are themean of two separate runs. Only one of the two experimentswith 25 HIV-1 RNA copies/ml yielded a measurable signal (65copies/ml).


assays, as expected from the stochastic distribu-tion of the lowest amount detectable with twoprocedures of equal sensitivity. The twofoldhigher values obtained with the ‘in-house’ RT-cPCR were not unexpected, since our standardassay was initially calibrated on the first AM-PLICOR HIV-1 MONITOR (version 1.0) assay,which has been consistently shown to yield valuesabout twofold higher than the QUANTIPLEXassay (Nolte et al., 1998; Aschbacher et al., 1999).The new set of RT and PCR primers used in the‘in-house’ ultrasensitive assay have been specifi-cally designed to detect gag RNA of all knownHIV-1 subtypes, including group O, similar to theredesigned primer set in the AMPLICOR HIV-1MONITOR assay (Triques et al., 1999). Collec-tion of samples from subjects infected with nonBHIV-1 subtypes is in progress in order to maxi-mally adapt the procedure to the present andfuture status of the HIV-1 pandemics.

Acknowledgements

This study was supported by II ProgrammaNazionale di Ricerca AIDS (grant 308.85) and byIII Progetto Terapia Antivirale AIDS (grant 880/25), Istituto Superiore di Sanita, Ministero dellaSanita, Rome, Italy and by Fondazione Montedei Paschi di Siena (grant ‘Diagnostica microbio-logica diretta mediante tecniche biomolecolari’).G. Venturi is the recipient of an AIDS fellowshipfrom the Istituto Superiore di Sanita, Ministerodella Sanita, Rome, Italy and L. Romano is therecipient of a fellowship from Glaxo-Wellcome.We thank Chiron Corp. for making the QUAN-TIPLEX 3.0 assay available for testing.

References

Aschbacher, R., Monari, P., Lolli, S., Donzelli, C., Colangeli,V., Vignoli, M., Ramazzotti, M., Furlini, G., Re, M.C.,1999. Evaluation of three different commercial proceduresfor quantifying human immunodeficiency virus type-1RNA levels. New Microbiol. 22, 1–9.

Coste, J., Montes, B., Reynes, J., Peeters, M., Segarra, C.,Vendrell, J.P., Delaporte, E., Segondy, M., 1996. Compar-ative evaluation of three assays for the quantitation of

human immunodeficiency virus type 1 RNA in plasma. J.Med. Virol. 50, 293–302.

Erali, M., Hillyard, D.R., 1999. Evaluation of the ultrasensi-tive Roche AMPLICOR HIV-1 MONITOR assay forquantitation of human immunodeficiency virus type 1RNA. J. Clin. Microbiol. 37, 792–795.

Hammer, S.M., 1996. Advances in antiretroviral therapy andviral load monitoring. AIDS 10 (Suppl 3), 1–11.

Korber, B., Hahn, B., Foley, B., Mellors, J.W., Leitner, T.,Myers, G., McCutchan, F., Kuiken, C.L. (Eds.), 1997.Human Retroviruses and AIDS 1997: A Compilation andAnalysis of Nucleic Acid and Amino Acid Sequences.Theoretical Biology and Biophysics Group, Los AlamosNational Laboratory, Los Alamos, NM.

Lin, H.J., Pedneault, L., Hollinger, F.B., 1998. Intra-assayperformance characteristics of five assays for quantificationof human immunodeficiency virus type 1 RNA in plasma.J. Clin. Microbiol. 36, 835–839.

Mellors, J.W., Rinaldo, C.R. Jr, Gupta, P., White, R.M.,Todd, I.A., Kingsley, L.A., 1996. Prognosis in HIV-1infection predicted by the quantity of virus in plasma.Science 272, 1167–1170.

Montaner, J.S., Hogg, R., Raboud, J., Harrigan, R., O’S-haughnessy, M., 1998. Antiretroviral treatment in 1998.Lancet 352, 1919–1922.

Nolte, F.S., Boysza, J., Thurmond, C., Clark, W.S., Lennox,J.L., 1998. Clinical comparison of an enhanced-sensitivitybranched-DNA assay and reverse transcription-PCR forquantitation of human immunodeficiency virus type 1RNA in plasma. J. Clin. Microbiol. 36, 716–720.

Ou, C.-Y., Kwok, S., Mitchell, S.W., Mack, D.H., Sninsky,J.J., Krebs, J.W., Feorino, P., Warfield, D., Schochetman,G., 1988. DNA amplification for direct detection of HIV-1in DNA of peripheral blood mononuclear cells. Science239, 295–297.

Powderly, W.G., Saag, M.S., Chapman, S., Yu, G., Quart, B.,Clendeninn, N.J., 1999. Predictors of optimal virologicalresponse to potent antiretroviral therapy. AIDS 13, 1873–1880.

Schuurman, R., Descamps, D., Weverling, G.J., Kaye, S.,Tijnagel, J., Williams, L., van Leeuwen, R., Tedder, R.,Boucher, C.A., Brun-Vezinet, F., Loveday, C., 1996. Mul-ticenter comparison of three commercial methods forquantification of human immunodeficiency virus type 1RNA in plasma. J. Clin. Microbiol. 34, 3016–3022.

Sun, R., Ku, J., Jayakar, H., Kuo, J.C., Brambilla, D., Her-man, S., Rosenstraus, M., Spadoro, J., 1998. Ultrasensitivereverse transcription-PCR assay for quantitation of humanimmunodeficiency virus type 1 RNA in plasma. J. Clin.Microbiol. 36, 2964–2969.

Trabaud, M.A., Audoly, G., Leriche, K., Cotte, L., Ritter, J.,Sepetjan, M., Trepo, C., 1997. Development of a reversetranscriptase PCR-enzyme-linked immunosorbent assayfor quantification of human immunodeficiency virus type 1RNA in plasma: comparison with commercial quantitativeassays. J. Clin. Microbiol. 35, 1251–1254.

Triques, K., Coste, J., Perret, J.L., Segarra, C., Mpoudi, R.,Reynes, J., Delaporte, E., Butcher, A., Dreyer, K., Her-man, S., Spadoro, J., Peeters, M., 1999. Efficiencies of four


versions of the AMPLICOR HIV-1 MONITOR test forquantification of different subtypes of human immunodefi-ciency virus type 1. J. Clin. Microbiol. 37, 110–116.

Zazzi, M., Romano, L., Catucci, M., Venturi, G., De Milito,

A., Valensin, P.E., 1999. Clinical evaluation of an in-housereverse transcription-competitive PCR for quantitation ofhuman immunodeficiency virus type 1 plasma RNA. J.Clin. Microbiol. 37, 333–338.

..

ultrasensitive in-house reverse transcription-competitive pcr for quantitation of hiv-1 rna in...

Documents