Effector Function Activities of a Panel of Mutants of a Broadly Neutralizing Antibody against Human Immunodeficiency Virus Type 1

ABSTRACT The human antibody immunoglobulin G1 (IgG1) b12 neutralizes a broad range of human immunodeficiency virus-type 1 (HIV-1) isolates in vitro and is able to protect against viral challenge in animal models. Neutralization of free virus, which is an antiviral activity of antibody that generally does not require the antibody Fc fragment, likely plays an important role in the protection observed. The role of Fc-mediated effector functions, which may reduce infection by inducing phagocytosis and lysis of virions and infected cells, however, is less clear. To investigate this role, we constructed a panel of IgG1 b12 mutants with point mutations in the second domain of the antibody heavy chain constant region (CH2). These mutations, as expected, did not affect gp120 binding or HIV-1 neutralization. IgG1 b12 mediated strong antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) of HIV-1-infected cells, but these activities were reduced or abrogated for the antibody mutants. Two mutants were of particular interest. K322A showed a twofold reduction in FcγR binding affinity and ADCC, while C1q binding and CDC were abolished. A double mutant (L234A, L235A) did not bind either FcγR or C1q, and both ADCC and CDC functions were abolished. In this study, we confirmed that K322 forms part of the C1q binding site in human IgG1 and plays an important role in the molecular interactions leading to complement activation. Less expectedly, we demonstrate that the lower hinge region in human IgG1 has a strong modulating effect on C1q binding and CDC. The b12 mutants K322A and L234A, L235A are useful tools for dissecting the in vivo roles of ADCC and CDC in the anti-HIV-1 activity of neutralizing antibodies.

The broadly neutralizing antibody immunoglobulin G1 (IgG1) b12, directed to an epitope overlapping the CD4 binding site of gp120, was originally isolated from a human immunodeficiency virus type 1 (HIV-1)-infected individual by means of phage-display library cloning (11). This antibody neutralizes T-cell-line-adapted (TCLA) viruses and a broad range of primary viruses in various in vitro assays (13,30,41,52). Several studies demonstrated that IgG1 b12 completely protects severe combined immunodeficiency (SCID) mice populated with human peripheral blood lymphocytes (hu-PBL-SCID mice) from infection with both TCLA and primary viruses (21,41). The protection against primary viruses is apparent even if the antibody is given several hours after viral challenge (21). Furthermore, IgG1 b12 also protected against vaginal challenge with a pathogenic R5 SHIV (simian immunodeficiency virus [SIV]/HIV chimera expressing HIV-1 envelope) in rhesus macaques (43). In addition to IgG1 b12, a second broadly neutralizing monoclonal antibody (MAb) to gp120 (2G12 [53]) and three broadly neutralizing MAbs to gp41 (2F5, Z13, and 4E10 [9,39,60]) have been described. Recent studies demonstrated that 2G12 and 2F5, alone or in combination with one another, can protect against intravenous and/or mucosal SHIV challenge in macaques (3,35,36). Sterile protection typically requires that high antibody serum concentrations be achieved (e.g., in vitro neutralization titers of 1:100 or greater) (40,42,43), although some exceptions have been noted. In vaginal challenge studies with SHIV 89.6PD in rhesus macaques, for example, MAb 2G12 protected at antibody serum concentrations close to the 90% virus neutralization titer (36).
The antiviral activity of antibodies can be mediated by the neutralization of free virions or by binding to virus-specific proteins expressed on the surface of infected cells and the recruitment of Fc-mediated effector function (40).The importance of Fc-mediated effector function in protection against HIV-1 infection, however, is unclear. In a recent study, Binley and colleagues infused serum immunoglobulins purified from SIV mac251 -infected macaques (SIVIG) into other SIV mac251infected macaques and measured the impact on plasma viremia of infected animals. The effects on viral load observed were very modest and transient, with kinetics which seemed inconsistent with the neutralization of free viruses as the mechanism driving the effect, and a role of Fc-mediated effector mechanisms was therefore suggested. An experiment using SIVIG F(abЈ) 2 fragments to address this hypothesis, however, was inconclusive (7).
Fc receptors expressed on human peripheral blood cells play an important role in stimulating a variety of cytotoxic, phagocytic, and inflammatory functions. Once a virus-infected cell is opsonized by IgG, it may cross-link Fc␥R on the cell surface of an effector cell and mediate a cytotoxic response (antibodydependent cellular cytotoxicity [ADCC]). Arrays of antibody Fc's presented on the surface of an infected cell may also activate the classical pathway of complement activation ultimately leading to cell lysis (complement-dependent cytotoxicity [CDC]).
Multiple sites on IgG have been proposed to interact with Fc␥R. Mutagenesis studies have shown that the lower hinge region (234-LLGGPS-239) of IgG plays an important role in the binding of IgG Fc receptors (14,19,28,33,34,38,47,58,59). High-affinity binding to Fc␥RI is most notably affected by mutation of L235. Mouse (m)IgG2b, which does not bind Fc␥RI, has a glutamic acid at this position, and substitution of this residue by leucine was shown to restore the binding affinity of mIgG2b to be comparable to that of hIgG1 (19). The binding affinity of IgG for Fc␥RII, in contrast, seems more sensitive to mutations of L234 than L235, indicating that the Fc␥R interaction sites are overlapping but not identical (33). Residues in the lower part of the hinge region itself and the lower CH2 domain in addition may have a modulating effect on Fc␥R affinity (15,33).
Complement activation via the classical pathway is activated through binding of C1q to the Fc domain of IgG or IgM, complexed with antigens (12,25). Duncan and Winter (18) showed alanine substitutions in mIgG2b at positions E318, K320, K322, and N297, the last leading to the removal of carbohydrate, resulting in mutants in which binding to human C1q was strongly reduced compared to the wild type and the ability to mediate CDC was abrogated. Since E318, K320, and K322 are conserved residues in human IgG and IgG of several other species, they were designated the binding site for C1q (18). However, this binding motif has been conserved in all four human IgG subclasses which, in apparent conflict, exhibit large differences in their C1q binding abilities. Several studies have indeed implicated additional residues in C1q binding and have suggested that the binding sites for C1q on mIgG2b and hIgG are not completely identical (8,27,38,51,59). In a recent study on rituxan, a chimeric MAb with hIgG1 constant domains used in the therapy of non-Hodgkin's B-cell lymphomas, Idusogie and colleagues demonstrated that alanine substitution at positions D270, K322, P329, and P331 but not at positions E318 and K320 significantly reduced the ability of the chimeric MAb to bind C1q and activate complement, suggesting that E318 and K320 are only of minor importance for complement activation by hIgG1 (27). Some studies have furthermore suggested that C1q binding and complement activation may be modulated by residues in the lower hinge region (38). Thus, there are species differences in C1q binding. Finally, it has been shown that C1q binding alone is not sufficient for complement activation and complement-mediated cell lysis (51). Intrinsic factors, such as segmental flexibility of the hinge region (10, 50) but also extrinsic factors such as antigen specificity and density (5,6), may play an important modulating role.
In the present study, we introduced point mutations in the heavy chain constant domain of IgG1 b12. Amino acid mutations were chosen on the basis of the studies discussed above. However, because of subtle species differences in C1q and Fc␥R binding and the possible modulation of ADCC and CDC by antigen specificity and density, the impact of these mutations on the biological activity of IgG against HIV-1-infected cells was not immediately clear. The mutants were compared for their ability to bind Fc␥R and C1q and to mediate ADCC and CDC. These mutants can now be used to examine the role of Fc-mediated effector function in protection against HIV-1 infection in vivo.

MATERIALS AND METHODS
Cell lines, viruses, and MAbs. Uninfected CEM-NKr and CEM-NKr cells chronically infected with HIV-1 MN were obtained from Shermaine Tilley (Public Health Research Institute, New York, N.Y.) (1). To obtain clones of the infected CEM-NKr cells with an increased level of envelope expression, we performed a limiting dilution and characterized for Env expression by flow cytometry. A clone with a higher level of Env expression (75% of cells expressing Env) was selected for our experiments. Adult human elutriated monocytes were obtained from Mutagenesis of heavy-chain constant domain. Mutatagenesis was performed on the heavy-chain constant region derived from the IgG1 b12-expression plasmid pDR12 (see below) (13). A SacI-SalI endonuclease restriction fragment from pDR12, containing the CH2 fragment, was subcloned into M13 mp18. Site-directed mutagenesis (32) was performed using the Muta-gene M13 in vitro mutagenesis kit (Bio-Rad, Hercules, Calif.). Five clones encoding the desired changes (K322A, L234A, L235E, G237A, and L234A, L235A) were identified by automated DNA sequencing. The mutated SacI-SalI fragments were then cloned back into in the pDR12 expression vector.
Expression and purification of antibodies. Recombinant antibody was expressed in the vector pDR12 (provided by Raju Koduri and Dean Sauer). It contains a b12 light chain and heavy-chain expression cassette in which transcription is driven from a human cytomegalovirus promoter. The heavy-chain expression cassette contains the genomic human IgG1 gene. Selection and amplification of the plasmid was done on the basis of expression of the gene for glutamine synthetase (4).
IgG1 b12 mutant DNAs, prepared as described above, were cut with SalI and transfected into Chinese hamster ovary cells (CHO-K1 cells; American Type Culture Collection, Manassas, Va.) using lipofectin reagent per the manufacturer's recommendations (Life Technologies, Grand Island, N.Y.). Cells were distributed in six-well tissue culture plates, and clones were selected with L-methionine sulfoximine ranging in concentration from 40 to 100 M (Sigma, St Louis, Mo.). Wells containing discrete colonies were assayed by enzyme-linked immunosorbent assay (ELISA) for antibody production. The highest producers were cloned by limiting dilution, expanded, and grown in 3-liter spinner flasks.
Recombinant IgG1 was expressed in CHO-K1 cells in glutamine-free Glasgow minimum essential medium (GMEM supplemented with 10% dialyzed fetal bovine serum [FBS]) (Tissue Culture Biologicals, Tulare, Calif.), MEM nonessential amino acids (Gibco-BRL, Grand Island, N.Y.), 1 mM MEM sodium pyruvate (Gibco-BRL), 500 M L-glutamic acid, 500 M L-asparagine, 30 M adenosine, 30 M guanosine, 30 M cytidine, 30 M uridine, 10 M thymidine (Sigma), 100 U of penicillin/ml, 100 g of streptomycin/ml, and 50 M Lmethionine sulfoximine (Sigma) in a 3-liter spinner flask. The supernatants were sterile filtered and purified over protein A-Sepharose Fast Flow (Pharmacia, Arlington Heights, Ill.). The antibody was eluted in 0.1 M citric acid, pH 3.0. The pH of the antibody solution was immediately brought to neutrality by the addition of 1 M Tris (pH 9.0), and the antibody was dialyzed against phosphatebuffered saline (PBS). Antibody concentrations were determined by the absorbance at 280 nm and confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Antibody yields using this method ranged from 5 to 25 mg/liter.
Recombinant gp120 ELISA. IgG1 b12 and the Fc mutants were tested for binding to gp120 in ELISA essentially as described previously (17). Briefly, recombinant monomeric gp120 JR-FL (provided by Paul Maddon and Bill Olson, Progenics, Tarrytown, N.Y.) was coated to the wells of a microtiter plate by incubating overnight at 4°C. The plates were washed four times with PBS containing 0.05% (vol/vol) Tween 20 and blocked with 3% bovine serum albumin (BSA). The blocking solution was removed, and serial dilutions of the antibody set were added in duplicate (diluted in PBS-1% BSA-0.05% Tween 20) and incubated for 1 h at 37°C. The wells were washed and incubated for 1 h at 37°C with alkaline phosphatase-labeled goat anti-human IgG F(abЈ) 2 fragments (Pierce, Rockford, Ill.) (1:500 dilution in PBS-1% BSA-0.05% Tween 20). The plates were washed and developed with nitrophenol substrate (Sigma), and the absorbance was read at 405 nm.
HIV-1 neutralization. Neutralization of HIV-1 primary isolates was assessed using a phytohemagglutinin (PHA)-activated peripheral blood mononuclear cell (PBMC)-based assay as described previously (60). PBMCs (from three CCR5 wild-type donors) were isolated and stimulated with PHA (5 g/ml) (Sigma) for 48 h, followed by PHA and interleukin-2 (40 U/ml) (obtained from the ARRRP, contributed by Hoffman-La Roche) for 72 h in RPMI 1640 medium containing 10% heat-inactivated FBS, 100 U of penicillin/ml, 100 g of streptomycin/ml, and 2 mM L-glutamine. The antibodies were diluted, and 50 l per well was pipetted into round-bottom microtiter plates, after which an equal volume containing 100 50% tissue culture infective doses of HIV-1 stock was added. The antibody-virus mixture was incubated for 1 h at 37°C. Next, 100 l of PHAactivated PBMCs (5 ϫ 10 5 /ml) was added to each well. After an overnight incubation, the cells were washed two times with tissue culture medium. On day 7, the cultures were collected and treated with 1% (vol/vol) Empigen (Calbiochem, La Jolla, Calif.). Triplicate samples were then tested for p24 content using an ELISA, as originally described by Moore et al. (37). In brief, sheep anti-p24 Ab D7320 (Aalo Bioreagents) was coated overnight on 96-well polystyrene enzyme immunoassay (EIA) plates (Costar) in 100 mM NaHCO 3 , pH 8.5. The plates were washed in PBS, and p24 was captured from serial dilutions of the HIV-1-containing samples in PBS-0.1% Empigen. After a 3-h incubation, unbound p24 was washed away and bound p24 was detected with alkaline phosphatase-labeled antibody BC1071 (International Enzymes) diluted 1:3,000 in PBS containing 20% sheep serum and 2% nonfat dry milk. After a 1-h incubation, the plates were washed and developed with an AMPAK kit (Dako Diagnostics) as recommended by the manufacturer. Production of p24 antigen in the antibody-containing cultures was compared to p24 production in cultures without antibody run in the same assay, and the antibody concentrations resulting in a 90% reduction in p24 content were determined.
Fc␥R binding assays. Fc␥R binding assays were performed essentially as described by Parren et al. (45). Binding of antibody to Fc␥R was assessed using antibody monomers or dimers as indicated. Antibody dimers were prepared by incubating overnight with F(abЈ) 2 fragments of mouse anti-human -light chain MAb K35 in a molar ratio of 1:1, which resulted in stable tetrameric complexes as detailed by Huizinga et al. (26). Fc␥R-transfected cells (3 ϫ 10 5 in 100 l of PBS-1% BSA) were incubated with 25 l of serial dilutions of antibody monomers or dimers for 45 min at 4°C, washed, and then incubated with fluorescein isothiocyanate-labeled F(abЈ) 2 fragments of goat anti-human IgG. After washing, cell-bound antibody was detected using flow cytometry. The assays were performed using IIA1.6 cells transfected with Fc␥RIA and ␥-chain, or Fc␥RIIa (H131), and Jurkat cells transfected with Fc␥RIIIa (24,45,54,55).
C1q ELISA. Antibody was diluted to 1.25 g/ml in PBS and coated overnight at room temperature onto EIA ELISA plates (Costar, Corning, N.Y.). Plates were washed three times with PBS-0.05% Tween 20 and a titration of human C1q (Calbiochem) prepared in PTG (PBS-0.02% Tween-0.1% gelatin) was added. After a 4-h incubation at room temperature, the plates were washed four times with PBS-0.05% Tween 20. A mixture of goat anti-human C1q (Calbiochem) and rabbit anti-goat IgG alkaline phosphatase conjugate (Sigma), both diluted at 1/1,000 in PTG, was added, and the plates were incubated for 1 h at room temperature. The plates were washed four times and developed using nitrophenol substrate (Sigma). Absorbance was measured at 405 nm. All data are expressed as means of triplicates.
ADCC with TCLA HIV-1. ADCC was assessed in standard chromium release assays (56). Effector cells were either PBMCs or adult human elutriated monocytes, as noted in the text. The PBMCs were isolated by centrifugation over Histopaque-1077 (Sigma). Cells were washed in PBS and resuspended in RPMI 1640 containing 10% FCS, 2 mM L-glutamine, penicillin (50 U/ml), and streptomycin (50 g/ml) at a density of approximately 4 ϫ 10 6 cells/ml and incubated overnight at 37°C prior to use as effector cells in ADCC. The following day, 10 6 target cells were labeled with Na 2 51 CrO 4 (Amersham, Arlington Heights, Ill.) for approximately 2 h in 50 l of fetal calf serum (FCS) containing 10 mM HEPES. These target cells were either uninfected CEM-NKr cells or CEM-NKr cells chronically infected with HIV-1 MN . After labeling, target cells were washed four times with RPMI 1640-10 mM HEPES. In each well of a microtiter plate, 10 6 washed 51 Cr-labeled target cells were incubated with antibody for 30 min at 37°C in a total volume of 150 l. Then, 10 6 PBMCs or 6 ϫ 10 5 human elutriated monocytes (monocytes were thawed rapidly at 37°C and washed once with RPMI 1640 before addition to target cells) were added as a source of effector cells in 50 l of assay medium, bringing the total volume to 200 l. The plates were spun at 1,000 rpm for 5 min in a Beckman GS-6R centrifuge to pellet the cells prior to a 4-h incubation at 37°C. At the end of this incubation period, plates were spun another time as described above, 100 l of supernatant was collected, and 51 Cr release was measured in a gamma counter (Packard, Meriden, Conn.). The percent specific lysis was calculated as follows: (experimental release Ϫ spontaneous release)/(total release Ϫ spontaneous release) ϫ 100%. Total 51 Cr release was determined by substituting 50 l of antibody for Empigen detergent (Calbiochem). All data are expressed as the means of triplicate determinations.
CDC assay. Target cells (CEM-NKr cells; uninfected or infected with HIV-1 MN ) were labeled by Na 2 51 CrO4 as described above. After three washes, the labeled cells were sensitized by adding wild-type or mutant IgG1 b12 to the cells at a final concentration of 10 g/ml and were incubated at 4°C for 1 h. After three washes, 2 ϫ 10 5 sensitized target cells were dispensed into 96-well U-bottom microtiter plates. Rabbit serum (Calbiochem) was used as a source of complement and was serially diluted with RPMI 1640. The latter was added to sensitized target cells at 100 l/well, and the plates were incubated at 37°C for 1 h. At the end of this time, plates were spun and 100 l of supernatant was used to measure the 51 Cr release as described above.

RESULTS
Antigen recognition and neutralization. Five IgG1 b12 mutants were constructed by introducing point mutations in the lower hinge (L234A, L235E, G237A, and double mutant L234A, L235A) and the N-terminal end of the CH2 domain (K322A). Although it was unlikely that these mutations would influence the affinity of the antibody for gp120, all five antibodies were tested for binding to recombinant HIV-1 JRFL gp120 in ELISA. All mutants bound similarly to gp120 as expected (Fig. 1a). In addition, we tested all mutants in a PHAactivated PBMC-based neutralization assays with HIV-1 JR-FL (Fig. 1b), HIV-1 JR-CSF , and HIV-1 89.6 ( Table 1). All mutants neutralized these primary isolates similarly ( Fig. 1b; Table 1).
Binding of wild-type and mutant IgG1 b12 to Fc␥RI, Fc␥RIIa, and Fc␥RIIIa. The ability of antibodies to mediate ADCC is dependent on the relative affinity of the antibody for Fc␥RI, -II, and -III. We first measured the ability of IgG1 b12 and the mutant antibodies described above to bind to Fc␥RI by flow cytometry using both monomeric and dimeric IgG1 (cross-linked with an anti-light chain MAb) (26). Binding of IgG by the high-affinity receptor Fc␥RI is usually studied using monomeric IgG. We, however, also included dimeric IgG which allowed us to also examine possible lower-affinity interactions. In these binding studies, we used Fc␥RIa-and ␥-chaintransfected IIA1.6 cells (54). Binding of both dimeric and monomeric wild-type and mutant IgG1 b12 to IIA1.6 cells revealed that the mutation at position 234 reduced the affinity for Fc␥RI about fivefold, whereas the other two mutations in the lower hinge (L235E and G237A) reduced IgG1 b12 affinity for Fc␥RI about 40-fold. Combining the 234 and 235 mutations completely abolished Fc␥RI binding, and binding was reduced to undetectable levels even in the IgG dimer-binding assay. The Fc␥RI binding affinity of K322A was not affected. The rank order of Fc␥RI-binding by the IgG1 b12 variants therefore is as follows: b12 ϭ K322A Ͼ L234A Ͼ Ͼ G237A, L235E, double-mutant L234A, L235A (Table 2).
Next, we tested the ability of our antibody panel to bind to the low-affinity receptors Fc␥RIIa and Fc␥RIIIa. The mutation in the CH2 domain (K322A) only slightly reduced the binding affinity to both Fc␥RIIa and Fc␥RIIIa compared to wild-type IgG1 b12 ( Table 2). All mutations in the lower hinge region in contrast abolished binding to both Fc␥RIIa and Fc␥RIIIa (Table 2). Therefore, the importance of L235 for Fc␥RI binding was confirmed, but binding could only be completely abrogated by introducing a double mutation at L234 and L235. In contrast to previous studies on hIgG3 (34), mutagenesis of L234 as well as L235 in hIgG1 abolished both Fc␥RII and Fc␥RIII binding.
ADCC. Antibody (IgG1 b12) bound to envelope expressed on the surface of HIV-1-infected cells may recruit effector cells, such as NK cells or monocytes, by interacting with specific Fc receptors and induce ADCC leading to lysis of the infected cells. In order to measure the ability of IgG1 b12 and the mutants described above to mediate ADCC of HIV-1infected cells, we incubated serial dilutions of these antibodies with 51 Cr-labeled HIV MN -infected CEM-Nkr cells in the presence of PBMCs and purified monocytes as effector cells. As shown in Fig. 2a and b, IgG1 b12 mediated specific cell lysis of HIV MN -infected CEM-NKr cells in the presence of both human PBMCs and purified monocytes, whereas ADCC by mutant K322A was reduced. The ability of IgG1 b12 mutants L235E, L234A, G237A, and double mutant L234A, L235A to mediate ADCC of HIV-1 MN -infected CEM-NKr cells was strongly reduced with PBMCs and abolished with monocytes as effector cells (Fig. 2a and b). The positive control serum, FDA-2, is a potently neutralizing serum with a 90% HIV-1 MN neutralization titer of 1:4,000 (44); ADCC of IgG1 b12 on noninfected CEM-Nkr cells was used as a negative control.
To verify the studies with the IgG1 mutant antibodies, we examined the IgG1 b12-mediated ADCC of HIV MN -infected CEM-NKr with PBMCs and monocytes in the presence and absence of anti-Fc␥R antibodies. F(abЈ) 2 fragments of an anti-Fc␥RIII antibody, 3G8, efficiently inhibited the PBMC-mediated ADCC of HIV MN -infected CEM-NKr (Fig. 3a). No inhibition of ADCC was observed with Fab fragments of anti-Fc␥RI 10.1 or anti-Fc␥RII IV.3. Thus, ADCC of HIV-infected cells by PBMCs is mediated through Fc␥RIIIa. This is in agreement with the strong reduction of ADCC in mutants in which binding to Fc␥RIIIa was abolished (Fig. 2a). In addition, the ability of mutant K322E to mediate ADCC was reduced about twofold compared to IgG1 b12, which corresponds with its reduction of binding to Fc␥RIIIa ( Fig. 2a; Table 2). As shown in Fig. 3b, anti-Fc␥RI (Fab 10.1) but not anti-Fc␥RII (Fab IV.3) or anti-Fc␥RIII [F(abЈ) 2 3G8], inhibited monocyte-mediated ADCC of HIV-1 MN -infected CEM-Nkr-MN cells. These results suggest that ADCC of HIV-infected cells by monocytes is mediated through Fc␥RI. A relatively small (fivefold) reduction in Fc␥RI binding affinity as observed for the mutant L234A is therefore sufficient to abrogate the monocyte-mediated ADCC of HIV-1-infected cells (Fig. 2b).
C1q binding and complement activation. The ability of b12 and the mutant antibodies to bind C1q and CDC of HIV-1-  a Data are expressed as the antibody concentration at which half maximal binding was achieved, K 50 of Ͼ50 g/ml indicate that this was not achieved. infected cells was investigated. As shown in Fig. 4a and b, wild-type IgG1b12 bound well to C1q and mediated a potent CDC of HIV-1 MN -infected CEM-Nkr cells in the presence of rabbit serum as a source for complement. The ability of the K322A mutant to bind C1q and mediate CDC was strongly reduced. All b12 variants with mutations in the lower hinge region (L235E, L234A, G237A and L234A, L235A) furthermore were reduced in their ability to mediate CDC of HIV-1infected cells as well ( Fig. 4a and b).
These results strongly suggest that for hIgG1, the residue at position 322 at the N-terminal end of the CH2 domain as well as residues in the lower hinge region are involved in C1q binding and consequent complement activation.

DISCUSSION
The role of effector function in the anti-HIV-1 activity of antibody is poorly understood. It has been shown in a number of passive antibody transfer studies that MAbs or polyclonal antisera capable of neutralizing the challenge virus can protect against HIV-1 infection (2, 3, 20-22, 35, 36, 40, 41, 46). Significantly, passive immunization studies using broadly neutralizing antibodies, including b12, have recently been used to protect rhesus macaques from intravenous and mucosal challenge with pathogenic primary HIV-1 isolate-derived SHIVs (35,36,43).
Neutralizing antibody may limit the dissemination of an enveloped virus, such as HIV-1, by at least two separate mechanisms. First, the antibody may interact with free virus and neutralize it by interfering with the attachment or fusion of its target cells, thereby protecting the cells from infection. Second, the neutralizing antibody may bind to envelope expressed on the surface of infected cells and induce cell lysis by the recruitment of effector functions, including ADCC and CDC (reviewed by Parren and Burton [40]). The relative roles of neutralization and Fc-mediated effects in HIV-1 infection have not yet been directly determined. In a recent study, Binley and colleagues (7) suggested that a transient effect on viral load by the infusion of anti-SIV antibodies was likely due to killing of SIV-infected cells, although this conclusion could only be drawn by inference.
We set out to directly study the role of Fc-mediated effector function in HIV-1 infection by the preparation of a panel of Fc  51 Cr for 2 h at 37°C. The labeled target cells were incubated with wild-type or mutant IgG1 b12 before addition of cultured PBMCs (a) or purified human monocytes (b) as effector cells. Wild-type and mutant IgG1 b12 were used at 12.5 g/ml. Serum from an HIV-1-seropositive patient (FDA-2 [44]) was used at a 1/4,000 dilution as a positive control. Uninfected CEM-NKr cells incubated with wild-type IgG1 b12 were included as a negative control. The assays were performed twice with similar results.  (38,48,57,59). We also showed that double mutation of amino acids 234 and 235 completely abolished binding to all Fc␥Rs.
We examined the nature of the Fc␥Rs involved in the ADCC of HIV-infected cells by using anti-Fc␥R antibodies.
Our data indicate that ADCC by PBMCs is inhibited only in the presence of anti-Fc␥RIII antibody, while ADCC by monocytes is abolished in the presence of anti-Fc␥RI antibody. Alsmadi and Tilley (1) proposed previously that ADCC of HIVinfected cells is exclusively mediated through Fc␥RIII expressed on the surface of CD56 ϩ NK cells. In their study, they measured ADCC activity using cultured PBMCs as effector cells which are likely reduced in monocyte content due to the adherence of these cells to plastic. Similarly, we found that PBMC-dependent ADCC was primarily mediated through Fc␥RIII, even though monocytes represented up to 8% of total cells in our PBMC preparations, as measured by fluorescenceactivated cell sorting (FACS) using a CD14 marker (data not shown). The effective ADCC by purified monocytes in vitro in our study, however, indicates that it is likely that monocytes or macrophages expressing Fc␥RI contribute to the ADCC of HIV-1-infected cells in vivo. Therefore, we suggest that ADCC of HIV-infected cells in vivo may be mediated by both Fc␥RI and Fc␥RIIIa.
Human IgG1 has the ability to bind C1q and lyse cells by activating complement through the classical pathway. The binding site for C1q on murine IgG2b was mapped to residues 318, 320, and 322 (18). We measured the ability of b12 and the mutants described to bind to C1q and to activate complementmediated cell lysis. We used heterologous rather than homologous complement, as this resulted in more efficient lysis of HIV-1-infected (human) cells (data not shown). Changing lysine 322 to alanine completely abolished binding to C1q and complement activation, which confirms a previous report (27). Less expected was the reduction of C1q binding and complement activation by mutations in the lower hinge. In an earlier study, Morgan and colleagues had reported on the reduction of C1q binding and complement activation by changes in residues 235 and 237 for a chimeric mouse/human antibody (38). We show that in a fully human IgG1, L235, G237, and, in addition, L234 in the lower hinge region appear to play a significant direct or indirect role in C1q binding.
A concern may be that the null phenotype of the L234A, L235A double mutation for Fc␥R as well as C1q binding is a result of more-drastic rearrangements of the IgG Fc structure of this mutant. However, in a recent study, we have shown that the binding of the L234A, L235A double mutant to FcRn was only slightly reduced (Ͻ25%) compared to wild-type b12 (57). Protein A and G binding furthermore were unaffected (not shown). Proteins A, G, and FcRn all bind to the CH2-CH3 interface. The retention of FcRn binding, in particular, is significant as this receptor has two important functions, namely the cross-placental transport of maternal IgG to the fetus and the protection of IgG from normal serum protein catabolism (23,29,49). The serum half-life of the L234A, L235A mutant should therefore not be significantly affected.
In summary, we demonstrated that manipulation of residues 234, 235, and 237 in the lower hinge region of hIgG1 modulates binding to Fc␥Rs. Our data also indicated that, in hIgG1, both the lower hinge and N-terminal end of the CH2 domain are involved in C1q binding and complement lysis. Furthermore, using both PBMCs and purified monocytes as effector cells, we found that ADCC of HIV-infected cells is mediated through both Fc␥RI and Fc␥RIIIa. As shown in our results, mutant K322A was only slightly, up to twofold, reduced in its ability to mediate ADCC and did not mediate CDC. On the other hand, the double mutant L234A, L235A mediated neither CDC nor ADCC. Both mutants were unchanged in their ability to neutralize primary HIV-1 isolates. Therefore, these mutants display specific valuable features that can be used in future in vivo studies. Our goal will be to use these mutants in in vivo passive antibody transfer-SHIV challenge studies in rhesus macaques to elucidate the relative roles of neutralization and of ADCC and complement activation in protection against HIV-1 infection and pathogenesis.