Share this article on:

Changes in host cell molecules acquired by circulating HIV-1 in patients treated with highly active antiretroviral therapy and interleukin-2

Abbate, Isabella; Dianzani, Ferdinandoa; Turriziani, Ombrettab; Antonelli, Guidob; D'Offizi, Giampieroc; Galati, Vincenzoc; Pierdominici, Marinad; Pandolfi, Francoe; Capobianchi, Maria R.c

Basic Science

Objective: To analyse cell membrane proteins (CMP) acquired by HIV-1 present in the plasma of asymptomatic patients, and their modifications after a cycle of highly active antiretroviral therapy (HAART) and interleukin (IL)-2.

Design and methods: Plasma samples from eight drug-naive asymptomatic subjects underwent immobilized antibody capture (IAC) to detect CMP on the surface of circulating HIV-1. The CMP considered were lymphocyte subset markers (CD45RA, CD45RO), activation markers (HLA-DR), adhesion molecules (LFA-3), costimulatory proteins (B7-2), lymph-node homing receptors (CD62L) and pro-apoptosis molecules (FasL). This analysis was repeated after one cycle of HAART + IL-2, after virus rebound.

Results: LFA-3, followed by CD45RO and HLA-DR, are the most represented CMP on the surface of circulating virions in naive asymptomatic patients; CD45RA, CD62L, B7-2 and FasL are detected only occasionally. After rebound, a significant reduction of CD45RO and HLA-DR, but not of LFA-3, is observed on virions, whereas CD45RA and CD62L, as well as other molecules, are not affected, remaining almost undetectable.

Conclusions: Assuming that CMP on HIV-1 reflect the cellular origin of virions, activated T cells expressing CD45RO, HLA-DR, and LFA-3 may be the main source of HIV-1 in asymptomatic patients. After a cycle of HAART + IL-2, followed by therapy interruption, CD45RA and CD62L are detected on virions rarely, indicating that even during virus rebound, expanded naive T cells do not become a major target of virus replication. Furthermore, the presence of HLA-DR on rebound HIV-1 is decreased, consistent with decreased activation of the HIV-producing cells. More extensive investigation may clarify the significance of these findings with respect to pathogenesis.

From the Laboratory of Microbiology, Sandro Pertini Hospital, the aLibera Università Campus Biomedico, the bDepartment of Experimental Medicine and Pathology, Section of Virology, University ‘La Sapienza’ the cNational Institute for Infectious Diseases IRCCS ‘L. Spallanzani', the dDepartment of Clinical Medicine, University ‘La Sapienza', and the eInstitute of Internal Medicine, Catholic University, Rome, Italy.

Received: 14 August 2000;

revised: 29 September 2000; accepted: 10 October 2000.

Sponsorship: Supported by Grants to M. R. C. and to F. D. from Istituto Superiore di Sanità, National AIDS Project, by Grants to F. P. from MURST and by Fondi Ricerca Corrente to L. Spallanzani I.R.C.C.S.

Requests for reprints to: Maria R. Capobianchi, National Institute for Infectious Diseases ‘L. Spallanzani’ IRCCS, Laboratory of Virology, Via Portuense, 292-00149 Rome, Italy.

Back to Top | Article Outline


It has been shown previously that HIV-1 virions, budding from infected cells, incorporate a number of functional cell membrane proteins (CMP) that represent a footprint of their cellular origin [1–13]. Furthermore, virus-bound CMP may help in the recognition of target cells and expand the tropism of the virus to CD4-negative cells expressing the natural ligands for the virus-bound adhesion molecules. It is also likely that virus-embedded cell activation molecules could participate in the activation of cells exposed to the virions, thus contributing to both enhanced virus replication and immune dysregulation [6–9]. However, most observations have been made on HIV grown in vitro[1–13], and the significance of virus-embedded CMP in disease pathogenesis remains to be determined. Recently, it has been shown that HLA-DR [14], as well as other cellular molecules, in particular CD26, CD36 and CD44 [15], are present on the envelope of HIV circulating in vivo, providing information on the cellular origin of the virions.

We have analysed the presence of a number of CMP present on the surface of HIV-1 virions circulating in the plasma of asymptomatic patients. Specifically, the CMP considered were markers of lymphocyte subset (CD45RA, CD45RO), markers of cell activation (HLA-DR) and molecules involved in intercellular adhesion (LFA-3), costimulatory function (B7-2), lymph node homing (CD62L) or apoptosis (FasL). Furthermore, because these patients underwent a cycle of highly active antiretroviral therapy (HAART) plus interleukin (IL)-2 followed by therapy interruption, the pattern of CMP present on virions before therapy was compared with that present on virions rebounding after therapy suspension.

Back to Top | Article Outline

Material and methods


A subgroup of asymptomatic patients was selected as a part of a larger controlled clinical trial that included 22 HIV-1 infected subjects, and has been described in part elsewhere [16,17]. The patients were antiviral drug-naive, with CD4 cell counts of 400–600 × 106/l and HIV-RNA load > 5000 copies/ml. All patients agreed to participate by signing an informed consent form and the study, conducted according to GCP standards, was approved by the Institutional Ethical Committee. The therapy [indinavir 2400 mg/day, stavudine 60–80 mg/day, and epivir 300 mg/day) plus recombinant IL-2 (rIL-2; Chiron, Emmeryville, CA) 106 IU subcutaneously, 5 days per week on alternate weeks] was administered for 28 weeks and then suspended. In one patient therapy was interrupted at week 16. The clinical results of the study will be detailed elsewhere (G. D'Offizi et al., unpublished data). For CMP determination on circulating virions by the immunocapture method (see below) a subgroup of patients (n = 8) was selected from those participating in the study (n = 22), showing both pre-therapy and rebound viraemia levels > 20 000 RNA copies/ml. In these patients, the range of viral load was 62 000–630 000 RNA copies/ml at t0 and 93 000–566 000/ml at t1.

Back to Top | Article Outline

Plasma viraemia

Plasma viraemia was monitored by using a commercially available quantitative reverse transcripton (RT)–PCR assay (Amplicor HIV-1 Monitor version 1.5; Roche, Milan, Italy), with a threshold level of 400 copies/ml, performed according to manufacturer's instruction.

Back to Top | Article Outline

Lymphocyte phenotyping

Lymphocyte phenotype analysis was performed at the times indicated on whole blood lysed with Ortho-mune lysing reagent (Ortho Diagnostic System Inc., Raritan, New Jersey, USA). Cells were washed and labelled with a cocktail of monoclonal antibodies for 30 min at 4°C. Naive T cells were identified with a cocktail of anti-CD4–cychrome, anti-CD45RA–fluoresceine isothiocyanate and anti-CD62L–phycoerythrin (PharMingen, San Diego, California, USA), whereas memory T cells were identified by a double staining with anti-CD4–cychrome plus anti-CD45RO–fluoresceine isothiocyanate (PharMingen). Stained cells were washed in phosphate-buffered saline containing 2% foetal calf serum and subjected to cytofluorimetric analysis performed on a FACS Calibur (Becton Dickinson, San Jose, California, USA) equipped with Cell Quest (Macintosh) software. For the analysis, total lymphocytes were first identified and gated by forward and side scatter. Then, cells were additionally gated for CD4 expression. A total of 10 000 gated events were collected for each sample.

Back to Top | Article Outline

CMP determination on virions

Plasma samples obtained before therapy (t0) and 4 weeks after therapy interruption (t1) were used to establish the CMP profile of circulating virions. CMP present on HIV-1 particles were detected by immobilized antibody capture (IAC), a method developed in this laboratory that is based on measurement of HIV-1 binding to monoclonal antibodies (MAb) immobilized to plastic wells [3]. By this and by a similar method, the extent of capture of HIV-1 by strain-specific anti-gp120 MAb [18] and by V3 loop-specific MAb [19] was 5–9% of the input virus. The original method was modified for application to circulating HIV-1 because it is generally found at lower concentrations in plasma as compared with in vitro grown virus. Specifically, 50 μl of clarified plasma samples, diluted as appropriate to contain 20 000–50 000 RNA copies/ml, were applied to wells coated with MAb to each CMP. The following MAb were used: anti HLA-DR (clone CR3/43; Dako, Milan, Italy), anti CD45RA (clone 4KB5; Dako) anti L-selectin (CD62L) (clone FMC46; Dako), anti CD45RO (clone UCHL1; Dako), anti FasL (clone NOK-1; PharMingen), anti B7-2 (CD86) (clone BU63; Serotec, Oxford, UK), anti LFA-3 (CD58) (clone BRIC-5; Serotec). Control MAb were anti-Trichomonas vaginalis, provided by P. L. Fiori, Sassari, Italy, as described [3], anti-cytomegalovirus (clone CCH2; Dako), or anti B7-1 (CD80) (clone DAL-1; Serotec). Duplicate wells were used for each MAb. The detection of MAb-bound virions was performed by measuring HIV-specific RNA, retained in the MAb-coated wells, by RT–PCR (Amplicor HIV-1 Monitor version 1.5). Specifically, MAb-retained virions in the duplicate wells were treated with the lysing solution and then pooled. Extraction and detection of viral RNA were performed according to the manufacturer's instructions. In each assay a concomitant re-titration of input virus was performed. For each MAb the intra- and inter-assay variation coefficient of the extent of virus capture was < 10% and < 30%, respectively. The results were expressed as proportion (%) of input RNA copy number specifically retained by each MAb, after subtraction of values obtained in the control MAb-coated well. The values of specific HIV-1 binding to anti-CMP MAb ranged between 0.2 and 19% of input virus, values that are lower than the corresponding values obtained by immunoprecipitation with Staphylococcus aureus[2,14], but that are consistent with several previous data obtained with plastic-immobilized MAb [3–6,9,14, 20,21].

Preliminary experiments were aimed at ruling out the possibility that soluble molecules present in the plasma could inhibit virus binding to CMP-specific MAb, thus masking the presence of such molecules on circulating virions [15]. Specifically, the presence of a number of CMP on ultracentrifuged preparations of HIV grown in cell culture was tested by using IAC either in the absence, or in the presence of plasma derived from one of the study patients, collected at the time of virus load nadir (undetectable levels) as the result of ongoing HAART. The results suggest that the pattern of CMP present on HIV grown in vitro was not altered by the presence of molecules derived from the plasma of patients; in particular, the extent of virus capture by MAb to HLA-DR was not reduced in the presence of the aviraemic plasma sample – this is different from observations made in another study that used different methods [15].

Back to Top | Article Outline


Plasma viraemia

At 12 weeks a 2 log10 decrease in plasma viraemia was reached in all subjects, and < 400 HIV-1 RNA copies/ml were reached in seven out of eight subjects at week 20. Therapy was suspended at week 28; only one patient interrupted treatment at week 16 because of hepatitis C virus superinfection. Starting from 1 week after therapy suspension, viral load rebound to pre-therapy levels was observed (Fig. 1).

Back to Top | Article Outline

Changes in CD4 T cell subset counts

The immunological and phenotypic analysis of lymphocytes of these subjects has been described in part elsewhere [17]. By comparing the data obtained at the two time points considered in the present study, 4 weeks after therapy suspension the patients showed a significant increase of both absolute and percentage values (mean ± standard error) of naive CD4/CD45RA/CD62L T cells as compared with pre-therapy values: 324 ± 27 × 106 cells/l (t1) versus 166 ± 28 × 106 cells/l (t0) (P = 0.002) and 52 ± 3% (t1) versus 30 ± 5% (t0) (P = 0.003). No significant differences in CD4/CD45RO T cells were observed in both absolute or percent values: 442 ± 63 × 106 cells/l (t1) versus 359 ± 42 × 106 cells/l (t0) (P > 0.05) and 67 ± 3% (t1) versus 64 ± 4% (t0) (P > 0.05).

Back to Top | Article Outline

Changes in CMP included in circulating virions after HAART plus IL-2 therapy

The extent of specific capture of circulating HIV-1 by MAb to CMP was consistent with previous results obtained on in vitro grown HIV-1 [3–6,9,10,20]. The results shown in Table 1 indicate that LFA3 was the most represented CMP on the surface of circulating virions. CD45RO and HLA-DR were also highly expressed on the surface of circulating HIV, whereas CD45RA, CD62L, B7-2 and FasL were detected only occasionally, and to a lower extent. Interestingly, when comparing the mean pre-therapy levels of CMP included in HIV envelope to the post-therapy levels, a significant reduction of both CD45RO and HLA-DR was observed, whereas the amount of virion-bound LFA3 was not affected significantly. When considering the individual changes observed for HLA-DR, CD45RO and LFA-3 shown in Fig. 2 it can be seen that for HLA-DR and CD45RO, approximately 50% of patients showed a strong reduction. Furthermore, the most relevant reduction was observed for those patients with the highest starting level, whereas the remaining showed either a weak decrease, or even a slight increase of CMP (one patient for HLA-DR and two patients for CD45RO). Concerning LFA-3, only two patients showed a detectable decrease, whereas the remainder showed either unchanged or slightly increased levels on circulating HIV. It should be noted that although the three CMP did not follow a consistent pattern of reduction/increase for each single patient, in the subject with increased HLA-DR on virions no increase of absolute CD4 number was detected after therapy. The pattern of the remaining CMP embedded to low level in the surface of circulating HIV remained virtually unchanged, although with a general tendency towards reduction.

Back to Top | Article Outline


In this paper we report the presence of several previously undescribed CMP on the surface of HIV-1 released into the plasma of patients, both before and after a cycle of HAART and IL-2. LFA3, CD45RO and HLA-DR were the molecules expressed most frequently. The presence of LFA-3 is consistent with previous findings, showing a strong presence of adhesion molecules on the surface of HIV-1 [3–7], whereas the presence of CD45RO is in apparent contrast with a recent report, showing that in vitro grown HIV-1 selectively excludes CD45 molecules from its envelope [21]. Although the discrepancy could be due to technical differences (i.e., the MAb used for virus capture), it is possible that in vivo circulating virions could display different cell-derived membrane molecules from those displayed on HIV grown in vitro. Further studies will be necessary to clarify this.

After controlled therapy suspension, viral rebound resulted in reduction of HLA-DR and CD45RO expression, but not of LFA-3. These findings have several implications. Assuming that the CMP present on the surface of circulating HIV can represent a footprint of their cellular source, as it has been demonstrated for HIV grown in vitro[4], the main host cell giving rise to viraemia in the asymptomatic phase of the infection appears to be activated memory T cells, as virions derived from plasma show a high presence of CD45RO and HLA-DR, together with a virtual absence of CD45RA and CD62L. This is in agreement with previous observations on the susceptibility of different lymphocyte subpopulations to HIV infection in vitro[22]. Furthermore, even if recent findings indicate that naive T cells can harbour replication-competent virus [23], not only were pre-therapy virions virtually deprived of CD45RA and CD62L, but rebound virus did not show any increase of CD45RA and CD62L included in the virus envelope either, indicating that no major changes of virus cell source occurred as a consequence of HAART plus IL-2. Nevertheless a significant increase of naive T lymphocytes (CD4/CD45RA/CD62L) in terms of both absolute and percentage values was observed. This supports the possibility that even if HAART plus IL-2 increased naive T lymphocytes (CD4/CD45RA/CD62L) significantly, these cells did not become a major target for HIV-1.

An intriguing aspect of the findings reported here is that a selective decrease of virus-bound HLA-DR and CD45RO was observed when comparing pre-therapy to rebound circulating virions, whereas HIV-embedded LFA-3, highly present in pre-treatment virions, was not modified significantly after HAART plus IL-2. As far as HLA-DR is concerned, the significant reduction of this molecule observed in HIV present in the circulation of patients after therapy interruption supports a reduction of the activation state of the cells actually involved in producing circulating virions, consequent to HAART, in agreement with previous observations [24]. Also consistent with a reduced activation state of HIV-producing cells is the general tendency to show reduced amounts of many other CMP on virions after HAART. On the contrary, whereas CD45RO in virions was reduced significantly after therapy interruption, the level of cellular expression of this molecule was not affected significantly at the time of virus rebound, both in terms of number of positive cells and mean channel fluorescence (data not shown). It is possible that subtle modifications of the cell population harbouring HIV-1 replication occur as a result of therapeutic intervention. Concerning the presence of LFA-3 in the virus envelope, it is tempting to speculate that it has a role in enhancing virus replication. In fact, triggering of cell membrane CD2 by interaction with virus-embedded LFA-3 could determine cell activation and protection from apoptosis [25].

More studies are necessary to clarify the clinical significance of the presence of CMP on circulating virions, as well as the implications of the selective reduction of CMP such as HLA-DR and CD45RO observed on rebound HIV after therapy suspension. Nevertheless, this is the first attempt to analyse the changes in the profile of cell-derived molecules acquired by circulating HIV in patients after controlled therapy interruption. In this respect, the observation that in the unique subject with increased HLA-DR on virions no increase of absolute CD4 number was detected after therapy warrants further investigation to establish the clinical significance of CMP variations in patients undergoing antiretroviral treatment. Furthermore, the observation that naive T cells do not become an alternative source of circulating virions, although a prominent increase of this cell population occurred at the time of analysis, could be of significance.

Back to Top | Article Outline


1. Gelderblom H, Reupke H, Winkel T, Kunze R, Pauli G. MHC-antigens: constituents of the envelopes of human and immunodeficiency viruses. Z Naturforsch 1987, 42: 1328 –1334.
2. Arthur LO, Bess JW, Sowder RC II Jr. et al. Cellular proteins bound to immunodeficiency viruses: implication for pathogenesis and vaccines. Science 1992, 258: 1935 –1938.
3. Capobianchi MR, Fais S, Castilletti C, Gentile M, Ameglio F, Dianzani F. A simple and reliable method to detect cell membrane proteins on infectious human immunodeficiency virus-1 particles. J Infect Dis 1994, 169: 886 –889.
4. Abbate I, Capobianchi MR, Fais S. et al. Host cell antigenic profile acquired by HIV-1 is a marker of its cellular origin. Arch Virol 1995, 140: 1849 –1854.
5. Fais S, Capobianchi MR, Abbate I. et al. Unidirectional budding of HIV-1 at the site of cell-to-cell contact is associated with co-polarization of intercellular adhesion molecules (ICAM)-1 and HIV-1 viral matrix protein. AIDS 1995, 9: 329 –335.
6. Castilletti C, Capobianchi MR, Fais S. et al. HIV type 1 grown on interferon γ-treated U937 cells shows selective increase in virion-associated intercellular adhesion molecule 1 and HLA-DR and enhanced infectivity for CD4-negative cells. AIDS Res Hum Retroviruses 1995, 11: 547 –553.
7. Rizzuto C, Sodroski JG. Contribution of virion ICAM-1 to human immunodeficiency virus infectivity and sensitivity to neutralization. J Virol 1997, 71: 4847 –4851.
8. Rossio JL, Bess J, Henderson LE, Cresswell P, Arthur LO. HLA Class II on HIV particles is functional in superantigen presentation to human T cells: implications for HIV pathogenesis. AIDS Res Hum Retroviruses 1997, 11: 1433 –1439.
9. Liao Z, Roos JW, Hildreth JEK. Increased infectivity of type 1 particles bound to cell surface and solid-phase ICAM-1 and VCAM-1 through acquired adhesion molecules LFA-1 and VLA-4. AIDS Res Human Retroviruses 2000, 16: 355 –366.
10. Bastiani L, Laal S, Kim M, Zolla-Pazner S. Host cell-dependent alterations in envelope components of human immunodeficiency virus type 1 virions. J Virol 1997, 71: 3444 –3450.
11. Saifuddin M, Tarlan H, Atkinson JP, Holguin MH, Parker CJ, Spear GT. Human immunodeficiency virus type 1 incorporates both glycosyl phosphatidylinositol-anchored CD55 and CD59 and integral membrane CD46 at levels that protect from complement-mediated destruction. J Gen Virol 1997, 78: 1907 –1911.
12. Tremblay MJ, Fortin JF, Cantin R. The acquisition of host-encoded proteins by nascent HIV-1. Immunol Today 1998, 19: 346 –351.
13. Poon DTK, Coren LV, Ott DE. Efficient incorporation of HLA class II onto human immunodeficiency virus type 1 requires envelope glycoprotein packaging. J Virol 2000, 74: 3918 –3923.
14. Saarloos M, Sullivan BL, Czerniewski MA, Parameswar KD, Spear GT. Detection of HLA-DR associated with monocytotropic, primary, and plasma isolates of human immunodeficiency virus type 1. J Virol 1997, 71: 1640 –1643.
15. Lawn SD, Roberts BD, Griffin GE, Folks TM, Butera ST. Cellular compartments of human immunodeficiency virus type 1 replicationin vivo: determination by presence of virion-associated host proteins and impact of opportunistic infection. J Virol 2000, 74: 139 –145.
16. Dianzani F, Antonelli G, Aiuti F. et al. The number of HIV DNA-infected mononuclear cells is reduced under HAART plus recombinant IL-2. Antiviral Res 2000, 45: 95 –99.
17. Pandolfi F, Pierdominici M, Marziali M. et al. Low-dose IL-2 reduces lymphocyte apoptosis and increases naive CD4 cells in HIV-1 patients treated with HAART. Clin Immunol 2000, 94: 153 –159.
18. Dianzani F, Castilletti C, Gentile M, Gelderblom HR, Frezza F, Capobianchi MR. Effects of IFNα on late stages of HIV-1 replication cycle. Biochimie 1998, 80: 745 –754.
19. Bastiani Lallos L, Laal S, Hoxie JA, Zolla-Pazner S, Bandres JC. Exclusion of HIV coreceptors CXCR4, CCR5, and CCR3 from the HIV envelope. AIDS Res Human Retroviruses 1999, 15: 895 –897.
20. Frank I, Kacani L, Stoiber H. et al. Human immunodeficiency virus type 1 derived from cocultures of immature dendritic cells with autologous T cells carriers T-cell-specific molecules on its surface and is highly infectious. J Virol 1999, 73: 3449 –3454.
21. Nguyen DH, Hildreth JEK. Evidence for budding of human immunodeficiency virus type 1 selectively from glycolipid-enriched membrane lipids rafts. J Virol 2000, 74: 3264 –3272.
22. Spina CA, Prince HE, Richman DD. Preferential replication of HIV-1 in the CD45RO memory cell subset of primary CD4 lymphocytes in vitro. J Clin Invest 1997, 99: 1774 –1785.
23. Ostrowski MA, Chun T, Shawn J. et al. Both memory and CD45RA+/CD62L+ naive CD4+ T cells are infected in human immunodeficiency virus type 1-infected individuals. J Virol 1999, 73: 6430 –6435.
24. Gea-Banacloche JC, Clifford Lane H. Immune reconstitution in HIV infection. AIDS 1999, 13: S25 –S38.
25. Tuosto L, Piazza C, Moretti S. Ligation of either CD2 or CD28 rescues CD4+T cells from HIV-gp120-induced apoptosis. Eur J Immunol 1995, 10: 2917 –2922.

cell derived molecules; immobilized antibody capture; virus envelope; structured therapy interruption

© 2001 Lippincott Williams & Wilkins, Inc.