|Home | About | Journals | Submit | Contact Us | Français|
Almost three decades ago murine IgG3 was proposed to interact with a different receptor than the other IgG subclasses, but the issue remains unresolved. The question of whether a specific receptor exists for IgG3 is critically important for understanding antibody-mediated immunity against Cryptococcus neoformans (Cn)3, where the different isotypes manifest profound differences in protective efficacy. In this study we revisited this question by analyzing IgG1-and IgG3-mediated phagocytosis with variable region-identical monoclonal antibodies (mAb) using mouse macrophages deficient in various receptors and in conditions of Fcγ receptor (FcγR) and complement receptor (CR) blockage with specific antibody. IgG3 was an efficient opsonin for Cn in FcγR-deficient and CD18-deficient cells and in the presence of blocking antibodies to FcγR and CR. Like IgG1, IgG3-mediated phagocytosis was associated with fungal residence in a mature phagosome that was followed by intracellular replication and exocytosis events. We conclude that a specific receptor for IgG3 exists in mice that is structurally different from the known FcγRs.
Phagocytosis is a receptor-mediated event where specific recognition of microbes by phagocytic cells, such as macrophages or dendritic cells, leads to microbial internalization and targeting to a phagolysosomal compartment for degradation and antigen presentation (1–3). During an adaptive immune response, antibody (Ab) is the primary mediator of this interaction, where microbes bound by specific IgG interact with Fcγ receptors (FcγR) on effector cells to promote clearance of infection (4, 5). Characterizing the receptor interactions during Ab-mediated phagocytosis is important for understanding the role of Ab generated during host defense as well as for assessing the mechanism of passive Ab therapy, where treatment with immune serum or specific monoclonal Abs (mAb) can ameliorate disease (6–8).
In mice, the activating FcγRs are FcγRI, FcγRIII and FcγRIV, all of which share a common γ chain containing an intracellular immunoreceptor tyrosine-based activating motif (ITAM) sequence necessary to mediate activation when Ab is bound (5). FcγRIIb, the inhibitory receptor, does not pair with the common γ chain but rather has an intracellular immunoreceptor tyrosine-based inhibitory motif (ITIM) sequence and mediates inhibitory signaling (9). The balance of positive and negative signals determines the outcome of the interaction of Ab-bound microbes with cells, as the threshold to trigger phagocytosis or other events is based on the ratio of activating to inhibitory receptor engagement (10). Because Ab isotypes have different affinities for the various FcγRs, they can trigger different effector functions based on their receptor binding specificity. It is critical to note that while in humans there is a similar system, mouse IgG isotypes and FcγRs are not synonymous with human ones, and while the specificities of human IgG isotypes for the various human FcγRs are well characterized (11, 12), the mouse system is different and less well understood. Various studies in mice have shown that IgG2a is the most promiscuous Ab, interacting with all FcγRs, whereas IgG1 is more selective and only binds to one activating receptor, FcγRIII (13). While much work has been done to study the interaction of the various Ab isotypes with the different FcγRs, major questions still remain. In this regard, three of the four mouse IgG subclasses, IgG1, IgG2a and IgG2b, have been rather well characterized in terms of their affinity and specificity for the different known FcγRs (14, 15), but the data for IgG3 has been inconsistent and most current reviews conclude that IgG3 interacts only very weakly with known FcγRs (13). Understanding the mouse FcγR system is important because mice continue to be the most commonly used animal system for immunological studies.
In 1981, Diamond and Yelton proposed that a unique IgG3 receptor existed, based on the spontaneous J774 cell variant that specifically lost the ability to phagocytose sheep red blood cells coated in IgG3, but retained the ability to phagocytose via the other antibody isotypes (16). A subsequent study by Gavin et al showed that the known receptor FcγRI, which has high affinity for IgG2a, could also interact with IgG3 (17). However, this study demonstrated only low-affinity binding of IgG3 to FcγRI-transfected cells. Additionally, in bone marrow derived macrophages from FcγRI-deficient mice, phagocytosis of IgG3-coated particles failed, compared to macrophages derived from wild type mice, in which they visualized internalization via IgG3. However, this study was not consistent with the original observation by Diamond and Yelton (16), where phagocytosis by the variant cell line via other Ab isotypes, such as IgG2a, was unchanged, which would not be true if FcγRI was the receptor lost by the J774 variant. Moreover, the Gavin study (17) did not examine the role of IgG3 phagocytosis in terms of a microbe-specific Ab interaction with host cells during infection. To date, the role of IgG3 in phagocytosis is not clear. In addition to these studies there is other literature evidence strongly supportive for the notion that IgG3 Abs engage a different receptor than other IgG subclasses: 1) IgG1 and IgG3 have very different protective efficacy in mice (18); 2) IgG1 is toxic whereas IgG3 is non-toxic in mice with chronic cryptococcal infection and high serum antigen levels (19, 20); 3) IgG3, but not the other IgG subclasses, mediates antigen-antibody complex enhancement of the antibody response in Fcγ-deficient mice (21).
In this study, we have revisited the problem of IgG3-mediated phagocytosis by comparing the opsonic efficacy of IgG1 and IgG3 for Cryptococcus neoformans (Cn). Many mAbs specific for Cn have been generated and used to study the effects of passive mAb treatment on cryptococcal infection (22–24). The in vitro system of Cn and macrophage interactions is ideal to study receptor-mediated phagocytosis because in the absence of opsonin, phagocytosis of Cn by macrophages is essentially zero, and yeast cells are easily identified and counted by microscopy (25–27). However, this system has some unique features that need to be taken into account, including the phenomenon of Ab-mediated phagocytosis proceeding through both FcγRs and complement receptors (CR) in the absence of complement. In this system, IgG can function as directly opsonizing Ab, where the Fc portions of Ab bound to the capsule of Cn are recognized by FcγRs on phagocytic cells. In addition to Fc-mediated phagocytosis through FcγRs, CRs can also mediate phagocytosis independently of complement through a mechanism whereby Ab binding to the capsule of Cn causes a conformational change that allows capsular polysaccharide to directly interact with CD18, the common component of the dimeric receptors CR3 and CR4 (28). Because complement can also be an opsonin for Cn, our experiments exclude all sources of complement to focus specifically on Ab-receptor interactions. Ab-mediated phagocytosis of Cn is a system that has been extensively studied, and the relevant parameters are so well understood that Cn phagocytosis by IgG1 monoclonal Ab (mAb) 18B7 has been subjected to mathematical modeling, producing a system where the efficiency of phagocytosis was accurately described by 11 differential equations (29). Using primary mouse macrophages and the 3E5 mAb isotype switch variants IgG1 and IgG3, specific for Cn capsular polysaccharide, our results show that IgG3 can promote internalization of Cn in the absence of the known phagocytic receptors, i.e. FcγRs and CRs, providing conclusive evidence that IgG3 interacts with a different receptor than IgG1.
All experiments were done with primary cells. Peritoneal macrophages were obtained from C57Bl/6 mice (parental strain, referred to here as FcR +/+) (Jackson Laboratory, Bar Harbor, ME), mice deficient for the common Fcγ chain (Fcγ −/−), mice deficient for both the common Fcγ chain as well as the inhibitory FcγRII (Fcγ/FcγRII −/−), mice deficient for the alpha chain of FcγRI (FcγRI −/−) which have been described previously (30), as well as CD18-null mice (gift from Dr. Chris Kevil, Louisiana State University). Briefly, mice were sacrificed by CO2 asphyxiation and the peritoneal cavity of each mouse was lavaged with 10 ml PBS. Lavage fluids were pooled for each group of mice, and peritoneal cells were collected by centrifugation at 300 g for 10 min at room temperature. Cells were then resuspended in 37°C RPMI media (Meidatech, Inc., Manassas, VA) containing 10% NCTC (Gibco), 1% non-essential amino acids and 1% penicillin-streptomycin (Mediatech). Cells were counted in a hemocytometer and plated in 96 well tissue culture plates (BD Falcon) at 1 × 105 cells/200 ul in each well. Cells were allowed to adhere, then washed with media to remove nonadherent cells, incubated overnight at 37°C in 10% CO2, and then used the next day in phagocytosis assays. In all experiments, mice were treated in accordance with institutional guidelines and the animal protocols were accepted by the institutional review committee.
Cryptococcus neoformans strain 24067 (serotype D) was obtained from the American Type Tissue Collection (Rockville, MD). Cryptococcal cells came from overnight cultures grown in Sabouraud media (Difco) at 30°C. Yeast cells were washed three times with sterile PBS, then counted on a hemocytometer and suspended at the appropriate cell density in RPMI media.
The 3E5 IgG1, IgG2a and IgG3 mAbs have been described previously (23). Briefly, mAb 3E5 was originally isolated as an IgG3 mAb following immunization of mice with GXM conjugated to tetanus toxoid, and the other isotypes were later generated via in vitro isotype switching. Ascites was generated by injecting hybridoma cells into the peritoneal cavity of pristine-primed BALB/c mice (National Cancer Institute) and harvesting the fluid. Antibodies were purified from ascites using a protein G column following the manufacturer’s instructions (Pierce), then dialyzed in PBS and quantified by ELISA with an isotype-matched standard to determine concentration. Abs were aliquotted and stored in PBS at −20°C and thawed once before use.
Phagocytosis assays were performed in 96 well plates containing primary peritoneal cells isolated one day prior to the experiment. In the case of CR or FcγR blocking experiments, antibodies to CR3 and CR4 (CD18, CD11b, and CD11c; BD Pharmingen) or the 2.4G2 antibody (BD Pharmingen) were allowed to bind for 30–60 minutes at 37°C. Then the IgG1 or IgG3 solution was added together with a Cn suspension in RMPI, for a final volume of 200 ul, with blocking Abs at 10 ug/ml, opsonizing IgG1 and IgG3 Abs at 10 ug/ml (unless otherwise noted), and Cn at an E:T of 1:1, with 1×105 Cn/well. Phagocytosis was allowed to proceed for 2 hours, at 37°C in 10% CO2. Cells were then washed 3x with PBS, fixed with methanol at −20°C for 30 minutes, washed again with PBS, and finally stained with Giemsa diluted 1:20 with sterile water. Cells were then analyzed under an inverted microscope, counting three fields/well, with at least 100 cells/field. Macrophages with internalized Cn were readily distinguishable from cells that had taken up no Cn, or where Cn was simply attached to the outside, due to the visible vacuole containing engulfed Cn. Experiments performed with Uvitex dye confirmed the accuracy of light microscopy determination of ingested cells (data not shown), where extracellular Cn are distinguishable from intracellular Cn by the exclusion of dye from Cn that have been internalized by macrophages, as described previously (31). Percent phagocytosis is calculated as the number of macrophages containing one or more Cn divided by the total number of macrophages visible in one field. Each experimental condition was done in triplicate and averaged, and t-tests were used for statistical comparisons.
Primary peritoneal cells were isolated as described above and grown on MatTek glass bottom plates (MatTek Corporation, Ashland, MA) and allowed to adhere overnight at 37°C in 10% CO2. Phagocytosis assays and fixation were carried out under the same conditions as above. Then cells were blocked with PBS + 1% BSA and stained with LAMP-1 Ab directly conjugated to FITC (BD Pharmingen) at a dilution of 1:100 for 1 hour at 37°C and finally washed and mounted in 0.1 M propyl gallate (Sigma) solution in 50% glycerol to minimize quenching. Cells were analyzed on a Zeiss microscope and compared to unstained cells at the same exposure time to account for autofluorescence, which was minimal.
Peritoneal macrophages were harvested from mice and resuspended in PBS in microcentrifuge tubes pre-blocked with BSA, in suspensions of 5×106 cells/ml for Condition I or at 3.5×106 cells/ml for Condition II. Condition I used cells from FcR +/+ and Fcγ −/− mice and Condition II used cells from FcR +/+ and Fcγ/FcγRII −/− mice. For both experiments, IgG1 and IgG3 were radiolabeled with 188Re eluted from 188Re/188W generator (Oak Ridge National Laboratory, Oak Ridge, TN) as described previously (32). 188Re-labeled IgG1 and IgG3 were added to the cells in 0.08–0.32 nM concentrations (equivalent to concentrations of 0.012–0.048 ug/ml). After incubation for 1 hr at 37°C (for Condition I) or at 4°C (for Condition II), the tubes were counted in a gamma counter, the cells were collected by centrifugation and the pellets were counted again. Scatchard analysis was used to compute the mAb binding constant (Ka) and binding sites per cell as described previously (33).
Student’s t-test was used to compare the averages of percent phagocytosis between experiments with the Bonferroni correction for multiple comparisons.
First we analyzed the opsonic efficacy of IgG1 and IgG3 in promoting phagocytosis of Cn as a function of Ab concentration with wild type C57Bl/6 (FcR +/+) macrophages. Although there was some variation from experiment to experiment, both IgG1 and IgG3 mAbs were opsonic for Cn and promoted ingestion by primary peritoneal macrophages in a dose-dependent manner (Fig. 1). Similar to prior in vitro Cn phagocytosis experiments, in the absence of opsonin, virtually no phagocytosis occurred (30). An antibody concentration of 10 ug/ml was the lowest concentration of Ab that gave optimal phagocytosis and statistical analysis showed that there was no significant difference between the efficacy of IgG1 and IgG3, thus this concentration was used throughout the rest of the experiments.
To explore the roles of the different receptors in IgG1 and IgG3-mediated phagocytosis, we used both receptor-blocking conditions as well as macrophages deficient in either FcγRs or CRs. In phagocytosis assays with FcR +/+ macrophages, mAb 2.4G2 was used to block FcγRs. 2.4G2 is specific for FcγRII and III, and was originally used to clone and identify the first FcγRs (34). The complement receptors, CR3 (CD18/CD11b) and CR4 (CD18/CD11c), have been shown to be involved in Ab-mediated phagocytosis of Cn, and Abs to CD18, CD11b and CD11c have been shown to be effective at inhibiting phagocytosis via these CRs (28). To block FcγRs or CRs, FcR +/+ macrophages were pre-incubated with specific Ab. Under CR block, IgG1 phagocytosis was reduced to about 31% of the level of phagocytosis with no block, and to about 8.5% under FcγR block (Fig. 2). With both FcγR and CR blocked, IgG1 phagocytosis was similar to the level of the control with no opsonin. However, IgG3 had high levels of phagocytosis under all conditions, and was apparently unaffected by the addition of FcγR block. The two conditions with CR block reduced phagocytosis via IgG3 by about 15–25% (Fig. 2). The difference between IgG1- and IgG3-mediated phagocytosis was not statistically significant under the condition with no block, but IgG1-mediated phagocytosis under CR block was significantly lower when compared to IgG1-mediated phagocytosis with no block or to IgG3-mediated phagocytosis under CR block. Although not indicated in Figure 2, IgG1-mediated phagocytosis was also significantly reduced under conditions of Fc block and CR + Fc block.
Because the blocking experiments indicated that IgG3 was able to induce phagocytosis while the known opsonic receptors were blocked, we next analyzed the interaction of Ab with cells deficient for certain receptors. Mice deficient in the common γ chain lack functional versions of all known activating FcγRs (Fcγ −/−). IgG1-mediated phagocytosis of Cn with Fcγ −/− cells compared to FcR +/+ cells was significantly reduced (Fig. 3). If cells were preincubated with CR blocking antibodies, IgG1-mediated phagocytosis was essentially abrogated in these cells. However, IgG3-mediated phagocytosis still occurred at comparable levels in Fcγ −/− cells, which were not significantly different than wild type levels (Fig. 3), indicating that this isotype was able to induce Cn phagocytosis even in the absence of all activating FcγRs and with blocked CR. Similarly, cells from mice deficient for both the common γ chain and the inhibitory receptor FcγRII (double knock-out Fcγ/FcγRII −/−) were able to efficiently phagocytose IgG3-coated Cn, even in the presence of CR blocking antibodies, whereas IgG1-mediated phagocytosis failed (Fig. 3). Finally, phagocytosis was evaluated in mice deficient for CD18, the common component of CR3 and CR4, the receptors involved in binding to the Cn capsule (28). In these cells, IgG3-mediated phagocytosis, both alone and in the presence of Fc blocking conditions, was essentially unchanged from wild type levels (Fig. 3). However, IgG1-mediated phagocytosis was significantly reduced in the CD18 −/− cells, at 55% of wild type levels, and when Fc block was present IgG1 phagocytosis failed.
Additionally, we addressed the possibility that FcγRI is responsible for binding and phagocytosis via IgG3, as posited by other researchers (17). In cells from mice deficient for the alpha chain of FcγRI (FcγRI −/−), IgG3 still had high levels of phagocytosis while CRs were blocked (Fig. 4), indicating that another receptor must be responsible for IgG3-mediated ingestion. Overall, levels of IgG1- or IgG3-mediated phagocytosis in FcγRI −/− cells were not different from the levels in wild type, when comparing the two cells types either with or without CR block for each Ab. To confirm that there was a reduction of functioning FcγRI in these knock-out mice, we also used IgG2a, a third isotype switch variant of mAb 3E5 (30). FcγRI is the high affinity receptor for IgG2a and probably accounts for the majority of IgG2a recognition by cells (35). In phagocytosis assays, IgG2a and IgG1 were similarly effective in promoting Cn uptake in FcR +/+ cells (Fig. 4). However, IgG2a was significantly less effective in promoting phagocytosis with FcγRI −/− cells, manifesting a reduction in phagocytosis of about 50%. When CRs were blocked, IgG2a phagocytosis was greatly reduced in FcγRI −/− cells, confirming that FcγRI is the main receptor responsible for IgG2a phagocytosis and that it was severely reduced in the knock-out cells. However, IgG3 phagocytic efficacy was unaffected by the absence of FcγRI (Fig. 4).
To account for this complex set of interactions, we propose a model that posits the existence of an IgG3-specific receptor (Fig. 5). Here, the known activating FcγRs (FcγRI, FcγRIII and FcγRIV) and the CRs known to interact with Ab-coated Cn (CR3 and CR4) are depicted, and their opsonic interactions with Cn are indicated with solid green lines. In the absence of Ab or other opsonin, no phagocytosis occurs. Phagocytosis of IgG1-coated Cn is attributable to CR and FcγRIII. IgG2a utilizes CR and mainly FcγRI, although the other FcγRs may also contribute, as IgG2a has been shown to interact with all activating FcγRs (possible interactions indicated by dotted lines). For both IgG1 and IgG2a, eliminating both CRs and FcγRs abrogates their phagocytic function. However, IgG3-mediated phagocytosis still occurs in the absence of these receptors, and we propose that its function can be explained by the presence of some unknown receptor (X). For the case of IgG3, CRs have been shown to be involved, and although phagocytosis in the absence of FcγRs still occurs normally, we cannot say that FcγRs do not interact with IgG3 at all, hence their interaction in terms of phagocytosis is questionable in our system (indicated by dotted black lines). However, in the absence of CRs and FcγRs, IgG3 is still an efficient opsonin, presumably through the X receptor.
To confirm that phagocytosis of Cn via IgG3 leads to internalization and maturation of phagocytic compartments, we performed IgG3 phagocytosis assays with Fcγ −/− cells under CR block, and then stained for Lamp-1, an intracellular marker of the phagolysosomal compartment (Fig. 6A). Cn opsonized with IgG3 in Fcγ −/− cells with CR blocked localized to Lamp-1 positive compartments. For comparison, in Fcγ −/− cells with CR blocked, no phagocytosis of Cn occurred via IgG1 and therefore Lamp-1 positive compartments showed no internalized Cn in all fields examined (Fig. 6B).
To investigate the outcome of opsonization through the different receptors, we evaluated Lamp-1 staining as a function of time after IgG1- and IgG3-mediated phagocytosis of Cn in both wild type and knock-out cells with CR blocked. Although there were some differences in intensity and pattern of Lamp-1 staining, overall the vast majority of internalized Cn localized to Lamp-1 positive compartments following phagocytosis at 2, 4 and 6 hours with both IgG1 and IgG3, and we found no significant differences between experimental conditions when quantifying these results (data not shown). Additionally, we analyzed the cryptococcal load of phagocytic macrophages by averaging the number of intracellular Cn in each macrophage under different conditions. The average numbers of Cn per macrophage for each condition were as follows: FcR+/+ CR blocked cells with IgG1 (2.6 +/− 0.8), FcR +/+ CR blocked cells with IgG3 (1.6 +/− 0.4) and Fcγ −/− CR blocked cells with IgG3 (2.3 +/− 0.6). There was no significant difference between these numbers, suggesting that phagocytosis via the different receptors led to a comparable Cn load.
By compiling light microscopy time lapsed images taken over 24 hours into videos, we were able to follow the interaction of macrophages and Cn and compare differences in the effects of IgG1 or IgG3 opsonization. Previous work has established that Ab-mediated phagocytosis of Cn by macrophages can result in such events as intracellular cryptococcal replication and exocytosis, where the macrophage expels internalized Cn, leaving both the macrophage and Cn viable (36). To assess IgG1 interactions with FcγR, we used FcR +/+ macrophages with CR blocked, to eliminate the possibility of CR promoting phagocytosis, and found that internalized Cn underwent both intracellular replication and exocytosis events (Fig. 7A). When IgG3 was analyzed with Fcγ −/− macrophages under CR block, eliminating the contribution of known FcγRs and CRs, we also observed successful phagocytosis followed by both intracellular fungal replication and exocytosis events (Fig. 7B). Hence, by all parameters studied, IgG1- and IgG3-mediated opsonization of Cn produced similar outcomes.
To evaluate the possibility of an unknown receptor for IgG3, we examined the interaction of radiolabeled Ab with primary cells. Scatchard analysis revealed that IgG3, but not IgG1, bound significantly to Fcγ −/− cells, consistent with the notion that a unique cellular receptor for IgG3 exists (Table I). The affinity of IgG3 for Fcγ −/− cells was 1.5 × 109 M−1, while the number of binding sites was determined to be 6 × 104 per cell. When IgG3 was added to FcR +/+ cells, binding was similar (2.0 × 109 M−1) as were the determined number of binding sites (7 × 104 per cell). There was low binding of IgG1 to FcR +/+ macrophages, consistent with previous data reported on IgG1 interacting with low affinity receptors (14). Similarly, IgG1 binding to Fcγ −/− macrophages was non-specific, as linear regression analysis showed that the slope was not significantly different from zero, consistent with and indicative of the absence of a receptor (Table I). The first Scatchard analysis was performed at 37°C to mimic in vivo conditions of Ab-cell interactions. In order to rule out the possibility that Ab binding could cause turnover of cellular receptors, we repeated this analysis at 4°C, using double knock-out cells deficient for both the inhibitory FcγR as well as the Fcγ chain (Fcγ/FcγRII −/−) and obtained similar results. Here, the affinity of IgG3 for Fcγ/FcγRII −/− cells was 2.3 × 108 M−1 and the number of binding sites was determined to be 1.3 × 105 per cell. Once again, IgG1 showed no significant interaction with Fcγ/FcγRII −/− cells.
We have revisited the issue of IgG3 engagement of FcγRs by looking at its functional role in phagocytosis using several new tools that have become available since this issue was last investigated (16, 17). Specifically, the availability of mice with selective deficiencies of FcγR expression in combination with a relatively clean assay of phagocytosis in the form of Cn interaction with macrophages, new serological reagents to block other opsonic receptors and a well characterized IgG1 and IgG3 isotype switch pair specific for Cn provided a new system to revaluate the old question of whether IgG3 engages a different receptor (16). The IgG3 isotype is the least studied isotype among murine IgG subclasses because it is relatively rare among mAbs and has a propensity to behave as a cryoglobulin (37). IgG3 has the remarkable capacity to self-aggregate after binding antigen thus providing for a mechanism for increased avidity (38). Indeed, IgG3 is the major isotype produced against polysaccharide antigens in mice, yet its role in the immune response is not fully understood. Our findings show that IgG3 is highly effective at promoting Cn phagocytosis both in mouse cells deficient for known FcγRs with CRs blocked and in cells deficient for CD18 with FcγRs blocked, whereas IgG1 is completely ineffective in these conditions. This observation necessarily implies that IgG3 is promoting phagocytosis through a different type of receptor than the classical FcγR receptors engaged by IgG subclasses. Furthermore, this phagocytosis appears to be functionally comparable to that observed with the other FcγRs, since the internalized Cn localize to mature, Lamp-1+ phagolysosomal compartments and exhibit similar behavior with regard to intracellular replication of internalized Cn and exocytosis events.
In revisiting the problem of IgG3 phagocytosis we considered the explanation proposed by Gavin et al that IgG3 can promote phagocytosis through FcγRI (17). However, our observations that macrophages from FcγRI −/− mice efficiently phagocytosed IgG3-opsonized Cn in conditions of CR blockage strongly argued that FcγRI is not the receptor responsible for promoting IgG3 phagocytosis. In fact, this conclusion is in agreement with the observations that led to the original proposal that IgG3 used a different receptor. When Diamond and Yelton concluded in 1981 that IgG3 bound to a unique cellular receptor, they based their conclusion on the observation that a spontaneous variant of the macrophage-like cell line J774 had lost the ability to phagocytose particles coated in IgG3 (16). However, the J774 cell variant used in that study still retained the capacity for phagocytosis of IgG2a, as well as the other IgG isotypes. Presumably, if IgG2a phagocytosis was normal then the function of FcγRI was also normal, and the lost activity of IgG3 must be due to some unidentified receptor.
If IgG3 does promote phagocytosis through a non-classical FcγR type, it must bind to macrophages that lack the known FcγRs. To explore this possibility, we used a radiolabelled ligand binding approach to determine if IgG3 could bind Fcγ −/− or Fcγ/FcγRII −/− primary cells. Macrophages deficient for gamma chain expression lack functional FcγRI, FcγRIII and FcγRIV receptors, and the double knock-out cells also lack the inhibitory receptor FcγRII. The approach taken was Scatchard analysis for both IgG1 and IgG3, and comparing macrophages from FcR +/+ mice to Fcγ −/− (Condition I) or to Fcγ/FcγRII −/− mice (Condition II). Additionally, the first experiment was performed at 37°C (Condition I) to mirror relevant Ab-receptor interactions in vivo, and the second experiment was performed at 4°C (Condition II) to reduce the level of receptor internalization and turnover after Ab binding. As expected, IgG1 showed no specific binding to macrophages from Fcγ −/− or Fcγ/FcγRII −/− mice. Similarly, IgG1 demonstrated binding to FcR +/+ (Ka 2.2 × 108 M−1 Condition I and 1.6 × 108 M−1 Condition II), consistent with reports that this isotype has low affinity for FcγR (14). In contrast, IgG3 showed higher affinity binding to macrophages from FcR +/+, Fcγ −/− and Fcγ/FcγRII −/− mice. Both the affinity of IgG3 for the new putative receptor in macrophages and the receptor number on cells from Fcγ −/− mice (1.5 × 109 M−1 and 8 × 104 binding sites per cell) or Fcγ/FcγRII −/− mice (2.3 × 108 M−1 and 1.3 × 105 binding sites per cell) were comparable to the range of affinities and receptor numbers reported for the other IgG-FcγR interaction and other FcγRs, respectively (14, 39). Of interest, the affinity of IgG3 was lower in Condition II (4°C) compared to Condition I (37°C), which is consistent with the lower temperature slightly decreasing the affinity of Ab for receptor. Also supporting this difference is the slightly higher number of IgG3 binding sites in Condition II (4°C), which indicated that the lower temperature did decrease the level of receptor turnover, hence slightly increasing the number of available receptors per cell. Taken together, these data indicate the presence of a high affinity receptor for IgG3 on the surface of mouse cells that is different from the other known FcγRs.
Additionally, the phagocytosis data provide insights into the proportion of involvement from FcγR or CR in IgG1-mediated phagocytosis. With CR blocked, IgG1 phagocytosis decreased by about one third in FcR +/+ cells, and a similar decrease was seen with IgG1 in CD18−/− cells, whereas in Fcγ−/− cells, IgG1 phagocytosis decreased by about two thirds relative to that observed with FcR +/+ macrophages, indicating that FcγRs appear to account for the majority of phagocytosis and CRs comprise the remainder.
Given our results implying that IgG3 promoted phagocytosis through a non-classical FcγR, we evaluated the functional outcome of IgG3-mediated opsonization of Cn. Cryptococcus neoformans is a facultative intracellular pathogen that replicates intracellularly in mature phagosomes after IgG1- or C-mediated phagocytosis (40, 41). IgG3-mediated phagocytosis of Cn by Fcγ −/− cells with CR blocked resulted in fungal cell ingestion into a membrane bound mature phagosome that was decorated by Lamp-1. Intracellular residence in Fcγ −/− macrophages was followed by fungal replication and occasional phagosome extrusion in a manner that was qualitatively similar to that observed for IgG1-mediated phagocytosis in FcR +/+ macrophages with CR blocked (36). Hence, in the Cn phagocytosis system, phagocytosis via IgG1 using classical FcγRs or via IgG3 using the non-classical, putative IgG3 receptor produced quantitatively and qualitatively similar outcomes for the fungal-macrophage interaction in manners studied thus far.
In summary, we provide strong evidence that murine IgG3 interacts with a cellular receptor that is structurally different than the classical FcγRs in that it lacks the γ chain. Our results are consistent with the original Diamond and Yelton proposal for a distinct IgG3 receptor and are supported by several studies showing major differences in the biological effects of IgG3 versus the other IgG subclasses (18–21, 30). Understanding Ab-receptor interactions is essential to fully grasp the role of isotype in Ab responses induced during infection or to discover the mechanism of Ab-mediated effects during passive mAb therapy. Given that different isotypes have been shown to have very different roles when administered in vivo (42), and given that passive Ab treatment is a growing field that is dependent on research and development occurring in the mouse model and ultimately leading to human therapies (7, 8, 43), it is essential to understand mouse IgG-receptor interactions for all isotypes. This research highlights the additional information we need to learn about this system. The results presented here suggest that future studies focus on the molecular and structural characterization of this receptor.
We would like to acknowledge Andre Nicola, Johanna Rivera and Antonio Nakouzi for their technical advice and assistance. We thank Dr. Chris Kevil for his kind gift of CD18−/− mice. We would like to thank Al Watford for his assistance in the animal facility. The data in this paper are from a thesis to be submitted by Carolyn Saylor in partial fulfillment of the requirements for the degree of doctor of philosophy in the Sue Golding Graduate Division of Medical Science, Albert Einstein College of Medicine, Yeshiva University, Bronx, NY.
1This work was supported by NIH grants AI33774, AI33142 and HL59842-01 to AC. CS was supported by NIH T32 AI07506. ED is Sylvia and Robert S. Olnick Faculty Scholar in Cancer Research and was also supported by NIH grant AI60507.
3Abbreviations used: Ab: antibody, mAb: monoclonal antibody, FcR: Fc receptor, FcγR: Fcγ receptor, CR: complement receptor, Cn: Cryptococcus neoformans, ITAM: immunoreceptor tyrosine-based activating motif, ITIM: immunoreceptor tyrosine-based inhibitory motif