Molecular Profile of Human Immunodeficiency ... - Journal of Virology

2 downloads 0 Views 2MB Size Report
Molecular Profile of Human Immunodeficiency Virus Type 1. Infection in Symptomless Patientsand in Patients with AIDS. PATRIZIA BAGNARELLI,1 STEFANO ...
Vol. 66, No. 12

JOURNAL OF VIROLOGY, Dec. 1992, p. 7328-7335

0022-538X/92/127328-08$02.00/0 Copyright © 1992, American Society for Microbiology

Molecular Profile of Human Immunodeficiency Virus Type 1 Infection in Symptomless Patients and in Patients with AIDS PATRIZIA BAGNARELLI,1 STEFANO MENZO,1 ANNA VALENZA,1 ALDO MANZIN,1 MAURO GIACCA,2 FAUSTO ANCARANI,3 GIORGIO SCALISE,3 PIETRO E. VARALDO,1 AND MASSIMO CLEMENTI1* Institute of Microbiology' and Clinic of Infectious Diseases,3 University of Ancona Medical School, I-60100 Ancona, and International Center for Genetic Engineering and Biotechnology, I-34100 Trieste, 2 Italy Received 13 April 1992/Accepted 9 September 1992

Recent molecular evidence indicates that active human immunodeficiency virus type 1 (HIV-1) infection is detectable in both symptomless and symptomatic infected patients. For this main reason, it has been pointed out that precise quantitative analysis of viral activity in vivo is necessary, firstly, for the pathogenetic investigation of the steps relevant to infection progression and, secondly, for better clinical management of HIV-1-infected patients. In this study, the presence of HIV-1 genomic RNA in plasma samples, specific HIV-1 transcripts in peripheral blood mononuclear cells, and proviral DNA sequences were assayed for 33 HIV-1-infected patients (including symptomless and symptomatic subjects) by using a competitive polymerase chain reaction method that allows quantitation of the RNA/DNA target sequences. The quantitative results obtained confirm that transcription of HIV-1 structural genes and complete viral replication occur in all the HIV-1-infected patients independently of the clinical stage. However, although sharp individual differences were detected, a high degree of correlation of the molecular parameters studied with both disease progression and a decrease in the number of CD4+ T lymphocytes was documented. Interestingly, despite the increasing viremia level associated with infection progression, the mean transcriptional activity of individual infected cells was found to be only moderately greater in AIDS patients than in asymptomatic infected subjects. In addition, it was noted that quantitation of HIV-1 genomic RNA in plasma samples and quantitation of specific HIV-1 transcripts in peripheral blood mononuclear cells appear to be more reliable and sensitive markers of viral activity than quantitative analysis of proviral HIV-1 sequences in peripheral lymphocytes.

(27) and the analysis of viral activity in infected individuals (5, 10, 18, 25, 30). Those studies have provided convincing evidence that substantial ongoing transcription of viral genes in peripheral blood mononuclear cells (PBMCs) occurs in almost all HIV-1-infected individuals independently of their clinical and immunological conditions; accordingly, a similar approach has allowed genomic HIV-1 RNA in plasma samples to be detected in both symptomatic and asymptomatic infected subjects (4, 26). These data have clearly indicated that sensitive molecular assays for the quantitation of HIV-1 viremia levels, specific HIV-1 transcripts in PBMCs, and proviral DNA sequences are urgently needed for the more correct clinical management of symptomatic and symptomless infected patients, before and during the course of specific anti-HIV-1 treatment. Furthermore, precise pathogenetic investigation of the steps leading to AIDS could benefit greatly from quantitative molecular assays (7). For these main reasons, we have recently developed a new quantitative reverse transcription (RT)-PCR method (23) based on competitive RT (cRT) and subsequent PCR amplification of a modified (internally deleted) RNA template. The competitive RNA template bears the same primer recognition sequences as do those of the wild-type template to be quantified, thus ensuring identical thermodynamics and amplification conditions for both templates. The technique allows quantitation of HIV-1 viremia levels, HIV-1 specific cellular transcripts, and, by using competitor DNA (competitive PCR [cPCR]), proviral DNA sequences in PBMCs. For this research, we have used this molecular approach to study a group of HIV-1-infected subjects, including asymptomatic and symptomatic patients. Thirty-three patients representing all stages of HIV-1 infection have been

The human immunodeficiency virus type 1 (HIV-1), first isolated from patients with AIDS or with clinical signs that frequently precede AIDS (8, 16, 28), has subsequently been associated with several clinical conditions, including primary acute illness and asymptomatic infection (14). It has recently been estimated that at least 50% of HIV1-seropositive patients develop AIDS within 10 years after infection (20). Although the presence of viral strains with different biological characteristics has been observed in asymptomatic HIV-1-infected subjects and AIDS patients (3, 9, 11, 13, 34), little is presently known about the mechanism involved in the progression of infection toward AIDS at the molecular level. Since understanding the HIV-1 sequence mutation is central to current evolution of the knowledge of AIDS pathogenesis, this particular aspect has been extensively investigated. However, despite the large amount of information on the genetic heterogeneity of HIV-1 isolates in the different phases of the infection (6, 12, 15, 22, 29, 32, 35), no evidence has been provided that the presence of a particular virus genotype is associated with disease onset and development. Similarly, by using molecular assays, detailed information about several factors that influence HIV-1 activity in vitro has been obtained during the last few years (see reference 17 for a review), but the actual pathogenetic importance of these cofactors lacks adequate in vivo confirmation. In the last few years, polymerase chain reaction (PCR) technology has supplied a unique means of performing both the highly sensitive molecular diagnosis of HIV-1 infection *

Corresponding author. 7328

VOL. 66, 1992

checked for HIV-1 viremia, viral gag transcripts in PBMCs, and proviral HIV-1 sequences. The quantitative results presented here confirm that HIV-1 is expressed in all stages of the natural infection and that high correlation exists between levels of viral activity and disease progression and indicate that quantitation of HIV-1 viremia levels by cRTPCR is a reliable index to study infection in these patients at the molecular level. Accordingly, this parameter may be important either in the pathogenetic study of the natural history of HIV-1 infection or in the clinical approach to these patients and in the management of specific antiviral therapies.

MATERIALS AND METHODS Patients. Thirty-three HIV-1-seropositive subjects (2 Centers for Disease Control [CDC] class I, 16 CDC class II, 7 CDC class III, and 8 CDC class IV) were selected for this study at the Clinic of Infectious Diseases, University of Ancona Medical School, Ancona, Italy. Clinical samples were collected during the period from July to September 1991. The average age of the patients was 30 years (standard deviation, 7 years), with no significant difference in ages among subjects in different CDC classes. The time since the first laboratory diagnosis of HIV-1 infection averaged 30 months for class II patients, 29 months for class III patients, and 55 months for class IV patients. The risk factors associated with HIV infection were distributed as follows: intravenous drug addiction, 21 subjects (14 males and 7 females); heterosexual contact with an HIV-1-infected partner, 6 (1 male and 5 females); and homosexual activity, 6. Three CDC class II and four class IV patients were under treatment with zidovudine at the time of enrollment in the present study. Treatment conditions are given in Table 1. The control population was 20 HIV-seronegative healthy blood donors. Clinical samples and nucleic acid purification. EDTAtreated peripheral blood was centrifuged over a Ficoll density gradient. Plasma was recovered from the upper phase after centrifugation, and PBMCs were recovered from the top of the Ficoll layer and washed three times with phosphate-buffered saline. The plasma was subsequently centrifuged to clear the platelets and cell debris (2,800 x g, 10 min), and 1 ml of supernatant, diluted in 9 ml of RPMI 1640 medium (Whittaker, Walkersville, Md.), was ultracentrifuged at 150,000 x g for 2 h in a swing-out rotor (Kontron Instruments, Zurich, Switzerland). RNA was extracted from the plasma virion pellet and from the PBMC pellet by the guanidinium thiocyanate method, as previously described (4). DNA was extracted from nuclei (to minimize unintegrated DNA contamination) (5). The following substrates were analyzed by using the quantitative cPCR and cRT-PCR assays: (i) HIV-1 genomic RNA from plasma, (ii) virusspecific intracellular RNA (from PBMCs), and (iii) proviral HIV-1 DNA from nuclei (from PBMCs). cRT-PCR and cPCR. The gag fragment (nucleotides 1551 to 1665) was analyzed with primer pair SK38 and SK39 (27) in each sample. Synthesis of cDNA and amplification were carried out as previously described (23). Competitive analysis was performed by using, as an internal competitor, the plasmid pSKAN, a derivative of plasmid pBS (Stratagene, La Jolla, Calif.) in which the gag fragment with an 18-bp deletion was inserted downstream from the T3 RNA polymerase promoter. Briefly, plasmid pSKAN was transcribed in vitro after linearization, and competitor RNA was purified, treated with DNase, and quantified by spectrophoto-

MOLECULAR PROFILE OF HIV-1 INFECTION

7329

metric reading and gel electrophoresis. Each RNA sample (10 ml, equivalent to 100 ,ul of plasma and 500,000 PBMCs) was reverse transcribed along with 2 pl of increasing-copynumber competitor RNA for 30 min at 42°C in the presence of 100 U of Moloney murine leukemia virus reverse transcriptase (Bethesda Research Laboratories, Inc., Gaithersburg, Md.)-200 nM SK39 primer-0.2 mM deoxynucleoside triphosphates-20 U of RNasin, and subsequently amplified after the addition of 2.5 U of Taq DNA polymerase (Perkin Elmer Cetus, Norwalk, Conn.)-300 nM SK39 primer-500 nM SK38 primer. Reverse transcription and amplification were performed by using the same 1 x reaction buffer (containing 50 mM NaCl, 10 mM Tris-HCl [pH 8.3], 1.5 mM MgCl2, and 0.01% gelatin). DNA samples (20 pl, equivalent to 100,000 nuclei) were amplified in reaction tubes containing 10 ,ul of the competitor plasmid pSKAN at increasing copy numbers. The PCR profile (15 s of denaturation at 94°C, 15 s of annealing at 60°C, 30 s of extension at 72°C) was repeated for 50 cycles by using a GeneAmp PCR System 9600 (Perkin Elmer). All of the samples were tested once in a series of four different reactions (see below); it had been previously observed (23) that, under these conditions, the coefficient of variation for the assay is 4%. Competition analysis. Five microliters of each 100-,ul reaction sample was run on a 10% polyacrylamide gel. The upper band, corresponding to the 115-bp wild-type amplification product, was clearly distinguishable from the 97-bp deleted competitor product (Fig. 1) after ethidium bromide staining. The peak areas of DNA bands fluorescence emission were measured by means of a CDC video densitometer (Ultra Violet Products Ltd., Cambridge, United Kingdom) directly over the transilluminator. Other methods. Serum p24 antigen was assayed by using a commercial enzyme-linked immunosorbent assay (Du Pont, Wilmington, Del.), according to the manufacturer's instructions. Specificity of positive results was confirmed by neutralization tests. Primers SK38 and SK39 used in this study were synthesized in our laboratory by using solid-phase phosphoramidite chemistry in a DNAsm synthesizer (Beckman Instruments Inc., Fullerton, Calif.). Statistical analysis. Because of the asymmetric distribution of data, the following nonparametric statistical methods were employed to analyze the molecular data obtained: the Spearman rank correlation coefficient and the Mann-Whitney-Wilcoxon test. RESULTS

Optimization of cPCR. Preliminary experiments were performed in order to find a convenient range of competitor copy numbers to test clinical samples. The choice of such a range was critical, since it represented a compromise between using a wide span and the smallest number of reactions to achieve an accurate analysis, in order to reduce time and assay costs. A set of four reactions containing 10, 20, 100, or 500 competitor DNA (pSKAN) molecules, respectively, was found to cover the wild-type proviral DNA range in most clinical samples (standardized to 100,000 cell equivalents). Figure 2 shows four competition patterns. In a few cases, the wild-type DNA copy number was less than 10 and additional competition against five pSKAN molecules was performed to obtain more accurate quantitation. Densitometric analysis was carried out to calculate peak areas: the deleted area (DA) was multiplied by 1.1855 to correct for its

7330

J. VIROL.

BAGNARELLI ET AL. TABLE 1. Clinical and quantitative molecular data from 33 HIV-1-infected patients

tit N.4 CDCDclass

no.a

.fofn

%offCD4cells

PBMCs/jli of blood

NDb

in PBMCs

1 2

I I

ND

ND ND

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Avg CV

II II II II II II II II II II II II II II II II II II

1,443 1,313 974 902 804 728 725 697 663 588 580 367 364 350 216 206 683 52

29.9 41.6 31.0 22.1 26.9 17.3 28.9 29.8 22.2 21.3 26.3 17.3 21.3 11.6 11.3 12.4 23.2

19 20 21 22 23 24 25 Avg CV

III III III III III III III III III

1,368 826 714 516 465 397 393 668 52

30.7 27.2 21.1 16.4 20.9 26.9 12.9 22.3

26 27 28 29 30 31 32 33 Avg CV

IVc2 IVc2

449 344 265 236 219 24 21 14 197 83

12.4 10.3 15.7 8.9 24.8 3.6 1.8 0.8 9.8

IVcl IVcl IVcl IVcl IVcl IVc2 IV IV

ProviralDNA gagmRNA p24 antigen HIV-1 of (coples/mI of Copies/5 x 10i Copies/105 CD4+ Copies/105 Copies/105 CD4+ (pg/ma plasma) plasma) PBMCs PBMCs PBMCs PBMCs genome

82 40

28

25

9 42 24

40 10 53 16 67

729,000 54,000

ND ND

230 490 4,320 9,650 830 31,050 1,050 1,310 4,950 19,900 35,960 166,250 9,350 22,030 3,030 32,090 21,406 190

15 695 45 170 3,775 565 ND 155 50 785 1,120 21,225 155 1,820 3,175 5,675 2,628 206

654 ND 104 45 736 850 24,552 144 3,146 5,600 9,116 3,218 200

2,370 94,940 26,370 5,390 117,870 43,600 8,210 42,679 109

55 4,655 3,510 795 7,815 115 1,155 2,586 112

136,250 126,150 762,880 10,320 502,110 178,210 288,490 156,600 270,126 91

11,075 935 6,330 450 16,145 13,235 23,235 2,745 9,269 87

ND ND

ND ND

ND ND

RNA/DNA

ratio

ND ND

28 72 4 26 10 34 20 83

40 72 94 106 416 19 225 88 273 107 103

0.75 9.93 0.64 1.31 94.38 6.28 ND 2.58 0.63 7.85 8.00 58.96 7.75 14.00 63.50 33.38 20.66 140

36 3,344 3,324 967 7,443 86 1,791 2,427 107

4 6 24 50 80 66 48 40 74

13 22 114 305 381 246 372 208 76

2.75 155.00 29.25 3.16 19.54 0.35 4.81 30.69 182

17,834 1,807 8,115 1,011 12,963 72,720 258,167 70,385 55,375

132 22

1,063

16.78 8.50 316.00 9.00 147.00 12.85 62.80 30.50 75.43 143

10 334 29 154 2,799

157

4 14 14 26 8 18 10 12 16 20

4 10 22 206 74 18 61 119

13 34 45 118 30 104 34

213 26 112 88 5,659 4,111 2,308 1,698 127

a Data for patients 1 and 2 were obtained from sera collected 1 month after exposure and data for the other parameters were not available. CV, coefficient of variation. The following patients were treated with specific anti-HIV-1 chemotherapy with zidovudine (500 mg/day) for the duration of treatment specified: patient no. 8, 4 months; patient no. 10, 12 months; patient no. 13, 24 months; patient no. 27, 17 months; patient no. 30, 14 months; patient no. 31, 12 months; and patient no. 32, 7 months. Twenty control samples (from 20 HIV-1-seronegative individuals) yielded negative results for all virological parameters (data not shown). b ND, not done.

lower molecular ethidium bromide molar incorporation (DAc) and the wild-type area (WA). The ratio between these areas (DAc/WA = y) was plotted against the corresponding competitor copy number (x) for each of the four points, and M

1

2

3 4 5 6

bhn 11 5 97

FIG. 1. Competition analysis. Lanes 1 to 6, coamplification of a constant amount of pHXB2 (60 molecules) with pSKAN at twofold

increasing copy numbers (25 to 800). The amplified wild-type products are seen as the upper 115-bp band, and the deleted competitor is the lower 97-bp band. Lane M, molecular size markers.

a linear regression curve was extrapolated, as shown in Fig. 2. The number of HIV-1-specific templates present in the clinical sample was calculated according to the regression curve expression for DAc/WA = 1. Optimization of cRT-PCR. Preliminarily, the amount of contaminating DNA in clinical RNA samples was determined. An aliquot (the same amount used for investigating RNA; see below) of both viral genomic and intracellular RNA from each patient sample was amplified along with 10 competitor DNA molecules (the RT step was omitted). RNA extracted from pelleted virions yielded no competition at all, and a subsequent amplification without competitor DNA showed that only 4 of 33 samples gave a positive result. These data indicate that the very low number of DNA molecules present in these four samples was unable to compete with 10 competitor molecules. These results prob-

MOLECULAR PROFILE OF HIV-1 INFECIION

VOL. 66, 1992

7331

30 20 10 0

O

3 l o3 o2 ol ol FIG. 2. Competitive PCR for quantitative detection of proviral HIV-1 DNA. Gel electrophoresis of coamplifications is shown above each graph. Lanes (left to right) contain increasing copy numbers (10 to 500) of competitor DNA and DNA equivalent to 100,000 PBMCs. The graphs represent densitometric analysis of the ratio between the peak area of the deleted band (corrected = DAc) and wild-type band (y axis) plotted against competitor pSKAN copy number for each lane (x axis). Ratios higher than 25 or lower than 0.04 were excluded as being unreliable. l

11 ol

1

i

ably depend on the presence of either some HIV-1 virions bearing intravirionally reverse-transcribed DNA sequences (21) or residual proviral sequences from lysed infected cells. In contrast, intracellular RNA samples (Fig. 3) yielded positive competition for the presence of contaminating DNA in 60% of the cases. In all but three of the positive samples, the copy number of specific viral DNA sequences was quantified to be less than 10 molecules, and, in a further competition assay, the copy number of these three never exceeded 50 (we did not investigate whether it was episomal rather than degradation products of integrated proviral DNA

1

2

3

4

5

6

7

8

9

bp -115 97

FIG. 3. Presence of specific HIV-1 DNA sequences in intracellular RNA extracts (500,000 PBMCs) from nine subjects (lanes 1 to 9, respectively). The number of HIV-1 DNA contaminating molecules may be precisely quantified by using a competitive strategy. Preliminary competition assay (results of which are shown in the figure) was performed for each RNA sample against 10 molecules of competitor DNA, without previous RT. In some cases, further competition analysis (against five competitor DNA molecules) was necessary to precisely define the amount of contaminating DNA to be subtracted from the copy number of HIV-1-specific transcripts detected by cRT-PCR. The 115- and 97-bp bands represent the amplified wild-type products and the deleted competitor, respec-

tively.

TTTTIT'

-

1 03 i o4 FIG. 4. Competitive RT-PCR from plasma RNA. Gel electrophoresis of coamplification is shown above each graph, lanes (left to right) contain increasing copy numbers (50 to 6,250) of RNA competitor and RNA equivalent to 100 ,ul of plasma. The graphs represent densitometric analysis: y axis, DAc/WA; x axis, competitor RNA copy number in each lane. 1 01

1 02

1

03

1

04

01

1

o2

sequences). However, the DNA copy number was always subtracted from the copy number of HIV-1-specific transcripts detected by cRT-PCR, thus avoiding DNase treatment. Assays for clinical RNA samples were standardized as follows: purified RNA from 100 jil of plasma and 500,000 PBMCs were used in each competition experiment. Four reactions containing fivefold increases in competitor RNA molecules (50, 250, 1,250, and 6,250) were found adequate for covering wild-type RNA molecules in most samples. The results of these reactions, together with those of respective densitometric analysis, are shown in Fig. 4 (plasma RNA) and 5 (specific gag transcripts in PBMCs). In a few cases, however, the wild-type RNA copy number exceeded 6,250 and an additional competition (against 31,250 competitor RNA molecules) was performed. Molecular data from HIV-1-infected patients. Thirty-three HIV-1-infected patients were studied, regardless of their clinical status, by using cPCR and cRT-PCR. Table 1 shows the results obtained for these subjects and their clinical conditions (31 were in CDC classes II to IV and 2 were studied at seroconversion). HIV-1 viremia levels, specific HIV-1 gag transcripts, and proviral DNA sequences were detected in all of the seropositive subjects but in none of the 20 HIV-1-seronegative control individuals. In spite of this homogeneous positivity of the qualitative results, a marked variability in the quantitative data was observed among patients at different CDC stages of the disease and, in several cases, also within the same CDC stage. However, signs of a progressively increasing level of HIV-1 activity were evident, regardless of antiviral therapy (three patients in class II and four patients in class IV were under zidovudine treatment at the time of this study). In fact, quantitative data on

J. VIROL.

BAGNARELLI ET AL.

7332

3l ¢-7---]T 21

/ II

Iv

''' ' -"1

*i2'

10

6jI

/

01

10

10 2

1 03

1

04

l

14 I *

!

~~~

~

8.1

4

2-

-i

12

I

I

421

1 01 1 0o 1 0? 1 o3 1 °4 l 0Q FIG. 5. Competitive RT-PCR for intracellular RNA. Gel electrophoresis of coamplifications is shown above each graph; lanes (left to right) contain increasing copy numbers (50 to 6,250) of competitor RNA and RNA from 500,000 PBMCs. The charts represent densitometric analysis: y axis, DAc/WA; x axis, competitor RNA copy number in each lane.

HIV-1 viremia ranged from 230 to 166,250 genomes per ml of plasma (average 21,406) for CDC class II patients, from 2,370 to 117,870 genomes (average 42,679) for class III patients, and from 10,320 to 762,880 (average 270,126) for CDC class IV patients. Transcription of the gag coding sequence (shown as the number of gag transcript molecules per 5 x 10' PBMCs) was also related to clinical progression, but in this case, the increase in viral transcriptional activity was evident for only class IV patients (mean gag transcript copy number per 5 x 105 PBMCs: 2,628 for class II patients, 2,586 for class III patients, and 9,269 for class IV patients). Since infected monocytes represent a negligible proportion of total infected PBMCs (2, 31), HIV-1-specific cellular transcripts and provirus copy number were also corrected, assuming that infection occurs in only CD4+ T lymphocytes. The CD4+ T-lymphocyte proportion of PBMCs was calculated, and the raw results were divided by this fraction, thus obtaining data related to CD4+ T lymphocytes. Under these

conditions, the differences between the asymptomatic and symptomatic patients studied were more evident. In fact, although the number of proviral DNA copies (expressed as provirus copy number per 105 PBMCs) showed only moderate mean variation among different groups of patients (20 for class II patients, 40 for class III patients, and 61 for class IV patients), evidence that clinical progression of infection is associated with an increased proportion of infected target cells was underlined when the data were corrected according to the absolute CD4+ cell number per cubic millimeter. As mentioned above, two patients with an acute clinical syndrome following primary HIV-1 infection were studied (Table 1). In both cases, we analyzed serum samples obtained 1 month after the putative exposure for the presence of genomic RNA. Results indicate that after resolution of the acute illness, viremia levels may remain elevated. This finding agrees with that of a previous investigation which found a higher frequency of virus isolation in the plasmas of patients in the acute phase of disease, comparable to those in plasma samples from AIDS patients (1). Statistical analysis of quantitative data. In order to evaluate the possible significance of the molecular and clinical parameters observed for HIV-1-infected patients, we compared the results obtained from subjects with the different CDC classes of infection. Table 2 shows the ratios of the rough mean values for each series of parameters (calculated in Table 1) along with the relative P value, calculated by the MannWhitney-Wilcoxon test. Significant differences were observed between HIV-1-infected symptomless subjects (CDC classes II and III) and symptomatic patients (CDC class IV) for all the virological parameters analyzed. Nevertheless, HIV-1 viremia levels (10 times higher in CDC class IV than in CDC class II and III patients) and virus-specific cellular transcripts (about 19 times higher in class IV patients than in CDC class II and III patients) appeared to be the molecular parameters of HIV-1 activity in vivo, which were more closely associated with the clinical progression of infection. In order to obtain a reliable index of mean transcriptional activity associated with each integrated provirus, a new parameter (the RNA/DNA ratio; Table 1) was calculated for each patient by dividing the number of transcripts by the number of proviral copies from 105 PBMCs. Interestingly, the values obtained showed considerable variability among patients in each CDC class, while the observed 3-fold increase in mean values between symptomless and symptomatic patients (Table 3) was minimal compared with the approximately 19-fold increase in total transcriptional activity (RNA in 105 CD4+ T cells). The progression of virological results with respect to the CDC stage (whose statistical analysis is reported in Table 2) is graphically shown in Fig. 6. Moreover, the number of

TABLE 2. Comparison of molecular results and CDC clinical stages of disease for HIV-1-infected patients

(P)0 HIV-1 pro*ses/

Mean value

CDC class

III/II IV/II IV/III IV/(II + III)

No. of CD4+ PBMCs/pJ of blood

0.98 0.29 0.29 0.29

No. of HIV-1 genomes/ml of plasma

1.99 (>0.1) (0.02) (0.0019) 12.62 (0.00085) 6.33 (0.0065) (0.003) (0.0005) 9.69 (0.0004)

No. of HIV-1 gag

transcripts/105 CD4+ PBMCs

0.75 17.21 22.81 18.67

(>0.1) (0.0033) (0.017) (0.0018)

105 CD4+

PBMCs

No. of CD4+

HIV-1

PBMCs/single HIV-1

proviruses/ml

0.96 (>0.1) 0.38 (0.025) 0.40 (>0.1) 0.39 (0.039)

1.88 (>0.1) 2.03 (>0.1) 1.08 (>0.1) 1.60 (>0.1)

1.94 (>0.1) 15.87 (0.032) 8.16 (>0.1) 12.30 (0.047)

RNA/DNA ratio

ratiosofblo 1.49 3.65 2.46 3.16

(>0.1) (0.035) (>0.1) (0.033)

a Mean values for data relative to classes II, III, and IV for the categories listed were compared by single division as defined in the first column. P values are

given for the Mann-Whitney-Wilcoxon comparison of data for the same parameters for the CDC classes.

MOLECULAR PROFILE OF HIV-1 INFECTION

VOL. 66, 1992

7333

TABLE 3. Spearman correlation between data for different molecular parameters Correlation coefficient (P)l No. of NNo.oofgenomes/ transcriptS/105 plasmea CD4+ cells

No. of

Parameter and patients or CDC class

proviruses/105

No. of ml of plasma

CD4+ cells

No. of CD4+ T lymphocytes/pl All cases II II + III IV

-0.60 (0.001) -0.52 (0.0527) -0.62 (0.0034) -0.50 (0.19)

-0.66 -0.65 -0.49 -0.29

No. of proviruses/105 CD4+ cells All cases II II + III IV

(0.0003) (0.015)

-0.72 -0.67 -0.59 -0.62

(0.0022) (0.449)

0.59 (0.0013) 0.86 (0.0012) 0.65 (0.0024) -0.19 (0.6143)

(0.071) (0.101)

0.62 (0.0008) 0.68 (0.0143) 0.57 (0.0038) 0.79 (0.0376)

No. of genomes/ml of plasma All cases II II + III IV a

(0.0001) (0.0157)

0.75 (0.0001) 0.59 (0.033) 0.56 (0.01) 0.40 (0.28)

Correlation coefficients and P values are given for the comparisons of the parameters listed.

infected target cells per single provirus copy was calculated assuming that there is approximately one copy of HIV-1 provirus per infected cell (31, 33). A higher number of infected cells in symptomatic patients was evident, with the mean values ranging from 2,143 CD4+ T lymphocytes per

106 A

0 0 0

0 0 0

10 o3

0 °

C

i

I

E

:* *

103o 4, lo I 0 0.

g *

102 -.

0

0

I

a.

I

CD

III

101

IV

S

10

. S 0

0

10'

* *

0 0

1°4IV1 ~~o~ 20 =S

0.

1o* -

0.l z~~~ 11 l S

1

10°

III

E

IV

III

IV

C.

* IIl l V

CDC CLASS 1i4

0

.

F

I * 1 I'

0 0

j.8

0

I 102

0 0

0

I 103 a.

03

0~~~~

I

I

11

IV

CDC CLASS

*

*

II

I 0

11

*

102

10 '

0

101

10 1

0

E

0

D

z

*

0

E 104

. .* 0

3.0 S

CDC CLASS

a

*a S 0

0

0

I

0

S

104

E

0

IV

B S

0

5

HIV-1 proviral DNA copy for patients in CDC classes II and III to 816 CD4+ T lymphocytes per HIV-1 provirus copy for patients in CDC class IV; P = 0.039 (Fig. 6E). No significant difference between symptomatic and asymptomatic patients (P, >0.1) (Fig. 6F) was observed for the concentration of

2

*

*

0

.h

0.

101.

10 1 11

III

11

IV

CDC CLASS CDC CLASS FIG. 6. Molecular parameters relative to the CDC stages of disease listed in Table 2 subjects.

are

III

IV

CDC CLASS presented graphically. Dots represent individual

7334

J. VIROL.

BAGNARELLI ET AL.

infected cells per milliliter of blood; this is not surprising, since it may directly depend on the decrease in target cells during the course of the infection. To investigate the degree of correlation between molecular results and CD4+ T-lymphocyte counts, data were analyzed by using the Spearman correlation coefficient (Table 3). As shown, a high degree of correlation was observed between CD4+ cell counts and the virological parameters analyzed. In CDC class IV patients, however, the P values were constantly higher than 0.1. Finally, a high degree of correlation was observed between all virological data compared with each other; however, statistical correlation between HIV-1 genomes in plasma and intracellular viral transcripts was less strong for class IV patients than for class II and III patients. DISCUSSION In the present study, we have used a cPCR-based method to examine the level of HIV-1 activity in symptomatic and symptomless infected patients. Using this molecular approach, we have obtained quantitative data on HIV-1 viremia levels, specific viral transcripts, and proviral sequences in infected cells; although sharp individual variability is evident, we have observed that complete viral replication occurs in all patients studied. In addition, quantitative virological data are correlated with disease progression. This mainly points out that the approach used in this research is capable of detecting differences in viral activity levels among these patients. Furthermore, our results indicate that all the quantitative virological parameters are correlated with the decrease in CD4+ T-lymphocyte counts. From a clinical point of view, these findings supply a virological and molecular basis to the hypothesis that early specific anti-HIV-1 therapy might be useful in most cases, whether or not clinical symptoms are present or the CD4+ T-lymphocyte count decreases. The quantitative data from CDC class III patients deserve particular attention. In this group of patients, no significant difference was observed for any of the virological parameters compared with those from CDC class II subjects (P value, invariably higher than 0.1), thus confirming the general clinical concept according to which these subjects are considered asymptomatic. Nevertheless, there is no significant difference in the provirus copy numbers of target cells between CDC class III and IV patients (P value, consistently higher than 0.1). On the contrary, when virological parameters indicating viral activity (free HIV-1 virions in plasma and HIV-1 transcripts in target cells) are considered, the results obtained for these groups of infected subjects proved to be quite distinct. From this point of view, the CDC class III patients indeed represent an intermediate group of HIV1-infected patients. This finding also indicates that RNA parameters of viral activity are more reliable and sensitive markers of infection progression than is quantitation of DNA proviral sequences integrated in target cells. These conclusions agree with previous quantitative analyses using virus isolation from blood and 50% tissue culture infective dose titration (19). Interestingly, the mean transcriptional activity calculated for each infected cell (the RNA/DNA ratio) is only moderately enhanced in the symptomatic patients. Although far from being definitive, this may suggest that the global increase in viral activity observed for AIDS patients is likely to depend on the greater number of infected cells revealed in

these stages and only partially on the higher transcriptional level for individual infected cells. This is in partial contrast to the results of a recent report (24) indicating a higher increase in transcriptional activity in the symptomatic phases on the basis of quantitative results obtained by RT-PCR and an external reference standard curve. Since, in the absence of an internal RNA competitor for RT-PCR, the possibility that sample-to-sample variability in both RT and PCR amplification reactions efficiency may influence quantitative results cannot be excluded, differences are probably due to technical reasons. However, depending on its importance in the better understanding of the mechanisms of AIDS pathogenesis, this aspect needs to be thoroughly clarified in the future by using quantitative molecular methods; similarly, the exact role of viral reservoirs other than CD4+ T cells in infection development deserves particular investigation. In fact, in CDC class IV patients, the lack of correlation between HIV-1 viremia levels and specific HIV-1 cellular transcripts (P, 0.28) or proviral sequences (P, 0.61) quantified in infected cells might be explained either by active viral replication in cells other than peripheral lymphocytes or by the inability of the immune system to play an efficient role in the clearance of free viruses in these highly immunodeficient patients. The present study extends the molecular investigation of HIV-1 infection to several virological parameters by using cRT-PCR and cPCR, thus allowing a more complete evaluation of viral activity in symptomless and symptomatic subjects. In this perspective, the approach described here seems to be a tool reliable enough to study infected individuals in long term follow-up and to obtain more information on the early phase of the infection. ACKNOWLEDGMENTS This research was supported by grants from the Ministero della Sanita, AIDS Project, and the Consiglio Nazionale delle Ricerche, target project Biotechnology and Bioinstrumentation (BTBS). S.M. is the recipient of a Ministero della Sanita, AIDS Project, fellowship.

1.

2.

3.

4.

5.

6.

REFERENCES Albert, J., H. Gaines, A. Sonnerborg, G. Nystrom, P. 0. Pehrson, F. Chiodi M. von Sydow, L. Morberg, K. Lidman, B. Christensson, B. k5j6, and E. M. Fenyo. 1987. Isolation of human immunodeficiency virus (HIV) from plasma during primary HIV infection. J. Med. Virol. 23:67-73. Aoki, S., R. Yarchoan, R. V. Thomas, J. M. Pluda, K. Marczyk, S. Broder, and H. Mitsuya. 1990. Quantitative analysis of HIV-1 proviral DNA in peripheral blood mononuclear cells from patients with AIDS and ARC: decrease of proviral DNA content following treatment with 2',3'-dideoxyinosine (ddl). AIDS Res. Hum. Retroviruses 6:1331-1339. Asjo, B., L. Morfeldt-Manson, J. Albert, G. Biberfeld, A. Karisson, K Lidman, and E. M. Fenyo. 1986. Replicative capacity of human immunodeficiency virus from patients with varying severity of HIV infection. Lancet ii:660-662. Bagnarelli, P., S. Menzo, A. Manzin, M. Giacca, P. E. Varaldo, and M. Clementi. 1991. Detection of human immunodeficiency virus type 1 genomic RNA in plasma samples by reversetranscription polymerase chain reaction. J. Med. Virol. 34:8995. Bagnarelli, P., S. Menzo, A. Manzin, P. E. Varaldo, M. Montroni, M. Giacca, and M. Clementi. 1991. Detection of human immunodeficiency virus type 1 transcripts in peripheral blood lymphocytes by the polymerase chain reaction. J. Virol. Methods 32:31-39. Balachandran, R., P. Thampatty, A. Enrico, C. Rinaldo, and P. Gupta. 1991. Human immunodeficiency virus isolates from

VOL. 66, 1992

7.

8.

9. 10.

11.

12.

13.

14.

15.

16.

17. 18. 19.

20.

21.

22.

asymptomatic homosexual men and from AIDS patients have distinct biologic and genetic properties. Virology 180:229-238. Baltimore, D., and M. B. Feinberg. 1990. Quantitation of human immunodeficiency virus in the blood. N. Engl. J. Med. 322: 1468-1469. Barre-Sinoussi, F., J. C. Chermann, F. Rey, M. T. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. VezinetBrun, C. Rouzioux, W. Rozenbaum, and L. Montagnier. 1983. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immunodeficiency syndrome (AIDS). Science 220: 868-871. Cheng-Mayer, C., D. Seto, M. Tateno, and J. A. Levy. 1988. Biologic features of HIV-1 that correlate with virulence in the host. Science 240:80-82. Clementi, M., A. Manzin, P. Bagnarelli, S. Menzo, G. Carloni, and P. E. Varaldo. 1992. Human immunodeficiency type 1 and hepatitis B virus transcription in peripheral blood lymphocytes from co-infected subjects. Arch. Virol. 126:1-9. Cloyd, M. W., and B. E. Moore. 1990. Spectrum of biological properties of human immunodeficiency virus (HIV-1) isolates. Virology 174:103-106. Delassus, S., R. Cheynier, and S. Wain-Hobson. 1991. Evolution of human immunodeficiency virus type 1 nef and long terminal repeat sequences over 4 years in vivo and in vitro. J. Virol. 65:225-231. De Rossi, A., M. Pasti, F. Mammano, L. Ometto, C. Giaquinto, and L. Chieco-Bianchi. 1991. Perinatal infection by human immunodeficiency virus type 1 (HIV-1): relationship between proviral copy number in vivo, viral properties in vitro, and clinical outcome. J. Med. Virol. 35:283-289. Fauci, A. S. 1988. The human immunodeficiency virus: infectivity and mechanisms of pathogenesis. Science 239:617-622. Fisher, A. G., B. Ensoli, D. Looney, A. Rose, R. C. Gallo, M. S. Saag, G. M. Show, B. H. Hahn, and F. Wong-Staal. 1988. Biologically diverse molecular variants within a single HIV-1 isolate. Nature (London) 334:444 447. Gallo, R. C., S. Z. Salahuddin, M. Popovic, G. M. Shearer, M. Kaplan, B. F. Haynes, T. J. Palker, R. Redfield, J. Oleske, B. Safai, G. Withe, P. Foster, and P. D. Markham. 1984. Frequent detection and isolation of cytopathic retroviruses (HTLV-III) from patients with AIDS and at risk for AIDS. Science 224:500504. Greene, W. C. 1991. The molecular biology of human immunodeficiency virus type 1 infection. N. Engl. J. Med. 324:308-317. Hart, C., T. Spira, J. Moore, J. Sninsky, G. Schochetman, A. Lifson, J. Galphin, and C.-Y. Ou. 1988. Direct detection of HIV RNA expression in seropositive subjects. Lancet ii:596-599. Ho, D. D., T. Moudgil, and M. Alam. 1989. Quantitation of human immunodeficiency virus type 1 in the blood of infected persons. N. Engl. J. Med. 321:1621-1625. Lemp, G. F., S. F. Payne, G. W. Rutherford, N. A. Hesslo, W. Winklestein, J. A. Wiley, A. R. Moss, R. E. Chaisson, R. T. Chen, D. W. Feigal, P. A. Thomas, and D. Werdegar. 1990. Progression of AIDS morbidity and mortality in San Francisco. JAMA 263:1497-1501. Lori, F., F. Veronese, A. Devico, P. Lusso, S. M. Reitz, and R. C. Gallo. 1991. HIV-1 virions carry viral DNA associated with reverse-transcriptase. Possible role of this complex. Abstr. VII Int. Conf. AIDS (Florence 16-21 June 1991), vol. 1, p. 59. Malim, M. H., D. McCarn, L. S. Tiley, and B. R. Cullen. 1991. Mutational definition of the human immunodeficiency virus type

MOLECULAR PROFILE OF HIV-1 INFECITION

7335

1 Rev activation domain. J. Virol. 65:4248-4254. 23. Menzo, S., P. Bagnarelli, M. Giacca, A. Manzin, P. E. Varaldo, and M. Clementi. 1992. Absolute quantitation of viremia in human immunodeficiency virus infection by competitive reverse transcription and polymerase chain reaction. J. Clin. Microbiol. 30:1752-1757. 24. Michael, N. L., M. Vahey, D. S. Burke, and R. R. Redfield. 1992. Viral DNA and mRNA expression correlate with the stage of human immunodeficiency virus (HIV) type 1 infection in humans: evidence for viral replication in all stages of HIV disease. J. Virol. 66:310-316. 25. Murakawa, G. J., J. A. Zaia, P. A. Spallone, D. A. Stephens, B. E. Kaplan, R. B. Wallace, and J. J. Rossi. 1988. Direct detection of HIV-1 RNA from AIDS and ARC patient samples. DNA 239:295-297. 26. Ottmann, M., P. Innocenti, M. Thenadey, M. Micoud, F. Pelloquin, and J.-M. Seigneurin. 1991. The polymerase chain reaction for the detection of HIV-1 genomic RNA in plasma from infected individuals. J. Virol. Methods 31:273-284. 27. Ou, C.-Y., S. Kwok, S. W. Mitchel, D. H. Mack, J. J. Sninsky, J. W. Krebs, P. Feorino, D. Warfield, and G. Schochetman. 1988. DNA amplification for direct detection of HIV-1 in DNA of peripheral blood mononuclear cells. Science 239:295-297. 28. Popovic, M., M. G. Sarngadharan, E. Read, and R. C. Gallo. 1984. Detection isolation and continuous production of cytopathic retroviruses (HTLV-III) from patients with AIDS and pre-AIDS. Science 224:500-504. 29. Saag, M. S., B. H. Hahn, J. Gibbons, Y. U, E. S. Parks, W. P. Parks, and G. M. Show. 1988. Extensive variation of human immunodeficiency virus type 1 in vivo. Nature (London) 334:

440-444. 30. Schnittman, S. M., J. J. Greenhouse, H. C. Lane, P. F. Pierce, and A. S. Fauci. 1991. Frequent detection of HIV-1-specific mRNAs in infected individuals suggests ongoing active viral expression in all stages of disease. AIDS Res. Hum. Retroviruses 7:361-367. 31. Schnittman, S. M., M. C. Psallidopoulos, H. C. Lane, L. Thompson, M. Baseler, F. Massari, C. H. Fox, N. P. Salzman, and A. S. Fauci. 1989. The reservoir for HIV-1 in human peripheral blood is a T cell that maintains expression of CD4. Science 245:305-308. 32. Simmonds, P., P. Balfe, C. A. Ludlam, J. 0. Bishop, and A. J. L. Brown. 1990. Analysis of sequence diversity in hypervariable regions of the external glycoprotein of human immunodeficiency virus type 1. J. Virol. 64:5840-5850. 33. Simmonds, P., P. Balfe, J. F. Peutherer, C. A. Ludlam, J. 0. Bishop, and A. J. L. Brown. 1990. Human immunodeficiency virus-infected individuals contain provirus in small numbers of peripheral mononuclear cells and at low copy numbers. J. Virol. 64:864-872. 34. Tersmette, M., R. E. Y. de Goede, B. J. M. Al, I. N. Winkel, R. A. Gruters, H. T. Cuypers, H. G. Huisman, and F. Miedema. 1988. Differential syncytium-inducing capacity of human immunodeficiency isolates: frequent detection of syncytium-inducing isolates in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. J. Virol. 62:2026-2032. 35. Wolfs, T. F. W., J.-J. de Jong, H. van den Berg, J. M. G. H. Tunagel, W. J. A. Krone, and J. Goudsmit. 1991. Evolution of sequences encoding the principal neutralization epitope of human immunodeficiency virus 1 is host dependent, rapid, and continuous. Proc. Natl. Acad. Sci. USA 87:9938-9942.