Evidence of Early B-Cell Dysregulation in Simian Immunodeficiency Virus Infection: Rapid Depletion of Naïve and Memory B-Cell Subsets with Delayed Reconstitution of the Naïve B-Cell Population

Abstract

Despite eliciting a robust antibody response in humans, several studies in human immunodeficiency virus (HIV)-infected patients have demonstrated the presence of B-cell deficiencies during the chronic stage of infection. While several explanations for the HIV-induced B-cell deficit have been proposed, a clear mechanistic understanding of this loss of B-cell functionality is not known. This study utilizes simian immunodeficiency virus (SIV) infection of rhesus macaques to assess B-cell population dynamics beginning at the acute phase and continuing through the chronic phase of infection. Flow cytometric assessment demonstrated a significant early depletion of both naïve and memory B-cell subsets in the peripheral blood, with differential kinetics for recovery of these populations. Furthermore, the altered numbers of naïve and memory B-cell subsets in these animals corresponded with increased B-cell activation and altered proliferation profiles during the acute phase of infection. Finally, all animals produced high titers of antibody, demonstrating that the measurement of virus-specific antibody responses was not an accurate reflection of alterations in the B-cell compartment. These data indicate that dynamic B-cell population changes in SIV-infected macaques arise very early after infection at the precise time when an effective adaptive immune response is needed.

Effective B-cell responses result in the generation of memory B-cell populations which are able to proliferate and produce antibodies that can control primary and secondary insults by microbial pathogens (2). Impaired maturation and timing of B-cell-mediated immune responses result in the production of ineffective antibodies, which are unable to control infection and may result in the persistence of the pathogen (36). Although human immunodeficiency virus (HIV) infection generally elicits high-titer antibodies, virus-specific titers do not correlate with delayed clinical progression, suggesting that antibodies produced during HIV infection are not sufficient to provide long-term viral control (6). Ineffective antibody production in the context of HIV infection could be a result of numerous T-cell and B-cell abnormalities induced either directly or indirectly through infection. B-cell perturbations, characterized during chronic infection, include hypergammaglobulinemia (11, 31), a diminished in vitro response to mitogenic stimulation (10, 37), diminished antibody responses to vaccination (15, 23), and loss of memory B-cell subsets (3, 10, 37). It is highly likely that these B-cell abnormalities are linked with the inability of HIV-infected individuals to form effective antibody responses to HIV and opportunistic pathogens.

B-cell perturbations during acute HIV infection may lead to dysfunctions observed during chronic infection. Despite numerous reports that hypothesized that B-cell phenotypic and functional abnormalities arise due to the effects of chronic infection, a limited number of acute infection studies have provided evidence that B-cell dysfunctions may be initiated much earlier. Studies by De Milito et al. and others have reported a decrease in CD27⁺ B cells associated with chronic HIV infection (3, 4, 10-12, 15, 30, 31, 36-38, 40). The reduction of this population may explain the diminished antibody responses to non-HIV antigens present in HIV-infected individuals. However, the mechanism for this loss of memory B cells during chronic infection is unclear. One possibility is that B-cell losses are related to reduced T-cell numbers. In a study by Titanji et al., a strong correlation between the number of CD4 T cells and the percentage of memory B cells was reported in chronic HIV infection (37). Conversely, others have reported that no correlation was found between CD4 numbers and memory B-cell numbers (3, 10). Interestingly, reductions in percentages of B cells, increased expression of Fas on B cells, increased total plasma IgG levels, a decreased percentage of IgM memory B cells, and decreased B-cell responses to antigenic stimulation have been shown to occur within 6 months of HIV infection (36, 37). Disruption of germinal centers in the gut during acute HIV infection may also compromise the humoral immune response (20). While these studies provide insight into virus-induced changes in the B-cell compartment during infection, it is difficult to ascertain precisely when these changes occur, due to limitations in sample size and numbers during this early period of infection. The conflicting reports reflect the high amount of variability present in human HIV infection and illuminate the need for a model to study B-cell populations in which experimental parameters can be more rigorously controlled. An understanding of the effects of HIV on the B-cell population during this critical early phase of infection is needed to determine how the initial interactions between virus and host immune system set the stage for long-term disease progression in the infected host. The simian immunodeficiency virus (SIV)/macaque model provides a system in which to ask these questions.

Studies in SIV-infected macaques have demonstrated that the number of total B (CD20⁺) cells in the periphery decreases dramatically during the acute phase of infection (13, 24). The loss of these cells coincides with a similar depletion of peripheral CD4 T cells and is associated with primary viremia. Interestingly, the loss of total B cells is greater in magnitude than the loss of CD4⁺ T cells (24). In order to understand how these cells are being depleted, it is necessary to characterize B-cell subsets during SIV infection in the macaque. The present study was designed to assess phenotypic changes in B-cell numbers during the acute phase of SIV infection, both in the total B-cell population as well as in B-cell subsets. Our results identified early, rapid changes in B-cell subsets that were not apparent in analysis of the total B-cell population. Specifically, we identified a significant depletion from the periphery of both the naïve (CD20⁺ CD27⁻) and memory (CD20⁺ CD27⁺) B-cell populations during acute infection and increased total B-cell population activation that may be related to ineffective antibody production commonly associated with SIV infection. Furthermore, the data demonstrate that measurement of envelope-specific antibody responses was not a sensitive reflection of SIV effects on B-cell subsets. These data provide novel information about the timing and dynamics of phenotypic changes in the B-cell compartment during SIV infection that may be associated with functional changes observed later in chronic infection. These results can be used to tailor therapeutic treatments designed to preserve the B-cell compartment early in SIV/HIV infection.

MATERIALS AND METHODS

Animals.

Ten colony-bred rhesus macaques (Macaca mulatta) of Indian origin were maintained and used in accordance with the guidelines of the Animal Care and Use Committee at the Oregon National Primate Research Center. Animals were infected on day zero with SIVmac239 intravenously with 10,000 infectious units. Beginning at day 105 postinfection all animals received daily antiretroviral therapy (ART) which included both tenofovir (PMPA; 30 mg/kg of body weight until day 134 and then 20 mg/kg subsequently) and emtricitabine (FTC; 50 mg/kg until day 134 and then 20 mg/kg subsequently). Complete blood counts were obtained at each blood draw by using a Coulter ACT 5 Diff open reader cell counter (Beckman Coulter, Fullerton, CA).

Viral quantification.

Assessment of plasma SIV RNA was carried out using a real-time reverse transcription-PCR assay (threshold sensitivity <100 SIV gag RNA copy equivalents/ml of plasma; interassay coefficient of variation ≤25%) (5, 21).

Flow cytometric analysis.

Flow cytometric analysis was performed on fresh lysed whole blood samples as described previously (39). Briefly, 100 μl of citrate treated whole blood was obtained in a blood collection tube (BD Diagnostics, Franklin Lakes, NJ). Biotinylated or directly fluorochrome-conjugated antibodies to extracellular targets were incubated with whole blood at room temperature for 60 min. Following incubation, cells were washed once with 4 ml of cold (4°C) phosphate-buffered saline (PBS) with 0.1% of bovine serum albumin and 0.02% sodium azide (wash buffer). Cells were then fixed and permeabilized by a 10-min incubation at room temperature with fluorescence-activated cell sorting (FACS) lysing solution (BD Biosciences, San Jose, CA), washed, and incubated twice with 0.5 ml of FACS permeabilizing solution (BD Biosciences, San Jose, CA) for 10 min at room temperature. Cells were washed twice with cold wash buffer and then stained with directly conjugated intracellular antibodies at room temperature for 30 min, washed once, and analyzed by flow cytometry. Freshly stained lymphocytes were differentiated using forward and side scatter characteristics on an LSRII apparatus (BD Biosciences, San Jose, CA).

The total B-cell population was differentiated using CD20 (eBioscience, San Diego, CA) expression and was further subdivided according to expression of CD27 (BD Bioscience, San Jose, CA) and IgD (Southern Biotech, Birmingham, AL). B cells were characterized as CD20⁺ CD27⁺ (memory) or CD20⁺ CD27⁻ (naïve). Further parsing of memory cells using IgD identified IgM-secreting (CD20⁺ CD27⁺ IgD⁺) cells as well as IgG/IgA-secreting (CD20⁺ CD27⁺ IgD⁻) cells. Cell populations were also assessed for surface expression of CD95 (BD Bioscience) and Ki67 expression levels (BD Bioscience, San Jose, CA) as measures of activation and proliferation. CD4 T-cell counts were obtained by analyzing expression of CD3-positive (BD Bioscience, San Jose, CA) and CD4-positive (BD Bioscience, San Jose, CA) cells. List-mode multiparameter data files were analyzed using the FlowJo software program (PC version 7.2.5; Tree Star Inc., Ashland, OR).

SIV envelope-specific antibody endpoint titer.

Antibody reactivities to detergent-disrupted SIVsmB7 envelope proteins (16) were determined in a conconavalin A (ConA) enzyme-linked immunosorbent assay (ELISA) as previously described (7). Briefly, SIVsmB7 viral envelope proteins (gp120 and gp41) were captured onto 96-well microtiter plates (Immulon 2HB; Dynex Technologies, Chantilly, VA) coated with 5 μg of ConA/well for 1 h at 25°C. After a washing step with PBS, nonspecific binding was blocked by the addition of 5% dry milk in PBS (blocking solution) to all wells and incubation for 1 h at 25°C. Heat-inactivated plasma samples were serially diluted in blocking solution and incubated in the SIVsmB7 envelope-coated wells for 1 h at 25°C. After an extensive washing, peroxidase-conjugated anti-monkey IgG (Nordic Immunology Laboratories, Tilburg, The Netherlands) was diluted in blocking solution, added to each well, and incubated for 1 h at 25°C and washed. Following the final wash step, all wells were incubated with TM Blue substrate (Seracare, Milford, MA) for 20 min at room temperature, color was developed by the addition of 1 N sulfuric acid, and colorimetric analysis of antibody binding to SIVsmB7 was performed at an optical density of 450 nm (OD₄₅₀) using a Spectra Max 340 PC (Molecular Devices, Sunnyvale, CA).

Statistical analyses.

Statistical analyses were performed using GraphPad Prism 4 (San Diego, CA). Paired t tests were performed, comparing values for individual and longitudinal time points to the time zero value within a population.

RESULTS

Acute-phase decline in total B-cell levels corresponds with a peak in viral titer and drop in CD4⁺ T cells.

Ten rhesus macaques were infected intravenously with SIVmac239 at the Oregon National Primate Research Center. Peripheral blood samples were analyzed longitudinally during the first 150 days postinfection. Viral titers peaked at day 10 postinfection (mean, 3.5 × 10⁷) and declined to an average set point at day 84 (mean, 7.3 × 10⁶) (Fig. 1). Two animals did not control viral replication, maintaining viral loads of 10⁷ to 10⁸ copies/ml for 100 days postinfection. Five animals demonstrated intermediate levels of control, with setpoint viral titers of 10⁶ to 10⁷ copies/ml. The last three animals had relatively effective viral control, with viral titers that continued to gradually decline for about 60 to 85 days postinfection, reaching setpoint titers of 10³ to 10⁵copies/ml.

FIG. 1.

Measurement of viral loads following SIV infection. Ten rhesus macaques were infected intravenously with SIVmac239 and analyzed for the number of viral copies per milliliter of blood from postinfection day (PID) 0 to 149. Viral loads for each animal are plotted individually. All animals were treated with PMPA (30 mg/kg from day 105 until day 134 and then 20 mg/kg subsequently) and FTC (50 mg/kg from day 105 until day 134 and then 20 mg/kg subsequently) (shaded area). Data for animals with high viral loads are delineated by dotted lines, intermediate viral loads are shown by solid lines, and low viral loads are shown by dashed lines.

CD4⁺ T cells decreased during the first 10 days postinfection as previously reported (24), reaching the lowest point of approximately 800 cells/μl, coincident with the peak in viral load (Fig. 2). Consistent with a prior study by Roederer et al., the peripheral CD20⁺ B-cell population also exhibited a substantial drop in cell numbers, from an average of 1,157 cells/μl preinfection to a nadir average of 378 cells/μl at day 10 postinfection (24). Following this initial drop, the number of peripheral CD20⁺ (total B) cells rebounded by 30 days postinfection to an intermediate plateau, then recovered to preinfection levels between 80 and 90 days postinfection. Interestingly, a spike in total B-cell numbers was observed following the initiation of ART.

FIG. 2.

Measurement of CD4⁺ T-cell and total CD20⁺ B-cell numbers following SIV infection. The average numbers of CD4⁺ T cells and total CD20⁺ B cells for all animals are shown. Data represent the averages of 10 animals, with error bars representing the standard errors between the animals for each time point. Asterisks denote a significant difference from a particular time point compared to the day zero value, using Student's t test (P < 0.05).

Naïve (CD20⁺ CD27⁻) B-cell numbers remained significantly lower than baseline levels longer than memory (CD20⁺ CD27⁺) B-cell numbers.

To further delineate naïve and memory B-cell subsets, total B cells were subdivided based on surface expression of CD27 (Fig. 3). For these studies, we took advantage of human antibodies that cross-reacted with the rhesus system, utilizing markers previously defined in humans. Thus, putative naïve cells were defined as cells that expressed CD20⁺ CD27⁻, while putative memory cells were defined as CD20⁺ CD27⁺ and will be referred to going forward as naïve and memory cells, respectively.

FIG. 3.

Differentiation of B-cell subsets using flow cytometric analysis. Representative plots of whole blood FACS staining and B-cell differentiation using antibodies to CD20, CD27, and IgD are shown. (a) CD20⁺ B cells were differentiated from the total lymphocyte population and delineated using side and forward scatter characteristics. (b) Expression of CD27 on B cells was used to separate CD27⁻ (naïve) from CD27⁺ (memory) B cells. (c) Memory B cells were further subdivided based on expression of IgD into CD20⁺ CD27⁺ IgD⁺ (IgM-secreting) and CD20⁺ CD27⁺ IgD⁻ (IgG/IgA-secreting) populations.

Memory (CD20⁺ CD27⁺) cell numbers (Fig. 4b) declined after day 3 postinfection to an average of 255.5 cells/μl by day 10 postinfection (Fig. 4c). The number of memory cells recovered to a level that was slightly reduced from preinfection levels (748.8 cells/μl) by day 30 postinfection. The number of naïve (CD20⁺ CD27⁻) cells (Fig. 4a) increased significantly (P < 0.005) from an average of 401.4 cells/μl at day 0 to 519.2 cells/μl at day 3 (Fig. 4c and Table 1). This initial increase was then followed by a significant decrease in cell numbers, reaching the lowest point by day 10 postinfection. The naive cell population remained significantly diminished compared to preinfection values out to day 40 postinfection, only returning to preinfection levels between days 80 and 90 postinfection.

FIG. 4.

Longitudinal analysis of naïve (CD20⁺ CD27⁻) and memory (CD20⁺ CD27⁺) B cells. Peripheral blood from rhesus macaques infected with SIVmac239 was obtained at the indicated time points after acute and early chronic infection. Cells were stained for surface expression of CD20 and CD27 and analyzed by flow cytometry to differentiate naïve and memory populations. Data representing the number of CD20⁺ CD27⁻ naïve B cells (a) or CD20⁺ CD27⁺ memory B cells (b) over longitudinal time points are shown. The mean value for all 10 animals in each group is indicated by the dark solid line. (c) Average numbers of naïve (CD20⁺ CD27⁻) and memory (CD20⁺ CD27⁺) B cells, with error bars representing standard errors for all 10 animals. Data during ART are indicated by the shaded area. Asterisks denote a significant difference from a particular time point compared to the day 0 value, using Student's t test (P < 0.05).

TABLE 1.

Changes in CD20⁺ CD27⁻ (naïve) and CD20⁺ CD27⁺ (memory) B-cell subsets compared to baseline following SIV infection

Time pointa	CD20+ CD27−		CD20+ CD27+
	No. of cells (SD)b	% Change from day zeroc	No. of cells (SD)d	% Change from day zeroe
0	401.97 (162.22)	0	750.59 (527.57)	0
3	521.3 (186.71)	29.69	773.85 (518.36)	3.10
7	230.65 (103.28)	−42.62	550.82 (389.29)	−26.62
10	125.83 (62.8)	−68.70	261.31 (200.73)	−65.19
14	260.42 (122.81)	−35.21	250.54 (164.8)	−66.62
21	251.4 (144.48)	−37.46	419.47 (270.85)	−44.11
28	329.61 (139.04)	−18.00	541.23 (364.12)	−27.89
35	242.05 (98.21)	−39.78	496.62 (312.34)	−33.84
42	276.65 (116.89)	−31.18	508.11 (276.16)	−32.31
56	329.37 (117.1)	−18.06	503.26 (267.61)	−32.95
71	304.84 (133.94)	−24.16	417.21 (261.9)	−44.42
84	389.86 (187.21)	−3.01	490.67 (373.33)	−34.63
96	484.08 (240.77)	20.43	595.27 (453.83)	−20.69
105f	412.8 (200.37)	2.69	562.34 (420.98)	−25.08
112	566.82 (249.95)	41.01	753.31 (647.36)	0.36
122	428.09 (171.65)	6.50	938.74 (650.33)	25.07
133	398.33 (160.9)	−0.91	780.75 (419.74)	4.02
149	300.1 (124.35)	−25.34	803.4 (544.29)	7.04

^aNumber of days postinfection with SIVmac239 (10,000 infectious units, intravenous).

^bNumber of CD20⁺ CD27⁻ (naïve) B cells present in peripheral blood at the indicated time point; values represent means of 10 animals.

^cPercent change in CD20⁺ CD27⁻ (naïve) B-cell number at each time point compared to the day zero preinfection time point.

^dNumber of CD20⁺ CD27⁺ (memory) B cells present in peripheral blood at indicated time points; values represent the mean of 10 animals.

^ePercent change in CD20⁺ CD27⁺ (memory) B cell number at each time point compared to the day zero preinfection time point.

^fInitiation of ART (PMPA and FTC).

Incomplete recovery of IgM-secreting (CD20⁺ CD27⁺ IgD⁺) cells.

Memory B cells were further differentiated based upon surface IgD expression (26, 27). In humans, CD20⁺ CD27⁺ IgD⁺ cells have been found to predominantly secrete IgM antibodies (34), while CD20⁺ CD27⁺ IgD⁻ cells have been found to predominantly secrete IgG or IgA antibodies (1). Using these markers in the rhesus system, both IgM-secreting (Fig. 5a) and IgG/IgA-secreting (Fig. 5b) cell populations decreased in total numbers of cells within 3 to 10 days postinfection. Following the initial decrease in cell number, IgG/IgA-secreting cells rapidly increased by day 28, to a mean of 303.8 cells/μl, where they remained steady until day 70 (Fig. 5c and Table 2). After day 70 this population once again increased, returning to preinfection values by day 96 postinfection. In contrast, IgM-secreting cells demonstrated a gradual increase in cell numbers between days 10 and 28 postinfection followed by a gradual decline after day 40 (Fig. 5c). Interestingly, IgM-secreting cells failed to rebound to preinfection values until the initiation of ART.

FIG. 5.

Longitudinal analysis of IgG/IgA-secreting (CD20⁺ CD27⁺ IgD⁻) and IgM-secreting (CD20⁺ CD27⁺ IgD⁺) cells. Peripheral blood from rhesus macaques infected with SIVmac239 was obtained at the indicated time points after acute and early chronic infection. Cells were stained with the surface markers CD20, CD27, and IgD and analyzed by flow cytometry to differentiate IgG/IgA-secreting and IgM-secreting cells. (a and b) Data representing the number of CD20⁺ CD27⁺ IgD⁻ IgG/IgA-secreting B cells (a) or CD20⁺ CD27⁺ IgD⁺ IgM-secreting B cells (b) over longitudinal time points. The mean value for all 10 animals in each group is indicated by the dark solid line. (c) Average numbers of IgG/IgA-secreting (CD20⁺ CD27⁺ IgD⁻) and IgM-secreting (CD20⁺ CD27⁺ IgD⁺) B cells, with error bars representing standard errors for all 10 animals. Data during ART are indicated by the shaded area. Asterisks denote a significant difference from the day zero value, based on Student's t test (P < 0.05).

TABLE 2.

Changes in CD20⁺ CD27⁺ IgD⁺ (IgM-secreting) and CD20⁺ CD27⁺ IgD⁻ (IgG/IgA-secreting) B-cell subsets compared to baseline following SIV infection

Time pointa	CD20+ CD27+ IgD+		CD20+ CD27+ IgD−
	No. of cells (SD)b	% Change from day 0c	No. of cells (SD)d	% Change from day 0e
0	367.05 (222.68)	0	381.34 (370.76)	0
3	387.86 (233.12)	5.67	382.31 (322.4)	0.25
7	239.84 (158.4)	−34.66	308.76 (267.68)	−19.03
10	96.61 (75.95)	−73.68	162.9 (137.29)	−57.28
14	118.51 (74.93)	−67.71	130.77 (93.63)	−65.71
21	173.45 (125.59)	−52.74	244.38 (154.81)	−35.92
28	237.28 (208.05)	−35.35	301.11 (175.27)	−21.04
35	247.08 (188.36)	−32.68	248.03 (145.52)	−34.96
42	258.25 (150.82)	−29.64	246.97 (133.81)	−35.24
56	225.98 (151.65)	−38.43	273.68 (127.14)	−28.23
71	173.37 (120.39)	−52.77	242.19 (156.1)	−36.49
84	173.03 (141.83)	−52.86	316.11 (247.01)	−17.11
96	223.57 (175.13)	−39.09	368.66 (291.33)	−3.33
105f	190.14 (159.56)	−48.20	370.48 (294.29)	−2.85
112	262.29 (189.73)	−28.54	484.12 (474.99)	26.95
122	374.26 (261.66)	1.96	562.67 (392.73)	47.55
133	326.22 (169.03)	−11.12	450.53 (258.54)	18.14
149	271.89 (171.05)	−25.93	530.46 (394.17)	39.10

^aNumber of days postinfection with SIVmac239 (10,000 infectious units, intravenously).

^bNumber of CD20⁺ CD27⁺ IgD⁺ (IgM-secreting) B cells present in peripheral blood at the indicated time point; values represent means of 10 animals.

^cPercent change in CD20⁺ CD27⁺ IgD⁺ (IgM-secreting) B cells at each time point compared to the day zero preinfection time point.

^dNumber of CD20⁺ CD27⁺ IgD⁻ (IgG/IgA-secreting) B cells present in peripheral blood at the indicated time point; values represent means of 10 animals.

^ePercent change in CD20⁺ CD27⁺ IgD⁻ (IgG/IgA-secreting) B cells at each time point compared to the day zero preinfection time point.

^fInitiation of ART (PMPA and FTC).

Dramatic and early increases in CD95 expression in all B-cell subsets.

To assess whether differences in B-cell recovery between memory subsets were due to increased cell activation, surface expression of CD95 was analyzed. All B-cell populations demonstrated increased CD95 surface expression following infection, despite differences in the basal level of CD95 expression. The magnitudes of the increases were also different between B-cell subsets (Fig. 6). For example, the lowest level of basal surface expression of CD95 (1.64%) was observed in the naïve B-cell population, and a significant increase in the percentage of these cells expressing CD95 (3.62%) (P < 0.05) was observed by day 7 postinfection (Fig. 6, closed circles). The percentage of CD95⁺ naïve B cells remained significantly elevated above baseline from day 7 through the entire study period. IgM-secreting cells expressed moderate basal levels of CD95 (17.9%) with a significant increase in the percentage of CD95-expressing cells to 27.0% by day 7 postinfection (P < 0.05). The percentage of IgM-secreting cells expressing CD95 remained significantly higher than preinfection levels for the duration of the study period, with the maximum percentage observed being 44.8% (Fig. 6, open triangles). Finally, 66.1% of IgG/IgA-secreting cells expressed CD95 preinfection (Fig. 6, closed squares), with a significant increase in expression (P < 0.05) by day 3 postinfection. As seen in the other populations studied, the increased expression remained significantly elevated for the duration of the study, with a maximum of 86.5% of IgG/IgA-secreting cells expressing surface CD95.

FIG. 6.

Expression of CD95 in B-cell subsets following SIVmac239 infection. The average percentages of IgG/IgA-secreting (closed squares), IgM-secreting (open triangles), and naïve (closed circles) B cells expressing CD95 were followed in longitudinal peripheral blood samples from 10 SIVmac239-infected rhesus macaques. Lines represent the average percentages of cells expressing CD95 for each cell population, with error bars representing standard errors across all 10 animals. Asterisks denote a significant difference from the day zero value based on Student's t test (P < 0.05).

Contrasting proliferative responses to SIV infection between B-cell subsets.

To assess whether decreased proliferation played a role in acute B-cell depletion, B-cell subsets were further analyzed for Ki67 expression. The proliferative responses observed were remarkably different when comparing the naïve, IgM-secreting and IgG/IgA-secreting populations (Fig. 7). The percentage of Ki67⁺ naïve cells slowly decreased for the entire duration of the study from 15.1% preinfection to 4.8% at day 149. ART did not alter the decline in the percentage of Ki67⁺ cells in this naïve B-cell population. In contrast, memory B cells exhibited a spike in the percentage of Ki67⁺ cells at day 21. The IgG/IgA-secreting population demonstrated a significant increase from 16.6% preinfection to 40.2% at day 21 (P < 0.005), followed by a moderate decline, with the average steady-state level of Ki67⁺ cells being maintained at a higher level than preinfection until the initiation of ART. A significant increase in the percentage of Ki67⁺ IgM-secreting cells was observed between preinfection and day 21 (19.5% to 32.2%, respectively; P < 0.05). This increase was followed by a significant decline, with the percentage of cells being maintained at a lower level than preinfection levels until initiation of ART.

FIG. 7.

Intracellular expression of Ki67 following infection with SIVmac239. Peripheral blood from rhesus macaques infected with SIVmac239 was obtained at the indicated time points during acute and early chronic infection. Lymphocytes from whole blood analysis were analyzed for intracellular Ki67 expression. Lines represent the average percentages of Ki67⁺ cells within each cell population present across all 10 animals. Asterisks denote a significant difference from the day zero time point based on Student's t test (P < 0.05).

Differential effects of ART treatment on B-cell subsets.

Treatment with PMPA (50 mg/kg) and FTC (30 mg/kg) was initiated at day 105 postinfection in all animals. The dosage of PMPA was reduced at day 134 to 20 mg/kg, and at the same time the dosage for FTC was reduced to 20 mg/kg. Following initiation of ART, all animals demonstrated an initial drop in viral setpoint, with the noncontrollers and intermediate controllers rebounding within an average of 15 to 25 days after treatment, respectively. In contrast, the three animals demonstrating the lowest viral setpoint had the most dramatic drop following initiation of ART, with two of the three animals reaching undetectable levels of virus within 10 days after treatment and the third animal's levels becoming undetectable within 45 days of treatment initiation (Fig. 1). Antiretroviral therapy resulted in a slight increase in the number of total B cells (Fig. 2). However, changes within B-cell subsets were more dramatic. The naïve B-cell population demonstrated an initial, moderate increase in cell numbers following the initiation of ART (Fig. 4), from 408 cells/μl at day 105 to 562.6 cells/μl at day 112. The increase in naïve B-cell numbers at the initiation of ART was then followed by a steady decline to 298 cells/μl by day 149. Memory B cells also demonstrated pronounced increases in cell numbers (Fig. 4), from 571 cells/μl at day 105 to 954.3 cells/μl at day 122. In contrast to the naïve B-cell population, the number of memory cells remained higher than pretreatment numbers out to day 149. Interestingly, ART did not elicit changes in the surface expression of CD95 on any of the B-cell populations studied (Fig. 6). While ART had no effect on the Ki67 expression in the naïve B-cell population, it did result in a transient increase in the percentage of IgM-secreting B cells expressing Ki67 and a sustained decrease in the percentage of IgG/IgA-secreting cells expressing Ki67 (Fig. 7).

Antibody production during early SIV infection.

To determine how analysis of B-cell subsets correlated with the production of SIV-specific antibody production, SIV envelope-specific antibody endpoint titers were measured in longitudinal serum samples in a concanavalin A ELISA. All animals with the exception of one rapid progressor demonstrated high-titer SIV-specific antibody by 4 weeks postinfection, and these titers were sustained for the duration of the study (Fig. 8). Samples from early time points (days 0 to 56) were not available for endpoint titer analysis. However, results were consistent with data previously published from our laboratory demonstrating that antibody titers from historical controls infected with SIV rapidly rise and peak within 4 to 12 weeks postinfection (6, 7).

FIG. 8.

Measurement of SIV-specific antibody endpoint titers by concanavalin A ELISA. Envelope-specific antibody titers were measured to concanavalin A-captured SIVsmB7 envelope proteins in longitudinal samples from rhesus macaques infected with SIVmac239. Endpoint titers are reported as the last 2-fold dilution above the cutoff of the assay. Lines represent the log₁₀ reciprocal endpoint titers from individual animals. Dashed lines are representative of animals with low viral setpoints (10³ to 10⁶), solid lines are representative of animals with intermediate setpoints (10⁶ to 10⁷), and dotted lines are representative of animals with high viral setpoints (10⁷ to 10⁸).

DISCUSSION

This study is, to our knowledge, the first detailed, longitudinal analysis of B-cell subsets during acute and early chronic SIV infection. Novel changes were identified very rapidly following SIV infection within specific B-cell subsets that have not been observed in previous studies when analyzing total B cells. While all B-cell subsets in this study demonstrated a similar and profound depletion in cell number concordant with peak viremia, differences in the timing of recovery to preinfection values were observed among the different populations. The depletion of multiple B-cell subsets from the periphery during acute infection may indicate a compromise in the early B-cell response to SIV. Further, these depletions may play a role in abnormal B-cell maintenance and functionality observed during later stages of infection For example, from these studies it is not clear whether the B cells are depleted due to direct or indirect viral effects or whether cells are trafficking from the periphery to tissues. Additional studies to determine the mechanisms for this depletion are needed.

It is likely that a combination of B-cell depletion and redistribution to lymphoid organs occurs during acute SIV infection. The current study only addressed B-cell populations in the periphery and could not directly demonstrate whether depletions from the periphery were reflected in the spleen or lymph nodes. Prior studies in cynomolgus and rhesus macaques indicated that redistribution of a portion of B cells to the spleen and lymph nodes occurs during acute infection (32, 42). Germinal centers have been shown to be reservoirs for virus and are areas where virus and viral proteins are likely to interact with B cells (42). As such, interactions in lymphoid organs may also lead to B-cell dysfunction, as there is evidence that the presence of proliferating B cells within the germinal centers of lymph nodes decreases as early as day 20 postinfection (42). Thus, the depletion of peripheral B cells from the periphery, either via redistribution to lymphoid organs or cell death, would have a detrimental effect on the B-cell population.

Several B-cell functional abnormalities have been associated with HIV infection, including hypergammaglobulinemia (3, 8, 10, 11, 31, 37, 41), increased basal activation accompanied by diminished reactivity to mitogenic stimulation (11, 14, 26-28, 31), and depletion of the memory B-cell subset (3, 10, 11, 15, 30, 31, 37). However, due to the difficulty in obtaining longitudinal samples from HIV-infected patients, especially during the early stages of infection, studies in HIV-infected patients have predominantly represented a snapshot of the B-cell repertoire at one or very few time points during the chronic phase of infection. Host-virus interactions that occur during the acute phase of HIV infection are known to heavily influence disease progression (36). Thus, the SIV/macaque model provides an effective means for a longitudinal B-cell analysis during acute infection.

The present study identified dynamic changes in the memory B-cell population during acute infection. The number of memory B cells within the periphery significantly dropped within 7 days of infection. The IgG/IgA-secreting B-cell population recovered to a level just below preinfection numbers within 30 days, while the IgM-secreting B-cell population recovered more slowly. The human B-cell population which is analogous to the IgM-secreting B-cell population in rhesus macaques has been shown to be critical for the production of antibodies to newly emerging viral mutants and opportunistic infections and is also very important in the production of T-cell-independent responses to pathogens, including pneumococcus (17-19, 22). Thus, a slow recovery in this B-cell subset has the potential to slow the initial response to SIV/HIV during acute infection and to render the host more susceptible to opportunistic infections later on. Defining the mechanism underlying the loss of this B-cell subset could aid in the design of targeted immune therapy to protect the host from opportunistic infections.

Measurement of total antibody production has been a gold standard by which investigators measure the functionality of the B-cell compartment. However, data from our lab and others have demonstrated that the production of quantitative levels of virus-specific antibodies may not necessarily indicate that the antibody produced is qualitatively effective at limiting virus replication (6, 7, 29, 35). Thus, the functional relevance of the antibody response in controlling infection may still be compromised despite measurement of a robust antibody titer. In a recent study by Scheid et al., IgG clones derived from memory B cells in HIV-infected patients demonstrating broad neutralizing activity were comprised of clonal responses to diverse envelope epitopes (33). In general, high-affinity antibody (binding) did not correlate with neutralization sensitivity (functionality). These data support our current study, where we demonstrated that fewer B cells were present during the initiation of the antibody response. Thus, it is likely that a small number of B-cell clones were responsible for the antibody response and that the breadth and potency of the response were limited. Additionally, the paucity of memory cells during acute infection when the initial antibody response is formed could have led to diminished efficacies of the antibodies produced. With a reduced memory population, antibody responses have to be formed de novo, effectively resulting in each successive SIV clone appearing as a new antigen to the antibody-mediated immune response. These data demonstrate that measurement of antibody titer alone is an insufficient means for assessment of B-cell activity and needs to be combined with in-depth analysis of the B-cell compartment.

The depletion of naïve B cells from the periphery during acute SIV infection is also likely to play a significant role in the slow development of the B-cell response. Naïve cells dictate the breadth and efficacy of the antibody response, requiring a multitude of signals, including B-cell receptor activation, CD40 stimulation, and cytokine signals to initiate activation, maturation, and proliferation (25). Thus, virally induced disruption of any of these signals could lead to decreased potency of the antibody response. The early depletion of naïve B cells renders the infected host more vulnerable to SIV during the initial, critical host-pathogen interaction. Further, the failure of the naïve population to rebound with ART suggests a limited ability for the host to recognize viral variants or new pathogens, resulting in a diminished B-cell response to both SIV and other opportunistic pathogens. Early antiviral treatment or therapeutic vaccination focused on preventing the loss of naïve B cells during acute infection would render this critical cell population more effective during the chronic stage of infection.

Finally, in addition to rapid alterations in population dynamics, increased activation was observed in all B-cell subsets. Although prior studies have demonstrated increased CD95 expression on B-cell populations during chronic infections (27, 38), the precise timing of this increase was unknown. Our data demonstrated that the increase in B cells expressing CD95 occurred almost immediately following SIV infection, was maintained through the acute phase of infection, and was unaffected by ART. This increased B-cell activation is of importance, as activation has also been implicated as a potential mechanism for altered B-cell activity, i.e., poor responses to B-cell-mediated vaccines (9). Therapeutic strategies to inhibit chronic activation, either via CD95 or other pathways, can be further explored using the SIV nonhuman primate model. In contrast to activation measured by CD95 expression, proliferative responses assessed using Ki67 were variable among B-cell subsets, indicating the potential for differential regulation between the memory and naïve B-cell subsets. Alterations in proliferative capacity may be related to how these specific subsets are able to respond to antigenic stimulation, and they warrant further functional studies.

Data from the present acute study and others have demonstrated that significant changes in total B-cell numbers occur in the periphery following acute SIV infection. The current study provides novel information about alterations in specific B-cell subsets during acute SIV infection that were not revealed when analyzing the total B-cell population. Furthermore, the data presented in this study clearly demonstrate that measurement of virus-specific antibody titer is not reflective of alterations in the subset composition in the B-cell compartment. It is important to monitor antibody specificity and functionality during SIV infection to fully understand the extent of the damage to the immune system. Further studies are warranted to identify potential mechanisms for these phenotypic alterations and to test whether concurrent functional changes in B-cell subsets occur during the same time frame.