Maintenance of HIV-Specific Memory B-Cell Responses in Elite Controllers Despite Low Viral Burdens

Abstract

Human immunodeficiency virus (HIV)–specific B-cell responses in infected individuals are maintained by active HIV replication. Suppression of viremia by antiretroviral therapy (ART) leads to quantitative and qualitative changes that remain unclear. Accordingly, B-cell responses were investigated in elite controllers (ECs), who maintain undetectable HIV levels without ART, and in individuals whose viremia was suppressed by ART. Despite a higher HIV burden in the ART group, compared with the EC group, frequencies of HIV-specific B cells were higher in the EC group, compared with those in the ART group. However, the initiation of ART in several ECs was associated with reduced frequencies of HIV-specific B cells, suggesting that responses are at least in part sustained by HIV replication. Furthermore, B-cell responses to tetanus toxin but not influenza hemagglutinin in the ART group were lower than those in the EC group. Thus, the superior HIV-specific humoral response in ECs versus ART-treated individuals is likely due to a more intact humoral immune response in ECs and/or distinct responses to residual HIV replication.

Human immunodeficiency virus (HIV) disease is associated with phenotypic and functional abnormalities of B cells that arise over the course of infection, several of which are driven by immune-activating effects of ongoing HIV replication [1, 2]. In the memory B-cell compartment, abnormalities arise early, intensify during the chronic phase of infection in viremic individuals, and can be reversed by early initiation of antiretroviral therapy (ART) [3]. Furthermore, HIV-specific B-cell responses in chronically HIV-viremic individuals are enriched within abnormal B-cell subsets and normalize with reduction of the viremia level by initiation of ART, suggesting that HIV-specific B-cell responses and associated cellular abnormalities are maintained by persistent HIV replication [4, 5].

A small percentage of HIV-infected individuals, referred to as elite controllers (ECs), spontaneously control plasma viremia, maintaining relatively normal CD4⁺ T-cell counts and levels of plasma viremia undetectable by conventional assays for years to decades without the need for ART [6]. ECs represent a heterogeneous group with diverse racial backgrounds and modes of HIV transmission, although genetic and immunologic characteristics, such as overrepresentation of certain HLA alleles, have been reported in this group of patients [7, 8]. The EC phenotype is likely a multifactorial phenomenon resulting from a combination of several host and/or viral factors, thus providing a unique opportunity to understand the mechanisms underlying the natural control of HIV infection. These studies may also provide clues for the development of therapeutic vaccines and the control of HIV replication in the absence of ART.

Several studies have addressed a variety of virologic and immunologic parameters in HIV infection by comparing ECs and HIV-infected individuals whose viremia is suppressed by ART. Most notably, while viral burdens of ECs and HIV-infected individuals who are receiving ART have been shown to be comparable [9], HIV-specific CD8⁺ T-cell responses are described as distinctly superior in ECs [10–13]. In addition, initiation of ART and the resulting suppression of viremia during the chronic phase of progressive infection fails to restore function [10, 12, 13], suggesting that the HIV-specific cellular response is a determinant, rather than a product of, control of viral replication. Last, although studies have shown evidence that ECs maintain immunologic control of viral replication for prolonged periods in the absence of disease progression, they nonetheless maintain higher levels of immune activation and chronic inflammation than HIV-infected individuals whose viral replication is suppressed by ART [14–17]. These studies support the hypothesis that highly functional HIV-specific CD8⁺ T cells may play a major role in maintaining the elite status of HIV-infected individuals while not fully eliminating the presence and effects of the virus.

The role of antibodies in achieving viral suppression in ECs is unclear. Serum levels of HIV-specific neutralizing antibodies in ECs have been described as either similar to or lower than those in viremic individuals [18, 19]. Other studies examining the HIV-specific antibodies in several groups of HIV-infected individuals, including ECs and individuals receiving ART, have demonstrated that a persistent but low amount of circulating HIV is necessary to produce and maintain broad HIV-neutralizing antibody activity [20], although ECs appear to have a superior capacity in terms of antibody polyfunctionality [21]. However, few studies have compared subsets together with antigen-specific functionalities of B cells in ECs and HIV-infected individuals receiving ART. Pensieroso et al demonstrated significantly lower and higher percentages of naive and activated memory B cells, respectively, in ECs, compared with healthy controls [22]. They also reported significantly higher frequencies of resting memory B cells in ECs as compared to HIV-infected individuals receiving ART [22].

Several studies have demonstrated ongoing viral replication in ECs, as evidenced by persistent residual viremia, immune activation, and genetic evolution of virus [23–26]. However, the effect of extremely low levels of viral replication on humoral immunity against HIV remains largely unknown. In this study, we investigated HIV-specific and other antigen-specific memory B-cell responses in 2 groups of HIV-infected individuals, ECs and chronically HIV-infected individuals whose viremia was suppressed by ART. In addition, the longitudinal effect of reducing residual viral replication by ART in ECs on the frequencies of HIV-specific B cells was also investigated. Our findings suggest that low levels of residual viral replication together with a more adequate humoral immune response contribute to higher frequencies of HIV-specific memory B cells in ECs, compared with chronically HIV-infected individuals receiving ART.

MATERIALS AND METHODS

Study Subjects

Leukapheresis products and blood specimens were obtained from study subjects following written informed consent and approval by the institutional review boards of the National Institute of Allergy and Infectious Diseases (Bethesda, Maryland), University of Maryland School of Medicine (Baltimore), and University of Toronto (Canada) and by the Office of Human Subjects Research at the National Institutes of Health (Bethesda). Cryopreserved peripheral blood mononuclear cells (PBMCs) were obtained from 2 cohorts, one based at the National Institutes of Health and the other at the Institute of Human Virology (Baltimore, Maryland). All HIV-infected individuals had undetectable HIV plasma viremia (<50 HIV RNA copies/mL). ECs had undetectable plasma viremia for at least a 2-year period in the absence of ART, as previously defined [27, 28]. Extended criteria for inclusion are described in the Supplementary Methods.

Phenotypic Analysis

PBMCs were obtained by density-gradient centrifugation and cryopreserved. Mature (CD10⁻) B cells were isolated from PBMCs by negative magnetic bead-based selection, as previously described [5]. Immunophenotyping was performed using the conjugated monoclonal antibodies listed in the Supplementary Methods. Analyses were performed on a FACSCanto II flow cytometer (BD Biosciences, San Jose, California), using FlowJo, version 9.8.5 software (Tree Star, Ashland, Oregon).

Enzyme-Linked Immunospot (ELISPOT) Assay

An ELISPOT assay was used to enumerate antigen-specific and total immunoglobulin G (IgG)–, immunoglobulin A (IgA)–, and immunoglobulin M (IgM)–secreting B cells, as described previously [4] and in the Supplementary Methods.

Quantitative Analyses of HIV Burden

The frequency of cells carrying HIV RNA and DNA was determined as described in the Supplementary Methods.

Statistical Analyses

Associations between continuous variables were evaluated by Spearman rank order correlation, and group comparisons were performed by nonparametric analysis (using Kruskal–Wallis and/or Mann–Whitney U tests), as described elsewhere [3].

RESULTS

Description of Cohorts

Three groups of individuals were studied (Table 1), including 9 ECs, 10 chronically HIV-infected individuals receiving ART with suppressed viremia, and 8 HIV-negative controls. The 3 groups were similar in age and sex distributions; however, CD4⁺ T-cell counts and percentage were significantly higher in the EC group as compared to the ART group (Table 1).

Table 1.

Profiles and Clinical Observations of Human Immunodeficiency Virus (HIV)–Infected Individuals Receiving Antiretroviral Therapy (ART), Untreated HIV-Infected Elite Controllers (ECs), and HIV-Negative Individuals (NEG)

Characteristic	ART (n = 10)	EC (n = 9)	NEG (n = 8)	P, ART vs EC	P, ART vs NEG	P, EC vs NEG
Age, y	50 (29–65)	55 (43–60)	54 (40–67)	NS	NS	NS
Sex, no.				NS	NS	NS
Male	8	5	6
Female	2	4	2
CD4+ T-cell count, cells/µL	562 (346–2229)	1054 (587–2285)	Not available	.013	NA	NA
Percentage of CD4+ T cells	40 (24–47)	49 (23–62)	Not available	.041	NA	NA
Plasma HIV RNA level, copies/mL	<50	<50	NA	NS	NA	NA
Duration of suppressive ART, mo	54 (22–93)	NA	NA	NA	NA	NA
Time after influenza vaccination, mo	6.0 (2–10)	4.5 (3–14)	6.0 (4–10)	NS	NS	NS
Time after tetanus vaccination, mo	19 (6–108)	29 (8–106)	52 (14–78)	NS	NS	NS

Data are median values (ranges), unless otherwise indicated. P values were assessed by the Kruskal–Wallis (age and vaccination data), Mann–Whitney U (CD4⁺ T-cell count and viremia level), or χ² (age) tests.

Abbreviations: NA, not applicable; NS, not significant.

B-Cell Subpopulations in Different Groups of HIV-Infected Individuals

Frequencies of B-cell subsets that commonly circulate in the peripheral blood of HIV-infected individuals, as recently reviewed [2], were evaluated. Consistent with previous findings [22], among the total B-cell population, the proportion of naive B cells was significantly higher (P = .004) in the ART group than in the HIV-negative group (Figure 1). The proportion of resting memory (RM) B cells was significantly higher (P = .005) in the HIV-negative group as compared to the ART group, whereas differences between the 2 HIV-infected groups and between the EC and HIV-negative groups were not significant. As expected, there were no significant differences in the other subsets that have been associated with ongoing viral replication and/or disease progression, namely immature/transitional B cells, for the latter [29], and tissue-like memory (TLM) B cells, activated memory (AM) B cells, and plasmablasts, for the former [3].

Figure 1.

Immunophenotyping of the B-cell subsets from chronically human immunodeficiency virus (HIV)–infected individuals receiving antiretroviral therapy (ART), untreated elite controllers (ECs), and HIV-uninfected individuals (NEG). The percentage of cells belonging to each of the 6 peripheral blood B-cell subsets identified was measured in all individuals described in Table 1. Differences among groups that were significant (P < .05) by the Kruskal–Wallis test prompted group-wise comparison by the Mann–Whitney U test (**P < .01 in subset with higher value).

HIV-Specific Memory B-Cell Responses in the 2 Groups of HIV-Infected Individuals

HIV-specific responses among IgG⁺ B cells from individuals in the EC and ART groups were evaluated by flow cytometry, using gp140 probes, as previously described [5, 30]. As shown in Figure 2A, the median frequency of HIV-specific B cells in the EC group was significantly higher than that in the ART group. One advantage of evaluating antigen-specific responses by multiparametric flow cytometry is that responses within subsets can also be measured without having to purify each subset, as required for ELISPOT analysis. When the distribution of HIV-specific responses was analyzed by B-cell subset, the majority of responses for both groups was contained within the RM subset (Figure 2B), consistent with previous findings showing that, as plasma viremia is suppressed by ART, the proportion of the response decreases among AM and TLM B cells as it increases among RM B cells [5]. When the distribution among subsets shown in Figure 2B was compared between the 2 groups, the ART group had higher proportions of total response within TLM B cells (P = .04) and intermediate memory (IM; CD27^-/IgG⁺) B cells (P = .004) than did the EC group.

Figure 2.

Frequencies of human immunodeficiency virus (HIV)–specific B cells in chronically HIV-infected individuals receiving antiretroviral therapy (ART) and untreated elite controllers (ECs), evaluated by flow cytometry with gp140 probe–binding frequencies among immunoglobulin G (IgG)⁺ B cells (A) and the distribution of gp140 binding among B-cell subsets (B), and measured by enzyme-linked immunospot assay as gp140-specific antibody-secreting cell (ASC) frequencies among cultured B cells (C) or as a percentage of all immunoglobulin (Ig) ASC (D). Significant differences are indicated by P values (by the Mann–Whitney U test). Horizontal bars represent median values. B-cell memory subsets as are follows: intermediate (IM), resting (RM), tissue like (TLM), and activated (AM).

Until recently, the ELISPOT assay was the only option for evaluating frequencies of antigen-specific memory B cells. In contrast to flow cytometry, memory B-cell responses evaluated by ELISPOT require 4–5 days of stimulation in vitro for memory B cells to differentiate into antibody-secreting cells (ASC). Despite these differences, the 2 methods can be compared and possibly used to strengthen observations, especially given that the same biotinylated gp140 probe is used to detect both types of responses. Accordingly, ELISPOT assays were performed as previously reported [4, 5]. Consistent with the flow cytometry–based evaluation, HIV-specific memory B-cell responses measured by ELISPOT were significantly higher in the EC group as compared to the ART group, whether reported as absolute numbers of HIV-specific ASC (Figure 2C) or as percentages of all ASC (Figure 2D). Similar findings were obtained with ECs and ART-treated individuals who met the HIV virologic but not immunization inclusion criteria (data not shown). These data are also consistent with another study that measured Gag- and Env-specific memory B-cell responses during disease progression [31]. Taken together, these data show that ECs have significantly higher frequencies of HIV-specific memory B cells than do chronically HIV-infected individuals receiving ART.

Non-HIV Antigen–Specific Memory B-Cell Responses as a Measure of General Immune Competency

Evaluation of antigen-specific memory B-cell responses against non-HIV antigens was used to determine whether the higher response against HIV in ECs was a reflection of better overall immune competency, compared with that among individuals in the ART group. Accordingly, responses were measured against 2 pathogens to which individuals are routinely immunized and/or infected, namely influenza virus and Clostridium tetani, and for which probes (influenza hemagglutinin [HA] and tetanus toxin, respectively) have recently been used in flow cytometry [5]. All 3 groups were included in these analyses and there were similar time periods between sample procurement and vaccination among the groups (Table 1). As shown in Figure 3A, flow cytometry did not detect significant differences in responses to influenza HA between the 3 groups. However, responses to tetanus toxin were significantly lower in the ART group as compared to both the EC and HIV-negative groups. Finally, as with responses against HIV, the distribution of the responses against non-HIV antigens was enriched within the resting memory B-cell subset in all 3 groups, for both influenza HA (Figure 3B) and tetanus toxin (Figure 3C). Of note, there were no differences when the distributions among subsets shown in Figure 3B and 3C were compared among the 3 groups. Collectively, these findings suggest that ECs respond to non-HIV antigens at levels and with profiles that are similar to those observed in HIV-negative individuals, whereas ART-treated individuals showed a deficiency in response to tetanus toxin.

Figure 3.

Frequencies of B cells specific for influenza HA (strain H1-CA09) and tetanus toxin C fragment (TTCF) in chronically HIV-infected individuals receiving antiretroviral therapy (ART), untreated elite controllers (ECs), and HIV-uninfected individuals (NEG). Frequencies of probe-binding immunoglobulin G (IgG)⁺ B cells measured by flow cytometry (A) and distribution of binding among B-cell subsets for influenza HA (B), and tetanus toxin (C). Differences among groups that were significant (P < .05) by the Kruskal–Wallis test prompted group-wise comparison by Mann–Whitney U test, with P values shown. Horizontal bars represent median values. B-cell memory subsets as are follows: intermediate (IM), resting (RM), tissue like (TLM), and activated (AM). Abbreviation: HA, hemagglutinin.

Correlations Between ELISPOT-Based ASC Frequencies and Antigen Binding, by Flow Cytometry

Next, analyses were performed to compare ASC frequencies measured by ELISPOT and antigen binding measured by flow cytometry, focusing exclusively on the HIV-infected groups, in which all 3 antigens were evaluated. As shown in Figure 4, there were significant direct correlations between HIV-specific and influenza HA-specific ASC frequencies and corresponding binding of antigen to IgG⁺ B cells. The correlation for HIV remained strong when the 1 outlier was removed (r = 0.7984; P = .0012). Tetanus was excluded from these analyses because tetanus toxin–specific ASC frequencies were very low in all 3 groups, possibly owing to lower levels of detection with rTTC protein as compared to the antigen included in tetanus vaccine [32]. Of note, while tetanus vaccine has been commonly used in evaluation of tetanus toxin–specific antibodies, it has become difficult to obtain commercially as a single agent, and thus we used rTTC protein. Taken together, these findings suggest that, despite the in vitro expansion required to measure ASC frequencies and the focus on IgG in flow cytometry but not ELISPOT, the 2 assays show strong correlations for at least HIV and influenza HA and as such, help solidify the observation that ECs have stronger HIV-specific responses than ART-treated individuals.

Figure 4.

Correlations between antigen-specific memory B cells (measured by flow cytometry) and antigen-specific antibody-secreting cells (ASC; measured by enzyme-linked immunospot assay). Data include individuals from human immunodeficiency virus (HIV)–infected groups described in Table 1, with elite controllers (ECs) shown in light gray circles and chronically HIV-infected individuals receiving antiretroviral therapy shown in gray circles. Spearman correlation coefficient r and P values are indicated in each graph. Abbreviations: HA, hemagglutinin; IgG, immunoglobulin G.

Frequencies of Cell-Associated HIV DNA and RNA in PBMCs From ECs and HIV-Infected ART Recipients and Effects of ART in ECs

Given previous observations that frequencies of HIV-specific memory B cells diminished in association with control of HIV plasma viremia by ART [5], viral burden was considered in the current study. Accordingly, levels of cell-associated HIV DNA (Figure 5A) and RNA (Figure 5B) were measured in PBMCs from the same sources used to measure HIV-specific B-cell responses. Consistent with findings of a recent study [33], the frequencies of PBMCs carrying HIV DNA and RNA were significantly higher in the ART group as compared to the EC group. These findings suggest that viral burden does not explain the increased frequencies of HIV-specific B cells observed in the EC group, especially given that the PBMCs from the EC group had more CD4⁺ T cells and thus more potential HIV targets than those from the ART group (Table 1). However, these were cross-sectional observations that may not fully capture viral dynamics. Another way to address the effects of viral burden in ECs is to perform longitudinal analyses following initiation of ART. In a previous study, a short course of ART in ECs led to a significant decrease in the infectious HIV burden and a subsequent rebound to pre-ART levels upon cessation of ART [26]. In the current study, B-cell responses were evaluated longitudinally in 2 individuals described in Table 1 and the 4 ECs previously reported [26]. As shown in Figure 6, HIV-specific B-cell responses, as measured by ELISPOT, decreased in all 6 individuals as the viral burden decreased with ART and in 3 of 4 individuals who later discontinued ART (S-1, S-3, and S-4), B-cell responses returned to or were greater than baseline levels in parallel with the return of the low level of viral burden. The decreases in HIV burden and B-cell responses before versus after ART were significant (P = .031), although correlations between these 2 parameters were not (data not shown). Collectively, these results suggest that, despite evidence from cross-sectional analyses that the viral burden alone did not explain the higher HIV-specific B-cell responses in the EC group as compared to the ART group, low levels of residual viral replication nonetheless likely, at least in part, drive such responses, as reflected from the longitudinal data in ECs receiving ART.

Figure 5.

Levels of cell-associated human immunodeficiency virus (HIV) DNA (A) and RNA (B) in peripheral blood mononuclear cells (PBMCs) from chronically HIV-infected individuals receiving antiretroviral therapy (ART) and untreated elite controllers (ECs). HIV proviral DNA (A) and HIV RNA (B) were measured by droplet digital polymerase chain reaction. Significant differences are indicated by P values (by the Mann–Whitney U test). The horizontal lines represent median values.

Figure 6.

Levels of human immunodeficiency virus (HIV) and frequencies of HIV-specific B cells in elite controllers (ECs) before and after initiation and discontinuation of antiretroviral therapy (ART). Levels of HIV DNA (EC-3 and EC-4 are from Table 1) or replication-competent HIV (S-1 through S-4 from the article by Chun et al [26]) in CD4⁺ T cells isolated from ECs at indicated time points. Frequencies of HIV-specific B cells were measured by enzyme-linked immunospot assay as in Figure 2D. B-cell cultures from S-2 and S-3 did not contain interleukin 21. The shaded area represents the period of ART, and open squares represent values below the limit of detection. Abbreviation: Ig, immunoglobulin.

DISCUSSION

ECs compose <1% of HIV-infected individuals, yet they are the focus of intense research efforts [34]. The identification of potential humoral immune-mediated mechanisms associated with the control of HIV viremia and the nonprogressive disease status in these individuals may lead to new approaches for preventing and managing HIV disease. It is generally agreed that the persistence of humoral responses to HIV in infected individuals, even those whose virus replication is well controlled by ART, is due to low levels of virus replication that continually stimulate B-cell responses [35–37]. However, the precise role of the residual viral burden in maintaining HIV-specific immune responses in ECs and the relationship of this interplay of humoral immunity and viral burden in ECs as compared to that in HIV-infected individuals who require ART to control viral replication are unclear. In this study, we addressed this issue in parallel in these 2 groups of individuals. We report that, despite a lower viral burden, as measured by HIV DNA and RNA levels in PBMCs, ECs displayed significantly higher frequencies of HIV-specific B cells than did chronically HIV-infected individuals receiving ART.

A difference in general immune competency was considered as a possible explanation for the differences in HIV-specific B cells between ECs and ART recipients. Lack of differences in response to non-HIV antigens, including influenza HA and tetanus toxin, between controllers and HIV-infected individuals receiving ART have been reported previously [31]. However, our findings were more nuanced in that influenza HA–specific but not tetanus toxin–specific B-cell responses were similar between the 2 HIV-infected groups and when compared to uninfected controls. The difference in findings may be explained by the patient groups investigated: we focused on ECs, while the previous study did not strictly confine their controller group to those with undetectable plasma viremia [31]. Of note, our findings were also similar to those of a recent study in which reduced HIV-specific B-cell responses were associated with impaired T-follicular helper function in ART-treated HIV-infected individuals as compared to ECs [38]. Our observations suggest that responses to frequently encountered pathogens (eg, influenza virus) are maintained in both HIV-infected groups, while responses to infrequently encountered recall antigens, such as tetanus toxin, may be diminished in the ART group. However, until larger cohorts are studied, these interpretations remain somewhat speculative. Nonetheless, given that HIV-specific B-cell responses are likely maintained by residual HIV replication, this state of viral persistence is more akin to exposure to influenza virus than to tetanus toxin in terms of frequency of antigen encounter. As such, the differences in the humoral response to HIV between ECs and ART-treated HIV-infected individuals, are reflective of HIV-specific B-cell responses and not to B-cell responses to recall antigens in general.

It is well known that B-cell subsets are altered during HIV disease and partially normalized after initiation of ART [3, 22, 31, 39]. Although enhanced broad HIV-neutralizing antibody activity has been associated with normalization of B-cell subsets [40], detectable viremia was necessary to drive the broad HIV-neutralizing antibody response, as other studies have demonstrated [19, 20]. In a previous study, we demonstrated that HIV-specific B-cell responses, as measured in the peripheral blood of infected untreated individuals, were concentrated within abnormal B-cell subsets—specifically, activated and tissue-like or exhausted memory B cells—which are largely absent in healthy uninfected individuals [5]. In addition, enrichment of HIV-specific responses within the resting memory B-cell subset, the predominant subset involved in responses to influenza HA and tetanus toxin, was associated with control of viremia and immune activation [5]. In the present study, similar observations were made for both HIV groups and all 3 antigens tested (as well as the uninfected group, for tetanus toxin and influenza HA), confirming that antigen-specific responses, including those against HIV, are enriched in resting memory B cells in the setting of low viral burden.

At first glance, Figure 5, which shows that the residual viral burden in ECs is much lower than that in ART-treated patients, suggests that the level of residual viral burden alone does not explain the significantly higher frequencies of HIV-specific B cells in the EC group. Certainly, the presence of residual virus in the EC group has an impact on the HIV-specific responses, as shown in Figure 6, where there was a reduction of HIV-specific memory B-cell responses following reduction of residual viral replication by ART in all 6 ECs that was reversed in 3 of 4 individuals who then discontinued ART. While this finding should be evaluated in a much larger cohort, the consistency of the decrease in B-cell response suggests that low but detectable levels of HIV replication may fuel humoral immunity against the virus in the EC group. In this regard, the reduction of HIV-specific responses following ART has been observed in other studies of B cells in viremic individuals with early or chronic HIV infection [3, 5] and of T cells in similar setting [41] and, to a more limited extent, in ECs [26]. It is clear that ECs harbor replication-competent virus and that residual viral replication likely occurs in the absence of ART and detectable viremia [26, 42–44]. While there may be debate as to whether viral burdens differ between ECs and ART-treated individuals, with some studies finding no differences [9, 13, 42, 45] and others finding smaller viral reservoirs in ECs [33, 46], these apparent discrepancies may simply be due to differences in cohort selection or the various assays used. Finally, it is possible that the decrease in HIV-specific B-cell responses in the ART group as compared to the EC group is due at least in part to the fact that the ART group had a higher proportion of the HIV-specific total response within the abnormal TLM B cells and IM (CD27⁻/IgG⁺) B cells than did the EC group.

In summary, our findings suggest that ECs have a superior B-cell response to HIV when compared to HIV-infected ART recipients. This superiority is unlikely to reflect a stronger capacity to respond to all pathogens, given the similarities among all 3 groups in the response to influenza HA. Superior HIV-specific B-cell responses were observed in ECs despite viral burdens lower than those in ART-treated individuals, although the longitudinal data suggest that viremia sustains these responses. Evidence suggests that antibodies do not play an important role in restricting viral replication in ECs [47, 48] and that ECs are not important sources of broad HIV-neutralizing antibodies [19]. Nonetheless, superior humoral memory responses, as evidenced by strong HIV-specific B cells despite a very low viral burden and a sustained response to infrequently encountered antigens (eg, those of C. tetani), adds to a mounting body of evidence that ECs maintain a strong immune function [10–13, 49, 50]. These findings thus provide insight into the interplay of virologic and immunologic factors in HIV infection, while illustrating the difficulties in elucidating which of these factors influence disease outcome.

Supplementary Data

Supplementary materials are available at http://jid.oxfordjournals.org. Consisting of data provided by the author to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the author, so questions or comments should be addressed to the author.