high levels of hiv-1 plasma during all stages of infection...
TRANSCRIPT
104 cells per well in round-bottomed, 96-wellplates and incubated with effectors for 4 hours intriplicate. Supernatant (30 gi) was removed fromeach well and counted in a Betaplate scintillationcounter (LKB-Wallac, Turku, Finland). Maximalcounts, released by addition of 6 M HCI, andspontaneous counts released without CTLs weredetermined for each target preparation. Percentspecific lysis was calculated as 100 x (experi-mental counts - spontaneous counts)/(maximalcounts - spontaneous counts).
16. 0. K. Rbtzschke et al., Nature 348, 252 (1990).17. L. A. Hawe et al., unpublished observations.18. P. Vieira and K. Rajewsky, Ant. Immunol. 2, 487
(1 990).19. J. J. Donnelly etal., J. Immunol. 145, 3071 (1990).20. Female BALB/c mice 10 weeks of age were
injected intraperitoneally with 0.5 ml of pooledserum [diluted in 2.0 ml of phosphate-bufferedsaline (PBS)] from mice that had been injectedthree times with 200 pg of NP DNA. Control micewere injected with an equal volume of poolednormal mouse serum or with pooled serum frommice that had recovered from infection with ANHK/68, also in 2.0 ml of PBS. The dose of NVHK/68immune serum was adjusted such that the en-zyme-linked immunosorbent assay (ELISA) titer ofanti-NP antibody was equal to that in the serumpooled from NP DNA-injected mice. Mice werechallenged unanesthetized in a blind fashion with104 median tissue culture infectious doses(TCID50) of ANHK/68 2 hours after serum injection,and a further injection of an equal amount ofserum was given 3 days later. Mice were killed 6and 7 days after infection, and viral lung titers inTCID50 per milliliter were determined as de-scribed (31).
21. Viral challenges were performed with a mouse-adapted strain of ANHK/68 and maintained subse-quently by in vivo passage in mice (I. Mbawuike,personal communication). The viral seed stockused was a homogenate of lungs from infectedmice and had an infectivity titer of 5 x 108TCIDJml on Madin-Darby canine kidney(MDCK) cells (American Type Culture Collection,Rockville, MD). For viral lung titer determinationsand mass loss studies, viral challenges wereperformed in blind fashion by intranasal instillationof 20 Fl containing 104 TCID50 on the nares ofunanesthetized mice, which leads to progressiveinfection of the lungs with virus but is not lethal inBALB/c mice (23). In survival experiments, micewere challenged by instillation of 20 Al containing1025 TCID50 on the nares under full anesthesiawith ketamine and xylazine; infection of anesthe-tized mice with this dose causes a rapid lunginfection that is lethal to 90 to 100% of nonimmu-nized mice [J. L. Schulman and E. D. Kilbourne, J.Exp. Med. 118, 257 (1963); G. H. Scott and R. J.Sydiskis, Infect. Immunity 14, 696 (1976); (23)].Viral lung titers were determined by tenfold serialtitration on MDCK cells in 96-well plates as de-scribed (31). Data are given as the logarithm ofinverse titers.
22. J. J. Donnelly et al., unpublished observations.23. R. A. Yetter et al., Infect. Immunity29, 654 (1980).24. S. Parker et al., unpublished observations.25. K. Tanaka et al., Annu. Rev. Immunol. 6, 359
(1988).26. Neutralizing (hemagglutination inhibiting) anti-
bodies were elicited in mice, ferrets, and rhesusmonkeys after injection of DNA encoding influen-za A hemagglutinin (22). Antibodies to humangrowth hormone were generated after administra-tion of DNA-coated gold microprojectiles intocells [D.-C. Tang et al., Nature 356, 152 (1992)].
27. D. Gray and P. Matzinger, J. Exp. Med. 174, 969(1991); S. Oehen et al., ibid. 176, 1273 (1992).
28. J. F. Young et al., in The Origin of PandemicInfluenza Viruses, W. G. Lauer, Ed. (Elsevier Sci-ence, Amsterdam, 1983), pp. 129-138.
29. B. S. Chapman et al., Nucleic Acids Res. 19, 3979(1991).
30. Abbreviations for the amino acid residues are asfollows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G,Gly; H, His; I, lie; K, Lys; L, Leu; M, Met; N, Asn; P,
Pro; Q, Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp;and Y, Tyr.
31. T. M. Moran et al., J. Immunol. 146, 321 (1991).32. PCR was performed as per instructions in the
GeneAmp kit (Perkin-Elmer Cetus).33. We thank J. Feigner (Vical) and D. DeFeo-Jones
[Merck Research Laboratories (MRL)] for assis-
tance and consultation on aspects of molecularbiology, A. Williamson (MRL) and H. Zweerink(MRL) for evaluation of the manuscript, and T.Moran (Mount Sinai University School of MedicineNew York) for advice on live virus challenges.
27 October 1992; accepted 30 December 1992
High Levels of HIV-1 in Plasma During All Stagesof Infection Determined by Competitive PCR
M. Piatak, Jr., M. S. Saag, L. C. Yang, S. J. Clark, J. C. Kappes,K.-C. Luk, B. H. Hahn, G. M. Shaw, J. D. Lifson*
Quantitative competitive polymerase chain reaction (QC-PCR) methods were used toquantify virion-associated human immunodeficiency virus type-i (HIV-1) RNA in plasmafrom 66 patients with Centers for Disease Control stage I to IVC1 infection. HIV-1 RNA,ranging from 100 to nearly 22,000,000 copies per milliliter of plasma (corresponding to 50to 11,000,000 virions per milliliter), was readily quantified in all subjects, was significantlyassociated with disease stage and CD4+ T cell counts, and decreased by as much as235-fold with resolution of primary infection or institution of antiretroviral therapy. Plasmavirus levels determined by QC-PCR correlated with, but exceeded by an average of60,000-fold, virus titers measured by endpoint dilution culture. Quantitation of HIV-1 inplasma by QC-PCR may be useful in assessing the efficacy of antiretroviral agents,especially in early stage disease when conventional viral markers are often negative.
The natural history of HIV-1 infection ischaracterized by a variable clinical course,with the development of acquired immun-odeficiency syndrome (AIDS) generally oc-curring after 7 to 11 years (1). A centralparadox of HIV disease involves the pro-gressive development of immunologic ab-normalities, beginning during the earlystages of infection when assays for circulat-ing p24 antigen and culturable virus inperipheral blood suggest minimal or absentlevels of viral replication (2-8). Recentstudies of HIV-1 DNA and RNA in clinicalsamples suggest that, as compared withperipheral blood cells, lymphoid tissue rep-resents a preferred and continuous site ofviral replication, although such studies havenecessarily been limited by the relative in-accessibility of the tissue compartment, es-pecially for repeated evaluation (9).
Previous polymerase chain reaction(PCR) studies of HIV-1 RNA in plasmahave generally been limited to qualitative(10) or semiquantitative analyses (1 1). Instandard PCR methods, the absoluteamount of product generated does not al-ways bear a consistent relation to the
M. Piatak, Jr., L. C. Yang, K.-C. Luk, J. D. Lifson,Division of HIV and Exploratory Research, GenelabsTechnologies Inc., Redwood City, CA 94063.M. S. Saag and S. J. Clark, Division of InfectiousDiseases, Department of Medicine, University of Ala-bama at Birmingham, Birmingham, AL 35294.J. C. Kappes, B. H. Hahn, G. M. Shaw, Division ofHematology-Oncology, Department of Medicine, Uni-versity of Alabama at Birmingham, Birmingham, AL35294.
*To whom correspondence should be addressed.
SCIENCE * VOL. 259 * 19 MARCH 1993
amount of target sequence present at initi-ation of the reaction, particularly for clini-cal specimens (12). Both the kinetics andefficiency of amplification of a target tem-plate are dependent on the starting abun-dance of that template and on the sequencematch of the primers and target templateand may also be affected by inhibitors pre-sent in the specimen (12). In PCR analysisof RNA samples, variable efficiencies inboth the reverse transcription and amplifi-cation steps are potential sources of vari-ability. For these reasons, comparison ofthe amount of specimen-derived PCR prod-uct to the amount of product from a sepa-rately amplified external control standard(11) does not provide a rigorous basis forabsolute quantitation. Normalization basedon coamplification of a heterologous "inter-nal control" target sequence (such asP-globin or actin) does not optimally ad-dress this problem, owing to different tem-plate abundances and priming efficienciesfor different primer-target combinations.
In the quantitative competitive PCR(QC-PCR) method for RNA quantitation(13) a competitive RNA template matchedto the target sequence of interest, but dif-fering from it by virtue of an introducedinternal deletion, is used in a competitivetitration of the reverse transcription andPCR steps, providing stringent internalcontrol (13-15). Increasing known copynumbers of competitive template are addedto replicate portions of the test specimen,and quantitation is based on determinationof the relative, not absolute, amounts of the
1749
differently sized amplified products derivedfrom the wild-type and competitive tem-plates, after electrophoretic separation. Toincrease the sensitivity and consistency ofamplification of only virion-associatedRNA, we pelleted virus from plasma byultracentrifugation, used procedures in-tended to maximize recovery of extractedRNA (13), and targeted a highly conservedsequence in HIV-1 gag (16), using oligonu-cleotide primers that incorporate neutralinosine residues at the few positions ofrecognized variability (13, 16, 17).
Sixty-six consecutively enrolled HIV-1-infected subjects representing all stages ofinfection [Centers for Disease Control(CDC) stages I to IV] (2) and ten HIV-1seronegative healthy donors were evaluatedfor virion-associated HIV-1 RNA by QC-PCR. Infected subjects were also tested forculturable virus and for p24 antigen withboth standard and immune complex disso-ciation (ICD) test procedures, as described(13, 17-19). For each subject, all virologicmeasurements were made with a single plas-ma sample, which was divided and frozen inreplicate portions. Figure 1 illustrates QC-PCR results from a control reconstructionexperiment, along with a representativeexperimental determination of plasmaHIV- 1 copy number for an infected patient.Virion-associated RNA was detected andquantified in plasma specimens from all 66HIV- 1-infected subjects (Table 1). Deter-mined RNA copy numbers ranged from1.00 x 102 to 2.18 x 107 HIV-1 RNAcopies per milliliter of plasma (correspond-ing to 0.50 x 102 to 1.09 x 107 virions permilliliter). Positive signals for wild-typeHIV-1 target sequences were not observedin specimens from any of ten uninfectedcontrol subjects or, when the reverse tran-scription step was omitted, for specimensfrom any of the HIV-1-infected patients.
The threshold sensitivity for RNA QC-PCR analysis was determined to be 100copies per reaction (13, 20). Althoughpositive signals could be detected for as fewas 10 to 20 copies, results were less consis-tent below 100 copies per reaction, andquantitation was less reliable. QC-PCRanalysis was typically performed with 0.5-ml plasma samples. However, pelleting thevirus from increased volumes of plasma forspecimens containing low numbers of viri-ons did allow for increased overall sensitiv-ity of detection. Thus, for one p24 antigen-negative, culture-negative CDC stage IIpatient (patient TIMI 0852, Table 1), ini-tial QC-PCR analysis of a 0.5-ml specimenportion gave a negative result. However,analysis of virus pelleted from 2.8 ml of areplicate plasma sample allowed unequivo-cal detection of the low amount of viruspresent in this patient's plasma (100 copiesper milliliter), whereas no HIV-1 RNA was
1750
seen with analysis of comparable volumes ofplasma from HIV-1 seronegative controls.In separate control studies, analysis of trip-licate portions of plasma samples from sixdifferent patients, with mean HIV RNAcopy numbers of 6.7 x 104 to 1.0 x 106 permilliliter, gave a mean SD of 22%. Analysisof replicate portions of the same HIV-1RNA preparation on six different days gavean SD of 15%. This reproducibility con-trasts with variability of up to 600 to 1000%reported for noncompetitive PCR proce-dures (11, 12, 14).
Whereas the QC-PCR method quanti-fied virion-associated HIV-1 RNA in all 66patients tested, virus culture and standardp24 antigen assays were much less sensitive,with positive results in 4/20 and 5/20 sub-jects with CD4' T cell counts >500 percubic millimeter, 6/18 and 7/18 subjectswith CD4' T cell counts of 200 to 500 percubic millimeter, and in 22/28 and 24/28subjects with CD4+ cells fewer than 200per cubic millimeter, respectively (Table1). In 30 patients with negative results instandard p24 antigen assays, the use of acidtreatment to dissociate p24 from immunecomplexes increased the frequency of detec-tion by only four patients, although abso-lute measured amounts of p24 were in-creased in 25/36 patients with detectablelevels of p24 before acid treatment.
QC-PCR-determined HIV-1 RNA lev-els in plasma differed between clinical stag-es, with levels for CDC stage II and IIIpatients (asymptomatic or persistent lym-phadenopathy) [mean, 78,200 copies permilliliter (n = 22)] significantly lower thanfor CDC stage IVC2 [AIDS-related com-plex (ARC)] [mean, 352,100 copies permilliliter (n = 23); P ' 0.05] (21), andRNA levels for CDC stage IVC2 patientssignificantly lower than for CDC IVC1(AIDS) patients [mean, 2,448,000 copiesper milliliter (n = 15); P ' 0.05] (21).CDC stage IVC1 (AIDS) patients had cir-culating virus levels comparable to thoseobserved during the peak of viral replica-tion in stage I (primary infection) patients[mean, 5,178,000 copies per milliliter (n =6); difference not significant] (21), imply-ing that late-stage disease is characterizedby a nearly complete loss of immunologicalcontrol of viral replication. There was alsoa significant correlation between increasingQC-PCR-determined HIV-1 RNA levelsand decreasing absolute CD4+ T cellcounts [Spearman rank correlation coeffi-cient r = -0.765, P < 0.0001] (22). Anonlinear regression analysis of these datayielded the equation log(RNA) = 4.43 +1.77exp[-0.0049(number of CD4+ Tcells)], with an r2 value of 0.56 (P <0.0001).
Figure 2 shows data for longitudinalspecimens collected from three individuals
SCIENCE * VOL. 259 * 19 MARCH 1993
[of six studied (Table 1)], beginning at thetime of presentation with acute HIV-1 in-fection (CDC stage I) and continuingthrough the establishment of chronic infec-tion (CDC stage II to IVC2). Each patientpresented with signs and symptoms of pri-mary infection and with no detectableHIV- 1-reactive antibodies as determinedby enzyme-linked immunosorbent screen-
A-RT O 1 2 3 4 5 6
2.
li1
20
-2
B
I02El
Fig. 1. QC-PCR quantitation of HIV-1 RNA.Video images are shown of electrophoreticallyresolved, ethidium bromide-stained PCR prod-ucts derived from wild-type target sequence(upper band, 260 bp) and the competitivetemplate (lower band, 180 bp), along with plotsused to determine copy numbers of targetsequence in the specimens (13, 17). (A) Re-construction experiment with in vitro transcriptsfrom plasmid (pQP1) containing full-length tar-get sequence (260-bp product) and plasmid(pQP1iA80) containing internally deleted com-petitive template (1 80-bp product). Actual copynumber added, 2000 copies pQP1 per reac-tion. (B) Quantitation of virion-associated HIV-1RNA from plasma of BECH 01 71, CDC stage 11(Table 1). Lanes are as follows: -RT, PCRamplification without reverse transcription (torule out contaminating DNA); lanes 0 to 6, 0,100, 500, 1,000, 5,000, 10,000, and 50,000copies of competing pQP1 A80 transcript perreaction, respectively. Note absence of signalin all -RT lanes and absence of 180-bp signalin all 0 lanes (no competitive template added).
~.y= -.4531 + 1.54D'IDr2 =Q0.981
Coplesrn = 2062
2 3 4 5~og(Competing copi.. added)
2 3 4
log(Compdting copies added)
1 ., c ACC4 * 4 CAC12-
1 .
Table 1. Virologic and clinical summary for 66 consecutively studied HIV-1-infected patients.
CD4+ HIV p24 Ag PlasmallPatient ID cells* HIV RNAt culture
(cells/ (copies/mi) ICDt Reg§ (TCID/mm3) (pg/mi) (pg/mi) ml)
920853817739358262
1080823760720705703678678644640627624616521458350323281270257231197
61553345643542542436033226724323614111710995826754292721107
CDC stage 1¶2,617,4001,485,0002,398,6002,427,900355,400
21,783,600CDC stage 11 or I1140,80036,0009,400
38,900113,40024,2001 001413,800
198,80078,7004,400
13,800128,30034,400
586,10067,20091,00046,1009,400
92,9008,600
84,900CDC stage IVC291,800114,50040,80049,10032,800192,500341,00033,30073,10094,700173,600
2,200,000223,00036,900
738,900655,500104,500469,900191,800448,200625,900687,900479,300
386892218275258
5,072
6601,250258360329
5,406
390
0
0
310
0
80
0
0
0
0
0
0
0
154183
0
5740
0
1256251#
1,0000**
1 0,000
0
0
0
0
0
0
0
70
0
0
0
0
0
0
0
53350
3500
0
0
0
350
0
0
5,0000
0
0
79740123361765179221209
0
173187396823
0
0
0
0
0
0
0
0
0
0
0
0
0
0
50
0
0
0
50
0
0
0
50
0
0
1,0700
0
0
0
2841011209011927200
15322640
238
0
0
0
0
0
0
1000
0
250
3,125125
0
0
525
3,1250
525255
CDC stage IVC1DEDA 0006 57 1,309,000 233 14 10,000VAST 0181tt 56 664,500 450 25 0DODO 0116 50 1,667,000 357 385 100,000SZHO 1173tt 37 808,300 631 195 625MILA 0284tt 32 815,100 0 0 0BIJA 0205 32 4,744,000 1,920 1,050 10,000NAPH 0073 14 1,804,000 480 480 1,000TIMI 0018 14 3,445,000 330 390 100WATI 0855 7 9,300,000 606 302 625MCMI 0063 7 232,000 221 12 625LENA 1029 <5 2,500,000 53 0 125FARO 1042 <5 4,800,000 199 204 625RUTH 1145 <5 424,800 230 143 125YOAL 0522 <5 2,600,000 205 5 125JOJI 0070 <5 1,606,000 2,460 885 100,000
*CD4+ cells determined by flow cytometry within 6 weeks of sampling for virologic assessments. tHIV RNAcopy number per milliliter of plasma, as determined by QC-PCR (13, 17). *immune complex-dissociated HIVp24 antigen (19). §HIV p24 antigen (19). IlTissue culture infectious doses of virus per milliliter of plasma(18). ¶Centers for Disease Control classification system for staging of HIV-1 infection (2). #Three dayspreviously, plasma cultures were positive at 10 TCID/ml, with HIV-1 RNA level of 1,350,600 copies/mI. **Eightdays previously, plasma cultures were positive at 1,000 TCID/ml, with HIV-1 RNA level of 216,400 copies/mI (Fig.2.). ttAZT or dideoxyinosine therapy at time of study. #4Extrapolated value.
SCIENCE * VOL. 259 * 19 MARCH 1993
ing assay or protein immunoblot, and eachsubsequently seroconverted with a full spec-trum of HIV-1-specific antibodies by day 50(7). Virion-associated HIV-1 RNA levelspeaked between 8 and 23 days after theonset of symptoms, reaching values be-tween 3.55 x 105 and 2.18 x 107 copiesper milliliter (corresponding to 1.78 x 105to 1.09 x 107 virions per milliliter). Cul-turable virus [10 to 10,000 tissue cultureinfectious dose (TCID) per milliliterl andp24 antigen (258 to 5,406 pg/ml) peaked atapproximately the same time as virionRNA levels, declining rapidly thereafter inparallel with virion RNA levels. Withinthe first 100 days after onset of symptoms,plasma RNA levels fell by between 20- and235-fold from peak levels but, in markedcontrast to p24 antigen and culturable vi-rus, remained continuously quantifiable forthe duration of follow-up in all patients.
Among three patients who presentedwith symptomatic acute infection (Fig. 2),there appeared to be associations betweenprofiles of viral load in plasma over time, asdetermined by RNA QC-PCR, and trendsin CD4' T cell counts and clinical status.For example, patient SUMA (Fig. 2A)showed an initial peak of HIV-1 RNA(1.49 x 106 copies per milliliter) thatdecreased by a factor of nearly 1 x 102within the first month of follow-up. ViralRNA reached a minimum (2.30 x 103copies per milliliter) at 278 days and re-mained at or below 1.50 x 104 RNA copiesper milliliter up to 473 days of follow-up.Plasma viral cultures and p24 assays re-mained negative after the initial peak. Thispatient maintained a normal CD4+ T cellcount (952 to 1108 per cubic millimeter)and remained entirely asymptomatic overthe period of follow-up. In contrast, patientFASH (Fig. 2B) showed the highest initialvirus peak (2.18 x 107 copies per millili-ter), as well as the highest persistent levelsof circulating virus, in the range of 3.26 x105 to 6.32 x 105 copies per milliliter over301 days of follow-up. The highest CD4+ Tcell count measured in this patient was 282per cubic millimeter, with a subsequentprogressive decline to 128 per cubic milli-meter and clinical progression to CDCstage IVC2, despite antiviral therapy. Al-though anecdotal, these observations sug-gest that higher levels of circulating virusand failure to effectively control virus rep-lication after initial infection may be asso-ciated with a negative prognosis. However,in addition to quantitative viral load, othervirologic or immunopathologic factors, in-cluding a syncytium-inducing viral pheno-type, likely contribute to variable rates ofCD4+ T cell depletion and clinical out-come (23). In this regard, patient WEAU(Fig. 2C), who experienced a sharp declinein CD4+ cells despite a peak level of viral
1751
HOBR 0961SUMA 0874BORI 0637INME 0632WEAU 0575FASH 1057
RIPH 0179HIDO 1099ATKA 0381HOJU 0143MAMA 0341ADFR 0194TIMI 0852ttBECH 0171MALE 0264ROJO 0331BELI 1233HAJO 0940ttADDI 0101FOMA 0784ttJUJA 0156WATH 0272ttBAMA 0037WAJO 1286BAST 0514ttSTMI 0862STBO 1287SLMI 0843
ALFR 0229DOBE 0859CALI 0950ttARLA 0846WOAL 0263ttGADA 1162ttMCSE 0176CHJI 0774EDWI 0817GRJO 0849MIWI 1278SMDO 0157ttHEMI 0562DAJO 0306SMST 1012DUSE 1021WHWI 1106NOWR 1192MIGE 1132ttDABE 0775EMJA 0809CLRA 0703LARO 0833
Is. .
Table 2. AZT treatment: Virologic and clinical data summary. Table shows values for patients beforeand after treatment with AZT for the indicated periods. ND, not determined.
HIV p24 Ag* Plasma Time onPatient ID HIV RNA culture AZT treatment(copies/mI) lCID Reg (TClD/mI) (weeks)
(pg/mI) (pg/mi)MCSE 0176 341,000 5,000 1,070 100 -
118,000 5,025 1,530 1 + 17NOWR 1192 469,900 209 20 3,125 -
82,400 41 0 ND + 1EMJA 0809 625,900 187 226 25 -
75,500 195 216 ND + 6TIMI 0018 3,450,000 330 390 100 -
235,000 35 35 0 + 2NAPH 0073 1,800,000 480 480 1,000 -
182,000 320 320 100 + 17MCMI 0063 232,000 221 12 625 -
134,300 160 7 ND + 6JOJI 0070 1,606,000 2,460 885 100,000 -
40,800 263 58 10 + 20SLMI 0843 84,900 0 0 0 - Week Ot
18,000 0 0 ND + Week 133,500 0 0 ND + Week 228,100 0 0 ND + Week672,700 0 0 0 - Week 7
ARLA 0846 49,100 0 0 0 - Week 07,300 0 0 ND + Week 16,500 0 0 ND + Week 211,200 0 0 ND + Week 658,400 0 0 0 - Week 7
MIWI 1278 173,600 79 0 0 - Week 021,900 28 0 ND + Week 110,900 24 0 ND + Week 29,200 31 0 ND + Week 6
136,300 47 0 25 - Week 7
*Parameters as for Table 1. tFor kinetic analysis of viral load by RNA QC-PCR over a 6-week period oftreatment with AZT, patients were studied before initiation of treatment (week 0), after 1, 2, and 6 weeks oftreatment, and 1 week after temporary discontinuation of treatment (week 7).
cell counts argue strongly for a direct rolefor HIV-1 replication in the pathogenesis ofHIV disease. These results confirm andextend work by Coombs (5), Ho (5), Pan-taleo (9), Schnittman (31), and Michael(32) who, by alternative approaches, alsofound evidence of continuous viral replica-tion related to clinical stage and diseaseprogression. The finding of a large propor-tion of circulating virus that is not cultura-ble raises the possibility that geneticallydefective (30) or otherwise noninfectiousvirus may contribute importantly to HIV-1pathogenesis, in keeping with precedents inanimal retrovirus systems (33). Numerousmechanisms by which noninfectious parti-cles might contribute to the pathogenesis ofHIV-1 infection have been proposed (34).Furthermore, recent observations have doc-umented the presence of HIV-1-reactivecellular immune responses in patients whohave been exposed to HIV-1 by sex or use ofshared needles but who are not demonstrablyinfected with the virus (35). Exposure to aninoculum consisting largely of noninfectiousparticles might produce such a result, even
without the establishment of productive in-fection of the exposed individual.
Finally, our data suggest that determina-tion of virion-associated HIV-1 RNA levelsin plasma by QC-PCR represents a marker
of viral replication with potential for wide-spread applicability in assessment of theactivity of antiretroviral therapy. There iscurrently a widely recognized need for newmarkers that allow timely assessment of thein vivo antiviral activity of new therapeuticapproaches and agents and that better pre-dict the ultimate clinical efficacy of new
treatments (36). Ideally, such markersshould bear a biologically plausible relationto the disease process, be present in most or
all patients, change rapidly in response toeffective therapy, be derived from readilyobtained clinical specimens, and correlatedirectly with eventual clinical outcome.
Quantitation of HIV-1 virion levels in plas-ma as determined by QC-PCR satisfies thefirst four of these requirements. Future stud-ies making use of QC-PCR methods willhelp to determine the relation betweenvirion-associated HIV-1 RNA levels andclinical outcome and the role of persistentviral replication and viremia in the patho-genesis of HIV infection and AIDS.
REFERENCES AND NOTES
1. R. Anderson, AIDS 1, 241 (1988); A. Munoz et al.,Am. J. Epidemiol. 130, 530 (1989); P. Bacchettiand A. R. Moss, Nature 338, 251 (1989).
2. Morbid. Mortal. Weekly Rep. 35, 334 (1986).3. A. S. Fauci, Science 239, 617 (1988); A. S. Fauci,
SCIENCE * VOL. 259 * 19 MARCH 1993
S. M. Schnittman, G. Poli, S, Koenig, G. Pantaleo,Ann. Intern. Med. 114, 678 (1991); H. C. Lane etal., N. Engi. J. Med. 313, 79 (1985); A. Amadori etal., AIDS Res. Hum. Retroviruses 6, 581 (1990).
4. G. Pantaleo, C. Graziosi, A. S. Fauci, N. Engi. J.Med. 328, 327 (1993).
5. R. W. Coombs etal., ibid. 321, 1626 (1989); D. D.Ho, T. Moudgil, M. Alam, ibid., p. 1621.
6. M. S. Saag etal., J. Infect. Dis. 164, 72 (1991).7. S. J. Clark et al., N. EngI. J. Med. 324, 954 (1991).8. E. S. Daar, T. Moudgil, T. Meyer, D. D. Ho, ibid, p.
961.9. G. Pantaleo etal., Proc. NatI. Acad. Sci. U.S.A. 88,
9838 (1991); C. Graziosi, G. Pantaleo, D. P.Kotler, A. S. Fauci, Clin. Res. 40, 333 (1992).
10. M. Ottmann et al., J. Virol. Methods 31, 273(1991); G. H. Murakawa et al., DNA 239, 295(1988).
11. M. Holodniy, M. A. Winters, T. C. Merigan, Bio-Techniques 12, 36 (1992); M. Holodniy, D. A.Katzenstein, D. M. Israelski, T. C. Merigan, J. Clin.Invest. 88, 1755 (1991); S. Aoki et al., AIDS Res.Hum. Retroviruses 8, 1263 (1992).
12. K. E. Noonan et al., Proc. Nati. Acad. Sci. U.S.A.87, 7160 (1990); R. E. Dickover et al., J. Clin.Microbiol. 18, 2130 (1990).
13. M. Piatak, K.-C. Luk, B. Williams, J. D. Lifson,BioTechniques 14, 70 (1993).
14. G. Gilliland, S. Perrin, K. Blanchard, H. F. Bunn,Proc. Natl. Acad. Sci. U.S.A. 87, 2725 (1990); G.Gilliland, S. Perrin, H. F. Bunn, in PCR Protocols:AGuide to Methods and Applications, M. A. Innis,D. H. Gelfand, J. J. Sninsky, T. J. White, Eds.(Academic Press, San Diego, CA 1990), pp. 60-69; M. Becker-Andre and K. Hahlbrock, NucleicAcids Res. 17, 9437 (1989).
15. M. Stieger, C. Demoliere, L. Ahlborn-Laake, J.Mous, J. Virol. Methods 34, 149 (1991); S. Menzoet al., J. Clin. Microbiol. 30, 1752 (1992); S.Jurriaans, J. T. Dekker, A. De Ronde, AIDS 6, 635(1992); D. T. Scadden, 0. Wang, J. Groopman, J.Infect. Dis. 165,1119 (1992); P. Bagnarelli et al.,J. Virol. 66, 7328 (1992).
16. Human Retroviruses and AIDS. A Compilationand Analysis of Nucleic Acid and Amino AcidSequences, G. Myers, B. Korber, J. A. Berzofsky,R. F. Smith, G. N. Paviakis, Eds. (Los AlamosNational Laboratory, Los Alamos, NM, 1991).
17. All patients were evaluated at the University ofAlabama, Birmingham, where approval from thehuman subjects review board and informed con-sent were obtained. Blood specimens were col-lected in acid-citrate-dextrose and processedwithin 3 hours of phlebotomy. After centrifugation(200g for 15 min), plasma was collected andcentrifuged again (1000g for 15 min) to ensurecell-free specimens. Replicate portions of plasmawere used immediately for virus culture or stored at<-700C until further analysis. For extraction ofvirion-associated RNA, plasma samples werethawed and subjected to ultracentrifugation (Beck-man type 70.1 rotor, 40,000 rpm, 1 hour) to pelletvirions. Pellets were resuspended in solution [20mM tris (pH 7.5), 150 mM NaCI, 2 mM EDTA],adjusted to contain 1 mg of proteinase K permilliliter and 0.1% SDS, and incubated at 37'C for1 hour. After repeated extraction with phenol:chlo-roform:isoamyl alcohol (24:24:1) followed by oneextraction with chloroform, samples were adjustedto contain -40 Ag of glycogen per milliliter (ascarrier) and 10 ng of 7.5-kb synthetic RNA permilliliter (BRL, Bethesda, MD) (as carrier and tonormalize total RNA content in the reverse tran-scription and PCR steps). RNA was then precipi-tated with ethanol at -20'C for 48 hours andpelleted by ultracentrifugation to maximize recov-ery. The RNA pellets were partially dried and thendissolved in 100 gi of sterile, ribonuclease-freewater. Samples were stored at -70'C until subse-quent analysis. The primers GAGO4 (CATICTATT-TGTTCITGAAGGGTACTAG) and GAG06 (GCIT-TIAGCCCIGAAGTIATACCCATG) have been de-scribed previously (13) and were designed toamplify an internal fragment of either 260 bp (fromwild-type HIV-1 target sequences) or 180 bp (fromthe pQP1A80 competitive template). To maintain
1753
-Wl. Bookseller
on
Nov
embe
r 10
, 200
7 w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from
equivalent and competitive priming efficiency withdivergent sequences, primers incorporated inos-ine residues at the few positions where divergencefrom the conserved consensus sequence hasbeen reported (13, 16). For QC-PCR analysis, twoplasmids were prepared, one containing the targetsequence (pQP1) and the other containing theidentical sequence except for an 80-bp internaldeletion (pQP1A80), sufficient to allow the derivedPCR products to be readily resolved by electro-phoresis (13). In vitro RNA transcripts were pre-pared with commercially available kits. Final prep-arations in water were determined to be essentiallyfree of degradation products by Northem (RNA)blot analysis and were quantified by measurementof absorbance at 260 nm. Portions were stored at-70'C until needed. PCR reaction conditions andprotocols were generally similar to those found incommercially available kits (Perkin-Elmer, Norwalk,CT). Each test sample was divided into eightreplicate portions and analyzed in the presence ofo to 50,000 copies per reaction of competitivetemplate. The initial reaction was performed in atotal volume of 30 gi and contained 5 gi of test RNA(corresponding to 5% or less of the total speci-men), 5 Fl of competing RNA preparation or water,and 30 U of cloned Moloney virus reverse tran-scriptase (BRL, Bethesda, MD). One portion fromeach specimen was analyzed without reverse tran-scription and in the absence of competitive tem-plate. After 10 min at room temperature to allow forpartial extension and stabilization of random hex-amer primers, conversion of RNA into cDNA wasallowed to continue for 30 min at 42"C. This reac-tion was then adjusted to contain primers andadditional buffer in a total volume of 60 jil. Ampli-fication was performed as described (13), with 45cycles (940C for 1 minm 50MC for 2 min, and 72"C for1 min), followed by a final incubation at 5'C for 5min. After amplification, approximately 7% of eachreaction product mixture was separated by elec-trophoresis in composite 2% Synergel (DiversifiedBiotech, Newton Center, MA)-1% agarose (FMCBioproducts, Rockport, ME) gels in 20 mM trisacetate (pH 7.8) and 1 mM EDTA. Gels werestained with ethidium bromide for visualizationunder ultraviolet illumination. Quantitation of fluo-rescence of both wild-type and competitive tem-plate product bands was performed on a Lynx4000 molecular biology workstation with matchedcustom software (Applied Imaging, Santa Clara,CA), as described (13). Competition equivalencepoints were determined by interpolation on plots ofthe logarithm of the calculated ratio of signal for thecompetitive template-derived product over thesignal for the wild-type target sequence-derivedproduct (corrected for molar ratio) versus the log-arithm of the copy number of added competitivetemplate (Fig. 1) (13).
18. Endpoint dilution cultures of plasma for cell-freeinfectious virus were determined for fresh speci-mens, generally in duplicate or quadruplicate, asdescribed (6, 7). Plasma samples were not fil-tered (to eliminate the possibility of inadvertentloss of virus).
19. Regular and ICD p24 antigen determinationswere performed in duplicate with the CoulterDiagnostics kit assay, according to the manufac-turer's recommendations (Coulter, Hialeah, FL).
20. M. Piatak et al., unpublished observations.21. Analysis of variance, Duncan's multiple range
test, and Tukey's Studentized range (HSD) test ascalculated by the SAS statistical analysis softwarepackage (SAS Institute, Cary, NC). Plasma RNAdata were subjected to logarithm transformationbefore statistical analysis.
22. Spearman correlation coefficients as calculatedby the SAS statistical analysis software package.Plasma RNA data were subjected to logarithmtransformation before statistical analysis.
23. M. Tersmette et al., Lancet 335, 983 (1989); M.Tersmette et al., J. Wrol. 63, 2118 (1989); D. S.Stein, J. A. Korvick, S. H. Vermund, J. Infect. Dis.165, 352 (1992).
24. Paired t test analysis on logarithm-transformeddata set.
25. S. P. Layne et al., Virology 189, 695 (1992); A. S.Bourinbaiar, Nature 349, 111 (1991); J. A. Mc-Keating, J. P. Moore, ibid., p. 660.
26. F. Barr6-Sinoussi et al., Science 220, 868 (1983);M. Popovic, M. G. Sarngadharan, E. Read, R. C.Gallo, ibid. 224, 497 (1984); J. A. Levy etal., ibid.225, 840 (1984).
27. To calculate ratios of total virions determined byRNA QC-PCR analysis to culturable virus, weassigned specimens that did not contain cultura-ble virus a TCID value of 1 to avoid denominatorsof 0. The ratios calculated therefore represent alower estimate of the true ratios of total circulatingvirions to culturable virus.
28. R. A. Weiss et aL, Nature 316, 69 (1985); M.Robert-Guroff, M. Brown, R. C. Gallo, ibid., p. 72;P. L. Nara, R. R. Garrity, J. Goudsmit, FASEB J. 5,2437 (1991).
29. J. P. Moore, J. A. McKeating, R. A. Weiss, Q. J.Sattentau, Science 250, 1139 (1990).
30. Y. U et al., J. Virol. 65, 3973 (1991); Y. U et al.,ibid. 66, 6587 (1992); J. Meyerhans etal., Cell58,901 (1989).
31. S. M. Schnittman, J. J. Greenhouse, H. C. Lane, P.F. Pierce, A. S. Fauci, AIDS Res. Hum. Retrovi-ruses 7, 361 (1991).
32. N. L. Michael, M. Vahey, D. Burke, R. R. Redfield,J. Virol. 66, 310 (1992).
33. J. Overbaugh, P. R. Donahue, S. L. Quacken-bush, E. A. Hoover, J. I. Mullins, Science 239, 906(1988); D. C. Aziz, Z. Hanna, P. Jolicoeur, Nature338, 505 (1989); S. K. Chattopadhyay, H. C.Morse Ill, M. Makino, S. K. Ruscetti, J. W. Hartley,Proc. Nati. Acad. Sci. U.S.A. 86, 3862 (1989).
34. J. B. Margolick, D. J. Volkman, T. M. Folks, A. S.Fauci, J. Immunol. 138, 1719 (1987); D. L. Mann
et al., ibid., p. 2640; M. R. Shalaby et al., Cell.Immunol. 110, 140 (1987); L. 0. Arthur et al.,Science 258, 1935(1992); L. Imberti, A. Sottini, A.Bettinardi, M. Puoti, D. Primi, ibid. 254, 860(1991); C. Terai, R. S. Kombluth, C. D. Pauza, D.D. Richman, D. A. Carson, J. Clin. Invest. 87, 1710(1991); A. G. Laurent-Crawford et al., Wrology185, 829 (1991); J. C. Ameisen and A. Capron,Immunol. Today 12, 102 (1991); H. Groux etal., J.Exp. Med. 175, 331 (1992); M. K. Newell, L. J.Haughn, C. R. Maroun, M. H. Julius, Nature 347,286 (1990); L. Meyaard et al., Science 257, 217(1992); R. S. Mittler and M. K. Hoffman, ibid. 245,1380 (1989); G. P. Unette, R. J. Hartzman, J. A.Ledbetter, C. H. June, ibid. 241, 573 (1988).
35. M. Clerici, J. A. Berzofsky, G. M. Shearer, C. 0.Tacket, J. Infect. Dis. 164, 178 (1991); M. Clericietal., ibid. 165, 1012 (1992).
36. J. Cohen, Science 258, 388 (1992).37. We thank the patients who provided specimens
for this study; R. Whitley for helpful discussions; A.Fauci for sharing work before publication; S. Camp-bell-Hill and P. Chopra for technical assistance; S.Chen for assistance with statistical analysis; and R.Cuevas and A. Chaprnan for assistance with prep-aration of the manuscript. Supported in part bygrants A125922 (to J.D.L), A127290 (to G.M.S.), andA132775 (to M.S.S.) from the National Institutes ofHealth; by grant DAMD 17-90-C-0064 (to G.M.S.)from the Department of Defense; and by the coreresearch facilities of the University of Alabarna, Bir-mingham, Center for AIDS Research, the Birming-ham Veterans Administration Medical Center, andthe General Clinical Research Center.
10 December 1992; accepted 5 February 1993
Requirement of the Carboxyl Terminus of a BacterialChemoreceptor for Its Targeted ProteolysisM. R. K. Alley, Janine R. Maddock, Lucille Shapiro*
The bacterium Caulobacter crescentus yields two different progeny at each cell division;a chemotactically competent swarmer cell and a sessile stalked cell. The chemotaxisproteins are synthesized in the predivisional cell and then partition only to the swarmer cellupon division. The chemoreceptors that were newly synthesized were located at thenascent swarmer pole of the predivisional cell, an indication that asymmetry was estab-lished prior to cell division. When the swarmer cell differentiated into a stalked cell, thechemoreceptor was specifically degraded by virtue of an amino acid sequence located atits carboxyl terminus. Thus, a temporally and spatially restricted proteolytic event was acomponent of this differentiation process.
Cell divisions that yield two different prog-eny cells are fundamental to developmentalprograms in all organisms (1). In Caulobac-ter crescentus, the generation of two distinctcell types at each cell division is due, inpart, to the asymmetric distribution of pro-teins in the cell before it has divided. Thepredivisional cell assembles a flagellum andseveral pili at one pole, which are thenpartitioned to the swarmer cell progeny. Inaddition to bearing a flagellum and polarpili, the swarmer cell is chemotacticallycompetent. The chemoreceptor McpA,
Department of Developmental Biology, Beckman Cen-ter, Stanford University School of Medicine, Stanford,CA 94305.*To whom correspondence should be addressed.
which is one of the family ofMCPs (methylaccepting chemotaxis protein), is synthe-sized only in the predivisional cell (2).Thus, the swarmer cell inherits McpA fromthe predivisional cell (3).We now describe our studies of the
spatial distribution of the newly synthe-sized chemoreceptor in the predivisionalcell and its subsequent fate throughout thecell cycle. Assays in vitro have shown thatthe methyl-accepting activity of thechemoreceptors, and the activities of themethyltransferase and methylesterase arelost during the transition of a swarmer cellinto a stalked cell (4). The McpA chemo-receptor is positioned at the flagellatedpole of the swarmer cell (3), and we nowshow that it is degraded during the transi-
SCIENCE * VOL. 259 * 19 MARCH 1993
limliliiiiiiiiiiiiiiii
1754
on
Nov
embe
r 10
, 200
7 w
ww
.sci
ence
mag
.org
Dow
nloa
ded
from