Journal Logo



Marsh, James E.2; Farmer, Christopher K. T.2; Jurcevic, Stipo2; Wang, Yi3; Carroll, Michael C.4; Sacks, Steven H.2,5

Author Information



Alloantibodies arise from prior transplantation, blood transfusion, or pregnancy. They can be detected in 12–60% of recipients of a kidney transplant and are associated with at least double the incidence of acute rejection, increased chronic rejection, and worse long-term graft survival (1). Furthermore, patients anticipating transplantation with pre-existing anti-MHC antibodies have increased waiting time. It is likely that strategies to inhibit the production of these antibodies may be therapeutically useful.

Pepys (2) originally demonstrated that mice depleted of complement with cobra venom factor had an impaired antibody response to a T-dependent antigen. Subsequent studies in humans (3,4), dogs (5), and guinea pigs (6,7) with spontaneous deficiencies of complement highlight a role for complement in the generation of T-dependent antibodies. The development of knockout mice deficient in the complement components C1q (8), C3, C4 (9,10), or the complement receptors CR1 and CR2 (11,12) has allowed these observations to be extended. Such studies have shown that complement deficiency has a major effect on antibody class switching and the generation of memory responses (6,7,10–12). Recently the cellular and molecular mechanisms underlying the influence of complement on immune function have been investigated. Complement-deficient mice have a defect in localization of antigen to lymph nodes and in antigen presentation (13–16). Dempsey et al. (17) have shown a profound effect of C3d-coupled antigen directly on the B-cell response in mice. However, despite these findings, the physiological significance of complement mediated enhancement of the antibody response is unclear, because in general the experiments described were performed with threshold doses of antigen, and it has been suggested that the effect of complement deficiency can be overcome by increasing the dose of antigen (6,7,10), thus questioning the influence of complement on antibody production for antigen doses encountered in real life.

Recent studies have drawn attention to the role of complement in acute allograft rejection. Animals treated with complement inhibitors, or those with a deficiency in different complement components, demonstrate prolonged survival of skin and primarily vascularized grafts (2,18–21). The mechanisms underlying the influence of complement are likely to be complex. Complement-inhibited or -depleted transplant recipients demonstrate reduced ischemia-reperfusion injury (22), vessel wall inflammation (18), and altered recruitment of effector cells (18). The effect on the production of donor-specific antibodies is less clear. Studies investigating the terminal complement components have failed to demonstrate any role in alloantibody production (21,23). In contrast, Pratt et al. (24) demonstrated reduced alloantibody production in rat renal allograft recipients treated with the human complement inhibitor, soluble complement receptor type 1. However, this effect was transient, lasting approximately 7 days, coinciding with the production of antibodies to the inhibitor, leaving some unanswered questions on the role of complement in the alloantibody response. First, what is the duration of any effect of complement on alloantibody production? Second, can the effect of complement deficiency be easily saturated? These issues should be clarified before any attempt at therapeutic complement inhibition is considered.

In this study, we have investigated the alloantibody response in complement-deficient mice immunized by fully major histocompatibility complex (MHC)-disparate skin grafts. We originally set out to investigate threshold responses, anticipating that a tissue graft would provide exposure to a large antigenic load over an extended period, which might saturate the effect of complement deficiency on the immune response. However, we report marked impairment of the allospecific immunoglobulin (Ig)G response in mice with deficient classical pathway function. We conclude that complement activation has an important influence on the alloantibody response against tissue antigens and that this may have repercussions regarding the control of alloantibody responses in man.


Skin Grafting

Tail skin from the donor animals was grafted onto the left flank of the recipients under enflurane anesthetic according to a modified version of the protocol used by Billingham and Medawar (25). The graft site was covered with paraffin embedded gauze and a dressing of dry gauze and clear tape. The dressing was removed on day 7, and the graft was assessed visually on a daily basis for signs of rejection. The time to rejection was determined as the day on which >90% of the graft area was necrotic.


The skin grafting was performed across a full class I and class II MHC barrier. Donor B10.BR (H2k), recipient C57BL/6 (H2b) and 129/Ola (H2b), and third party BALB/c (H2d) mice were purchased from Harlan, UK. In addition, a cross of C57BL/6×129/Ola at the F2 generation were bred at our animal facility to produce animals of a mixed genetic background (wild type F2 mice). The C3-/- and C4-/- (H2b) mice were derived by homologous recombination in embryonic stem cells as previously described (9,10). These mice have undetectable levels of C3 or C4, respectively (9,10). B10.D2nsn and B10.D2osn (H2d) mice were purchased from Jackson Laboratories, Bar Harbor, ME. Female, 6- to 8-week-old mice were used throughout the experiments.

Flow Cytometry

Blood was obtained from recipient mice by tail bleeding, allowed to stand at room temperature for 2 hr, spun in a microcentrifuge, and serum collected and stored at −20°C until use. The presence of donor specific alloantibodies was assessed by flow cytometry using donor strain T cells as the target cells. For this, spleens were harvested from naive donor strain mice and splenocytes prepared by separation over a ficoll gradient (Sigma, Poole, UK). The splenocytes were resuspended in 50 μl of Hanks Balanced Buffered Solution (HBBS) at a concentration of 1×107/ml and incubated with 10 μl of recipient serum. Optimal antibody binding was achieved with neat serum, which was used throughout the experiments unless specifically indicated. The splenocytes were then incubated with secondary antibodies specific for IgM (Jackson Immunoresearch Laboratories, West Grove, PA), IgG, IgG1, IgG2a, IgG2b, or IgG3 (Serotec, Kidlington, UK), conjugated with fluorescein, followed by incubation with a phycoerythrin-conjugated antibody specific for CD3 (Serotec). Each incubation was performed at 4°C for 30 min and followed by washing three times in 2 ml of phosphate-buffered saline (PBS) with 1% bovine serum albumin. The cells were then fixed in 400 μl of 1% paraformaldehyde in PBS and analyzed by two-color flow cytometry (FACScan flow cytometer, Becton Dickinson, Mountain View, CA) gated on the T cells. The median fluorescence of each test sample was compared with that obtained from serum of a naive unprimed mouse, and the relative median fluorescence was determined. Specificity of alloantibody binding was assessed by testing the serum against third party strain splenocytes (BALB/c). A positive control was obtained using monoclonal anti-Kk antibody (Serotec) as the primary antibody in place of recipient serum. The titre of each sample was assessed by serial dilution of the test serum.

Lymphocytotoxicity Assay

Functional binding of alloantibody was assessed by a lymphocytotoxicity assay according to a modified version of the method described by Mittal et al. (26). In brief, donor strain splenocytes were suspended in HBBS at a concentration of 1.5×107/ml. Terasaki plates were coated with liquid paraffin and 1 μl of the cells added to each well and incubated with 1 μl of doubling dilutions of recipient serum for 15 min. A further incubation with 5 μl of rabbit complement (Quest Biomedical, Solihul, UK) for 15 min was performed, and the cells were then double stained with ethidium bromide and acridine orange to assess cell viability. The proportion of dead cells was counted by viewing under ultraviolet illumination.

Hemolytic Complement Assay

Functional activity of the alternative pathway of complement activation was assessed by a hemolytic complement assay according to the method described by Van Dijk et al. (27) and Klerx et al. (28). Fifty microliters of rabbit erythrocytes (Harlan Sera, UK) were suspended in 100 μl of veronal buffered saline at a concentration of 1.5×108/ml and incubated in a 96-well plate with serial dilutions of mouse serum and zymosan (Sigma) for 1 hr at 39°C. The plate was then centrifuged for 10 min, the supernatant collected, and the optical density (OD) at 405 nm assessed in a spectrophotometer (Dynatech, UK). The percentage of hemolysis at each serum concentration was determined according to the following formula:MATH

By exposing the red cells to water instead of veronal buffered saline, 100% hemolysis was obtained, and 0% hemolysis was determined by exposing the erythrocytes to zymosan in the absence of serum.

Inhibition of the Terminal Pathway of Complement Activation

Murine monoclonal antimouse C5 antibody (BB5.1) was used to inhibit the production of both C5a and C5b-9 (29,30). An isotype-matched (IgG1) monoclonal antibody (135.8) was used as control (29,30).

Lymphocyte Proliferation Assay

The proliferative response of alloreactive T cells was assessed by a mixed lymphocyte culture. Spleens were harvested from the recipient mice 11 days after skin grafting, and the splenocytes were isolated and cultured in 96-well plates in Roswell Park Memorial Institute (RPMI)-1640 medium supplemented with 10% fetal calf serum (FCS) (Helena Bioscience, Sunderland, UK), 50 μM of 2-mercaptoethanol, 100 U/ml of penicillin, 100 μg/ml of streptomycin, and 10 μM of HEPES at 2×105 cells/well. At 37°C, 5% CO2 in a humidified atmosphere, 2×105 irradiated (2000 Rad) stimulator cells were co-cultured with these cells. After 72 hr of incubation, 0.5 μCi of [3H]thymidine (Amersham, Amersham, UK) was added to each well, and 24 hr later, the cells were harvested and the incorporated radioactivity measured on a liquid scintillation counter. Each sample was analyzed in triplicate.

Assay for Interferon-γ Producing T Cells

After skin grafting, the number of T cells producing interferon-γ (IFNγ) after stimulation with MHC-disparate splenocytes was assessed by an ELISPOT assay. 96-well filtration plates (Multiscreen, Millipore, Watford, UK) were coated with 100 μl of monoclonal rat antimouse anti-IFNγ antibody (R&D Systems, Abingdon, UK) at a concentration of 5 μg/ml. After washing with sterile PBS, the plates were blocked with 1% bovine serum albumin/PBS. Splenocytes from donor, recipient, and third party strains (BALB/c) were prepared and suspended at a concentration of 2×106/ml in RPMI-1640 medium supplemented with 10% FCS, 50 μM of 2-mercaptoethanol, 100 U/ml of penicillin, 100 μg/ml of streptomycin, and 10 μM of HEPES. Irradiated (2000 Rad) stimulator cells and 2×105 responder cells were placed in each well and incubated for 48 hr at 37°C, 5% CO2. The plates were then washed once with distilled water to achieve cell lysis and then six times in PBS/0.05% Tween. A biotinylated polyclonal antimouse IFNγ antibody (R&D Systems) was then added (100 μL/well at a concentration of 200 ng/ml), and the plates were incubated at 4°C overnight. The plates were further washed, and 100 μl of streptavidin alkaline phosphatase was added. BCIP/NBT alkaline phosphatase (Sigma) substrate was added to each well, and the resultant spots counted by using a ×5-dissecting microscope. Each sample was analyzed in triplicate. A positive control was obtained by incubating the responder cells with concanavalin A, and negative controls were obtained by exposing the responder cells medium alone.

Statistical Analysis

The analysis of skin graft survival was assessed by the Mann-Whitney U test and relative antibody production by analysis of variance.


Complement-Deficient Mice have an Impaired Allospecific IgG Response

We first examined the overall impact of the complement cascade on the alloimmune response by grafting C3-/- mice and C3 sufficient C57BL/6 mice with allogeneic donor skin (B10.BR). The C3-/- mice and C4-/- mice exhibited a marked, sustained, and almost identical reduction in amplitude of the donor-specific IgG response up to 84 days after skin grafting, at which point the mice were killed (Fig. 1). This indicates that the classical pathway of complement activation was important for augmenting alloantibody production.

Figure 1
Figure 1:
Donor-specific IgG alloantibody response in C57BL/6, C3-deficient, and C4-deficient mice after a B10.BR skin graft. Flow cytometric analysis was gated for T cells. The reactivity of each neat serum sample is expressed as the mean value (±95% confidence interval) of the relative median fluorescence of the test serum compared with naive C57BL/6 serum against donor strain (B10.BR). A positive control anti-Kk antibody diluted at 1/10 produced a relative median fluorescence of 45. Control samples using test serum or anti-Kk antibody with third party (BALB/c) or recipient (C57BL/6) strain target cells produced a relative median fluorescence of 1. The wild type sera had a higher relative median fluorescence at days 28, 56, and 84 (P <0.0001) than either the C3-/- or C4-/- sera.

Because the C3-/- mice and C4-/- mice were derived from 129×B6 chimeras, it was possible that the introduction of 129 genes into the B6 strain led to the reduction in the antibody response observed in the complement-deficient mice. Although 129 mice share the same H2 type as B6 mice (H2b), they differ at minor histocompatibility loci. We therefore directly compared the response in 129/Ola and B6 mice. In fact, 129/Ola recipients displayed an exaggerated antidonor IgG response compared with B6 recipients (Fig. 2A). Furthermore, complement sufficient 129×B6 F2 generation mice also demonstrated elevation of the antibody response compared with B6 recipients of B10.BR skin grafts (Fig. 2B), suggesting that the addition of 129 genes conferred hyperresponsiveness compared with the B6 strain. We conclude that the hyporesponsiveness displayed by C3-/- mice was most unlikely to be due to contamination of the B6 strain by 129 genes.

Figure 2
Figure 2:
Donor-specific IgG response in recipients of B10.BR skin grafts. Neat serum from different recipient strains of mice was incubated with B10.BR splenocytes. Binding was assessed by two-color flow cytometry at different time points after grafting. The relative median fluorescence (±standard error) is expressed as the median fluorescence obtained from incubating B10.BR splenocytes with recipient serum and compared with that obtained using naive C57BL/6 serum. (A) C57BL/6 mice compared with 129/Ola mice. (B) C57BL/6 mice compared with 129×B6 F2 mice.

The F2 129×B6 mice probably represent the closest control strain for the C3- or C4-deficient mice, because the F2 progeny possess on average 50% B6 genes and 50% 129 genes, in a random combination. As shown in Figure 3, direct comparison of C3+/+ F2 129×B6 with C3-/- recipients again showed that a reduction of the allospecific antibody response was associated with C3 deficiency.

Figure 3
Figure 3:
Donor-specific IgG response in recipients of B10.BR skin grafts. Serum from C3-sufficient (C57BL/6×129 F2; n=11), C3-deficient (n=10), and C4-deficient (n=13) recipients was incubated with naive B10.BR splenocytes. Binding was analyzed by two-color flow cytometry at different time points after skin grafting. The relative median fluorescence is expressed as the median fluorescence obtained from incubating B10.BR splenocytes with recipient serum compared with that obtained using naive C57BL/6×129 F2 serum. (A) The mean±standard error of the relative median fluorescence for each serum sample diluted 1/10. A second B10.BR skin graft was placed on the opposite flank on day 35 after the first graft. The response in complement sufficient mice was greater at each time point than in either C3- or C4-deficient mice (P <0.0001). There was no significant difference between the responses in the C3- and the C4-deficient mice. The relative median fluorescence obtained from incubating B10.BR (H2k) splenocytes with anti-Kk monoclonal antibody before incubation with secondary antibodies was 45. The relative median fluorescence obtained from incubating third party BALB/c naive splenocytes with recipient serum or anti-Kk antibody was 1.0. (B) Titration of wild type and C3-/- sera at 14 days after grafting. (C) Titration of wild type and C3-/- sera at 21 days after grafting.

Furthermore, there was a profound deficiency in second set responses in C3-/- and C4-/- mice. Five weeks after receiving their first B10.BR allograft, a second B10.BR graft was placed on the opposite flank. The wild type mice demonstrated an accelerated and augmented IgG response after the second graft, but IgG production remained subdued in the complement-deficient mice (Fig. 3). Even when the exposure to alloantigen is increased, complement still has a profound effect, suggesting that the effect of complement deficiency on alloantibody production is not easily saturated.

The IgM Response is not Impaired by Complement Deficiency

Because complement may play an important role in antibody class switching, we analyzed the antidonor IgM response. Despite the marked impairment of allospecific IgG in complement-deficient mice, the IgM response at days 14 and 21 after grafting was not significantly decreased, as shown in Figure 4. This concurs with other studies, in which complement seemed to have greater effect on the generation of immunological memory than on the primary immune response (6–8,10,12).

Figure 4
Figure 4:
The allospecific IgM response in mice transplanted with MHC-disparate skin. Neat recipient serum was incubated with donor strain splenocytes, and two-color flow cytometry was performed to detect donor-specific IgM binding to T cells. The mean±standard error of the relative median fluorescence is shown. The response in B6, C3-/-, and C4-/- mice was compared. No significant difference in binding was observed at 10 or 14 days.

Influence of Complement on Circulating Antidonor IgG Isotype

The experiments described above indicate that complement deficiency led to a defect in B-cell class switching during maturation of the immune response against donor antigen. In view of the functional importance of IgG isotype, we investigated the isotypes generated in C3-deficient animals, compared with complement sufficient (129×B6) F2 controls. Whereas wild type mice demonstrated several circulating antidonor IgG isotypes, in particular the strongly complement-fixing IgG2a, only IgG1 could be detected with any certainty in the C3 mutant mice (Fig. 5). In contrast to IgG2a, murine IgG1 has little or no complement-fixing ability (31–33).

Figure 5
Figure 5:
IgG isotypes produced in C3-sufficient and C3-deficient mice. Neat serum from the recipient mice was assessed for binding to donor splenocytes by two-color flow cytometry 21 days after MHC-disparate skin grafting. The data represent the mean±standard error of the relative median fluorescence for each set of serum compared with binding of naive mouse serum with donor strain splenocytes. The relative median fluorescence of all sera tested for all the isotypes against third party BALB/c splenocytes was 1.0. There was no significant difference in the production of IgG1 between the two groups (P =0.32), nor of IgG2b. The complement sufficient animals produced more IgG2a and IgG3 than the C3-/- mice (P <0.0001).

Functional Differences in Circulating Antidonor IgG

To evaluate the functional implications of the reduced binding level of serum IgG in complement-deficient mice, we examined antibody mediated cell lysis in the presence of heterologous complement (Fig. 6). The serum dilution, at which 50% lysis was observed, was 1 in 11 for the C3-deficient mice and 1 in 194 for the C3 sufficient mice. This difference is likely to reflect both the reduced total amount of IgG and defective class switching to complement-fixing IgG isotypes (particularly IgG2a) in the complement-deficient mice.

Figure 6
Figure 6:
Lymphocytotoxicity against donor strain splenocytes, in the presence of normal rabbit complement, with recipient serum 21 days after a B10.BR skin graft. The wild type animals were the 129×C57BL/6 F2 mice. Pooled serum from all the recipients (n=9 in both groups) was used and analyzed in triplicate. The C3-deficient serum is shown with the dotted line and the wild type serum with the solid line. The mean±standard error of percent hemolysis is shown for each dilution. The normal serum was obtained from a pool of naive 129×C57BL/6 F2 serum.

C5 is Unimportant in Augmenting the Alloantibody Response

Whereas previous work has concentrated upon the role of the C3 and C4 in the regulation of the immune response, the role of C5 and its products have been less well documented. The congenic mouse strains B10.D2osn and B10.D2nsn are respectively C5-deficient and C5-sufficient (34). We assessed the alloantibody response of these strains of mice (H2d) after grafting with B10.BR skin. At 14 and 21 days after grafting, there was no reduction in IgM, or IgG binding (data not shown), with donor strain splenocytes in the C5-deficient mice. Therefore, it is unlikely that the terminal pathway components are important in alloantibody production.

It is possible that in B10.D2 mice, the antibody response was less complement-dependent compared with the B6 strain. This strain-combination difference could have explained the lack of effect of C5 on the antidonor response. To control for this, we examined the effects of C5 inhibition in the 129×B6 F2 strain, using BB5.1 monoclonal antibody directed against mouse C5. BB5.1 completely inhibits the production of C5a and C5b-9, by binding to C5 (29). A single intraperitoneal injection of 0.75 mg of BB5.1 completely inhibited the total hemolytic complement activity of the F2 mice for at least 6 days (data not shown), confirming the effect reported in other strains (29,35). In contrast, an isotype-matched control antibody (135.8) had no effect.

C3 sufficient 129×C57BL/6 F2 were grafted with B10.BR skin, and treated twice weekly with injections of 0.75 mg BB5.1 or isotype control from the day before skin grafting until sacrifice at day 28. None of the C5-inhibited mice had hemolytic complement activity at the end of the experimental period (data not shown). Despite this, the allospecific IgG response was not significantly different in those mice receiving the active or control antibody, at any of the time points assessed (data not shown). These data confirm the results in B10.D2 mice; namely that C5a and C5b-9 do not have a major role in the regulation of the antidonor antibody response.

Complement Deficiency is Associated with an Impaired Antidonor T cell Response

As a prelude to further studies, we explored the possibility of an impaired T cell response in C3-deficient animals, by evaluating the T cell cytokine response (Fig. 7) and the T cell proliferative response (Fig. 8), after re-stimulation with donor antigen. At 8 days after skin grafting, C3 +/+ mice yielded nearly twice as many IFNγ-producing T cell clones, compared with C3 -/- mice, after re-stimulation by donor strain splenocytes in vitro. Similarly, the proliferative response showed evidence of defective T cell priming in C3 -/- mice, compared with 129×C57BL/6 F2 controls. Thus, it is implied that the T cell response against donor antigen is less well primed in the presence of defective complement activation and consequently that T cell cytokine production is diminished.

Figure 7
Figure 7:
Lymphocyte production of IFNγ after stimulation with allogeneic cells. Eight days after a B10.BR skin graft, splenocytes were harvested from 129×C57BL/6 F2 wild type or C3-/- mice and restimulated with either donor strain (B10.BR) or third party (BALB/c) splenocytes in culture. The production of IFNγ by alloreactive T cells was assessed by ELISPOT assay, with each IFNγ-producing clone resulting in a spot visible under a ×5 microscope. The graphs show the number of spots counted after 48 hr of culture. Each experiment was performed in triplicate, and the mean count±standard error is shown. Cells not exposed to any stimulator cells produced no spots. Those stimulated with concanavalin A produced >100 spots/well.
Figure 8
Figure 8:
Lymphocyte proliferation upon restimulation of recipient splenocytes in culture. Eleven days after receiving a B10.BR skin graft, recipient mice were killed and splenocytes were harvested and restimulated in vitro with donor strain or third party (BALB/c) irradiated splenocytes. Proliferation was measured by [3H]thymidine uptake. Data bars show the proliferative responses of splenocytes from wild type mice (129×C57BL/6 F2) and C3-/- mice, compared with those from naive mice and with concanavalin A-stimulated cells. Unstimulated splenocytes gave a background count of 3865.

Complement Deficiency is Associated with Prolonged Acceptance of Skin Grafts

It was not the primary intention of this study to evaluate graft survival, because the mechanisms contributing to graft survival are complex and complement is likely to have a direct effect on the graft as well as an influence on the production of antidonor antibodies. In all of the experiments, the C3- and C4-deficient recipients demonstrated a small (2–3 days) but statistically significant (P <0.0001) prolongation of graft survival (data not shown) compared with the complement sufficient controls. However, C5-deficient or -inhibited mice displayed similar graft survival to their respective controls, not supporting an effect of either C5a or C5b-9 on the rejection response against skin grafts. Furthermore, after a second skin graft, wild type mice demonstrated accelerated graft rejection, but C3-deficient mice failed to demonstrate altered kinetics compared with their first graft.


The findings reported here show that mice unable to activate complement beyond C3 have a severely impaired antibody response against donor tissue antigens. Defective classical pathway activation (C4 deficiency) was sufficient to explain the degree of impairment. Although these findings do not exclude an effect of the alternative pathway effect, the lack of additional effect of deficiency of C3 over that of C4, and the lack of effect in C5-deficient or -depleted animals, indicates that the alternative and terminal pathways are unlikely to contribute a major effect to the response described.

The striking observation from these experiments is the magnitude and the duration of the effect of complement on the antidonor response. Studies of other T-dependent antibody responses against proteins, such as against ovalbumin and hen egg lysozyme, similarly have shown an influence of complement. However, at increased antigen doses, the influence exerted by complement was often lost. For instance, Bottger et al. (6,7) noted that the profound impairment in memory responses in complement-deficient guinea pigs could be partially overcome by increasing the antigen dose by only a factor of 2 and completely abrogated by increasing it by a factor of 10. Although the dose of antigen cannot be accurately described in our experiments, skin grafting provides a stringent antigenic stimulus leading to rapid and complete rejection. It is therefore remarkable that the response was markedly complement dependent, as shown in both functional (hemolytic) and binding assays. Furthermore, the response to a second graft remained complement dependent.

The results in this study, therefore, not only extend (to alloantigen) the range of T-dependent antigens against which the immune response is complement mediated but also provide strong experimental evidence of relevance in a pathological setting. There are at least three ways in which the profound effect of complement in these experiments might be explained. First, the inflammatory response against the graft might have served to enhance the specific immune response, as predicted by Janeway (36) and later Matzinger (37). That is, local inflammation mediated by complement might have increased the costimulatory capacity of antigen-presenting cells (APC) and have led to a more vigorous immune response. Second, antigen leaving an inflamed site might be more strongly coated by complement fragments and thus more efficiently targeted to the follicular dendritic region of the secondary lymphoid compartment (16) and/or more effectively engage with antigen specific B cells (13). Thus, compared with the delivery of soluble antigen administered with or without adjuvant, antigen delivery from an inflamed site might employ complement at more than one step of immune activation. Third, the route of antigen delivery may have made a difference to the immune response. Whereas previous studies have used bolus dosing of soluble antigen, it takes several days for the skin graft to become vascularized. Exposure of the innate and specific components of the recipient immune system to foreign antigen will therefore differ in skin grafts. Moreover, there may be differences in the mode of presentation of alloantigen and conventional T-dependent antigen. Alloantigen may also directly activate the recipient immune system, whereas conventional antigen relies upon antigen processing and subsequent presentation. It is possible that unprocessed alloantigen present on donor cells (direct presentation) is more readily opsonized by complement, compared with conventional antigen fragments processed and presented by the host cells (indirect presentation).

The main defect of the antibody response in complement-deficient mice seemed to be in the ability of donor-specific B cells to undergo normal antibody class switching. We observed no defect in the initial production of serum IgM antibodies. It is therefore possible that this represents a maturation arrest of B cells in the absence of full stimulation, that is, they remain in an IgM secreting phase. Although we found a marked reduction in the IgG2a and IgG3 antidonor responses, there was no significant impairment in the production of the IgG1 isotype in C3-deficient mice. IgG1 has little or no complement fixing activity in mice. Thus, in the absence of C3- or C4-mediated stimulation of the B cells, antibody generation is directed toward a noncomplement fixing isotype. This is consistent with data obtained by Cutler et al. (8) who observed that C1q-deficient mice had a defect in the production of IgG2a and IgG3 antibodies but not IgG1 after injection with keyhole limpet hemocyanin. The mechanism for this differential effect on IgG subclass generation is not clear. It may be speculated that because IgG1 effector function is independent of complement, the production of IgG1 is not linked to a complement-sensitive mechanism.

In addition to an effect on B cells, skin transplantation elicited a marked depression of T cell priming in C3-deficient animals. This is consistent with a stimulatory role of complement on alloimmune T cell response, similar to that described for other T-dependent antigens (6–9). Cooperation between B cells and T cells offers a possible explanation. B cells function as APC, and their expression of costimulatory molecules, such as B-7, is increased by coligation of the CR2 and B cell antigen receptors (38). Accordingly, interaction of C3 fragments with CR2 on B cells might enhance the capacity of B cells to stimulate donor-specific T cells. Thus, the depression of the specific alloreactive T cell response in the absence of complement could reflect a deficit in B cell activation and the ability of B cells to recruit T cell help (38,39). Alternatively, there could be a primary defect of T cell activation, as mentioned in the preceding discussion.

Our finding of a depressed T cell cytokine and proliferative response is consistent with work by Cutler et al. (8), who showed that primed T cells from classical pathway (C1q)-deficient mice produced less IFNγ on exposure to keyhole limpet hemocyanin. However, T cell priming against the phage antigen φX 174 was unaffected in C3- and C4-deficient mice (10). The reason for this discrepancy is unclear. However, it may partly reflect differences in the route of antigen delivery and immune stimulation, as well as differences in the extent of local inflammation, as proposed above.

In conclusion, we have shown that the generation of anti-MHC antibodies against skin graft antigens is strongly dependent on the classical pathway of complement activation. Moreover, the results strongly suggest that this requirement for complement (which may be exaggerated by local inflammation, in which complement is a participant) may be relevant in a pathological setting. This has important implications regarding the prevention of alloimmunity in clinical transplantation, where the production of anti-HLA antibodies has a significant impact on patient selection, management, and treatment outcome. These studies provide additional impetus to investigate the therapeutic use of complement inhibitors in transplantation, with particular regard to the prevention of sensitization of naïve recipients.


1. McKenna RM, Takemoto SK, Terasaki PI. Anti-HLA antibodies after solid organ transplantation. Transplantation 2000; 69: 319.
2. Pepys MB. Role of complement in induction of the allergic response. Nat New Biol 1972; 237: 157.
3. Jackson CG, Ochs HD, Wedgwood RJ. Immune response of a patient with deficiency of the fourth component of complement and systemic lupus erythematosus. N Engl J Med 1979; 300: 1124.
4. Finco O, Li S, Cuccia M, Rosen FS, Carroll MC. Structural differences between the two human complement C4 isotypes affect the humoral immune response. J Exp Med 1992; 175: 537.
5. O’Neil KM, Ochs HD, Heller SR, Cork LC, Morris JM, Winkelstein JA. Role of C3 in humoral immunity: defective antibody production in C3-deficient dogs. J Immunol 1988; 140: 1939.
6. Bottger EC, Metzger S, Bitter-Suermann D, Stevenson G, Kleindienst S, Burger R. Impaired humoral immune response in complement C3-deficient guinea pigs: absence of secondary antibody response. Eur J Immunol 1986; 16: 1231.
7. Bottger EC, Hoffmann T, Hadding U, Bitter-Suermann D. Influence of genetically inherited complement deficiencies on humoral immune response in guinea pigs. J Immunol 1985; 135: 4100.
8. Cutler AJ, Botto M, van Essen D, et al. T cell-dependent immune response in C1q-deficient mice: defective interferon gamma production by antigen-specific T cells. J Exp Med 1998; 187: 1789.
9. Wessels MR, Butko P, Ma M, et al. Studies of group B streptococcal infection in mice deficient in complement component C3 or C4 demonstrate an essential role for complement in both innate and acquired immunity. Proc Natl Acad Sci U S A 1995; 92: 11490.
10. Fischer MB, Ma M, Goerg S, et al. Regulation of the B cell response to T-dependent antigens by classical pathway complement. J Immunol 1996; 157: 549.
11. Ahearn JM, Fischer MB, Croix D, et al. Disruption of the Cr2 locus results in a reduction in B-1a cells and in an impaired B cell response to T-dependent antigen. Immunity 1996; 4: 251.
12. Molina H, Holers VM, Li B, et al. Markedly impaired humoral immune response in mice deficient in complement receptors 1 and 2. Proc Natl Acad Sci U S A 1996; 93: 3357.
13. Qin D, Wu J, Carroll MC, Burton GF, Szakal AK, Tew JG. Evidence for an important interaction between a complement-derived CD21 ligand on follicular dendritic cells and CD21 on B cells in the initiation of IgG responses. J Immunol 1998; 161: 4549.
14. Fischer MB, Goerg S, Shen L, et al. Dependence of germinal center B cells on expression of CD21/CD35 for survival. Science 1998; 280: 582.
15. Croix DA, Ahearn JM, Rosengard AM, et al. Antibody response to a T-dependent antigen requires B cell expression of complement receptors. J Exp Med 1996; 183: 1857.
16. Papamichail M, Gutierrez C, Embling P, Johnson P, Holborow EJ, Pepys MB. Complement dependence of localisation of aggregated IgG in germinal centres. Scand J Immunol 1975; 4: 343.
17. Dempsey PW, Allison ME, Akkaraju S, Goodnow CC, Fearon DT. C3d of complement as a molecular adjuvant: bridging innate and acquired immunity. Science 1996; 271: 348.
18. Pratt JR, Hibbs MJ, Laver AJ, Smith RA, Sacks SH. Effects of complement inhibition with soluble complement receptor-1 on vascular injury and inflammation during renal allograft rejection in the rat. Am J Pathol 1996; 149: 2055.
19. Rother U, Ballantyne DL Jr, Cohen C, Rother K. Allograft rejection in C’6 defective rabbits. J Exp Med 1967; 126: 565.
20. Glovsky MM, Ward PA, Fudenberg HH. Role of complement in guinea pig skin allograft rejection. 1. Effect of cobra venom C3 inactivator and fumaropimaric acid on rejection. Clin Immunol Immunopathol 1973; 1: 165.
21. Brauer RB, Baldwin WM III, Ibrahim S, Sanfilippo F. The contribution of terminal complement components to acute and hyperacute allograft rejection in the rat. Transplantation 1995; 59: 288.
22. Weisman HF, Bartow T, Leppo MK, et al. Soluble human complement receptor type 1: in vivo inhibitor of complement suppressing post-ischemic myocardial inflammation and necrosis. Science 1990; 249: 146.
23. Qian Z, Jakobs FM, Pfaff-Amesse T, Sanfilippo F, Baldwin WM3rd . Complement contributes to the rejection of complete and class I major histocompatibility complex: incompatible cardiac allografts. J Heart Lung Transplant 1998; 17: 470.
24. Pratt JR, Harmer AW, Levin J, Sacks SH. Influence of complement on the allospecific antibody response to a primary vascularized organ graft. Eur J Immunol 1997; 27: 2848.
25. Billingham RE, Medawar PB. The technique of free skin grafting in mammals. J Exp Biol 1951; 28: 385.
26. Mittal KK, Mickey MR, Singal DP, Terasaki PI. Serotyping for homotransplantation. 18. Refinement of microdroplet lymphocyte cytotoxicity test. Transplantation 1968; 6: 913.
27. Van Dijk H, Rademaker PM, Klerx JP, Willers JM. Study of the optimal reaction conditions for assay of the mouse alternative complement pathway. J Immunol Methods 1985; 85: 233.
28. Klerx JP, Beukelman CJ, Van Dijk H, Willers JM. Microassay for colorimetric estimation of complement activity in guinea pig, human and mouse serum. J Immunol Methods 1983; 63: 215.
29. Wang Y, Rollins SA, Madri JA, Matis LA. Anti-C5 monoclonal antibody therapy prevents collagen-induced arthritis and ameliorates established disease. Proc Natl Acad Sci U S A 1995; 92: 8955.
30. Frei Y, Lambris JD, Stockinger B. Generation of a monoclonal antibody to mouse C5 application in an ELISA assay for detection of anti-C5 antibodies. Mol Cell Probes 1987; 1: 141.
31. Jansen JL, Gerard AP, Kamp J, Tamboer WP, Wijdeveld PG. Isolation of pure IgG subclasses from mouse alloantiserum and their activity in enhancement and hyperacute rejection of skin allografts. J Immunol 1975; 115: 387.
32. Spiegelberg HL. Biological activities of immunoglobulins of different classes and subclasses. Adv Immunol 1974; 19: 259.
33. Neuberger MS, Rajewsky K. Activation of mouse complement by monoclonal mouse antibodies. Eur J Immunol 1981; 11: 1012.
34. Hammer CH, Gaither T, Frank MM. Complement deficiencies of laboratory animals. In: Gershwin MB, Merchant B, eds. Immunological defects in laboratory animals. New York: Platinum Press, 1981: 207.
35. Wang Y, Hu Q, Madri JA, et al. Amelioration of lupus-like autoimmune disease in NZB/WF1 mice after treatment with a blocking monoclonal antibody specific for complement component C5. Proc Natl Acad Sci U S A 1996; 93: 8563.
36. Janeway CA. Approaching the asymptote? Evolution and revolution in immunology. Cold Spring Harbor Symposia on Quantitative Biology, 1989: 1.
37. Matzinger P. Tolerance, danger, and the extended family. Annu Rev Immunol 1994; 12: 991.
38. Boackle SA, Morris MA, Holers VM, Karp DR. Complement opsonization is required for presentation of immune complexes by resting peripheral blood B cells. J Immunol 1998; 161: 6537.
39. Kozono Y, Abe R, Kozono H, Kelly RG, Azuma T, Holers VM. Cross-linking CD21/CD35 or CD19 increases both B7-1 and B7-2 expression on murine splenic B cells. J Immunol 1998; 160: 1565.
© 2001 Lippincott Williams & Wilkins, Inc.