Persistence of HIV-1 structural proteins and glycoproteins in lymph nodes of patients under highly active antiretroviral therapy

  1. Mikulas Popovic*,†,
  2. Klara Tenner-Racz‡,
  3. Colleen Pelser*,
  4. Hans-Jurgen Stellbrink§,
  5. Jan van Lunzen§,
  6. George Lewis*,
  7. Vaniambadi S. Kalyanaraman,
  8. Robert C. Gallo*,†, and
  9. Paul Racz‡
  1. *Institute of Human Virology, University of Maryland Biotechnology Institute, University of Maryland, Baltimore, MD 21201; ‡Bernhard-Nocht Institute for Tropical Diseases, 2000 Hamburg 4, Germany; §University Hospital Eppendorf, D-20246 Hamburg, Germany; and Advanced BioScience Laboratories, Kensington, MD 20895

  1. Contributed by Robert C. Gallo, August 12, 2005

 
Next Section

Abstract

Here we report a long-term persistence of HIV-1 structural proteins and glycoproteins in germinal centers (GCs) of lymph nodes (LNs) in the absence of detectable virus replication in patients under highly active antiretroviral therapy (HAART). The persistence of viral structural proteins and glycoproteins in GCs was accompanied by specific antibody responses to HIV-1. Seven patients during the chronic phase of HIV-1 infection were analyzed for the presence of the capsid protein (HIV-1p24), matrix protein (HIV-1p17), and envelope glycoproteins (HIV-1gp120/gp41), as well as for viral RNA (vRNA) in biopsy specimens from LNs obtained before initiation of therapy and during HAART that lasted from 5 to 13 months. In parallel, these patients were also monitored for viremia and specific anti-HIV-1 antibody responses to structural proteins and glycoproteins both before and during treatment. Before-therapy viral levels, as determined by RT-PCR, ranged from 3 × 103 to 6.3 × 105 copies of vRNA per ml, whereas during treatment, vRNA was under detectable levels (<25 copies per ml). The pattern of vRNA detection in peripheral blood was concordant with in situ hybridization results of LN specimens. Before treatment, vRNA associated with follicular dendritic cells (FDCs) was readily detected in GCs of LNs of the patients, whereas during therapy, vRNA was consistently absent in the GCs of LN biopsies of treated patients. In contrast to vRNA hybridization results, viral structural proteins and glycoproteins, evaluated by immunohistochemical staining, were present and persisted in the GC light zone of LNs in abundant amounts not only before initiation of therapy but also during HAART, when no vRNA was detected in GCs. Consistent with immunohistochemical findings, specific antibody responses to HIV-1p17, -p24, and -gp120/gp41, as evaluated by ELISA and virus neutralization, persisted in patients under therapy for up to 13 months of follow-up. The implications of these findings are discussed in relation to HIV-1 persistence in infected individuals and the potential role of chronic antigenic stimulation by the deposited structural proteins in GCs for AIDS-associated B cell malignancies.

HIV infection is characterized by a severe impairment of both cellular and humoral immunity. Both T and B cell compartments are profoundly altered (1, 2). Parallel with immune-persistent activation of these compartments (1-4), HIV-infected individuals show decreased humoral responses to antigens (5-7). The alteration of B cells is manifested by hypergammaglobulinemia (1, 2), increased spontaneous antibody secretion in vitro (8), enhanced levels of autoantibodies (9), and increased incidence of B cell lymphomas (10). The widespread use of highly active antiretroviral therapy (HAART) has substantially modified the natural history of HIV-1 infection. The effects of this therapy are manifested by a strong suppression of viral replication in the peripheral blood and in lymphoid tissue in individuals infected with HIV-1 (11, 12). As a result, CD4+T cell counts increase, T cell activation decreases, and antigen-specific and nonspecific T cell function improves (13-18). Similarly, B cell responses are normalized, although this reversal of profound alteration of immune system is a slow and incomplete process in a number of long-term treated patients (19-22).

Earlier in vitro studies explored interactions of mononuclear cells from peripheral blood with native or recombinant HIV-1 structural proteins and glycoproteins, with the matrix protein HIV-1p17 (23, 24), and particularly with the HIV-1Env (gp120/160) (25). These extensive studies of B and T cell interactions with HIV-1Env and HIV-1p17 showed a broad spectrum of changes in cell surface markers, cytokine production, B cell maturation, and increased T cell proliferation and HIV-1 replication in the virus-infected T cell cultures (23-25). However, the significance of these studies has been in question because of the absence of clear-cut in vivo evidence demonstrating persistence of HIV-1 structural proteins and glycoproteins accessible to mononuclear cells.

Observations from earlier studies demonstrated that, in untreated individuals infected with HIV-1, the gag proteins (the capsid HIV-1p24 and -p17) can be consistently detected in germinal centers (GCs) of the lymphoid tissue (26-30). Double immunolabeling for the HIV-1gag proteins and for either markers of follicular dendritic cells (FDCs) or IgM revealed colocalized staining on the surface of FDCs (28). Because IgM is only bound to and not produced by FDCs, the finding indicates that the HIV-1gag protein in the GC is located on FDCs extracellularly and very likely, these antigen-antibody complexes are accessible to mononuclear cells. Importantly, long-term retention of antigen-antibody complexes on FDCs was documented in experimental animal studies during immunization (31).

Since the introduction of HAART, HIV-1 infection has been effectively controlled in a large number of patients for years. In treated patients, HIV-1 RNA [viral RNA (vRNA)] in blood frequently declines below detectable levels (11, 12). Although it is well established that HIV-1 infection and replication take place mainly in the lymphatic tissue, a limited number of systematic studies have been performed analyzing HIV-1 status in lymph node (LN) biopsies from patients before and during long-term therapy. In these studies, the persistence of HIV-1p24 was observed in GCs of LNs in patients chronically infected with HIV-1 who were under HAART and exhibited vRNA below detectable levels in blood and LN specimens (32-36). These studies did not address the persistence of other HIV-1 structural proteins and glycoproteins in HIV-1 infected patients under long-term antiretroviral therapy. Taking into consideration earlier in vitro studies on the capacity of HIV-1p17 and -1Env to induce a broad spectrum of changes in mononuclear cells (23-25), it was important to establish the longterm persistence of these viral structural protein and glycoproteins in patients under HAART. In this paper, we provide evidence demonstrating the long-term persistence of HIV-1p17, -p24, and -gp120/gp41 in GCs of LNs in seven HIV-1-infected patients under HAART. These viral structural proteins and glycoproteins persist in LNs in the absence of detectable HIV-1 replication and were accompanied by the presence of specific anti-HIV-1 antibodies in patients' sera.

Previous SectionNext Section

Materials and Methods

Patients. A total of seven patients chronically infected with HIV-1 were included in this study. The patients' code numbers, dates of procurement of blood, plasma, sera, and biopsy specimens from LNs obtained before treatment and during HAART are listed in Table 1, which is published as supporting information on the PNAS web site. Five patients were from the clinical study “COSMIC” (35) and were treated by the following antiviral drugs: stavudine, lamivudine, nelfinavir, saquinavir, and recombinant IL2. Results of this clinical trial have been reported (35). Two patients were from the clinical study “PEGI” and were treated with zidovudine, laminvudine, neviraprine, and nelfinavir. The cases for the study were selected from these clinical trials according to the following criteria: (i) the availability of paraffin-embedded as well as frozen tissue specimens in liquid nitrogen; (ii) the presence of a marked follicular hyperplasia (FH) in LNs at the baseline; (iii) the follow-up biopsies should show the persistence of FH in some patients and normalization of the follicles in others. From the PEGI clinical trial, one case was with FH and one with nonenlarged GCs. Biopsies of axillary LNs were performed in patients before initiation of treatment and during therapy lasting 5-13 months, when plasma viremia exhibited <25 copies per milliliter. As controls, stored LN tissues with FH not related to HIV-1 infection were selected from our files.

Lymphocyte Subsets. Lymphocyte subsets were evaluated on fresh whole blood samples from these patients by flow cytometry (FACSCalibur, Becton Dickinson) by using monoclonal antibodies directed against CD3+, CD4+, and CD8+ (Becton Dickinson), as described (35). Values of the lymphocyte subsets determined before and during HAART are listed in Table 1.

HIV-1 RNA Assay. Viremia was measured in plasma samples by detection of HIV-1 RNA using Roche Amplicor PCR (Roche Diagnostics, Mannheim, Germany). Before initiation of the anti-retroviral therapy as well as subsequently in the course of therapy, HIV-1 RNA copy numbers were monitored in a single laboratory from 6 to 13 months. When values fell below 400 copies of vRNA per milliliter, an ultrasensitive modification of the assay was applied by using 1 ml of plasma (Cobas Amplicor HIV-1 Monitor, Version 1.5, ultrasensitive protocol, Roche Diagnostics). The detection limit of vRNA was <25 viral copies per milliliter of plasma. Detected HIV-1 copy numbers before and during therapy are listed in Table 1.

Light Microscopy. Biopsies of LN tissues were fixed in 4% neutral-buffered formalin overnight, embedded in paraffin, and stained with hematoxylin/eosin, Giemsa, or Gomori silver impregnation for routine histology. Portions of LNs were embedded in tissue-freezing medium (Leica, Nussloch, Germany), snap-frozen in liquid nitrogen, and stored at -70°C until use.

Immunohistochemistry. Before staining, frozen sections of LNs were cut, fixed in 4% paraformaldehyde, and incubated with primary antibodies against HIV-1 proteins: p17 (BT-2, a mouse monoclonal antibody, Advanced BioScience Laboratories) and p24 (DAKO) at room temperature for 30 min. To detect HIV-1Env tissue, sections of LNs were incubated with ant-gp120 mouse monoclonal antibodies M85-2 or M777B3-4 (Advanced BioScience Laboratories) at 4°C overnight. Antibody binding was visualized by using the alkaline phosphatase antialkaline phosphatase technique, with New Fuchsin as red chromogen (30, 32).

In Situ Hybridization. HIV-1 RNA detection was performed on paraffin and frozen sections using a 35S-labeled single-stranded (antisense) RNA probe (Lofstrand Laboratories, Gaithersburg, MD). The probe was from 1.4- to 2.7-kb fragments representing ≈90% of the HIV-1 genome (37). The in situ hybridization procedure was described in detail elsewhere (32, 37). From each biopsy, 18-50 sections were hybridized, because it is known that in patients who respond well to HAART, cells expressing HIV-1 RNA are rare (38). As a positive control, cytospin preparations of H9 cells infected with HIV-1 were hybridized with the same probe. As a negative control, sections from each LN specimen were hybridized with a radiolabeled RNA sense-strand probe. Slides were dipped into photo emulsion (NTB2; Kodak) and exposed in the dark at 4°C for 7 days. The slides were developed in a developer (D19, Kodak), fixed, counterstained with hematoxylin, and mounted. Examination of the slides was performed with a Zeiss Axiophot microscope equipped with epiluminescent illumination, a 3CD camera, and a personal computer-based image analysis system (KS 4000; Kontron, Esching, Germany). Cells expressing HIV-1 RNA were counted, the area of the section measured, and the frequency of RNA-producing cells per mm2 of tissue section calculated.

HIV-1 Antibodies to HIV-1p17, HIV-1p24, and HIV-1env Detected by ELISA. Antibody titers to HIV-1IIIBp17, -p24, and -gp120/160 were determined by ELISA according to the manufacturer's instructions (Advanced BioScience Laboratories). Briefly, Nunc Maxi-sorp (Nalge Nunc) plates were coated with protein (1 μg/well) in 100 μl of coating buffer [Protein Detector horseradish peroxidase (HRP) ELISA kit, Kirkgaard & Perry Laboratories] overnight at 4°C. The plate was then emptied and blocked with 300 μl of blocking buffer (BSA) for 1 h at room temperature. After blocking, sera were tested at dilution starting from 1:100 in blocking solution for 1 h at 37°C. The plate was then washed five times with wash buffer and incubated with the secondary antibody (goat anti-mouse, rabbit, or human HRP antibody, 1 μg/ml in 100 μl of BSA diluents) for 1 h at 37°C. After incubation, the plates were washed again five times with wash buffer, and 100 μl of a 1:1 solution of HRP substrate/H2O2 was added. The substrate was incubated for 10 min at room temperature before adding the stop solution, and the reading was done at the absorbance 450 nm. Values were considered positive when the readings of the absorbance in the tested samples were 2-fold higher that those in control samples.

Neutrialization of R5 and X4 HIV-1. Neutralization of R5 and X4 was tested in a β-galactosidase chemiluminescent assay that has been described (39). Briefly, U373/CD4/MAGI cells expressing either CCR5 or CXCR4 were allowed to attach overnight to 96-well flat-bottom tissue culture plates. The next day, the medium was replaced with serial dilutions of antibody in culture medium and 50 TCID50 (tissue culture 50% of infective dose) per well of either the HIV-193BR020 (R5) or the HIV-1MN (X4) laboratory isolates. After 24 h, the cells were washed and maintained in fresh culture medium for 6-7 days. Infectivity was then determined by a chemiluminescence assay (Applied Biosystems) according to the manufacturer's protocol.

Previous SectionNext Section

Results

Detection of HIV-1 RNA in Patients Before and During Treatment with HAART. As shown in Table 1, HIV-1 viremia was monitored by RT-PCR detection of vRNA in plasma samples obtained from seven patients before initiation of antiretroviral therapy and during HAART. The detected base-line values in these patients for HIV-1 in plasma before therapy were in the range from 3 × 103 to 6.3 × 105 vRNA copies per ml. During antiretroviral therapy, vRNA in all seven patients was consistently below detectable levels (<25 copies per ml).

Detection of HIV-1 RNA in LNs by in Situ Hybridization. In contrast to HIV-1 antigens, the amounts and distribution of HIV-1 RNA detected by in situ hybridization showed marked differences before initiation of treatment and during HAART. As shown in Fig. 1A, heavy diffuse autoradiographic signals representing FDC-bound HIV-1 RNA were seen in the GCs of LNs from patients before treatment. In patients under HAART, the diffuse labeling was absent even in GCs with well developed light zones (see Fig. 1B). Before treatment, single cells showing HIV-1 gene expression were always present, mainly but not exclusively in the GCs. The frequency of these cells per mm2 of tissue varied between 0.7 and 3.7. This contrasted sharply with follow-up biopsies in which cells with vRNA were very rare (0-0.005 HIV-1 RNA-positive cells per mm2 of tissue section) and showed no preferential localization within the microenvironment of the LN. Furthermore, all productively infected cells showed fewer silver grains than cells in the first LN biopsies from patients before therapy. Because the amount of hybridization signals parallels the amount of HIV-1 RNA produced by cells (11), the finding suggests that under HAART, only low levels of HIV-1 are produced per cell. Thus, the in situ hybridization data, namely the detection of HIV-1 vRNA in GCs of LNs prior to the therapy of these patients, coincided with viremia.

Fig. 1.
View larger version:
Fig. 1.

Changes in HIV RNA burden in the LNs during HAART. In situ hybridization for vRNA (green) before HAART (A) shows diffuse hybridization signals in the GC, representing FDC-bound vRNA and many virus-producing cells (arrow). The former is under limit of detection in the follow-up biopsy (B). Cells with HIV-1 gene expression are rare (arrow).


Morphology and Immunohistochemical Analyses of Patients' LNs. As described above, the vRNA detections, LN biopsies from the seven patients with chronic HIV-1 infection, were also investigated for the presence of viral structural proteins and glycoproteins before treatment and during HAART (see Table 1). In five patients from one study (COSMIC), LN biopsies were obtained while patients were under antiretroviral therapy that lasted 6-13 months. In two patients from the second study (PEGI), biopsies were obtained after 5 months of therapy. The conventional histology of LN biopsies showed a marked FH at the baseline. The large irregularly shaped follicles occupied up to two-thirds of the cross-sectional area of the LNs. The dark zone of the GC was broad and contained numerous centroblasts, many of them in mitosis, and tangible body macrophages. In the follow-up biopsies, the histology was unchanged or improved toward normal LN structure that manifested itself by a slight enlargement of one or two regularly shaped GCs. In two subjects, the FH resolved. In these cases, the LNs showed active but nonenlarged and regularly shaped GCs.

Immunohistochemical staining of tissue specimens from LNs for detection of HIV-1 structural proteins and glycoproteins was performed on frozen sections by using highly specific mouse monoclonal antibodies. In all seven patients, the major capsid HIV-1p24, the matrix HIV-1p17 proteins, and the viral envelope protein HIV-1gp120 were readily detected in the LN tissues obtained before as well as during treatment (see also Table 1). The staining patterns of these HIV-1 proteins in tissue specimens of LNs from patients before and under HAART are depicted in Fig. 2 A-D. The detection of HIV-1gp120 in the tissue specimen is shown from a patient under treatment in parallel with a control tissue exhibiting FH not related to HIV-1 infection (see Fig. 2 E and F). Large deposits of the HIV-1 antigens were present in the light zone of the GCs containing centrocytes, some CD4+ T cells, and a dense mesh of FDC. As expected, the staining for structural proteins and glycoproteins showed a reticular pattern resembling the FDC network in LNs. In follow-up biopsies, HIV-1 antigens were always detected, even in cases that showed remission of the FH. The dark zone of GC showed no staining for HIV-1 antigens, and preparations were negative for these antigens so long as tissue sections did not include the light zone. As shown in Fig. 2, the intensity of staining and its pattern for the HIV-1 proteins (-p24 and -p17) and also for the envelope glycoprotein HIV-1gp120 (untreated not shown) did not noticeably differ in tissue specimens taken from patients either before initiation of therapy or during treatment. Thus, the results from the immunohistochemical analyses strongly suggest that HIV-1 structural proteins and glycoproteins persist in GCs of LNs in patients under HAART for a long time in the absence of detectable virus replication, because both the GCs of LNs and peripheral blood samples were consistently negative for vRNA.

Fig. 2.
View larger version:
Fig. 2.

Persistence of HIV-1 structural proteins in the GC of patients under HAART. Frozen sections were immunolabeled for the capsid p24 (A and B), the matrix p17 (C and D), and the envelope glycoprotein gp120 (E and F). The amounts of HIV-1p24 and -p17 deposited in the GCs show no differences at baseline (A and C) and in the follow-up biopsy (B and D). The HIV-1gp120 was also detected in the follow-up biopsy (E) but not in GCs of tonsils obtained from an HIV-1 antibody-negative patient (F).


Antibody Responses to HIV-1 Structural Proteins and Glycoproteins. To substantiate the findings of long-term persistence of HIV-1 structural proteins and glycoproteins in GCs of LNs in these patients under HAART, the presence of specific antibodies directed to HIV-1p17, -p24, and Env (gp160) was assessed in sera by ELISA. Antibody titers were determined by end-point dilution to each of these HIV-1 antigens in sera samples obtained before treatment and during therapy that lasted from 5 to 13 months (see also Table 1). As shown in Fig. 3, significant differences in antibody titers were found in sera samples among the patients as well as in antibody responses of a patient to each HIV-1 antigen. Antibodies directed against p17 antigen exhibited the lowest titers in these patients and were in the range from 1:100 to 1:3,200 (see Fig. 3A). Antibody titers against p24 exhibited a very broad range from 1:400 to 1: 4,000,000 (see Fig. 3B). The strongest antibody responses in a more narrow range, from 1:512,000 to 1:2,048,000, were found against HIV-1gp160 (see Fig. 3C). In most cases, however, sera samples from patients obtained before treatment with high antibody titers against gp160 had also higher titers against p24 and p17. A serum sample (code no. 6-B70) from one patient under HAART was negative for anti-p17 antibodies and exhibited lower titer to p24 (1:200) and higher titer to gp160 (1:256,000). In this particular case, serum was available only from the time of therapy. In general, there was a tendency toward a decline in antibody titers in sera obtained from patients under HAART. As shown in Fig. 3, in most cases, an ≈4-fold decrease was observed in antibody titers against p24 and p17 and less to gp160 (≈2-fold). As pointed out by others in earlier studies, the observed differences between p24 and gp160 antibody titers and kinetics may reflect different mechanism in the regulation of anti-p24 and -gp120 responses. It is possible that the generation of anti-HIV-1p24 and also -HIV-1p17 antibodies is helper T cell-dependent, whereas anti-HIV-1gp120 antibody production is helper T cell-independent and more directly responsive to the amount of viral antigen (19, 40).

Fig. 3.
View larger version:
Fig. 3.

Detection-specific anti-HIV-1 antibodies directed against the matrix protein p17 (A), the capsid protein p24 (B), and the envelope glycoproteins (gp160) (C). Note the persistence of the HIV-1 antibody titers evaluated by ELISA. There was an ≈2- to 4-fold decrease in antibody titers against these HIV-1 structural proteins and glycoproteins under HAART.


Presence of HIV-1 Neutralizing Antibodies. We have further characterized gp160 antibodies for their capacity to neutralize HIV-1 R5 and X4 isolates. The results from virus-neutralization experiments are shown in Fig. 4. The HIV-193BR020 isolate of R5 phenotype and the HIV-1MN isolate of X4 phenotype were used in these experiments, respectively. The neutralization capacity of sera was lower for the HIV-1 R5 isolate and reached its plateau for most sera samples at dilution 1:80, whereas neutralizing titers for the HIV-1 X4 isolate were at 1:320 dilution with these sera. These data are in agreement with earlier observations that an HIV-1 isolate of X4 phenotype adapted for growth in T cell lines is more sensitive to neutralizing antibodies (41, 42). Comparing neutralizing capacity of paired sera samples obtained from patients before and during HAART, the same pattern of decline was found as in the case of anti-HIV-1gp160 antibody detection by ELISA. The neutralizing titers exhibited 2- to 4-fold decreases in those sera obtained from patients during treatment.

Fig. 4.
View larger version:
Fig. 4.

Detection of virus neutralization antibodies against the R5 (A) and the X4 (B) HIV-1 isolates. Also the virus neutralizing antibodies persisted during the long-term HAART.


In conclusion, patients under long-term antiretroviral therapy not only retained HIV-1 structural proteins and glycoproteins deposited in GCs of lymph-nodes but also maintained specific antibody responses to these HIV-1 antigens in the absence of detectable virus replication.

Previous SectionNext Section

Discussion

The results described in this paper provide evidence that, in addition to the capsid protein HIV-1p24, the matrix protein HIV-1p17 and the envelope glycoproteins HIV-1Env (gp120/gp41) are also accumulated in the GC light zone of LNs in patients who were treated with antiretroviral drugs for the entire period of this study, up to 13 months. Previous studies focused on immunohistochemical detection of only one viral structural protein, namely p24, in patients under HAART (32-36). Although limited to the detection of one viral structural protein for a short period, these studies have already suggested the persistence of HIV-1 proteins and glycoproteins in GCs of LNs, because vRNA diminished to undetectable levels in patients' plasma and LN biopsies within 2-3 months from initiation of therapy (33-36). We extended this observation and firmly established that all major components of viral structural proteins and glycoproteins persist in GCs of LNs in patients chronically infected with HIV-1 in the absence of detectable virus replication under HAART for long periods of time. The persistence of these HIV-1 antigens in GCs of LNs was accompanied by the presence of specific antibodies recognizing HIV-1p17, -p24, and -gp120/gp41, respectively.

In this study, all seven patients responded well to HAART. The results of the in situ hybridization coincided with the data of the viral load detection in plasma. Because in these patients the peripheral virus burden was consistently below the limit of detection for at least 6 months, HIV-1 antigens could not be captured from the blood circulation. Although residual HIV-1 replication occurs in patients with undetectable vRNA both in plasma and in GC LNs (35, 38), it is unlikely that few productively infected cells found in LN biopsies of these patients (in the present study, a maximum of six RNA positive cells per 50 sections) provided the amounts of viral proteins and glycoproteins that could replenish the FDC antigen-antibody pool (see also Results). All LNs examined were consistently negative for FDC-bound vRNA. It is most likely that HIV-1 structural proteins and glycoproteins detected in the follow-up biopsies represent immune complexes captured by FDCs mainly before the initiation of the therapy. Thus, the GCs of LNs positive for HIV-1 antigens are long-lived where the viral structural proteins and glycoproteins persist for a long time.

Long-lived GCs with persisting antigens were also observed in mice after immunization with a recombinant vesicular stomatitis virus (VSV)-glycoprotein. VSV-specific B cells were associated with the persisting antigen and also produced immunoglobulins, hence strongly suggesting that FDC-bound viral antigens in the long-lived GCs most probably are involved in the maintenance of serological IgG memory (43, 44). The persisting humoral response to HIV-1 structural proteins and glycoproteins in patients under HAART very likely represent “conserved” antibodies, because they were generated and maintained mainly by the FDC antigenic pool from the virus before its replication was suppressed by therapy. Consequently, this “conserved” humoral response could provide an advantage for emergence of new HIV-1 variants resistant to the “historical” neutralizing antibodies. It will be important to compare patients' sera samples obtained before and during HAART and characterize in detail the reactivity of these “conserved” antibodies against HIV-1gp120/gp41, particularly in those cases when HIV-1-replicating variants emerge after long-term therapy. More detailed analyses of these antibodies can contribute to a better understanding of HIV-1 persistence in vivo.

A characteristic feature of HIV infection is B cell hyperactivity, manifested by marked hypergammaglobulinemia (1, 2). In the peripheral circulation of HIV-1-infected patients, B cells secrete abundant amounts of immunoglobulins, a fact that was also demonstrated by HIV-1 antibody production in vitro (8) and expression of B cell activation markers (3, 4). Several mechanisms have been suggested for excessive B cell activation in HIV-1 disease, including the direct stimulatory effects of viral proteins on B cells and induction of B cell stimulatory cytokines associated with HIV-1 infection (25). In addition, enhanced T cell help has been implicated in the induction of polyclonal B cell activation (45, 46). It has been demonstrated that the envelope glycoproteins induce differentiation of B lymphocytes from normal volunteers into Ig-secreting cells (20, 25). The induction of B cell differentiation is a complex process and is T cell-dependent, requiring contact between T cells exposed to HIV-1Env and B cells (20). Although several cell-surface molecules have been implicated into cell-to-cell contact-dependent interaction in HIV-1-induced B cell activation, the exact nature of the cell-surface molecules involved is not clear. In this context, the recent work of Hunziker et al. (47) using an animal model contributes to a better understanding of the mechanism of hypergammaglobulinemia in virus infection caused by pathogens that do not infect B cells. In mice infected with lymphocytic choriomeningitis virus (LCMV), a polyclonal hypergammaglobulinemia developed. More than 90% of the IgG-producing cells were nonspecific for LCMV. Hypergammaglobulinemia did not depend on IL-6 or B cell receptor specificity. The increase in IgG production resulted from switching natural IgM specificities to IgG in preimmune B cells with receptors that were not specific for the LCMV antigen. The process depended on CD4+T cells and the CD40L molecule. Hypergammaglobulinemia did not develop in mice deficient either in TCR or LD40L molecules. It is well established that CD40L is critical for GC formation and the Ig isotype switch that takes place mainly in the GC of LNs (48, 49). It is noteworthy that both viruses, HIV-1 and LCMV, show tropism for lymphoid organs and macrophages, and hypergammaglobulinemia develops during infection with these viruses when high amounts of viral antigens are generated.

Viremia in HIV-1-infected people is associated with generalized B cell dysfunction, resulting in the appearance of a phenotypically distinct subpopulation of B cells (20). This B cell subpopulation with plasmacytoid morphology and reduced expression of CD21 is a poor responder to B cell stimuli and secretes high levels of immunoglobulins (20). During effective antiretroviral therapy, levels of serum immunoglobulins and frequency of Ig-secreting cells are normalized in HIV-1-infected patients, thereby strongly suggesting that viremia drives B cell hyperreactivity in vivo (20, 50). Despite effective elimination of viremia, the immune system in a number of patients is not fully recovered to normal, and there is a discrepancy between the rapid decline of circulating Ig-secreting cells and a slow decline of IgG in the sera of patients during therapy. Patients successfully treated with anti-HIV-1 drugs generate specific antibodies to HIV-1 in the absence of detectable vRNA, suggesting persistent immune system activation by retention of HIV-1 antigens over long periods (19, 21). Based on these observations, Morris et al. (50) postulated that production of polyclonal IgG molecules in some compartment that is not in equilibrium with other components of the immune system is sustained for some time. Residual HIV-1 replication and the presence of persistent antigen trapped on FDCs were suggested as complementary factors sustaining immune response under HAART. Our study has established that HIV-1 proteins and glycoproteins are trapped and persist in GCs of LNs in patients before and after containment of virus by therapy.

In recent years, there has been discussion of the role of FDC-bound antigens in specific B cell responses, mainly in the regulation of serum IgG. Some researchers question the role of FDC-bound antigen in the GC reaction and maintenance of serological memory (51). A majority of researchers in this field, however, emphasize the importance of FDC and the GC reaction in maintaining serological memory (44, 48, 49). In this context, it should be pointed out that HIV-1 transgenic rats harboring an HIV genome deleted in p24 generate enlarged GCs in lymphatic tissues, morphologically almost indistinguishable from those in HIV-1 infected individuals (52). Consequently, we have focused our efforts on studies of p17 and Env persistence in GCs of LNs. These findings may be relevant to lymphopathogenesis involving GC formation and B cell activation, because both these viral structural protein and glycoproteins interact with mononuclear cells and induce and/or enhance a number of cytokines and chemokines that are critical in activation, recruitment, and cell proliferation (23-25).

Introduction of HAART has significantly prolonged and improved the quality of life in individuals infected with HIV-1. This potent therapy has effectively controlled not only HIV-1 load and replication but also the occurrence of AIDS-associated malignancies such as Kaposi sarcoma and B cell lymphoma of the CNS (10, 53, 54). It is well established that B cell stimulation and prolonged immune deficiency are main risk factors for non-Hodgkin's lymphoma (NHL) in people with AIDS (10). Given that at least the viral structural glycoproteins HIV-1gp120/gp41 and the matrix protein HIV-1p17 possess a capacity to actively interact with cells of the immune system (23-25), the long-term presence in GCs of LNs points to the fact that at least one risk factor for NHL was not eliminated by antiretroviral therapy. Earlier observations showed that the serum-soluble CD23 level correlates with the subsequent development of AIDS-related NHL (55, 56), and colocalization of CD23 expression in FDCs bounding the HIV-1 antigen-antibody complexes in GCs of LNs was also demonstrated (32).

Previous SectionNext Section

Acknowledgments

This paper is dedicated to Dr. Jan Svoboda, D.Sc, on the occasion of his 70th birthday. We thank Irene Kalisz, Sonya Geborgi, Petra Meyer, Gudrun Grossschupff, and Birgit Raschdoff for excellent technical assistance. This investigation was supported by the National Institutes of Health, National Institute of Allergy and Infectious Diseases Grant 1R03AIo55345-01; Concerned Parents for AIDS Research, Cigarette Restitution Funds; Fondazione Cassa di Risparmio di Perugia (Italy); and the German Ministry of Education and Research (Bundesministerium für Bildung und Forschung), Contract KompNet 01KI0211.

Previous SectionNext Section

Footnotes

  • ↵ † To whom correspondence may be addressed. E-mail: popovic{at}umbi.umd.edud or gallo{at}umbi.umd.edu.

  • These data were presented at the International Meeting of the Institute of Human Virology, Sept. 29-Oct. 3, 2003, Baltimore, MD.

  • Abbreviations: LN, lymph node; GC, germinal center; FDC, follicular dendritic cell; FH, follicular hyperplasia; HAART, highly active antiretroviral therapy; vRNA, viral RNA.

Previous Section
 

References

  1. ↵
    Lane, H. C., Masur, H., Edgar, L. C., Whalen, G., Rook, A. H. & Fauci, A. S. (1983) N. Engl. J. Med. 309 , 453-458.pmid:6224088
    Abstract
  2. ↵
    Pahwa, S. G., Quilop, M. T., Lange, M., Pahwa, R. N. & Greico, M. H. (1984) Ann. Intern. Med. 101 , 757-763.pmid:6388451
    MedlineISI
  3. ↵
    Martinez-Maza, O., Crabb, E., Mitsuyashu, R.T., Fahey, J. L. & Giorgi, J. V. (1987) J. Immunol. 138 , 3720-3724.pmid:2953790
    Abstract
  4. ↵
    Rosenber, Y. J., Lewis, M. G. & Kosco-Vilbois, M. H. (1998) Dev. Immunol. 6 , 61-70.pmid:9716906
    Medline
  5. ↵
    Amman, A. J., Schiffman, G., Abrams, D., Volberding, P., Ziegler, J. & Connant, M. (1984) J. Am. Med. Assoc. 251 , 1447-1449.
    Abstract
  6. Bernstein, L. J., Ochs, H. D., Wedgwood, R. J. & Rubinstein, A. (1985) J. Pediatr. 107 , 352-357.pmid:4032129
    CrossRefMedlineISI
  7. ↵
    Ballet, J. J., Sulcebe, G., Couderc, L. J., Danon, F., Rabian, C., Lanthrop, M., Clauvel, J. P. & Seligmann, M. (1987) Clin. Exp. Immunol. 68 , 479-487.pmid:3652522
    MedlineISI
  8. ↵
    Katz, I. R., Krown, S. E., Safai, B., Oettgen, H. F. & Hoffmann, M. K. (1986) Clin. Immunol. Immunopathol. 39 , 359-367.pmid:2938858
    CrossRefMedlineISI
  9. ↵
    Shirai, A., Cosentino, M., Leitman-Klinman, S. F. & Klinman, D. M. (1992) J. Clin. Invest. 89 , 561-566.pmid:1737846
    MedlineISI
  10. ↵
    Gaidano, G., Capello, D. & Carbone, A. (2000) Semin. Oncol. 27 , 431-441.pmid:10950370
    Medline
  11. ↵
    Cavert, W., Notermans, D. W., Staskus, K., Wietgrefe, S. W., Zupancic, M., Gebhard, K., Henry, K., Zhang, Z. Q., Mills, R., McDade, H., et al.(1997) Science 276 , 960-964.pmid:9139661
    Abstract/FREE Full Text
  12. ↵
    Notermans, D. W., Jurriaans, S., de Wolf, F., Foudraine, N. A., de Jong, J. J., Cavert, W., Schuwirth, C. M., Kauffmann, R. H., Meenhorst, P. L., McDade, H., et al. (1998) AIDS 12, 167-173.pmid:9468365
    CrossRefMedlineISI
  13. ↵
    Autran, B., Carcelain, G., Li, T. S., Blanc, C., Mathez, D., Tubiana, R., Katlama, C., Debre, P. & Leibowitch, J. (1997) Science 277 , 112-116.pmid:9204894
    Abstract/FREE Full Text
  14. Pakker, N. G., Notermans, D. W., de Boer, R. J., Roos, M. T., de Wolf, F., Hill, A., Leonard, J. M., Danner, S. A., Miedema, F. & Schellekens, P. T. (1998) Nat. Med. 4 , 208-214.pmid:9461195
    CrossRefMedlineISI
  15. Zhang, Z-Q., Notermans, D. W., Sedgewick, G., Cavert, W., Wietgrefe, S., Zupancic, M., Gebhard, K., Henry, K., Boies, L., Chen, Z., et al.(1998) Proc. Natl. Acad. Sci. USA 95 , 1154-1159.pmid:9448301
    Abstract/FREE Full Text
  16. Zanussi, S., Simonelli, C., Bortolin, M. T., D'Andrea, M., Crepaldi, C., Vaccher, E., Nasti, G., Politi, D., Barzan, L., Tirelli, U., et al. (1999) Clin. Exp. Immunol. 116 , 486-492.pmid:10361239
    CrossRefMedlineISI
  17. Notermans, D. W., Pakker, N. G., Hamann, D., Foudraine, N. A., Kauffmann, R. H., Meenhorst, P. L., Goudsmit, J., Roos, M. T., Schellekens, P. T., Miedema, F., et al. (1999) J. Infect. Dis. 180 , 1050-1056.pmid:10479130
    CrossRefMedlineISI
  18. ↵
    Landay, A. L., Spritzler, J., Kessler, H., Mildvan, D., Pu, M., Fox, L., O'Neil, D., Schock, B., Kuritzkes, D., Lederman, M. M., et al. (2003) J. Infect. Dis. 188 , 1444-1454.pmid:14624369
    CrossRefMedline
  19. ↵
    Notermans, D.W., de Jong, J. J., Goudsmit, J., Bakker, M., Roos, M. T. L., Nijholt, L., Cremers J., Hellings, J.-A., Danner, S. A. & de Ronde, A. (2001) AIDS Res. Hum. Retroviruses 17 , 1003-1008.pmid:11485617
    CrossRefMedline
  20. ↵
    Moir, S., Malaspina, A., Ogwaro, K. M., Donoghue, E. T., Hallahan, C. W., Ehler, L. A., Liu, S., Adelsberger, J., Lapointe, R., Hwu, P., et al. (2001) Proc. Natl. Acad. Sci. USA 98 , 10362-10367.pmid:11504927
    Abstract/FREE Full Text
  21. ↵
    Fournier, A.-M., Baillat, V., Alix-Panabieres, C., Fondere, J.-M., Merle, C., Segondy, M., Huguet, M.-F., Reynes, J. & Vendrell, J.-P. (2002) AIDS 16 , 1755-1760.pmid:12218386
    CrossRefMedline
  22. ↵
    Johansson, C. C., Bryn, T., Yndestad, A., Eiken, H. G., Bjerkeli, V., Froland, S., Aukrust, P. & Tasken, K. (2004) AIDS 18 , 171-179.pmid:15075533
    CrossRefMedlineISI
  23. ↵
    De Francesco, M. A., Caruso, A., Fallacara, F., Canaris, A. D., Dima, F., Poiesi, C., Licenziati, S., Corulli, M., Martinelli, F., Fiorentini, S., et al. (1998) AIDS 12 , 245-262.pmid:9517986
    CrossRefMedlineISI
  24. ↵
    De Francesco, M. A., Baronio, M., Fiorentini, S., Signorini, C., Bonfanti, C., Poiesi, C., Popovic, M., Grassi, M., Garrafa, E., Bozzo, L., et al. (2002) Proc. Natl. Acad. Sci. USA 99 , 9972-9977.pmid:12105273
    Abstract/FREE Full Text
  25. ↵
    Chirmule, N. & Pahwa, S. (1996) Microbiol. Rev. 60 , 386-406.pmid:8801439
    Abstract/FREE Full Text
  26. ↵
    Armstrong, J. A. & Horne, R. (1984) Lancet II , 370-372.
  27. Tenner-Rácz, K., Rácz, P., Dietrich, M. & Kern, P. (1985) Lancet I , 105-106.
  28. ↵
    Tenner-Racz, K., Racz, P., Bofill, M., Schulz-Meyer, A., Dietrich, M., Kern, P., Weber, J., Pinching, A. J., Veronese-Dimarzo, F., Popovic, M., et al. (1986) Am. J. Pathol. 123 , 9-15.pmid:3008562
    Abstract
  29. Biberfeld, P., Chayt, K. J., Marselle, L. M., Bieberfeld, G. & Gallo, R. C. (1986) Am. J. Pathol. 125 , 436-442.pmid:3642987
    Abstract
  30. ↵
    Tenner-Rácz, K., Rácz, P., Dietrich, M., Kern. P., Janossy, G., Veronese-Dimarzo, F., Klatzmann, D., Gluckman, J. C. & Popovic, M. (1987) AIDS 1 , 95-104.pmid:3130085
    Medline
  31. ↵
    Mandel, T. E., Phipps, R. P., Abbot, A. & Tew, J. G. (1980) Immunol. Rev. 53 , 29-58.pmid:6162778
    CrossRefMedlineISI
  32. ↵
    Tenner-Racz, K., Stellbrink, H.-J., van Lunzen, J., Scneider, C., Jacobs, J.-P., Raschdorff, B., Grossschuupff, G., Steinman, R. M. & Racz, P. (1998) J. Exp. Med. 187 , 949-959.pmid:9500797
    Abstract/FREE Full Text
  33. ↵
    Orenstein, J. M., Feinberg, M., Yoder, C., Schrager, L., Mican, J.-A.M., Schwartzentruber, D. J., Davey, R. T., Jr., & Walker, R .E. (1999) AIDS 13 , 2219-2229.pmid:10563707
    CrossRefMedline
  34. Kuster, H., Opravil, M., Schlaepfer, E., Fischer, M., Günthard, H. F., Lüthy, R., Weber, R. & Con, R. W. (2000) Am. J. Pathol. 156 , 1973-1986.pmid:10854220
    Abstract/FREE Full Text
  35. ↵
    Stellbrink, H.-J., van Lunzen, J., Westby, M., O'Sullivan, E., Schneider, C., Adam, A., Weitner, L., Kuhlmann, B., Hoffmann, C., Fenske, S., et al. (2002) AIDS 16 , 1479-1487.pmid:12131185
    CrossRefMedline
  36. ↵
    Schacker, T. W., Nguyen, P. L., Martinez, E., Reilly, C., Gatell, J. M., Horban, A., Bakowska, E., Berzins, B., van Leeuwen, R., Wolinsky, S., et al. (2002) J. Infect. Dis. 186 , 1092-1097.pmid:12355359
    CrossRefMedlineISI
  37. ↵
    Fox, C. H., Tenner-Racz, K., Racz, P., Fripo, A., Pizzo, P. A. & Fauci, A. S. (1991) J. Infect. Dis. 164 , 1051-1057.pmid:1955708
    MedlineISI
  38. ↵
    Zhang, L., Ramratnam, B., Tenner-Racz, K., He, Y., Vesanen, M., Lewin, S., Talal, A., Racz, P., Perelson, A. S., Korber, B. T., et al. (1999) N. Engl. J. Med. 340 , 1605-1613.pmid:10341272
    Abstract/FREE Full Text
  39. ↵
    Fouts, T., Godfrey, K., Bobb, K., Montefiori, D., Hanson, C. V., Kalyanaraman, V. S., DeVico, A. & Pal, R. (2002) Proc. Natl. Acad. Sci. USA 99, 11842-11847.pmid:12192089
    Abstract/FREE Full Text
  40. ↵
    Binley, J. M., Klasse, P. J., Cao, Y., Jones, I., Markowitz, M., Ho, D. D. & Moore, J. P. (1997) J. Virol. 71 , 2799-2809.pmid:9060635
    Abstract
  41. ↵
    Reitz, M. S., Jr., Wilson, C., Naugle, C., Gallo, R. C. & Robert-Guroff, M. (1988) Cell 54 , 57-63.pmid:2838179
    CrossRefMedlineISI
  42. ↵
    Spenlehauer, C., Kirn, A., Aubertin, A.-M. & Moog, C. (2001) J. Virol. 75 , 2235-2245.pmid:11160727
    Abstract/FREE Full Text
  43. ↵
    Bachman, M. F., Odermatt, B., Hengartner, H. & Zinkernagel, R. M. (1996) J. Exp. Med. 183 , 2259-2269.pmid:8642335
    Abstract/FREE Full Text
  44. ↵
    Bachmann, M. F. (1998) Immuol. Res. 17 , 329-344.
  45. ↵
    Schnittman, S. M., Lane, H. S., Higgins, S. E., Folks, T. & Fauci, A. S. (1986) Science 233 , 1084-1086.pmid:3016902
    Abstract/FREE Full Text
  46. ↵
    Chirmule, N., Oyaizu, N., Kalyanaraman, V. S. & Pahwa, S. (1992) Blood 79 , 1245-1254.pmid:1346976
    Abstract/FREE Full Text
  47. ↵
    Hunziker L., Recher M., Macpherson A. J., Ciurea A., Freigang S, Hengartner H. & Zinkernagel R. M. (2003) Nat. Immunol. 4 , 343-349.pmid:12627229
    CrossRefMedlineISI
  48. ↵
    Zinkernagel, R. M. (2002) Curr. Opin. Immunol. 14 , 523-536.pmid:12088689
    CrossRefMedlineISI
  49. ↵
    Kosco-Vilbois, M. H. (2003) Nat. Rev. Immunol. 3 , 764-769.pmid:12949500
    CrossRefMedlineISI
  50. ↵
    Morris, L., Binley, J. M., Clas, B. A., Bonhoeffer, S., Astill, T. P., Kost, R., Hurley, A., Cao Y., Markowitz, M., Ho, D. D., et al. (1998) J. Exp. Med. 188 , 233-245.pmid:9670036
    Abstract/FREE Full Text
  51. ↵
    Herberman, A. M. & Shlomik, M. J. (2003) Nat. Rev. Immunol. 3 , 757-764.pmid:12949499
    MedlineISI
  52. ↵
    Reid, W., Sadowska, M., Denaro, F., Rao, S., Foulke, J., Hayes, N., Jones, O., Doodnauth, D., Davis, H., Sill, A., et al. (2001) Proc. Natl. Acad. Sci. USA 98 , 9271-9276.