Our study demonstrates that binding of complement-opsonized HIV to complement receptor type 1 on human erythrocytes (E) via C3b fragments is followed by a rapid normal human serum-mediated detachment of HIV from E. The release was dependent on the presence of factor I indicating a conversion of C3b fragments to iC3b and C3d on the viral surface. This in turn resulted in an efficient binding of opsonized HIV to CR2-expressing B cells, thus facilitating B cell-mediated transmission of HIV to T cells. These data provide a new dynamic view of complement opsonization of HIV, suggesting that association of virus with E might be a transient phenomenon and the factor I-mediated processing of C3b to iC3b and C3d on HIV targets the virus to complement receptor type 2-expressing cells. Thus, factor I in concert with CR1 on E and factor H in serum due to their cofactor activity are likely to be important contributors for the generation of C3d-opsonized infectious HIV reservoirs on follicular dendritic cells and/or B cells in HIV-infected individuals.

Human immunodeficiency virus infection results in the activation of the complement system even in the absence of HIV-specific Abs (1), which results in deposition of C3 fragments on the viral surface both in vitro (2, 3) and in vivo (4). HIV bound extracellularly to the follicular dendritic cells (FDC)3 in germinal centers of lymph nodes represent by far the largest viral reservoir in HIV-infected individuals (5, 6). The binding of this infectious pool of HIV in the germinal centers depends mainly on interactions of complement receptor type (CR) 2 expressed on FDC (or B cells) with C3d fragments on the viral surface (4, 7). In addition, an association of complement-opsonized HIV with peripheral B cells through CR2-C3d interactions was described in HIV-infected individuals (8). CR2-C3d interactions between B cells and HIV were demonstrated to be critical for efficient B cell-mediated transmission of complement-opsonized HIV to T cells (9, 10).

The presence of C3d fragments on the surface of HIV bound to FDC in the germinal centers and to peripheral B cells in vivo suggests a sequential processing of C3 fragments, from C3b to iC3b and C3d, on the surface of HIV. As a first step for the generation of C3d-coated HIV particles, C3b fragments have to be generated on the surface of HIV. Binding of such C3b-opsonized immune complexes (IC) to CR1 (CD35) on erythrocytes (E), a process referred to as immune adherence, has an important role in eliminating IC by transporting them to the liver and spleen, where they are cleared by phagocytes. Formation of HIV containing IC (HIV-Ig) in vitro by incubation of HIV in the presence of envelope-specific Abs and normal human serum (NHS) facilitates the binding of complement-opsonized HIV-Ig to the surface of E through CR1-C3b interactions (11, 12). In addition, in the absence of HIV-specific Abs, complement opsonization of HIV was able to target HIV to E (11, 12). More importantly, an infectious pool of HIV associated with E was demonstrated in vivo in peripheral blood of HIV-infected individuals even with undetectable plasma viral load (13). Although both C3b- and C3d-coated HIV particles were detected in HIV-infected individuals, the exact mechanism by which C3b fragments on HIV turn into iC3b and C3d generating an infectious HIV reservoir on FDC remained unknown.

Complement-opsonized bacteria, as Streptococcus pneumoniae and Staphylococcus aureus bound to E through CR1-C3b interactions, was detached from E in the presence of factor I (fI), suggesting a degradation of C3b fragments on the bacterial surface into iC3b and C3d, which have lower affinity for CR1 (14, 15, 16). The absence of C3d fragments in sera of fI-deficient patients underlines the importance of fI in this process (17). These data suggest that the association of C3b-opsonized HIV with human E might only be a temporary phenomenon due to the fI-mediated generation of iC3b and C3d fragments on the viral surface. The cofactor activity of CR1 on E for fI has been proposed to contribute to the processing of C3b fragments on HIV-Ig (18).

In this study, we demonstrate that binding of opsonized HIV to human E is followed by a fI-dependent NHS-mediated release of HIV from E suggesting a conversion of C3b fragments to iC3b and C3d on the viral surface. As a consequence of the generation of C3d fragments, released HIV binds efficiently to CR2-expressing cells, which facilitates the B cell-mediated transmission of opsonized HIV to T cells. This fI-mediated sequential processing of C3 fragments from C3b to C3d on the surface of HIV is likely to have in vivo relevance since C3d-coated viruses are predominantly found on FDC in germinal centers, which represent by far the largest viral reservoir in HIV-infected individuals (4, 5, 7).

Sera from at least 10 healthy volunteers, referred to as NHS, were pooled and stored at −70°C in small aliquots until use. Some serum aliquots were heat-inactivated (hiNHS) by incubation of the serum for 30 min at 56°C. For some experiments, sera were depleted from fI (referred to as ΔfI) by incubation with anti-fI beads. Depletion was confirmed by dot plots (data not shown). Sera of chronically HIV-1-infected participants of a prospective clinical trial (ARCHY), naive to antiretroviral therapy, were involved in this study. All patients provided written informed consent before entering the clinical trial.

HIV-specific Abs for opsonization were purified from a serum-pool of at least 10 HIV-positive individuals. In virus capture assays, mAbs against human IgG (BD Pharmingen) and C3d-recognizing C3b, iC3b, and C3d fragments (Quidel) or isotype controls (BD Pharmingen) were used.

Protease inhibitor mixture, 4-(2-aminoethyl)-benzenesulfonyl fluoride (AEBSF), bestatin, phenantrolin, E-64, leupeptin, and aprotinin were purchased from Sigma-Aldrich and used in a concentration as recommended by the manufacturer.

Virus stocks of HIV-1 strain BaL, primary isolates 92BR030 (B subtype, R5-tropism), and 93BR020 (BF subtype, X4R5-tropism) were grown in IL-2 and PHA-prestimulated PBMC. Virus was opsonized with NHS in a final concentration of 1/10 in the absence (HIV-C) or presence of HIV-specific Abs (HIV-Ig-C) for 60 min at 37°C. Opsonization was also done in the presence of fI-depleted NHS (referred as HIV-Ig-C(ΔfI)). As nonopsonized control (designated as HIV), virus was incubated with heat-inactivated NHS (hiNHS). To remove NHS, 5 volumes of RPMI 1640 medium were added after the opsonization, virus preparations were ultracentrifuged, and the virus pellet was resuspended in RPMI 1640 without any supplement. Virus preparations were aliquoted and stored at −70°C until use. The results in this article represent experiments with the primary isolate 92BR030, if not indicated otherwise. Obtained results were confirmed using HIV-BaL and the primary isolate 93BR020. All experiments were repeated at least three times. The presence of complement fragments and Abs on the surface of HIV was determined by a virus capture assay as described elsewhere (19).

Human erythrocytes were isolated from peripheral blood of healthy individuals. Briefly, peripheral blood was diluted 1/5 in PBS and centrifuged with 1400 rpm for 10 min, then plasma and buffy coat were removed. The erythrocyte fraction was then washed three times with RPMI 1640, each time removing the upper layer of the cells with the supernatant. E suspensions were counted and adjusted to 1 × 109 E/ml in RPMI 1640. The final contamination of E with other cells was reduced to ≤0.01% as determined by FACS.

In brief, 1 × 108 E in 100 μl of RPMI 1640 were incubated with HIV, HIV-C, and HIV-Ig-C (or HIV-Ig-C(ΔfI)) at a concentration of 5 ng of HIV p24/ml for different periods of time (2, 15, 30, 60, 120, and 180 min) at 37°C in a shaker. After the incubation, E were washed two times with 1 ml of ice-cold RPMI 1640 to remove nonbound virus, and the pellet of E was lysed with 100 μl of 1% Igepal in RPMI 1640. The amount of HIV bound to E was determined by HIV p24 ELISA. HIV-binding assays were also done in the presence of hiNHS (or hiNHS(ΔfI)) in a final concentration of 1/10 or with different concentrations of AEBSF.

To investigate the release of HIV from E, 1 × 108 E in 100 μl of RPMI 1640 were loaded with 5 ng of p24/ml HIV and HIV-Ig-C (or HIV-Ig-C(ΔfI)) for 30 min at 37°C. After the incubation, E were washed two times with ice-cold RPMI 1640. The pellet of E was resuspended in 100 μl of RPMI 1640 and incubated for different periods of time (0, 15, 30, 60, 120, and 180 min) at 37°C. After the incubation, E were centrifuged and the pellet of E was lysed by the addition of 100 μl of 1% of Igepal in RPMI 1640. The amount of HIV in the pellet was determined by HIV p24 ELISA. The HIV release assay was also performed in the presence of hiNHS (or hiNHS(ΔfI)) in a final concentration of 1/10 or with AEBSF.

To determine whether fI is associated with HIV after opsonization, HIV, HIV-C, and HIV-Ig-C were centrifuged overnight through a sucrose gradient (500 μl of 10, 15, 20, 25, 30, 40, 50, 60, 70, 80, and 90% sucrose) with 27,000 rpm at 4°C. After centrifugation, 500-μl fractions were collected and lysed with ice-cold 1% Triton X-100 containing protease inhibitor mixture (Sigma-Aldrich) for 30 min at 4°C. Alternatively, 25 μl of each sample without lysis was incubated with E as described above. To determine HIV-containing fractions, slot blot analysis was performed with 100-μl lysates of each fraction using biotinylated Ab against HIV p24 (clone 37G12; provided by Dr. Kattinger BOKU, Vienna, Austria). Binding of biotinylated 37G12 Ab was visualized by HRP-conjugated streptavidin and ECL (Pierce Biotechnologies). To detect fI in the fractions, 100-μl lysates were separated on 10% SDS-PAGE under nonreducing conditions and transferred to a nitrocellulose membrane. Membranes were incubated with mouse anti-human fI mAb (Serotec). Binding of the primary Ab was visualized by HRP-conjugated goat anti-mouse IgG Ab and ECL.

To investigate whether E are involved in the B cell-mediated HIV transmission to T cells, HIV and HIV-Ig-C at a concentration of 1, 5, or 10 ng/ml p24 were preincubated with or without 5 × 107 E in the presence or absence of hiNHS (1/10) for 60 min at 37°C. After the preincubation, 2 × 105 T cells and B cells isolated from tonsillar tissue of healthy individuals were added and plates were incubated for 3 h at 37°C. Input virus was removed by washing the plates three times with RPMI 1640 and the cells were cultivated in RPMI 1640 supplemented with 10% FCS. Culture supernatants were collected after 2, 4, 7, 11, and 14 days after infection, and the amount of HIV in the supernatants was determined by HIV p24 ELISA.

Differences between the values were analyzed using GraphPad Prism software. A value of p < 0.05 in the unpaired Student’s t test was scored as significant (p < 0.05, <0.01, and <0.001 are symbolized on the figures with ∗,∗∗, and ∗∗∗, respectively).

Compared with nonopsonized HIV, a deposition of C3 fragments on the viral surface can be detected after incubation of HIV with NHS (Fig. 1,A, HIV-C). The presence of HIV-specific Abs during the opsonization enhanced the deposition of C3 fragments on the viral surface (Fig. 1,A, HIV-Ig-C). Subsequently, the binding of HIV, HIV-C, and HIV-Ig-C to human E was investigated. During the 30-min incubation, human E bound a significantly higher amount of HIV-C compared with nonopsonized HIV (Fig. 1,B, w/o (without), 0 min). As expected, the enhanced complement activation and C3 deposition on HIV in the presence of HIV-specific Abs resulted in increased binding of HIV-Ig-C compared with HIV-C (Fig. 1,B, ▪). Both HIV-C and HIV-Ig-C were released from E upon an additional 60-min incubation of HIV-loaded E in medium alone (Fig. 1,B, w/o), which is referred to as “spontaneous” release. Incubation of both HIV-C- and HIV-Ig-C-loaded E for 60 min in the presence of hiNHS (as source of fI) resulted in a strong reduction of bound HIV on the E surface (Fig. 1,B, hiNHS). Relative to the bound HIV-Ig-C after a 30-min incubation, the average amount of virus remaining on E from eight individuals after 60 min was 50.4 ± 7.7% (mean ± SEM) in medium and 26.6 ± 5.6% (mean ± SEM) in the presence of hiNHS (Fig. 1 C, w/o and hiNHS, 60 min).

FIGURE 1.

Complement opsonization mediates the binding of HIV and HIV-Ig to human E. A, As shown by virus capture assay, opsonization of HIV with NHS resulted in significant deposition of C3 fragments on the viral surface (HIV-C), which was further enhanced in the presence of HIV-specific Abs (HIV-Ig-C). Data represent a typical result of a virus capture assay from opsonized HIV performed in duplicates. B, E bound complement-opsonized HIV (HIV-C) compared with nonopsonized virus after 30 min of incubation (w/o, 0 min). The presence of HIV-specific Abs during the opsonization resulted in an enhanced binding of HIV-Ig-C to E (▪, w/o, 0 min). After several washing steps, incubation of HIV-loaded E for an additional 60 min in medium without any supplement resulted in a spontaneous detachment of both HIV-C and HIV-Ig-C from human E (w/o, 60 min), which was further enhanced in the presence of hiNHS (hiNHS, 60 min). C, Data from eight independent experiments performed in duplicate represent the binding of HIV-Ig-C to E and its release in the presence or absence of hiNHS after 60 min of incubation.

FIGURE 1.

Complement opsonization mediates the binding of HIV and HIV-Ig to human E. A, As shown by virus capture assay, opsonization of HIV with NHS resulted in significant deposition of C3 fragments on the viral surface (HIV-C), which was further enhanced in the presence of HIV-specific Abs (HIV-Ig-C). Data represent a typical result of a virus capture assay from opsonized HIV performed in duplicates. B, E bound complement-opsonized HIV (HIV-C) compared with nonopsonized virus after 30 min of incubation (w/o, 0 min). The presence of HIV-specific Abs during the opsonization resulted in an enhanced binding of HIV-Ig-C to E (▪, w/o, 0 min). After several washing steps, incubation of HIV-loaded E for an additional 60 min in medium without any supplement resulted in a spontaneous detachment of both HIV-C and HIV-Ig-C from human E (w/o, 60 min), which was further enhanced in the presence of hiNHS (hiNHS, 60 min). C, Data from eight independent experiments performed in duplicate represent the binding of HIV-Ig-C to E and its release in the presence or absence of hiNHS after 60 min of incubation.

Close modal

To investigate the reason for the spontaneous release of HIV-Ig-C from human E, release experiments were performed in the presence of different protease inhibitors. A protease inhibitor mixture was found to inhibit the release of HIV-Ig-C from E up to 90% (Fig. 2,A). The one component of the protease inhibitor mixture responsible for the release reduction was a serine protease inhibitor, AEBSF (Fig. 2,A). The inhibition of spontaneous release of HIV from human E by AEBSF was dose dependent (Fig. 2 B).

FIGURE 2.

A serine protease inhibitor, AEBSF, inhibits spontaneous release of HIV-Ig-C from human E. A, The spontaneous release of HIV-Ig-C from human E was inhibited by applying a protease inhibitor mixture (π-mixture). The component identified for the reduced release was AEBSF, a serine protease inhibitor. B, The inhibition of the HIV release by AEBSF was dose dependent. Data are representative of three independent experiments.

FIGURE 2.

A serine protease inhibitor, AEBSF, inhibits spontaneous release of HIV-Ig-C from human E. A, The spontaneous release of HIV-Ig-C from human E was inhibited by applying a protease inhibitor mixture (π-mixture). The component identified for the reduced release was AEBSF, a serine protease inhibitor. B, The inhibition of the HIV release by AEBSF was dose dependent. Data are representative of three independent experiments.

Close modal

Studying the kinetics of the binding of HIV-Ig-C to E, we found that HIV-Ig-C bound rapidly to E, reaching the peak binding level after 15 min of incubation (Fig. 3,A, w/o). Reduction of the amount of bound virus after 15 min was observed although a large amount (5 ng of p24/ml) of HIV was present in the supernatant. Relative to the peak binding of HIV on E, the amount of virus remaining on E after 180 min in medium varied between 23 and 71%, with a mean of 44.5 ± 18.3% from seven individuals (data not shown). Applying AEBSF during the incubation period inhibited this spontaneous loss of HIV-Ig-C from E without affecting maximum binding (Fig. 3,A, AEBSF). Binding of HIV-Ig-C to E in the presence of hiNHS reached the peak level, which was always lower than the maximum binding of HIV kept in medium alone, already after 2 min and was followed by a rapid reduction of bound HIV on E (Fig. 3 A, hiNHS). After 60 min, only a background binding could be observed.

FIGURE 3.

Kinetics of binding and release of HIV-Ig-C to and from human E. A, Incubation of HIV-Ig-C with human E in medium alone resulted in a rapid association of the virus with human E followed by detachment of HIV from E. The binding of HIV-Ig-C in the presence of hiNHS reached its peak already after 2 min, followed by a rapid reduction of the amount of bound HIV on E (hiNHS). AEBSF inhibited the spontaneous detachment of HIV-Ig-C. B, The spontaneous release of HIV-Ig-C in medium alone (w/o) was blocked by AEBSF. The presence of hiNHS led to a rapid detachment of HIV-Ig-C from human E. Data represent a typical result (performed in duplicates) from seven independent experiments.

FIGURE 3.

Kinetics of binding and release of HIV-Ig-C to and from human E. A, Incubation of HIV-Ig-C with human E in medium alone resulted in a rapid association of the virus with human E followed by detachment of HIV from E. The binding of HIV-Ig-C in the presence of hiNHS reached its peak already after 2 min, followed by a rapid reduction of the amount of bound HIV on E (hiNHS). AEBSF inhibited the spontaneous detachment of HIV-Ig-C. B, The spontaneous release of HIV-Ig-C in medium alone (w/o) was blocked by AEBSF. The presence of hiNHS led to a rapid detachment of HIV-Ig-C from human E. Data represent a typical result (performed in duplicates) from seven independent experiments.

Close modal

To further characterize the release of HIV from E, cells were loaded with HIV-Ig-C in medium alone for 30 min. After the washing steps, the amount of HIV bound to the pellet of E was measured and referred to as 100% bound virus. HIV-loaded E were then incubated either in medium, in the presence of AEBSF or hiNHS and the amount of E-associated HIV was determined in a time-dependent manner. In medium alone, HIV dissociated relatively slowly from E (Fig. 3,B, w/o). After 60 min of incubation, ∼50% of HIV remained associated to E. The presence of AEBSF during the incubation abrogated the spontaneous release of HIV from E (Fig. 3,B, AEBSF). HIV-Ig-C dissociated quickly from E in the presence of hiNHS (Fig. 3 B, hiNHS). After 60 min of incubation with hiNHS, nearly all E-associated HIV was released into the supernatant. We analyzed whether the release of opsonized HIV from E in medium or in hiNHS was associated with loss of CR1 from the surface of E and therefore with reduced binding capacity of E. Therefore, HIV-Ig-C-loaded E and E without HIV were incubated for 180 min in the presence or absence of hiNHS and after washing they were again incubated with HIV-Ig-C. The similar binding of newly added HIV-Ig-C to all E samples suggested that E did not lose their capacity to bind HIV-Ig-C during the HIV release (data not shown). FACS analysis of CR1 expression on E confirmed these data, since the amount of CR1 on the surface of E was not altered when compared before and after incubation of E with HIV (data not shown).

To investigate the role of fI in HIV release from E, HIV was opsonized either with NHS (HIV-Ig-C) or fI-depleted NHS (HIV-Ig-C(ΔfI)) in the presence of HIV-specific Abs. E were incubated with HIV-Ig-C or HIV-Ig-C(ΔfI) for 30 min, and after the washing steps the amount of E-bound HIV (∼450 pg/ml HIV p24 for HIV-Ig-C and 400 pg/ml HIV p24 for HIV-Ig-C(ΔfI), respectively) was set as 100% of bound virus (Fig. 4, A and B, w/o 0 min). HIV-loaded E were then incubated in medium in the presence of hiNHS as the source of fI or in fI-depleted hiNHS. After 60 min of incubation, E were washed and the amount of E-associated HIV was measured from the pellet of E. Spontaneous release of HIV-Ig-C from E reached ∼50% of initial viral load after 60 min (Fig. 4,A, w/o 60 min), whereas the amount of E-associated HIV-Ig-C(ΔfI) was reduced only up to 20% (Fig. 4,B, w/o 60 min), similar to results obtained in the presence of AEBSF (data not shown). The presence of hiNHS during the incubation of HIV-loaded E reduced the amount of both HIV-Ig-C and HIV-Ig-C(ΔfI) on E considerably (Fig. 4, A and B, hiNHS 60 min). However, if E loaded with HIV-Ig-C or HIV-Ig-C(ΔfI) were incubated in the presence of fI-depleted hiNHS, the release of HIV was reduced to the level of the spontaneous release in both cases (Fig. 4, A and B, hiNHS(ΔfI) 60 min).

FIGURE 4.

Both spontaneous and hiNHS-mediated detachment of HIV-Ig-C from human E was dependent on the presence of fI. HIV (BaL) was opsonized in the presence of HIV-specific Abs with NHS (HIV-Ig-C) or fI-depleted NHS (HIV-Ig-(ΔfI)C). E loaded with (A) HIV-Ig-C or (B) HIV-Ig-(ΔfI)C (w/o, 0 min) were incubated in medium, hiNHS, or hiNHS(ΔfI) for an additional 60 min. Spontaneous release (w/o) of HIV-Ig-C was decreased if HIV-Ig was opsonized with fI-depleted NHS (HIV-Ig-(ΔfI)C). NHS-mediated detachment of HIV-Ig-C and HIV-Ig-(ΔfI)C from E was abolished in the presence of fI-depleted hiNHS (hiNHS(ΔfI)). Data represent mean values from two independent experiments performed in duplicate. (C) NHS-mediated detachment of HIV-Ig-C from E was significantly more efficient in the presence of hiNHS from HIV-positive individuals after 10 min of incubation, compared with sera of healthy individuals.

FIGURE 4.

Both spontaneous and hiNHS-mediated detachment of HIV-Ig-C from human E was dependent on the presence of fI. HIV (BaL) was opsonized in the presence of HIV-specific Abs with NHS (HIV-Ig-C) or fI-depleted NHS (HIV-Ig-(ΔfI)C). E loaded with (A) HIV-Ig-C or (B) HIV-Ig-(ΔfI)C (w/o, 0 min) were incubated in medium, hiNHS, or hiNHS(ΔfI) for an additional 60 min. Spontaneous release (w/o) of HIV-Ig-C was decreased if HIV-Ig was opsonized with fI-depleted NHS (HIV-Ig-(ΔfI)C). NHS-mediated detachment of HIV-Ig-C and HIV-Ig-(ΔfI)C from E was abolished in the presence of fI-depleted hiNHS (hiNHS(ΔfI)). Data represent mean values from two independent experiments performed in duplicate. (C) NHS-mediated detachment of HIV-Ig-C from E was significantly more efficient in the presence of hiNHS from HIV-positive individuals after 10 min of incubation, compared with sera of healthy individuals.

Close modal

Interestingly, hiNHS from HIV-positive patients detached more opsonized HIV from E immediately after 10 min of incubation, compared with sera of healthy individuals (Fig. 4 C).

To test whether the spontaneous release is mediated by a direct association of fI to the virus, HIV, HIV-C, and HIV-Ig-C were ultracentrifuged overnight through a sucrose gradient to remove soluble fI contamination. Virus-containing fractions were determined by slot blot analysis revealing that HIV was presented in fractions from 5 to 8, corresponding to ∼30–50% sucrose (Fig. 5,A, slot blot). Western blot analysis for fI did not show any detectable amount of fI in any of the fractions (data not shown). However, the detection limit was ∼3 ng of fI as determined by serial dilutions with NHS (data not shown). Incubation of 25 μl from each fraction of HIV-Ig-C with E for 30 min at 37°C resulted in a binding of HIV to E from the virus-containing fractions (Fig. 5,A, ▪). After the washing steps, further incubation of HIV-loaded E for 60 min did not lead to any spontaneous release of HIV-Ig-C from E (Fig. 5,A, ▴). The absence of spontaneous release from HIV-containing fractions after sucrose gradient centrifugation allowed us to investigate the effect of fI on release of opsonized HIV from E depending exclusively on hiNHS. For this, E were incubated with 25 μl of HIV-Ig-C-containing fraction 6 for 30 min at 37°C. After the washing steps, HIV-loaded E were incubated with different concentrations of hiNHS resulting in a concentration-dependent release of HIV from E (Fig. 5,B). After 60 min of incubation with 1/10 dilution of hiNHS, ∼50% of HIV remained associated to E (Fig. 5 B).

FIGURE 5.

Effect of sucrose gradient centrifugation of HIV-Ig-C on the release of HIV from E. A, E bind HIV-Ig-C from virus-containing fractions (slot blot) after sucrose gradient centrifugation. The spontaneous release was abolished after gradient centrifugation. B, Detachment of HIV-Ig-C from E loaded with HIV derived from fraction 6 by incubation with different dilutions of hiNHS.

FIGURE 5.

Effect of sucrose gradient centrifugation of HIV-Ig-C on the release of HIV from E. A, E bind HIV-Ig-C from virus-containing fractions (slot blot) after sucrose gradient centrifugation. The spontaneous release was abolished after gradient centrifugation. B, Detachment of HIV-Ig-C from E loaded with HIV derived from fraction 6 by incubation with different dilutions of hiNHS.

Close modal

Both spontaneous and NHS-mediated release of opsonized HIV from E were fI dependent, indicating that C3b fragments were further processed into iC3b and C3d fragments on the surface of HIV-Ig. Thus, the binding of opsonized HIV to CR2-expressing Raji cells was studied. Therefore, E were preincubated with HIV-Ig-C at 5 ng p24/ml in medium alone, in the presence of hiNHS as fI source, or AEBSF for 120 min. Then the supernatant of E was transferred to 2 × 105 Raji cells. After 30 min of incubation, Raji cells were washed and the amount of bound HIV was measured. HIV-Ig-C incubated without E for 120 min (Fig. 6, HIV) bound to Raji cells similar to HIV-Ig-C incubated with E in medium alone or in the presence of AEBSF (Fig. 6, w/o and AEBSF). In contrast, incubation of E with HIV-Ig-C in the presence of hiNHS led to a significant enhancement of binding of HIV-Ig-C to Raji cells (Fig. 6, hiNHS).

FIGURE 6.

Enhanced binding of HIV released from E to Raji cells. Preincubation of HIV-Ig-C in medium alone in the absence (HIV) or presence (w/o) of E resulted in a similar association of HIV to Raji cells. The presence of hiNHS during the preincubation of E with HIV-Ig-C significantly enhanced the binding of HIV-Ig-C to Raji cells. Shown is a representative result of two independent experiments (performed in duplicate).

FIGURE 6.

Enhanced binding of HIV released from E to Raji cells. Preincubation of HIV-Ig-C in medium alone in the absence (HIV) or presence (w/o) of E resulted in a similar association of HIV to Raji cells. The presence of hiNHS during the preincubation of E with HIV-Ig-C significantly enhanced the binding of HIV-Ig-C to Raji cells. Shown is a representative result of two independent experiments (performed in duplicate).

Close modal

Finally, the consequences of fI-mediated processing of C3 fragments on HIV for the B cell-mediated transmission of complement-opsonized HIV to T cells was studied, since C3d-CR2 interactions were demonstrated to be crucial in this process (8, 9, 10). According to previous studies, opsonization of HIV resulted in an enhanced infection of T cells in T cell-B cell coculture compared with nonopsonized HIV (Fig. 7,A, cf HIV-Ig-C/B + T and HIV/B + T). Preincubation of HIV-Ig-C with human E further triggered B cell-mediated transmission of opsonized HIV to T cells (Fig. 7,A, cf HIV-Ig-C + E/B + T). In contrast, we did not observe a productive infection of T cells after preincubation of both HIV and HIV-Ig-C with E in the absence of B cells (Fig. 7,A, cf HIV + E/T and HIV-Ig-C + E/T). To further characterize the effect of processing C3 fragments on HIV, HIV-Ig-C was incubated with or without E in the presence or absence of hiNHS before addition of T and B cells. Preincubation of HIV-Ig-C with hiNHS caused a productive infection of T cells in T cell-B cell cocultures similar to infection with HIV-Ig-C in the absence of hiNHS (Fig. 7,B, HIV-Ig-C/T + B, w/o and hiNHS). In contrast, preincubation of HIV-Ig-C with E resulted in enhanced infection of T cells in coculture experiments, which was further triggered if HIV-Ig-C was incubated with E in the presence of hiNHS before addition of T and B cells (Fig. 7 B, HIV-Ig-C/T + B E and hiNHS + E).

FIGURE 7.

Enhanced B cell-mediated transmission of HIV-Ig-C to T cells in the presence of human E. A, Opsonization of HIV (93BR020 isolate) resulted in an enhanced infection of T cells in T cell-B cell cocultures compared with nonopsonized HIV (HIV-Ig-C/B + T and HIV/B + T). Preincubation of HIV-Ig-C with human E further triggered B cell-mediated transmission of opsonized HIV to T cells (HIV-Ig-C + E/B + T). B, Preincubation of HIV-Ig-C (93BR020 isolate) with hiNHS resulted in a similar productive infection of T cells in T cell-B cell cocultures compared with the infection with HIV-Ig-C in the absence of hiNHS (w/o and hiNHS, ▪). A significant enhancement of T cell infection in T cell-B cell coculture experiments with HIV-Ig-C preincubated with E was observed (E, ▪). This enhancement of infection was further increased if the virus was preincubated with E in the presence of hiNHS (E + hiNHS, ▪). Shown is a representative result of three independent experiments.

FIGURE 7.

Enhanced B cell-mediated transmission of HIV-Ig-C to T cells in the presence of human E. A, Opsonization of HIV (93BR020 isolate) resulted in an enhanced infection of T cells in T cell-B cell cocultures compared with nonopsonized HIV (HIV-Ig-C/B + T and HIV/B + T). Preincubation of HIV-Ig-C with human E further triggered B cell-mediated transmission of opsonized HIV to T cells (HIV-Ig-C + E/B + T). B, Preincubation of HIV-Ig-C (93BR020 isolate) with hiNHS resulted in a similar productive infection of T cells in T cell-B cell cocultures compared with the infection with HIV-Ig-C in the absence of hiNHS (w/o and hiNHS, ▪). A significant enhancement of T cell infection in T cell-B cell coculture experiments with HIV-Ig-C preincubated with E was observed (E, ▪). This enhancement of infection was further increased if the virus was preincubated with E in the presence of hiNHS (E + hiNHS, ▪). Shown is a representative result of three independent experiments.

Close modal

In this study, we provide evidence for a fI-mediated processing of C3b fragments into iC3b and C3d on opsonized HIV, which in turn resulted in an efficient binding of opsonized HIV to CR2-expressing B cells, thus promoting B cell-mediated transmission of HIV to T cells.

According to previous investigations, we showed that in vitro opsonization of HIV with complement targets HIV-C to human E (11, 12, 13). The presence of HIV-specific Abs during the opsonization further triggers complement activation resulting in enhanced association of HIV-Ig-C with E. Under our experimental conditions, we determined a maximum binding capacity of HIV-Ig-C to E at 10–15% of input virus (p24 Ag), which differed from the amount of E-associated HIV in previous reports. Montefiori et al. (11) observed very low binding of opsonized HIV IC to CR1 on CR1-transfected cell lines and E (∼1% of input virus), which is most likely due to differences in the experimental settings. In contrast, Horakova et al. (12) observed up to 80% of input virus associated with E. The inactivation of HIV at 56°C in the presence of 0.02% formaldehyde, which was demonstrated to increase the reactivity of Abs directed against certain viral epitopes, could explain the strikingly high binding of HIV to E (20).

Kinetic experiments studying the binding of HIV-Ig-C to E revealed a rapid association of opsonized HIV with E within a few minutes, reaching the peak level after 1–5 min in the presence of hiNHS. Longer incubation of opsonized HIV-Ig in the presence of hiNHS led to a fI-dependent dissociation of virus from E. A similar time-dependent optimum binding of opsonized S. pneumoniae followed by a detachment from human E was demonstrated in the presence of NHS in a fI-dependent manner (14). Interestingly, we observed a dissociation of HIV-Ig-C from E in the absence of hiNHS, which could be inhibited by AEBSF up to 90% and is referred to as spontaneous release in this study. Of note, we observed a detachment of HIV-Ig-C from E from 10 to 20% in the presence of AEBSF, which may be explained by other mechanisms unrelated to fI-dependent proteolytic cleavage such as shedding of receptors from E, vesiculation, or shedding of viral envelope proteins (21, 22). The inhibition of spontaneous release by AEBSF suggested the involvement of a serine protease in the spontaneous detachment of HIV-Ig-C from E, pointing at fI. Accordingly, the release of HIV-Ig-C from E was abolished if we used HIV, which was opsonized in the presence of fI-depleted serum. The absence of spontaneous release of HIV opsonized in the presence of fI-depleted serum suggested that fI activity might be associated with the viral surface. However, we could not detect fI from any of the fractions after sucrose gradient centrifugation of HIV-Ig-C, which might be explained with limits in sensitivity of the assay used. fI could bind to the viral surface through C3b fragments. Because of possible conformational changes in C3b, fI and C3b interactions are described in the presence of factor H (23, 24). The formation of factor H-C3b-fI complexes leads to the cleavage of C3b into iC3b (24). The previously described association of factor H (fH) with opsonized HIV through both C3b-fH and HIV gp120/gp41-fH interactions might also support the formation of factor H-C3b-fI complexes (25). However, the low-affinity constant of such interactions could explain the absence of spontaneous release after overnight centrifugation through a sucrose gradient. The relatively long period of time will favor dissociation of fI from the virus. A putative interaction of fI with Igs on HIV after opsonization with HIV-specific Abs could be excluded because of the spontaneous release of HIV-C from E (Fig. 1 B). Since no dissociation was observed when E were incubated with virus opsonized in the presence of fI-depleted serum, an association of fI to E is unlikely. The spontaneous release of HIV-Ig-C from E was unaltered after repeated washing of HIV-loaded E, indicating that this detachment is unrelated to contaminations of soluble serum components (i.e., fI). Thus, an association of fI with HIV is most likely, but is currently not proven due to the putative low affinity of fI to opsonized HIV. The definite mechanism for a supposed association of fI with the surface of opsonized HIV remains to be elucidated.

Additionally, we found a higher capacity of hiNHS derived from HIV-positive individuals to detach opsonized HIV from E (Fig. 4 C). This interesting but unexplained observation requires further studies and is currently being investigated in our laboratory.

For the efficient processing of C3 fragments, several proteins like CR1 and fH serve as cofactors for fI (26). Thus, in addition to the binding of opsonized HIV through C3b fragments, CR1 expressed on E might contribute to the detachment of opsonized HIV after fI-mediated cleavage of C3b fragments on the viral surface. Cofactor activity of CR1 expressed on B cells was also assumed to contribute to the conversion of C3b fragments and therefore to the deposition of C3d-carrying HIV through CR2 on the same cells (27). These data suggest that the cofactor activity of CR1 expressed by E for the generation of C3d fragments on HIV might dominate in the blood stream, whereas CR1 expressed on other cells could be relevant for this in tissues. fH might also participate in the fI-mediated conversion of C3 fragments on the surface of HIV due to its cofactor activity. The contributions of fH in the serum and CR1 on E as cofactors for the fI-mediated processing of C3 fragments on HIV require further investigations.

The fI-mediated cleavage of C3b fragments to iC3b and C3d indicates that immune adherence of HIV might be a transient phenomenon in vivo (25) meant to generate iC3b and C3d fragments on the viral surface and thus to target the virus to CR2-expressing cells, e.g., FDC. CR2-C3d interactions were demonstrated to be important in the formation of the FDC-associated infectious HIV pool in the germinal centers of lymph nodes (4). This pool represents by far the largest viral reservoir in HIV-infected individuals. Thus, fI-mediated processing of C3b fragments might be an important component in the generation of C3d-coated viral particles deposited on FDC. fI-mediated processing of C3 fragments could also be important for the long-term persistence of HIV-1 structural proteins and glycoproteins on FDC described in HIV patients even under highly active antiretroviral therapy (28). Furthermore, peripheral B cells from HIV-infected individuals were described to carry complement-opsonized HIV particles in vivo (8, 10). C3d fragments on the surface of HIV interacting with CR2 expressed on B cells were shown to be critical for the B cell-mediated transmission of opsonized HIV to T cells (8, 9). Thus, the fI-mediated processing of C3b fragments to iC3b and C3d on HIV is a prerequisite for an effective interaction of C3d-coated virus with CR2-expressing cells. Accordingly, opsonized HIV released from E in the presence of hiNHS showed a higher capacity to bind to CR2-expressing Raji cells. Preincubation of HIV-Ig-C with E both in the presence and absence of hiNHS enhanced the infection of T cells in T cell-B cell coculture experiments, suggesting that HIV-Ig-C released from E have a higher capacity to assemble on B cells and subsequently to infect T cells. In contrast to the direct transmission of HIV by E to CD4-positive HeLa cells as demonstrated by Hess et al. (14), we did not observe productive infection of unstimulated primary T cells in the absence of B cells after preincubation of HIV-Ig-C with E either in the presence or absence of hiNHS. The lack of productive T cell infection by HIV-Ig-C in the absence of B cells indicated that interactions of B cells with HIV-Ig-C released from E were crucial for the efficient transmission of opsonized HIV to unstimulated T cells. Thus, the fI-mediated processing of C3b fragments on HIV to C3d may have in vivo relevance in the generation of C3d-coated HIV particles with high affinity for CR2 expressed on B cells (or on FDC) promoting an effective transmission of opsonized HIV to T cells. Of note, recent studies have demonstrated complement-mediated enhancement of productive HIV infection in human monocyte-derived dendritic cells (DC) (29). Therefore, processing of C3b fragments to iC3b on HIV might also be involved in the complement-mediated enhancement of HIV infection of monocyte-derived DC through binding of iC3b-coated virus to CR3 and CR4 expressed on DC.

Our findings provide a new dynamic view of complement opsonization of HIV. In the first round, complement opsonization of HIV and HIV-Ig results mainly in the generation of C3b fragments on the viral surface allowing the interaction of HIV with CR1 on E. Association of HIV with E might participate in viral spread by transporting infectious viral particles. However, the rapid fI-dependent spontaneous and NHS-mediated dissociation of HIV from E strongly limits the time available for this process. The rapid dissociation of HIV and HIV-Ig from E might also rescue HIV from elimination of HIV-Ig by phagocytes in the liver and spleen. In contrast, the fI-mediated processing of C3b fragments to iC3b and C3d on the viral surface allows the interactions of opsonized HIV with CR2-expressing cells, thereby facilitating B cell-mediated transmission of opsonized HIV to T cells or generation of the FDC-associated HIV pool in the germinal centers probably leading to the same consequences. Detailed understanding of the process and of the subsequent consequences of this sequential C3 processing on HIV in viral pathogenesis might offer the possibility to develop new therapeutic strategies, for instance, by finding ways to avoid C3b processing on HIV or to reduce virus binding to CR2 in another way.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by Grants P14661 (to M.P.D.) and P17914 (to H.S.) from the Fonds zur Förderung der Wissenschaftlichen Forschung, Grants QLK-CT-2002-00882 (to H.S.), and TIP/VAC-012116 (to H.S.) from the 5th and 6th framework of the European Union, grants from the Ludwig-Boltzmann-Institute for AIDS Research, grants from the German Federal Ministry of Research (to H-J. S., Competence Network HIV, 01KI0211), and grants from the Government of Tyrol.

3

Abbreviations used in this paper: FDC, follicular dendritic cell; CR, complement receptor; IC, immune complex; fI, factor I; NHS, normal human serum; hiNHS, heat-inactivated NHS; AEBSF, 4-(2-aminoethyl)-benzenesulfonyl fluoride; fH, factor H; w/o, without; DC dendritic cell; E, erythrocyte.

1
Stoiber, H., M. Pruenster, C. G. Ammann, M. P. Dierich.
2005
. Complement-opsonized HIV: the free rider on its way to infection.
Mol. Immunol.
42
:
153
-160.
2
Boyer, V., C. Desgranges, M. A. Trabaud, E. Fischer, M. D. Kazatchkine.
1991
. Complement mediates human immunodeficiency virus type 1 infection of a human T cell line in a CD4- and antibody-independent fashion.
J. Exp. Med.
173
:
1151
-1158.
3
Doepper, S., H. Stoiber, L. Kacani, G. Sprinzl, F. Steindl, W. M. Prodinger, M. P. Dierich.
2000
. B cell-mediated infection of stimulated and unstimulated autologous T lymphocytes with HIV-1: role of complement.
Immunobiology
202
:
293
-305.
4
Bánki, Z., L. Kacani, P. Rusert, M. Pruenster, D. Wilflingseder, B. Falkensammer, H. J. Stellbrink, J. van Lunzen, A. Trkola, M. P. Dierich, H. Stoiber.
2005
. Complement dependent trapping of infectious HIV on human lymphoid tissues.
AIDS
19
:
481
-486.
5
Burton, G. F., B. F. Keele, J. D. Estes, T. C. Thacker, S. Gartner.
2002
. Follicular dendritic cell contributions to HIV pathogenesis.
Semin. Immunol.
14
:
275
-284.
6
Smith, B. A., S. Gartner, Y. Liu, A. S. Perelson, N. I. Stilianakis, B. F. Keele, T. M. Kerkering, A. Ferreira-Gonzalez, A. K. Szakal, J. G. Tew, G. F. Burton.
2001
. Persistence of infectious HIV on follicular dendritic cells.
J. Immunol.
166
:
690
-696.
7
Kacani, L., W. M. Prodinger, G. M. Sprinzl, M. G. Schwendinger, M. Spruth, H. Stoiber, S. Dopper, S. Steinhuber, F. Steindl, M. P. Dierich.
2000
. Detachment of human immunodeficiency virus type 1 from germinal centers by blocking complement receptor type 2.
J. Virol.
74
:
7997
-8002.
8
Moir, S., A. Malaspina, Y. Li, T. W. Chun, T. Lowe, J. Adelsberger, M. Baseler, L. A. Ehler, S. Liu, R. T. Davey, Jr, et al
2000
. B cells of HIV-1-infected patients bind virions through CD21-complement interactions and transmit infectious virus to activated T cells.
J. Exp. Med.
192
:
637
-646.
9
Dopper, S., D. Wilflingseder, W. M. Prodinger, G. Stiegler, C. Speth, M. P. Dierich, H. Stoiber.
2003
. Mechanism(s) promoting HIV-1 infection of primary unstimulated T lymphocytes in autologous B cell/T cell co-cultures.
Eur. J. Immunol.
33
:
2098
-2107.
10
Jakubik, J. J., M. Saifuddin, D. M. Takefman, G. T. Spear.
1999
. B lymphocytes in lymph nodes and peripheral blood are important for binding immune complexes containing HIV-1.
Immunology
96
:
612
-619.
11
Montefiori, D. C., B. S. Graham, J. Y. Zhou, J. T. Zhou, J. M. Ahearn.
1994
. Binding of human immunodeficiency virus type 1 to the C3b/C4b receptor CR1 (CD35) and red blood cells in the presence of envelope-specific antibodies and complement.
J. Infect. Dis.
170
:
429
-432.
12
Horakova, E., O. Gasser, S. Sadallah, J. M. Inal, G. Bourgeois, I. Ziekau, T. Klimkait, J. A. Schifferli.
2004
. Complement mediates the binding of HIV to erythrocytes.
J. Immunol.
173
:
4236
-4241.
13
Hess, C., T. Klimkait, L. Schlapbach, V. Del Zenero, S. Sadallah, E. Horakova, G. Balestra, V. Werder, C. Schaefer, M. Battegay, J. A. Schifferli.
2002
. Association of a pool of HIV-1 with erythrocytes in vivo: a cohort study.
Lancet
359
:
2230
-2234.
14
Hament, J. M., H. van Dijk, A. Fleer, P. C. Aerts, M. Schoenmakers, M. W. de Snoo, B. H. Dekker, J. L. Kimpen, T. F. Wolfs.
2003
. Pneumococcal immune adherence to human erythrocytes.
Eur. J. Clin. Invest.
33
:
169
-175.
15
Cunnion, K. M., P. S. Hair, E. S. Buescher.
2004
. Cleavage of complement C3b to iC3b on the surface of Staphylococcus aureus is mediated by serum complement factor I.
Infect. Immun.
72
:
2858
-2863.
16
Cunnion, K. M., E. S. Buescher, P. S. Hair.
2005
. Serum complement factor I decreases Staphylococcus aureus phagocytosis.
J. Lab. Clin. Med.
146
:
279
-286.
17
Leitao, M. F., M. M. Vilela, R. Rutz, A. S. Grumach, A. Condino-Neto, M. Kirschfink.
1997
. Complement factor I deficiency in a family with recurrent infections.
Immunopharmacology
38
:
207
-213.
18
Hess, C., J. A. Schifferli.
2003
. Immune adherence revisited: novel players in an old game.
News Physiol. Sci.
18
:
104
-108.
19
Frank, I., H. Stoiber, S. Godar, H. Stockinger, F. Steindl, H. W. Katinger, M. P. Dierich.
1996
. Acquisition of host cell-surface-derived molecules by HIV-1.
AIDS
10
:
1611
-1620.
20
Grovit-Ferbas, K., J. F. Hsu, J. Ferbas, V. Gudeman, I. S. Chen.
2000
. Enhanced binding of antibodies to neutralization epitopes following thermal and chemical inactivation of human immunodeficiency virus type 1.
J. Virol.
74
:
5802
-5809.
21
Sadallah, S., C. Hess, S. Miot, O. Spertini, H. Lutz, J. A. Schifferli.
1999
. Elastase and metalloproteinase activities regulate soluble complement receptor 1 release.
Eur. J. Immunol.
29
:
3754
-3761.
22
Miot, S., J. Marfurt, E. Lach-Trifilieff, C. Gonzalez-Rubio, M. Lopez-Trascasa, S. Sadallah, J. A. Schifferli.
2002
. The mechanism of loss of CR1 during maturation of erythrocytes is different between factor I deficient patients and healthy donors.
Blood Cells Mol. Dis.
29
:
200
-212.
23
DiScipio, R. G..
1992
. Ultrastructures and interactions of complement factors H and I.
J. Immunol.
149
:
2592
-2599.
24
Soames, C. J., R. B. Sim.
1997
. Interactions between human complement components factor H, factor I and C3b.
Biochem. J.
326
:
553
-561.
25
Stoiber, H., R. Schneider, J. Janatova, M. P. Dierich.
1995
. Human complement proteins C3b, C4b, factor H and properdin react with specific sites in gp120 and gp41, the envelope proteins of HIV-1.
Immunobiology
193
:
98
-113.
26
Lambris, J. D., Z. Lao, T. J. Oglesby, J. P. Atkinson, C. E. Hack, J. D. Becherer.
1996
. Dissection of CR1, factor H, membrane cofactor protein, and factor B binding and functional sites in the third complement component.
J. Immunol.
156
:
4821
-4832.
27
Leslie, R. G., W. M. Prodinger, C. H. Nielsen.
2003
. Complement receptors type 1 (CR1, CD35) and 2 (CR2, CD21) cooperate in the binding of hydrolyzed complement factor 3 (C3i) to human B lymphocytes.
Eur. J. Immunol.
33
:
3311
-3321.
28
Popovic, M., K. Tenner-Rácz, C. Pelser, H-J. Stellbrink, J. van Lunzen, G. Lewis, V. S. Kalyanaraman, R. C. Gallo, P. Rácz.
2005
. Persistence of HIV-1 structural proteins and glycoproteins in lymph nodes of patients under HAART.
Proc. Natl. Acad. Sci. USA
102
:
14807
-14812.
29
Bajtay, Z., C. Speth, A. Erdei, M. P. Dierich.
2004
. Cutting edge: productive HIV-1 infection of dendritic cells via complement receptor type 3 (CR3, CD11b/CD18).
J. Immunol.
173
:
4775
-4778.