Lymphoid organs are a major reservoir of HIV-1 and primary sites for HIV-1 replication [1,2]. By using in situ hybridization and other quantitative methodologies, it has been possible to characterize and quantify the different pools of HIV-1 in the lymphoid organs of infected individuals . Quantitative virologic analyses have shown that reduction of plasma HIV-1 RNA to undetectable levels by antiretroviral therapy is accompanied by a significant reduction in the amount of HIV-1 in lymphoid tissues . However, ongoing HIV-1 replication in lymphoid tissues occurs in patients taking antiretroviral therapy who have undetectable plasma HIV-1 RNA levels [5–7] and certainly occurs in treated patients with persistence of detectable plasma HIV-1 RNA . This ongoing viral replication in the presence of selective antiretroviral pressure can select for resistant HIV-1 containing mutations in the reverse transcriptase and/or protease regions  and, ultimately, be associated with therapeutic failure.
Detection of resistant HIV-1 is usually accomplished by analysis of the genotype (reverse transcriptase and protease sequences) of cell-free virus circulating in plasma . However, changes in the genotype of the virus in the peripheral blood may not necessarily reflect the situation in other anatomic sites such as lymphoid organs. Studies comparing the genotype of HIV-1 in lymphoid tissues with that of the virus circulating in the plasma of the same individual are limited by the difficulty and invasiveness of sampling procedures. Previous work by our group has shown the feasibility of using biopsies of the palatine tonsil as a procedure to obtain appropriate samples of lymphoid tissue for qualitative and quantitative HIV-1 studies [3,10]. To assess how representative the genotype of HIV-1 circulating in plasma is of the genotype of the virus present in lymphoid tissue, we analyzed HIV-1 sequences in paired plasma and tonsillar tissue samples from patients with various levels of plasma HIV-1 RNA who were receiving antiretroviral therapy.
Materials and methods
Patients and samples
Informed consent was obtained from patients in accordance with the Health Partners Research Foundation Institutional Review Board, and the University of Minnesota Committee on Use of Human Subjects in Research. Paired samples of plasma and lymphoid tissue were collected from HIV-1-infected patients followed in the HIV clinic at Regions Hospital (St Paul, Minnesota, USA). Samples of lymphoid tissue were obtained by biopsies of the pallatine tonsil under local anesthesia, as described elsewhere . Tissue specimens were immediately placed on dry ice and stored at −70°C until analyzed. Plasma samples were prepared by centrifugation of blood collected in tubes containing ethylenediaminetetraacetic acid (EDTA) as anticoagulant. Samples of plasma were separated within 6 h of blood collection and stored frozen at −70°C until analyzed.
Nucleic acid extraction from tissue and plasma
RNA was extracted from tonsillar tissue (6–65 mg fragments) by using the RNA STAT-60 extraction kit (Tel-Test Inc., Friehdwood, Texas, USA). After homogenization of tissue fragments in 1 ml RNA STAT-60 buffer, 0.2 ml chloroform was added followed by centrifugation (12 000 × g) for 15 min at 4°C. The RNA-containing aqueous phase was then transferred to 1.5 ml tubes, followed by precipitation of the RNA in isopropanol and incubation at 4°C for 30 min. RNA was pelleted by centrifugation (12 000 × g) for 10 min at 4°C, washed in 75% ethanol, and centrifuged (7500 × g) for 5 min at 4°C. RNA pellets were then air-dried, resuspended in diethyl pyrocarbonate-treated RNAase-free water and incubated at 55°C for 10 min. Whole genomic DNA was extracted from tonsillar tissue by incubating overnight the interphase of tissue homogenates in a lysis buffer containing 0.5% sodium dodecyl sulfate (SDS) and protein kinase. DNA was then purified by phenol : chloroform : isoamyl alcohol extraction and ethanol precipitation, using standard molecular procedures. Tissue RNA and DNA preparations were stored at −70°C and 4°C, respectively, until analyzed. Samples (1 ml) of plasma were quickly thawed and centrifuged at 10 000 × g for 30 min at 4°C. RNA was extracted from plasma by using a commercial method (QIAamp Viral RNA kit, Qiagen, Hilden, Germany), as described elsewhere . Plasma RNA preparations were stored at −70°C until analyzed.
Amplification of HIV-1 protease and reverse transcriptase sequences
Fragments representing codons 1–99 of the HIV-1 protease and codons 1–250 of the HIV-1 reverse transcriptase were amplified from tissue RNA preparations and from plasma RNA by a nested polymerase chain reaction (PCR) method, as described elsewhere . This method allows amplification of HIV sequences from plasma samples with mean HIV RNA levels of 848 copies/ml . Positive and negative controls were included in each run. Samples of an RNA preparation from tonsillar tissue of an untreated HIV-1-infected individual were used as positive controls. Reverse transcriptase PCR (RT-PCR) reactions using water instead of target RNA were used as negative controls. Contamination of RNA preparations with HIV DNA sequences was assessed by nested PCR without a previous reverse transcriptase step.
For amplification of HIV-1 sequences from tissue DNA preparations, 0.5–1 μg genomic DNA was used in a nested PCR with the same primers used to amplify HIV sequences from plasma and tissue RNA preparations . PCR conditions consisted of 40 cycles (first PCR) or 30 cycles (second PCR) at 94°C for 30 s, 50°C for 30 s, and 72°C for 30 s, followed by a final elongation step of 10 min at 72°C.
Analysis of HIV-1 protease and reverse transcriptase sequences
Amplified products from nested PCR were filter-purified and sequenced using a commercial kit (ThermoSequenase, Thermo Sequenase Dye Terminator Cycle Sequencing kit, v 2.0; Amersham Life Sciences, Cleveland, Ohio, USA), and sequencing primers described elsewhere . Sequences were edited using the GCG (Genetics Computer Group, Wisconsin Package version 9.1, Madison, Wisconsin, USA) and Staden (Medical Research Council Laboratory of Molecular Biology, MRC Center, Hills Road, Cambridge, UK) software packages. Results were compared with consensus amino acid sequences of reference HIV-1 strains MN (GeneBank accession no. M17449) and HXB2 (GeneBank accession no. K03455). Changes from the consensus sequences associated with resistance of HIV-1 to antiretroviral drugs were taken from published data . A total of 19 positions in the HIV-1 protease region and of 27 positions in the HIV-1 reverse transcriptase region associated with drug resistance were analyzed . The percentage of mutant sequences at each codon was quantified by measuring peak-on-peak heights in sequence chromatograms .
Patients and samples
Paired plasma samples and tonsillar tissue specimens were obtained from nine patients receiving combination antiretroviral therapy . At the time when the samples were obtained, the patients’ mean (±SD) CD4 count was 233 (± 117) × 106 cells/l (median, 251 × 106; range, 38–519), and their mean (±SD) plasma HIV-1 RNA level was 3.3 ± 1.2 log10 copies/ml (median, 3.2; range, 1.7–5.2).
Analysis of HIV-1 protease sequences in plasma and lymphoid tissue
HIV-1 protease sequence data were available for comparison from nine plasma/tissue RNA pairs from seven patients (Table 1). The degree of concordance found between HIV-1 protease amino acid sequences in these plasma/tissue RNA pairs was 100%. When the percentage of mutant sequences at drug resistance codons was compared, only one discrepancy was found (sample I2, Table 1). In this case, the percentage of mutant sequences at codon 82 of the HIV-1 protease was higher in plasma as compared with tissue RNA.
HIV-1 protease sequence data were available for comparison from nine tissue RNA/tissue DNA pairs obtained from seven patients (Table 1). The degree of concordance between HIV-1 protease amino acid sequences in these tissue pairs was 100%. Discrepancies in the percentage of mutant sequences at drug resistance codons were found at six positions in three of the tissue RNA/tissue DNA pairs (Table 1): codons 46 and 82 (sample B1); codons 71 and 77 (sample E2); and codons 82 and 90 (sample I2). In all these discrepant pairs, a higher percentage of mutant viral sequences was present in tissue RNA compared with tissue DNA (Table 1).
HIV-1 protease sequences were compared in 11 plasma/tissue DNA pairs obtained from eight patients (Table 1). The concordance between HIV-1 protease amino acid sequences in these plasma/tissue DNA pairs was 100%. When the percentage of mutant sequences at drug resistance codons was compared, discrepancies were found at seven positions in four tissue RNA/tissue DNA pairs (Table 1): codons 46 and 82 (sample B1), codons 71 and 77 (sample E2), codons 90 (sample I1), and codons 82 and 90 (sample I2). In all discrepant cases, a higher percentage of mutant viral sequences was present in plasma than in tissue DNA (Table 1).
Analysis of HIV-1 reverse transcriptase sequences in plasma and lymphoid tissue
HIV-1 reverse transcriptase sequence data were available from 10 plasma/tissue RNA pairs obtained from nine patients (Table 2). The concordance between HIV-1 reverse transcriptase amino acid sequences in these plasma/tissue RNA pairs was 100%. When the percentage of mutant sequences at drug resistance codons was compared, discrepancies were found at nine positions in five of the plasma RNA/tissue RNA pairs: codon 70 (sample A1); codon 215 (sample B1); codons 67 and 215 (sample D2); codons 67 and 70 (sample E1); and codons 67, 70, and 184 (sample E2). In five (55%) of these positions, the percentage of mutant sequences was higher in tissue RNA than in plasma RNA (Table 2).
HIV-1 reverse transcriptase sequence data were available for comparison from 10 tissue RNA/tissue DNA pairs obtained from eight patients (Table 2). The degree of concordance between HIV-1 reverse transcriptase amino acid sequences in these tissue RNA/tissue DNA pairs was 99%. The two discrepant sequences were found in the same pair (sample B1): the zidovudine resistance mutation at codon 41 was found in the tissue DNA but not in the tissue RNA and the zidovudine resistance mutation 215F was found in the tissue RNA whereas the tissue DNA contained the 215Y mutation. When the percentage of mutant sequences at drug resistance codons was compared, discrepancies were found at eight positions in five tissue RNA/tissue DNA pairs (Table 2): codon 70 (samples A1 and B1); codon 215 and 219 (sample D2); codon 67 (sample E1); and codons 67, 70, and 184 (sample E2). In seven (88%) of the discrepant sequences, the percentage of mutant sequences was higher in the tissue RNA than in the tissue DNA (Table 2).
HIV-1 reverse transcriptase sequence data were available for comparison from 12 plasma/tissue DNA pairs obtained from nine patients (Table 2). The concordance between HIV-1 reverse transcriptase amino acid sequences in these plasma/tissue DNA pairs was 97%. The discrepancies found include the following: presence of a 41L mutation in tissue DNA but not in plasma (sample B1); presence of the mutation 215Y in both the tissue and plasma DNA (sample B1); and presence of multiple mutations (41L, 67N, 74V/I, 181C, 184V, 215Y) in plasma but not in tissue DNA (sample I1). All the last group of mutations were present in samples of plasma, tissue RNA and tissue DNA obtained later from the same individual (sample I2).
When the percentage of mutant sequences at drug resistance codons was compared, discrepancies were found at 13 positions in six plasma/tissue DNA pairs (Table 2): codon 70 (samples A1, B1); codons 67, 181, 184, and 219 (sample C1); codons 67 and 219 (sample D2); codons 67 and 70 (sample E1); and codons 67, 70, and 184 (sample E2). In nine (69%) of these positions, a higher percentage of mutant viral sequences was present in plasma than in tissue DNA (Table 2).
This study showed a high degree of homology between the reverse transcriptase and protease sequences of HIV-1 circulating in plasma and the sequences of the virus present in RNA and DNA extracts of lymphoid tissue from individuals with detectable plasma HIV-1 RNA levels during antiretroviral therapy. However, when interpreting the results of this study, it is important to emphasize that population PCR products were sequenced and, therefore, the sequences compared represented the predominant viral population present in the samples studied. Because of the heterogeneity of HIV-1 sequences in infected individuals, it is possible that viral minorities with a genetic background that differed from that of the predominant HIV-1 population could have been missed .
The high degree of similarity observed between viral sequences in plasma and tissue RNA extracts is not surprising. Quantitative studies have suggested a close correlation between the concentration of HIV-1 in plasma and the production of virus in lymphoid tissues [3,15]. These studies have also shown that the predominant form of HIV-1 in lymphoid tissues is complete virions associated with follicular dendritic cells [1–3,16]. Therefore, lymphoid tissue RNA extracts would contain HIV-1 sequences representing the predominant viral genomes present in this compartment. Because of the close correlation between plasma and lymphoid tissues, analysis of plasma HIV-1 sequences would provide pertinent information on the sequence of the predominant virus in lymphoid tissues. The results of our study confirm that this is the case, and that the genotype of plasma HIV-1 represents the genotype of the virus in the lymphoid tissues of HIV-1-infected individuals. This finding has practical importance since HIV-1 resistance studies might be indicated in patients who have detectable HIV-1 RNA levels in plasma while receiving antiretroviral therapy . Because lymphoid tissue samples are difficult to obtain, HIV-1 resistance studies are most commonly carried out on virus circulating in the plasma. Our study suggests that, when using population sequencing, HIV-1 genotypic resistance studies in plasma provide the same information as genotypic studies of lymphoid tissue RNA preparations
Although the overall degree of concordance between HIV-1 amino acid sequences in plasma and tissue RNA or DNA extracts was high, there were a few discordances when the percentages of mutant viral sequences at drug resistance codons were compared. In all instances, the percentage of resistant mutant sequences in the reverse transcriptase and the protease region of the virus was higher in tissue RNA extracts or plasma fractions than in tissue DNA extracts. Similar findings when comparing HIV-1 reverse transcriptase and protease sequences in DNA and RNA extracts of lymphoid tissue and peripheral blood mononuclear cells (PBMC) have been reported by others . In that study, the discordances observed were primarily caused by the detection of wild-type HIV-1 sequences in DNA of lymphoid tissue or PBMC at the same time that mutant sequences were found in RNA preparations from the same compartments . Our findings support previous observations indicating that the appearance of resistance mutations in HIV-1 circulating in plasma precedes the appearance of the same mutations in cell-associated HIV-1 DNA, and that the delay in the appearance of drug-resistant HIV-1 mutants in tissue DNA preparations compared with plasma or tissue RNA is a consequence of the archival nature of most HIV-1 DNA [17–19].
Most of the HIV-1 DNA in PBMC and lymphoid tissues of HIV-1-infected individuals represents preintegrated, non-viable viral DNA; only a minority represents proviral, transcription-ready HIV-1 DNA . In our study, the amplification method used did not differentiate preintegrated from proviral HIV-1 DNA, and the amplified products sequenced represented pooled HIV-1 DNA populations. As indicated earlier, it is possible that these methods could miss viral minorities with a genetic background that differed from the pooled HIV-1 population sequenced .
In summary, the genotype of HIV-1 circulating in the plasma is very close to the genotype of the virus present in lymphoid tissue. In patients with detectable plasma HIV-1 RNA levels during antiretroviral therapy, for whom resistance studies might be considered, characterization of plasma HIV-1 sequences would provide useful information regarding the genotype of HIV-1 in lymphoid tissue.
1. Embretson J, Zupancic M, Ribas JL. et al
. Massive covert infection of helper T lymphocytes and macrophages by HIV during the incubation period of AIDS. Nature 1993, 362: 359 –362.
2. Pantaleo G, Graziosi C, Demarest JF. et al
. HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 1993, 362: 355 –358.
3. Haase AT, Henry K, Zupancic M. et al
. Quantitative image analysis of HIV-1 infection in lymphoid tissue. Science 1996, 274: 985 –989.
4. Cavert W, Notermans DW, Staskus K. et al
. Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection. Science 1997, 276: 960 –964.
5. Natarajan V, Bosche M, Metcalf JA, Ward DJ, Lane HC, Kovacs JA. HIV-1 replication in patients with undetectable plasma virus receiving HAART. Lancet 1999, 353: 119 –120.
6. Furtado MR, Callaway DS, Phair JP. et al
. Persistence of HIV-1 transcription in peripheral-blood mononuclear cells in patients receiving potent antiretroviral therapy. N Engl J Med 1999, 340: 1614 –1622.
7. Ruiz L, van Lunzen J, Arno A. et al
. Protease-inhibitor-containing regimens compared with nucleoside analogues alone in the suppression of persistent HIV-1 replication in lymphoid tissue. AIDS 1999, 13: F1 –F8.
8. Wong JK, Günthard HF, Hvlir DV. et al
. Reduction of HIV-1 in blood and lymph nodes following potent antiretroviral therapy and the virologic correlates of treatment failure. Proc Natl Acad Sci USA 1997, 94: 12574 –12579.
9. Hirsch MS, Conway B, D'Aquila RT. et al
. Antiretroviral drug resistance testing in adults with HIV infection. J Am Med Assoc 1998, 279: 1984 –1991.
10. Faust RA, Henry K, Dailey P. et al
. Outpatient biopsies of the palatine tonsil: access to lymphoid tissue for assessment of human immunodefciency virus RNA titers. Otolaryngol Head Neck Surg 1996, 114: 593 –598.
11. Niubó J, Li W, Henry K, Erice A. Recovery and analysis of human immunodeficiency virus type 1 (HIV) RNA sequences from plasma samples with low HIV RNA levels. J Clin Microbiol 2000, 38: 309 –312.
12. Hirsch MS, Brun-Vézinet F, D'Aquila RT. et al
. Antiretroviral drug resistance testing in adult HIV-1 infection. J Am Med Assoc 2000, 283: 2417 –2426.
13. Larder BA, Kohli A, Kellam P, Kemp SD, Kronick M, Henfrey RD. Quantitative detection of HIV-1 drug resistance mutations by automated DNA sequencing. Nature 1993, 365: 671 –673.
14. Martínez-Picado J, Sutton L, de Pasquale MP, Savara AV, D'Aquila RT. Human immunodeficiency virus type 1 cloning vectors for antiretroviral resistance testing. J Clin Microbiol 1999, 37: 2943 –2951.
15. Cohen OJ, Pantaleo G, Holodniy M. et al
. Decreased human immunodeficiency virus type 1 plasma viremia during antiretroviral therapy reflects downregulation of viral replication in lymphoid tissue. Proc Natl Acad Sci USA 1995, 92: 6017 –6021.
16. Pantaleo G, Cohen OJ, Schacker T. et al
. Evolutionary pattern of human immunodeficiency virus (HIV) replication and distribution in lymph nodes following primary infection: implications for antiretroviral therapy. Nat Med 1998, 4: 341 –345.
17. Günthard HF, Wong JK, Ignacio CC. et al
. Human immunodeficiency virus replication and genotypic resistance in blood and lymph nodes after a year of potent antiretroviral therapy. J Virol 1998, 72: 2422 –2428.
18. Kozal MJ, Shafer RW, Winters MA, Katznstein DA, Merigan TC. A mutation in human immunodeficiency virus reverse transcriptase and decline in CD4 lymphocyte numbers in long-term zidovudine recipients. J Infect Dis 1993, 167: 526 –532.
19. Smith MS, Koerber KL, Pagano JS. Zidovudine-resistant human immunodeficiency virus typ1 genomes detected in plasma distinct from viral genomes in peripheral blood mononuclear cells. J Infect Dis 1993, 167: 445 –448.
20. Chun T-W, Carruth L, Finzi D. et al
. Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection. Nature 1997, 387: 183 –188.