Antibody-Mediated Immunomodulation: a Strategy To Improve Host Responses against Microbial Antigens

The immunoregulatory properties of antibody have been recognized since the earliest passive immunization experiments, and the potential to modulate an immune response by deliberate immunization with antigen bound by antibody has been demonstrated in numerous instances over the decades (2, 6, 15, 18, 19, 37, 42, 52, 61, 62, 64, 75, 77, 89, 92, 100, 103, 105, 110, 115-117, 120, 127, 130). The concept of using a combination of antibody and antigen to improve the host response is not new and has been rediscovered for different applications many times over. In his later years, the first Nobel laureate, Emil von Behring, expanded on the approach of passive immunization and sought to induce immunity against diphtheria in children by administering a combination of diphtheria toxin and antitoxin (43). Enhancing or suppressive effects of antibody have been documented depending on the particular antigen-antibody interaction, and the underlying molecular mechanisms by which antibody can alter an immune response are being elucidated.

Historically, the connotation of immune response activation via immune complexes has generally been perceived as negative, and a plethora of literature regarding pathological associations abounds. However, the benefit of utilizing antibody in combination with antigen to achieve a desirable immune response is far less appreciated and is the focus of this minireview. There is increasing recognition that exogenously administered antibody may exert a therapeutic effect by redirecting the host response rather than by playing a purely passive role (16, 18, 26, 45, 53, 55, 56, 84, 90, 93, 100, 114, 129). Both polyclonal and monoclonal reagents, administered either alone or in combination with antigen, have been used to up-regulate beneficial or protective immune responses against infectious agents and malignant tumors as well as to down-regulate deleterious responses associated with inflammation, autoimmunity, and hypersensitivity (8, 55, 57, 58, 84, 102, 110). In light of a growing body of literature, the practicality of employing preformed antibody to manipulate an immune response toward a desired end is becoming more apparent and will broaden the strategies for active and passive immunization approaches against infectious disease.

IMMUNIZATION WITH IMMUNE COMPLEXES

Examples with individual antigens.

Immunization with immune complexes (IC) has been used to enhance immunogenicity of soluble molecules, to increase the number of monoclonal antibody (MAb) producing hybridomas against an antigen, and to elicit antibodies specific for poorly immunogenic epitopes. MAbs against human alpha-2-macroglobulin (36) as well as complement components (35) have been generated against IC composed of proteins immunoprecipitated with conventionally produced polyclonal antisera. Murine humoral (75) and T-cell (76, 77) responses against human serum albumin were stronger when the antigen was administered as an IC with syngeneic antibodies. To facilitate production of MAbs against weakly immunogenic regions of human thyrotropin (9) and follitropin (10), mice were immunized with IC containing MAbs against immunodominant epitopes in a successful effort to block the response against those sites. Antihapten immunoglobulin G2a (IgG2a) and IgG2b, but not IgG1, IgM, or IgA, complexed with trinitrophenol- or fluorescein-conjugated keyhole limpet hemocyanin (KLH) increased the primary antibody response in mice against the carrier protein by 20- to 1,000-fold, depending on the antigen-antibody combination, after a single injection of antibody-complexed haptenated KLH (32). Secondary responses were enhanced approximately threefold following boosting with IgG2-complexed antigen rather than free antigen. In a series of studies, Bouige et al. demonstrated that immunization with IC containing MAbs and several different types of antigens, including human secretory IgA (sIgA), bacterial polysaccharide from Escherichia coli, and a structural protein from hepatitis B virus, could result in the recognition of new epitopes that differed in location from those recognized by the modulating MAbs (15-17). For example, the pre-S2 region of hepatitis B surface antigen (HbsAg) was rendered more immunogenic when a MAb was bound to the S region of the HbsAg. Not all anti-HbsAg MAbs resulted in enhancement of the response, and indifference or negative modulations could also be demonstrated depending on the selection of the MAb (16).

In some cases immunogenicity has been studied using IC containing immunoglobulin-binding proteins in addition to antigen and antibody. For example, the presentation of a snake toxin in IC containing MAbs and antigen was reported to be increased by the incorporation of bacterial Fc-binding proteins such as protein A and protein G in the complexes (69), and MAbs recognizing novel epitopes have been generated against the feline homolog of CD4 with solid matrix antigen-MAb complexes containing formalin-fixed Staphylococus aureus (128).

While most published studies have evaluated changes in immunogenicity of protein antigens contained within IC, there is documentation that an antibody response against a nonprotein antigen can also be altered by using this approach. Unresponsiveness to pneumococcal cell wall polysaccharide (PnC) was reversed by immunization of transgenic mice, 90% of whose B cells express Ig specific for a phosphorylcholine (PC) determinant, with IC of PnC and anti-PC myeloma antibodies TEPC-15 and MOPC-603 (30). The effect was eliminated by treatment with anti-CD4, suggesting a mechanism engaging helper T cells. Interestingly, enhancement of the anti-PnC response varied depending on the fine specificity and variable light chain (V_L) gene usage of the three IgA myeloma proteins tested. Anti-PC MOPC-167 expressing the same heavy chain variable (V_H) and V_L genes used to encode the transgene antibody was not effective. Enhancement was also dependent on the ratio of antigen to antibody in the immune complexes. Whereas TEPC-15 markedly enhanced the anti-PnC response when it was incorporated into IC in 10-fold antigen excess, it had previously been shown to suppress the anti-PnC response when IC were prepared in 10-fold antibody excess (29).

Applications for infectious disease.

Because of the recognized immunomodulatory potential of antibody, immunization with IC containing either polyclonal or monoclonal reagents has now been explored in a number of studies in successful attempts to elicit beneficial responses against human and animal pathogens, including viruses and bacteria. Complexes of a formalinized Venezuela equine encephalitis vaccine and specific IgG at antigen-antibody equivalence enhanced the immune responses of rhesus monkeys to the vaccine (54). Antibodies elicited against the complex were predominantly IgG, compared to IgG and IgM, against the vaccine alone, and a more rapid secondary response was observed in monkeys primed with IC. Sustained protection was observed in mice 24 h after immunization with IC compared to that 8 days after administration of antigen alone. In a study of simian virus 5 (SV5) paramyxovirus, MAbs against viral proteins were coupled to solid matrices of fixed S. aureus or protein A-Sepharose and used to purify antigens from infected tissue culture cells. The resulting solid matrix antigen-antibody (SMAA) complexes were used as immunogens that induced specific humoral and cytotoxic T-cell responses (96). Decreased SV5 virus replication was demonstrated within infected lungs in a mouse model system, with protection found to correlate with the induction of cytotoxic T cells (97). A similar protocol was used to purify recombinant simian immunodeficiency virus (SIV) proteins and to utilize the SMAA complexes to elicit high titers of antisera against p17, p27, and reverse transcriptase (48, 95). A subsequent study showed that the complexes induced higher antibody levels than did antigen alone against some, but not all, SIV proteins (47), again suggesting that enhancement, indifference, or suppression of the host response depends on the particular antigen-antibody combination used for immunization. Equine herpes virus 1 glycoproteins C and D have also been incorporated into SMAA complexes and used to induce neutralizing and complement-activating antibodies and T-cell proliferative responses in BALB/c and C3H mice, with the glycoprotein D complex resulting in induction of protection against intranasal challenge (2). An IC of antigenic subunits of Newcastle disease virus (NDV) and specific polyclonal antibodies was used to generate a high-titer anti-NDV antibody response and to protect chickens against live viral challenge (90). Another poultry pathogen, infectious bursal disease virus (IBDV), has been targeted by an IC vaccination approach as well. Complexes of live infectious virus and hyperimmune chicken serum resulted in substantially improved protective responses compared to immunization with uncomplexed vaccine (44, 59). More germinal centers (GC) were induced in the spleen following IC immunization with larger amounts of antigen localized on splenic and bursal follicular dendritic cells (DC) (59).

As stated earlier, immunization with IC containing certain MAbs and HbsAg resulted in the formation of antibodies directed against novel epitopes of the vaccine immunogen (16). This finding may be of direct clinical applicability, considering the relatively high percentage of individuals who do not seroconvert following immunzation with HBsAg alone (22) in that additional target epitopes would be available and potentially recognizable by the nonresponding population. IC of HBsAg and MAbs have been reported to stimulate proliferation of immune T lymphocytes more efficiently than did antigen alone (31), and promising therapeutic efficacy of an IC vaccine in the treatment of chronic hepatitis B patients has been reported (125, 130). MAbs administered as part of IC have also been demonstrated to influence the humoral immune response against the P1 surface adhesin of the dental pathogen Streptococcus mutans (18, 87, 99,100). Both the specificity and the isotype of anti-P1 antibodies elicited in immunized mice were altered by the MAbs. The changes were influenced by the specificity of the MAbs, by the antigen-antibody ratios within the IC, and by the route of administration. In several cases, antibodies elicited against IC inhibited bacterial adherence significantly better than did antibodies elicited against antigen alone.

Applications for tumor immunity.

Modulation of immune responses by antibody is also now being examined for the development of therapeutic cancer vaccines. To test whether the immunogenicity of tumor antigens could be augmented by antibody following in situ IC formation, a mouse melanoma cell line was engineered to express alpha-galactosyl epitopes (67, 68). Mice with a deletion of alpha-1,3-galactosyltransferase demonstrating natural anti-Gal IgG were immunized with irradiated modified tumor cells prior to challenge with live parental melanoma tumor cells lacking the engineered antigen. Significant protection was observed compared to that by immunization with the irradiated parental tumor cell line, suggesting that the antibody-antigen interaction potentiated the immune response against the parental tumor that lacked alpha-galactosyl epitopes. This represents another example whereby the immunogenicity of epitopes other than those recognized by the modulatory antibody was affected. To test whether exposure of antigen presenting cells to IC compared to antigen alone would influence the resultant T-cell response, a study was undertaken in which DC were exposed to prostate-specific antigen (PSA) or PSA combined with an anti-PSA MAb (13). Both CD4- and CD8-T-cell responses were detected following DC stimulation with the IC, whereas a CD4 response predominated when DC were stimulated with PSA alone. Pulsing DC with human leukocyte antigen (HLA) allele-restricted peptides suggested that PSA alone was processed primarily through pathways favoring major histocompatibility complex (MHC) class II presentation while PSA and anti-PSA complexes were processed through both class I and class II pathways. Other studies have also utilized antibody to optimize tumor antigen presentation by DC. Rafiq et al. (92) demonstrated that tumor immunity specific for ovalbumin-expressing tumors could be achieved by immunization of C57BL mice with wild-type DC but not Fc receptor (FcR)-deficient DC loaded with IC containing ovalbumin and antibody. Tumor protection was eliminated when DC lacked β₂-microglobulin, transporters associated with antigen processing (TAP), or MHC class II, indicating that FcR-mediated uptake of antigen in IC resulted in both class I- and class II-restricted immune responses. Akiyama et al. (1) also targeted tumor antigens via IC to FcγR on DC. In experiments with the murine thymoma cell line E.G7, apoptotic tumor cells (ATC) were used as a source of tumor antigens. IC-containing ATC were more efficient than ATC alone at inducing cytotoxic T cells and tumor rejection. Again, using DC from FcγR-deficient mice, uptake of IC-containing ATC was shown to be FcγR dependent and IC-containing ATC were more effective than ATC alone in promoting maturation of DC, as evidenced by increased expression of CD86 and MHC class II.

The clinical benefit to humans of exposure to IC containing an immunomodulatory antibody has now been recognized. During clinical trials of the murine IgG1κ MAb OvaRex against the tumor antigen CA125, it was observed that treatment of ovarian cancer patients with Technitium^99m-labeled MAb in the presence of circulating CA125 could result in the induction of both specific B- and T-cell responses and unexpectedly favorable outcomes (86). Anti-MHC class I and II antibodies blocked secretion of IFN-γ by patients' peripheral blood mononuclear cells (PBMC) stimulated with CA125 or autologous tumor, indicating the induction of specific helper and cytotoxic T lymphocytes in a number of individuals receiving OvaRex (84). Further support that antibody therapy may enhance effective tumor immunity was provided by Dhodapkar et al., who demonstrated that coating tumor cells with antitumor antibodies promoted cross-presentation of myeloma-derived cellular antigens and the induction of tumor-specific cytotoxic T cells by DC in an FcγR-dependent manner (34).

SEQUENTIAL OR COADMINISTRATION OF ANTIBODY AND ANTIGEN

It has been shown that the dissociation rate of antigen from antibody can be slower than the time for antigen capture, endocytosis, and processing by professional antigen-presenting cells; thus, when antigen and antibody are simultaneously present in the host, the substrate for processing may often in actuality be an antigen-antibody complex (121). This is important to keep in mind, especially during passive immunization studies in which IC could easily form in situ even though exogenous antibody and challenge microorganisms are administered separately. Unintentional exposure of the host immune system to an immunogen in the form of an IC may explain the contribution of a “passive” antibody to an active adaptive immune response whose effects can be measured after the clearance of the exogenously administered antibody. The possibility that passively administered antibody used therapeutically to treat a streptococcal infection may work via an immunomodulatory mechanism was suggested by Ramisse et al. (93). Human plasma-derived Ig (IVIG) administered either intravenously or intranasally prior to challenge with Streptococcus pneumoniae was protective in a murine model of pneumonia. Compared to untreated mice, mice protected with IVIG or F(ab′)₂ fragments developed higher levels of measurable antibodies against pneumolysin and acquired greater resistance to subsequent reinfection even after clearance of the human antibodies, strongly suggesting that the treatment of the animals with IVIG potentiated the development of protective adaptive immunity. A similar approach had been used previously to demonstrate the efficacy of passive immunotherapy with IVIG or F(ab′)₂ fragments in a mouse model of staphylococcal pneumonia (94). The participation of exogenous antibody in an adaptive immune response was also suggested by Stenbaek (111). Passive immunization of BALB/c mice with hyperimmune sera against Actinobacillus pleuropneumoniae prior to administration of a mixture of serotypes resulted in the generation of numerous unique MAbs against the microorganisms in the mixture (111).

The contribution of passively administered antibody to the development of a protective antiviral immune response was demonstrated by Haigwood et al. (45). These investigators had shown previously that administration of polyclonal immune globulin with a high neutralizing titer against simian immunodeficiency virus (SIVIG) administered to macaques early in infection significantly improved the health of treated animals (46). Four of 6 treated animals maintained low or undetectable levels of viremia for more than a year after the clearance of the passive antibody compared to 1 of 10 controls, while the two rapid progressors controlled viremia only while the immune globulin was present. Interestingly, humoral immunity in long-term surviving animals that had received passive immunotherapy was found to be notably different from that of controls. Despite an 8-week delay in the formation of Env-specific antibodies in SIVIG-treated animals, production of neutralizing antibodies was significantly accelerated in animals that received SIVIG compared to that in control animals. Comparable neutralizing titers were measured at 12 versus 32 weeks in treatment versus control groups, respectively, and this accelerated response was associated with more-rapid control of post-acute-phase viremia and a delay in the onset of disease.

Taken together, the results of numerous studies reinforce the importance of considering the potential implications of modulation by antibody during development of both passive and active immunization approaches.

EFFECTS OF ANTIBODY ON ANTIBODY-MEDIATED IMMUNITY

Optimal antibody-mediated protection against a pathogen would depend on timely and sufficient induction of host antibodies of the correct specificity and isotype. There are numerous pathogens that can persist in the face of measurable immune responses and examples in which the response not only may be nonprotective but also may contribute to host damage (28). The studies documenting immunomodulation by antibody outlined above have important practical implications in that a humoral immune response against epitopes that are dominant but ineffective for protection may be decreased or eliminated, while hierarchically minor epitopes that are more relevant for protection may in turn become dominant. Depending on the system, an antibody interaction with an antigen has been shown to be able to influence the rapidity (45, 54, 61, 62), intensity (15, 16, 18, 32, 47, 75), specificity (17, 18, 32, 100, 118), and immunoglobulin isotype composition (18, 54, 63) of the subsequent antibody response to that antigen. Results are often, but not always, related to the antigen/antibody ratio, with greater amounts of antibody usually associated with a suppressive rather than enhancing effect (7, 18, 29, 73, 104). A prozone-like phenomenon has been described in passive immunization studies of MAbs against the fungal pathogen Cryptococcus neoformans. Depending on the MAb and its isotype, results were protective, nonprotective, or disease enhancing, with higher levels of antibody associated with a less-than-favorable outcome (113, 114).

Based on reports published to date, it is difficult to predict whether a given antibody will have an enhancing or suppressive effect on the magnitude or efficacy of the subsequent immune response to the antigen. Due to negative feedback mechanisms, high levels of IgG antibodies have often been associated with inhibitory effects (50, 53, 104) but enhancing effects can also be observed depending on the nature of the antigen and the interaction of antibody within IC with inhibitory versus activating FcRs (51, 104). Other factors reported to potentially, but not universally, influence outcome of immune responses have included antibody specificity (30, 82, 87), isotype (20, 32, 37, 82), affinity (53), and route of administration (18, 118).

While the effect of immunomodulation by antibody on the kinetics, amount, and isotype of an elicited humoral response may be easily measured and readily apparent, the effect on resultant antibody specificity may be less obvious and more likely overlooked. The ability to increase the number of MAb-producing hybridomas against antigen contained in an IC is an obvious indication that the specificity of the antibody response has been altered. However, when a polyclonal response against an antigen or pathogen is evaluated, the effect of the immunomodulatory antibody on specificity may be obscured by the large composite of antibodies in the mixture unless a directed effort is made to dissect their reactivities against different epitopes. In a study of murine antibody responses against the P1 surface protein of S. mutans, a candidate vaccine immunogen for dental caries, comparable levels of serum IgG and mucosal secretory IgA against S. mutans and full-length P1 were measured in certain experimental groups following immunization with and without an immunomodulatory anti-P1 MAb. It was not until P1 was subjected to partial proteolysis that a MAb-mediated shift in recognition of determinants contained in carboxy- versus amino-terminal fragments was demonstrated (18, 100). The changes in specificity and efficacy of the elicited anti-P1 response varied depending on the P1 epitope recognized by five different immunomodulatory MAbs (87). Interestingly, MAbs that were themselves inhibitory of S. mutans adherence promoted a less effective adherence inhibition response, while MAbs that themselves did not inhibit bacterial adherence promoted the formation of antibodies that did.

EFFECTS OF ANTIBODY ON CELL-MEDIATED IMMUNITY

Effective adaptive immunity and optimal host protection depend on induction of appropriate effector mechanisms with conventional thought directed toward a division between antibody-mediated humoral immunity and protection against extracellular organisms and toxins, while protection against intracellular pathogens and cancer would be expected to depend on the induction of cytotoxic CD8-positive T lymphocytes. However, antibody has demonstrated surprisingly beneficial effects against cancer and infectious agents in which cell-mediated immunity would be assumed to represent the operative protective mechanism (24, 55, 81, 86, 110). Again, the interaction of antibody with antigen appears to influence the subsequent immune response to that antigen, in this case helping to shape the nature and specificity of the cell-mediated response. Therefore, antibody itself may serve as a coordinating link between the humoral and cell-mediated branches of the adaptive immune system. An example in which a CD8-T-cell response was dependent on the presence of a natural IgM complement-activating antibody was demonstrated following vaccination of mice against visceral leishmaniasis (110). In this study, interleukin-4 secretion by CD11b⁺ CD11c^lo phagocytes, required for priming of vaccine specific cytotoxic T lymphocytes, did not occur in antibody- or complement-deficient animals and function was restored with serum from normal mice but not from Btk immune-deficient mice. Antibody has also been shown to contribute to a protective response against genital reinfection with the obligate intracellular bacterium Chlamydia trachomatis (80). In this case, ascending infections were increased in FcγR knockout mice compared to that in immunocompetent wild-type controls. In addition to promoting macrophage killing of infected epithelial cells by antibody-dependent cellular cytotoxicity, antichlamydial antibodies were shown to participate in protection by enhancing the induction of a Th1 response in FcR^+/+ mice compared to FcR^−/− mice. In vitro, these antibodies increase the rate of TH1 activation by Fcr^+/+, but not FcR^−/−, antigen-presenting cells.

The presentation of peptides derived from exogenous rather than cytosolic antigens by MHC class I is referred to as cross-presentation and is widely regarded as the mechanism for inducing a cytotoxic CD8-T-cell response against pathogens that do not necessarily infect antigen-presenting cells. Several studies have now pointed to the contribution of antibody to DC maturation and MHC class I-restricted presentation of antigen after uptake of immune complexes. Following uptake of IC containing ovalbumin and polyclonal IgG, Regnault et al. observed FcγR-dependent DC maturation indicated by increased expression of MHC class II, CD86, and CD40 (98). Unlike MHC class II-restricted antigen presentation following IC uptake, presentation of exogenous antigens in IC by MHC class I was limited to DC and was not observed using B cells as antigen-presenting cells. MHC class I-restricted presentation of peptides derived from the IC required the γ chain of FcγR, proteosomal degradation, and functional TAP1-TAP2. More recently, Gil-Torregrosa et al. reported that while IC are taken up much more efficiently than soluble antigen by DC, enabling cross-presentation of antigens at far lower and more physiologic concentrations, the process is tightly regulated and effective during only a limited time window (41). In this system, FcγR-mediated cross-presentation of IC containing OVA was enhanced during early DC maturation but down-regulated as the DC fully matured. The potential therapeutic relevance of increased cross-presentation has now been demonstrated. Enhancement of antigen processing and cross-presentation by nonneutralizing antibody resulted in the generation of a more beneficial CD8-T-cell response during a lentiviral infection compared to antigen alone (119). MHC class I cross-presentation following IC uptake was increased and skewed toward a known protective epitope of SIV Gag p55 in SIV-infected rhesus macaques by anti-p55 IgG in an FcR-dependent manner. The enhancement required both proteosomal and endosomal pathways and was inhibited by CD4 T-cell depletion.

POTENTIAL MECHANISMS OF IMMUNOMODULATION BY ANTIBODY

Although the factors that dictate whether an antibody complexed with antigen will change an immune response are not fully understood, numerous potentially overlapping mechanisms have been suggested (53, 55, 84, 110). Immunomodulation by antibody can be Fc-dependent or independent and can include increased uptake of antigen via FcR on antigen-presenting cells (66, 73), differential engagement of stimulatory versus inhibitory FcR (40, 51, 53, 104, 126), FcR-dependent enhancement of MHC class I-restricted cross-presentation (41, 98, 119), alterations in proteolysis and antigen processing (6, 72, 73, 103, 121), a shift in presentation of class II-restricted T-cell determinants (4, 5, 70), changes in cytokine expression by antigen-presenting cells and/or T cells (3, 4, 11, 12), masking of dominant epitopes by antibody (6, 9, 10, 14, 72, 121), exposure of cryptic epitopes induced by antibody binding (61, 103, 121), enhanced germinal center formation and generation of strong recall responses (53, 59, 60, 62, 64, 91, 108), changes in usage of germline-encoded V_H genes (85, 109), and induction of somatic hypermutation (85, 108, 109).

In addition to the effect of antibody on FcR-dependent cross-presentation by MHC class I outlined earlier, exogenous antibody can also substantially influence the induction of CD4-T-cell responses via MHC class II. Although the proteases and processing sites are not well understood (122), proteases perform two key functions in the class II MHC antigen-processing pathway, initiation and removal of the invariant chain chaperone for MHC class II and generation of peptides from foreign and self peptides for capture and display to T cells (123). Native and destabilized protein antigens vary with regard to immunogenicity (23, 33, 83, 106, 107), and destabilization of structure and increased susceptibility to proteolysis are associated with exposure of cryptic epitopes and a stronger and broader helper T-cell response (23, 33, 107). An immumodulatory MAb has now been shown to increase the rate and degree of antigen proteolysis in vitro (100). Presentation of particular antigen-specific T-cell determinants can be enhanced or suppressed as a direct consequence of antibody modulation of antigen processing (5, 6, 73, 103, 121). In one example, T-cell determinants within tetanus toxoid were substantially altered as a direct consequence of antibody modulation of antigen processing in human B lymphoblastoid cells, and a single bound antibody or its Fab fragment simultaneously enhanced presentation of one T-cell determinant by more than 10-fold while strongly suppressing presentation of a different T-cell determinant (103). Alteration of only a single processing site was found to shift remarkably the spectrum of tetanus toxoid epitopes displayed to the T-cell repertoire (5). The prediction by Lanzavecchia (65) that changes in the specificity of T-cell epitopes would modulate the fine specificity of an antibody response was borne out in studies of conformational epitopes of E. coli β-galactosidase (74). Therefore, changes at the T-cell level would be expected to influence the spectrum of antibodies elicited during a polyclonal response and would include not only antibodies that recognize linear epitopes but also more-complex conformation-dependent determinants.

The formation of GC in which memory B cells are generated is facilitated by trapping of IC and activation of complement in the network of follicular DC and/or B cells in the lymphatic follicles (53). Therefore, the immunomodulatory potential of an antibody would also be related to its ability to activate complement. MHC class II antigen presentation is influenced by the endocytic compartment used for processing of internalized antigen. Opsonization of antigen by C3b and uptake via complement receptors have now been shown to alter intracellular trafficking of internalized antigen compared to uptake via the B-cell receptor in B lymphocytes (88). Again, the resultant change in epitopes favored for display to CD4 T cells by MHC class II would ultimately influence the nature and specificity of the elicited response. Immunization with IC has been reported to accelerate the development of B memory cells, the formation of GC, and the maturation of antibody affinity compared with immunization by antigen alone (62). In studies designed to analyze how immunization with IC can alter the repertoire of antigen-reactive B cells at the molecular level, the rearranged Ig heavy chain variable (V_H) genes from mouse splenic GC were examined (85, 109). While most of the GC B cells in mice that received antigen alone expressed a single variable region gene, B cells of mice immunized with an antigen-MAb complex demonstrated heterogeneous V_H gene expression, including nine different germ-line segments. The frequency of somatic mutations within individual GC from mice primed with IC was also greater than that from antigen immunized mice. Thus, evidence continues to mount that an antibody's interaction with an antigen can serve to change and increase the diversity of the response to that antigen.

SUMMARY AND PERSPECTIVE

When an antigen is encountered as part of an immune complex with antibody, quantitative and qualitative changes in the response occur compared to exposure to the antigen alone. In addition to using immunization with IC as a model to study the regulatory effects of antibody, researchers are now capitalizing on immunomodulatory properties of antibody to generate useful laboratory reagents and, more exciting, to redirect host defense against infectious agents and tumors toward increased efficacy and improved protection. Modern hybridoma and genetic engineering technologies and the ability to humanize antibodies will undoubtedly facilitate the generation of antibodies with desired characteristics and alleviate cross-species concerns. There are still many diseases for which active vaccination remains elusive, likely in part because the immunodominant response against the agent is not optimal for its clearance. Effective vaccines do not necessarily replicate the natural immune response to a pathogen (27), and immunomodulation by antibody represents a versatile tool to shift the balance in the host's favor. There is also a large resurgence of interest in passive immunization-based therapies (21, 25, 38, 39, 49, 71, 78, 79, 101, 112, 124); therefore, understanding the way exogenous antibodies influence an immune response either deliberately or inadvertently is of broad and practical clinical relevance. In summary, a growing number of studies point to the utility of employing antibody to redirect and optimize antimicrobial host responses and place us on the threshold of expanding this potentially powerful and often unrecognized application.