By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Please be advised that we experienced an unexpected issue that occurred on Saturday and Sunday January 20th and 21st that caused the site to be down for an extended period of time and affected the ability of users to access content on Wiley Online Library. This issue has now been fully resolved. We apologize for any inconvenience this may have caused and are working to ensure that we can alert you immediately of any unplanned periods of downtime or disruption in the future.
Mast cells (MC) have been shown to mediate regulatory T-cell (Treg)-dependent, peripheral allograft tolerance in both skin and cardiac transplants. Furthermore, Treg have been implicated in mitigating IgE-mediated MC degranulation, establishing a dynamic, reciprocal relationship between MC and Treg in controlling inflammation. In an allograft tolerance model, it is now shown that intragraft or systemic MC degranulation results in the transient loss of Treg suppressor activities with the acute, T-cell dependent rejection of established, tolerant allografts. Upon degranulation, MC mediators can be found in the skin, Treg rapidly leave the graft, MC accumulate in the regional lymph node and the Treg are impaired in the expression of suppressor molecules. Such a dramatic reversal of Treg function and tissue distribution by MC degranulation underscores how allergy may causes the transient breakdown of peripheral tolerance and episodes of acute T-cell inflammation.
Mast cells (MC) are best known for their role in allergies and protecting our body from parasitic, bacterial and viral infection (1). IgE antibody against the allergen or infectious agent binds to the high-affinity IgE receptor on MC. Subsequent encounter with allergen leads to the release of the MC granular content with heightened inflammation. While the historical role of MC has been as regulators of inflammatory responses, MC have been recently been implicated as regulators of tolerance (2,3). The striking contrast in MC function is exemplified by their pro-inflammatory roles in nematode infection and allergies (1) or their role in mediating suppression, as in UV-B damage (4), mosquito bites (5) and graft tolerance (6,7). Recently, the pivotal role of MC in the establishment of acquired tolerance to an allograft was shown (6,7). It was hypothesized that the secretion of immunosuppressive mediators by MC was critical for sustaining tolerance. Recently, the dynamic and reciprocal nature of MC-Treg interactions was shown. It was reported that Treg can suppress IgE-mediated degranulation through OX40-OX40L interactions (8,9) thereby showing a natural mechanism for MC stabilization. It is clear that the release of inflammatory mediators as a consequence of MC degranulation results in inflammation. How this impacts peripheral tolerance and Treg function is not known. Therefore, studies were designed to determine if degranulation of MC present in tolerant allografts would impact on allograft survival. Data presented show that IgE-mediated degranulation, either within the graft or systemically, breaks established peripheral tolerance and leads to T-cell-mediated, acute rejection of the allograft. Degranulation causes the release of MC intermediaries, a rapid migration of both Treg and MC from the graft as well as a transient demise in the expression of Treg suppressive cytokines. Such a dramatic reversal of Treg function and tissue distribution by MC degranulation underscores how allergy may cause the transient breakdown of peripheral tolerance and episodes of acute T-cell inflammation.
Material and Methods
C57Bl/6, CB6F1 (C57Bl/6xBALB/c hybrid), C57BL/6-Ly5.2+, and C57BL/6-Rag−/− mice were purchased from the Jackson Laboratory. FoxP3/GFP reporter mice were provided by Dr. A. Rudensky (University of Washington School of Medicine, Seattle, WA) (23).
Skin graft model
Skin grafting was performed as described previously (43). For dual grafting, the first graft was placed on the back near the base of the tail, whereas the second graft was placed on the back close to the neck 2 weeks later. Grafts were monitored for rejection for 30 days postdegranulation and were considered rejected when 80% of the original graft disappeared or became necrotic.
Chemical degranulation was done by application of 50 μL Compound 40/80 (1 mg/mL, Sigma) directly under the graft. ‘Active’ immunization was achieved by 100 μg of OVA/Alum (Pierce) i.p., 37 days prior to grafting or passive by transfer of IgE. For passive immunization, 2 or 5 μg of either OVA-specific IgE (clone 2C6; Serotec) or TNP-specific IgE (clone A3B1; cross-reactive with NP) was given intravenously 24 h prior to degranulation, respectively. Degranulation was induced either locally (50 μL of 1 mg/mL OVA in PBS) or systemically (500 μL of 1 mg/mL OVA in PBS intraperitoneal) for mice that received OVA-IgE or were active immunized. Degranulation in mice that received NP-IgE was done by injecting 20 ng of NP17-OVA/NP23-BSA locally. Blocking of degranulation was done by subcutaneous injection of 100 μL of Cromolyn Sodium Salt (39 mM in PBS, Sigma-Aldrich) 30 min prior to degranulation.
Cytokine profile of the graft
Grafts were collected, cut to small pieces in HBSS (6 grafts/mL) 18 h postdegranulation and incubated for 1 h at 37°C. Cytokines in the supernatants were determined by multiplex analysis (Biorad) and verified by ELISAs (IL4, IL6, IL10 (Pharmingen), IL9 (PeproTech) and TNFα(eBioscience)).
Induction of inflammation
Mice were either treated with 200 μL complete Freund's adjuvant (Sigma), 50 μg TLR4 agonist (LPS, E. coli 055:B5, Sigma) or 50 μg TLR9 agonist (CpG; ODN-1826, Eurogentech) by intraperitoneal injection. For induction of local inflammation, 8 μL of 5 mg/mL FITC (Sigma) in 1:1 acetone (Fisher scientific)/dibutylphthalate (Sigma) was applied directly onto the preshaved graft. Positive controls included degranulation after passive immunization or intraperitoneal injection of 50 μg of agonistic αCD40 (FGK; BioExpress).
Tissues were snap-frozen and 8 μm sections were fixed with methanol prior to staining with α-CD117-A647. Hoechst was used for nuclear staining. Slides were analyzed by a confocal microscope (LSM510Meta, Zeiss) at a 100 × magnification. To analyze granulated MC, slides were stained with toluidine blue (Sigma). Toluidine blue-positive MC numbers in skin were determined by counting the number of stained MC in seven randomly chosen fields at 100× magnification by two independent observers using an Olympus DP70 camera system operating on a BX60 transmitted microscope.
Cell isolation and analysis
Naïve T-cells were isolated from pooled spleen and LN. Prior to cell sorting, T-cells were preenriched by CD4 negative selection (biotin-selection kit, StemCell). Isolation of graft T-cells from FoxP3-GFP mice was performed by digestion with DNAse, Liberase and Collagenase (10 mg/mL, Roche) for 45 min at 37°C after weighing the grafts. Analysis was performed after co-staining with CD4 (RM4–5, Pharmingen) by flow-cytometry. For determination of the total number of MC in the dLN, samples were stained with CD117 (clone: 2B8) and FcεRI (MAR-1, eBioscience) and analyzed by flow-cytometry.
T-cell studies in vivo
At day 27 and 35, mice were treated with 300 μL intraperitoneal and 50 μL locally at the site of the graft with mixture of αCD4 (GK1.5) and αCD8 (2.43) antibodies (kind gift from Dr. MJ Turk, Dartmouth, NH) to deplete T-cells. Depletion was confirmed at day 37. For the transfer of tolerance experiments, grafts dLN were collected at day 35 and T-cells purified by two rounds of CD4 negative depletion (StemCell). RAG−/− received an F1 graft 2 weeks prior to adoptive transfer of 1×106 purified T-cells by i.v. injection.
Functionality of regulatory T-cells in vitro and in vivo
One day postdegranulation, FoxP3-GFP+ T-cells from the graft dLN were FACS sorted. For in vitro studies, Ly5.2+ naïve T-cells were purified by FACS sorting and labeled with carboxyfluorescein succinimidyl ester (5 μM, CFSE). In total 5 × 104 were cultured with T-depleted irradiated Ly5.1+ splenocytes (3000 rad) at different ratios of Treg:Tnaïve. After 3 days CFSE dilution, profiles were analyzed within the Ly5.2+ population. For in vivo functionality, GFP+/Ly5.1+ Treg were mixed at different ratios with naïve, polyclonal T-cells and adoptively transferred into F1 pregrafted RAG−/−. Phenotypic analysis of Treg was performed by staining T-cells of the dLN after CD4 enrichment. All samples were stained with FoxP3 in either FITC or PE depending on the co-staining. FoxP3+ cells were analyzed for the expression of Ki67 (Ki67 staining kit, BD), CTLA4(UC10-4F10-11, Pharmingen), ICOS (15F9, eBioscience), CD62L(MEL-14, BioLegend), CD73(clone TY/11.8, eBioscience), CD103(2E7, eBioscience), GITR(YGITR 765, BioLegend), and IL7Rα(CD127; A7R34, eBioscience). For RT-PCR analysis, FoxP3+ Treg from the grafted FoxP3-GFP mouse were flow sorted. The primers used for estimates of gene expression were: TGF-β1 forward: 5′-TTGCTTCAGCTCCACAGAGA-3′, TGF-β1 reverse: 5′-TGGTTGTAGAGGGCAAGGAC-3′, IL-10 forward: 5′-GGTTGCCAAGCCTTATCGGA-3′, IL-10 reverse: 5′-ACCTGCTCCACTGCCTTGCT-3′, Ebi3 forward: 5′-AGCAGCAGCCTCCTAGCCT-3′, Ebi3 reverse: 5′-GCGGAGTCGGTACTTGAGAG-3′, granzymeB forward: 5′-CTCCACGTGCTTTCACCAAA-3′, granzymeB reverse: 5′-AGGATCCATGTTGCTTCTGTAGTTAG-3′.
All statistics have been calculated using the prism 4.03 software package (GraphPad Software, San Diego, CA). Survival data were analyzed using the Kaplan–Meier method, with the Wilcoxon rank test and the log-rank test. Luminex data were analyzed by the immune monitoring laboratory of the Norris Cotton Cancer Center/Dartmouth Hitchcock Medical Center and are shown as mean ± SD whereas other data are expressed as mean ± SEM. The latter were analyzed by one-tailed ANOVA and posttested by Tukey analysis. Statistical significance was calculated for a 95% confidence interval (p < 0.05). The exact p-values are denoted in the figures.
MC degranulation results in the rejection of an established tolerant allograft
Previously we showed that in mice tolerized with αCD154 and donor-specific transfusion (DST) both Treg and MC accumulate in the tolerant allograft and are functionally critical for allograft persistence (7). Herein, it was asked if IgE-mediated MC degranulation and the release of inflammatory mediators would alter Treg function, distribution and impact on allograft persistent in the tolerant host.
Both tolerized and syngeneic control mice were treated as shown in Figure 1A, unless stated differently (10–12). Within 15 days, all tolerized mice bearing an allogeneic skin graft in which MC had been degranulated lost their grafts. Syngeneic mice maintained their grafts for the duration of the experiment regardless of treatment (Figure 1B). These results indicate that degranulation of MC facilitates the rejection of allogeneic skin grafts in tolerized hosts. Both active immunization with OVA-Alum 30 days prior to grafting, followed by challenge with OVA, as well as chemically induced (compound 40/80) (13,14) degranulation showed the same results as with passive immunization (Supporting Figure S1A and B, respectively). Cytokines known to be abundant after IgE-mediated MC degranulation were increased in groups where MC had been degranulated (Figure 1C), confirming that the MC released their granular content and established a pro-inflammatory Th2-biased micro-environment similar to that observed in allergies (15).
Comparison of different inflammatory mediators in inducing graft rejection of tolerant allografts
A battery of inflammatory mediators was used to evaluate if any inflammatory insult under tolerizing conditions would inevitably lead to graft rejection. Mice were treated with inflammatory doses of the Toll-like receptor (TLR) 4 agonist, LPS; the TLR9 agonist, CpG-ODN 1826; Complete Freund's Adjuvant (CFA) or an agonistic αCD40 antibody. Finally, local, chemical-induced Th2-biased inflammation was induced using acetone/dibutylphthalate/FITC applied directly onto the graft (16–18). No acute rejection was observed in any of the groups tested except for the tolerant mice administered αCD40 antibody, as previously reported (19). These data demonstrate that the acute rejection observed after degranulation was not due to the introduction of overt inflammation per se (Figure 1D) but due to the nature of the inflammatory microenvironment created by degranulation.
Mast cell stabilization leads to prolonged allograft survival
To seek further evidence that breakdown of allograft tolerance was the effect of MC degranulation, degranulation was blocked by local intragraft injection of sodium cromoglycate (cromolyn) (20). The observed acute rejection after MC degranulation was completely blocked with the administration of cromolyn (Figure 2A) thereby confirming that the acute allograft rejection was the result of degranulation.
In this system of allograft tolerance, peripheral tolerance begins to decay at day 70 postgrafting in tolerant mice (21). To address if limited allograft survival in tolerized mice was due to ‘low-grade’ MC degranulation, cromolyn was applied to the allograft at day 30 and/or day 60 after engraftment. Single treatment did not impact graft survival. However, when performed on both days, allografts were maintained for >120 days even when allergic exacerbations were mimicked right after cromolyn treatment (Figure 2B). These results show that allograft survival in tolerized mice can be extended by the administration of cromolyn and suggests that MC degranulation may contribute to the natural decay of tolerance in this system.
MC degranulation leads to an accumulation of MC in the draining lymph nodes
To define the cellular events occurring upon MC degranulation, tolerant and syngeneic skin grafts were collected at day 1 and 5 after degranulation. Degranulation leads to disappearance of MC from the allograft, as measured with by the number of c-Kit+ cells present within the graft (Figure 3A) and inversely correlated with the number of MC in the corresponding allograft-draining axillary and brachial LN as quantified by the number of c-KithighFcεRIhigh cells (Figure 3B). MC numbers were confirmed by toluidine blue staining of both graft and dLN (Supporting Figure S2). Taken together, these data suggest that upon degranulation, some MC degranulate and other MC egress from the skin and enter the draining lymph.
MC degranulation-mediated rejection is dependent on T-cells
In order to address whether the rejection observed after MC degranulation was dependent upon T-cells, CD4+ and CD8+ T-cells were eliminated with depleting antibodies. MC degranulation in the absence of T-cells did not lead to graft loss (Figure 4A). Similar results were obtained using compound 40/80 to degranulate the MC (Supporting Figure S3). This demonstrates that T-cells were mediating graft rejection after degranulation.
MC degranulation results in a breakdown of T-cell tolerance
It is known that in this system a balance of allogeneic effector T-cells (Teff) and Treg controls graft rejection and suggests that MC degranulation has an impact on this balance. To directly measure the intrinsic ability of T-cells from tolerant mice to mediate graft rejection or not, an adoptive transfer study was designed. Purified dLN T-cells from the different groups were transferred into RAG−/− mice bearing an allograft. Adoptive transfer of CD4+ T-cells from tolerized mice after local MC degranulation rejected the allografts in the secondary recipients (Figure 4B). These data show that MC degranulation restores alloreactive T-cell responses in a tolerized mouse.
Regional degranulation results in the systemic decay of peripheral tolerance
Studies were designed to address whether local MC degranulation would facilitate a breakdown in allograft tolerance at the systemic level. To test this, tolerized mice received two skin allografts, one at day 0 and one at day 14. The first allograft (graft #1) was draining into the inguinal lymph nodes. The second graft (graft #2) drained to the brachial and axillary lymph nodes (day 14) (22). At day 30, graft #1 was challenged after passive immunization and rejected at the expected kinetics. Interestingly, graft #2 was also rejected after a delay of about 1–2 days (Figure 4C). As a control to ensure local degranulation of graft #1, a cohort of mice received a syngeneic graft #1 and an allogeneic graft #2. Local degranulation of the syngeneic graft did not lead to rejection of the allogeneic graft confirming that allergen was not reaching the other graft. Together, these data demonstrate that local MC degranulation leads to a systemic breakdown of peripheral tolerance.
Degranulation facilitates Treg and MC loss from the graft and impairs Treg function
MC and Treg are present in tolerant allografts and are required to maintain allograft tolerance (7). It was hypothesized that degranulation may change Treg composition and functionality, thereby altering graft longevity. Following degranulation, Treg were enumerated in the grafts by using Foxp3-GFP mice (23) as hosts. At day 1 following MC degranulation, a dramatic reduction of CD4+Foxp3+ was observed in allografts (Figure 5A). At this time point, MC were still present in the allograft (Figure 3A), demonstrating that the efflux of Treg precedes the efflux of MC from the skin allograft following MC degranulation.
In addition to the altered distribution of Treg, the in vitro and in vivo functions of Treg were also evaluated following MC degranulation. Treg function was measured in vitro by a suppressor assay. Treg from both tolerized, allografted mice and from those bearing syngeneic grafts suppressed WT Teff proliferation, as measured by CFSE dye dilution. However, Treg from mice whose MC were degranulated 1 day prior were less effective at suppressing Teff proliferation (Figure 5B). The impairment in Treg suppression was transient, because Treg recovered at day 5 postdegranulation, regained their ability to suppress Teff indistinguishably from the non-degranulated controls. These results indicate that MC degranulation leads to a transient reduction in Treg function. Furthermore, Treg function was also impaired after MC were degranulated in mice bearing syngeneic grafts, suggesting that this is a general mechanism of immune regulation mediated by MC. Of note, Teff in the same groups as tested in Figure 5B could still be suppressed by WT-Treg (data not shown).
To examine whether Treg after MC degranulation are impaired in vivo, an in vivo Treg suppressor assay was performed. Treg were mixed with WT-Teff and adoptively transferred into pregrafted RAG−/− mice. The majority of mice, which received Treg from non-degranulated mice at a 1:10 Teff:Treg ratio, retained their allogeneic grafts for the duration of experiment. However, grafts were rejected when Treg were isolated after MC degranulation at the same ratio of Teff:Treg (Figure 5C). In sum, these data indicate that after MC degranulation, Treg transiently lose their ability to mediate graft survival.
MC degranulation leads to a reduced expression of key Treg immunomodulatory molecules
To define the underlying molecular mechanisms for the loss in Treg function after degranulation, the expression of key surface molecules and cytokines was analyzed. Comparative FACS analysis revealed no changes in the expression of markers involved in maturation/migration (CD62L, CD103), peripheral homeostasis (IL7αR), proliferation (Ki67) and contact-dependent suppression (CTLA4, ICOS, CD73, GITR). Additionally, the level of FoxP3 expression was not changed as shown by histogram overlay and bar diagram of the MFI for the individual conditions (Figure 6A). Together, these results indicate that the prominent Treg markers noted are not affected by MC degranulation.
The expression of suppressive mediators molecules (TGFβ, IL10, Granzyme B(GzB) and EBI3 (Ebstein Bar virus Induced gene 3)) secreted by FACS-purified FoxP3-GFP Treg were determined by RT-PCR. TGFβ mRNA expression was reduced over 50% in Treg recovered 24 h after MC degranulation but was restored at day 5 postdegranulation when compared to Treg from non-degranulated controls. The mRNA levels of IL10 and GZB were nearly absent at 24 h postdegranulation, with a modest recovery of expression for IL10 at 5 days postdegranulation. EBI3, however, was not impacted by MC degranulation at 24 h postchallenge yet declined at day 5 (Figure 6B). Taken together, these data suggest that MC degranulation leads to the loss of multiple suppressive mediators utilized by Treg which may account for the loss in peripheral tolerance upon degranulation.
The studies presented significantly advance our knowledge of MC-Treg interactions and how degranulation impacts on the behavior and function of Treg. The data show (1) that antibody or chemically induced MC degranulation, either regionally or systemically, leads to a T-cell-dependent loss of tolerant skin allografts; (2) the ‘natural’ loss of tolerant allografts can be reversed by the blockade of MC degranulation in vivo; (3) degranulation of MC causes a rapid change in the cellular composition of the tolerant microenvironment, with a loss in both MC and Treg; (4) regional degranulation ultimately results in the systemic breakdown of T-cell tolerance and (5) MC degranulation causes a transient loss in Treg function.
Prior studies have documented the persistence of MC in tolerant allografts and their functional involvement in allograft survival. Histological examination indicated that these intragraft MC were not degranulated (6,7). Based on these observations, it was suggested that the secretion of immunosuppressive mediators by MC was critical for maintaining the allograft (6,7). In light of these findings, the impact of MC degranulation on the delicate balance of Treg and Teff under tolerant conditions was investigated. The various methods used for degranulation all led to acute T-cell-dependent allograft rejection.
MC-mediated graft loss could be blocked by using the MC stabilizing agent cromolyn providing additional proof that degranulation imparts an immunostimulatory impact leading to a breakdown in tolerance. Cromolyn has been effectively and safely used for over 30 years as a prophylactic treatment in allergies (24). A recent retrospective study demonstrated that blocking MC degranulation by a single application of cromolyn is effective in preventing allergies later in life. Most likely a permanent change in the composition of T-cell compartment was induced by the brief stabilization of the MC (25). Also preservation of transplant organs is prolonged by cromolyn by protecting the tissues from ex vivo, MC-mediated, ischemia (26,27). In the study presented, while a single application of cromolyn did not impact graft survival, a sequential treatment on day 30 and 60 postengraftment prolonged allograft tolerance over 120 days. We suspect that in the absence of MC stabilization, erosion of allograft tolerance over time as seen in this model could be partially explained by low levels of MC degranulation occurring even in the absence of allergens.
The regional or systemic delivery of allergen results in the tempered degranulation of MC with a significant proportion of MC retaining their granule content. MC degranulation resulted in the sequential egress of Treg and MC from the allograft. Unresolved is the role of MC graft egress and putative LN accumulation to the process of graft rejection. However, the lack of complete degranulation of the MC upon administration of allergen may be the result of the immunosuppressive impact that Treg exert on MC degranulation. Recent evidence has shown that Treg can suppress MC degranulation by contact-dependent OX40-OX40L interactions as well as by a contact-independent manner through IL10 secretion in vitro (8,28,29). However, even in the absence of Treg, MC degranulation is likely never complete since only a 50% reduction of granulated MC has been reported after C40/80 treatment (30).
It has been previously reported that the adoptive transfer of Treg from untreated animals can suppress ongoing allergy (32,33). This observation suggests a functional defect in the Treg compartment during allergy. No phenotypic differences were observed, and the level of FoxP3 expression was similar at all analyzed time points. This is in contrast to a recent report that during rejection, levels of FoxP3 of Treg are downregulated (34). Either this could be induced by using a different transplant model or degranulation leads to a different type of rejection as normal allograft rejection based on minor histocompatibility differences. Both in vitro and in vivo analyses revealed a temporary impaired ability of the Treg to suppress Teff responses after degranulation. This suggests a regulatory role for MC in Treg functionality. The underlying cause for the loss of Treg suppression may be due to the loss of the expression of GZB as well as TGFβ, and/or IL10. Many studies have examined the function of Treg in atopic patients, but the findings from these studies are inconclusive (34–39). Some of the discrepancies from these studies can be explained, however, by the observation herein that there is only a brief window of time where the Treg lose their suppressive function after degranulation.
The findings over the past 5 years underscore the enormous plasticity of the MC compartment and their impact on acquired immunity and acquired immune privilege (40). While MC have been extensively described in type I hypersensitivity diseases as being pro-inflammatory, it is evident that MC control the development and persistence of peripheral tolerance and immunity (6,7). Unlike T-cells, which can be decisively segregated into multiple effector and regulatory subsets, there are no defined subsets of functionally diverse MC (2). Emerging studies implicate MC as accessory cells for both Teff and Treg alike to mediate inflammation and suppression, respectively. The findings of this study are the first to show the reciprocal impact of MC degranulation on Treg function. The implications in allergy and in peripheral tolerance are significant. The clinical relevance for the presented data in regard to allograft transplant recipients becomes evident when the data from a retrospective study in patients who received a kidney transplant are taken into consideration. Atopic individuals, in this case allergic rhinitis, with a kidney transplant show more severe and acute episodes of rejection than individuals without any history of allergy (41). These data are concordant with our observation that MC degranulation breaks peripheral tolerance systemically.
In closing, with the increasing prevalence of allergies in Western countries (42), the durability of allografts in atopic patients may suffer increasing jeopardy. Long-term MC stabilization in transplant recipients could be beneficial in maintaining intact MC and functional Treg at the site of the graft and preventing temporal impairment of suppression due to degranulation. Since the regulation of Treg function seems to be a one component of MC function, defining MC products leading to downregulation of suppressive pathways in Treg will be a great asset for developing new treatment strategies in a variety of diseases where either loss of suppression needs to be restored or ablated. Altogether, MC seem to be able to regulate the immune response in general and Treg function in particular, either by supporting tolerance via secretion or by inducing inflammation via degranulation.
The authors would like to thank Gary A. Ward for assistance in cell sorting, Jose R. Conejo-Garcia for the expertise in bright field microscopy and Kenneth A. Orndorff for assistance in confocal miscroscopy. V.d.V. conducted most of the experiments and wrote the manuscript. A.W. performed the toluidine blue staining and quantified the MC in histological slides. K.B. prepared and assisted with the surgeries. T.W. purified and supplied Np-IgE. M.B., R.E. and R.N. provided discussions and final editing of the manuscript. V.d.V. and R.N. designed and discussed the experiments. All authors have read and commented on the manuscript prior to submission. This work was supported by a grant from the National Institutes of Health (NIH AI048667).
Conflict of Interest Statement
The authors have no conflicting financial interests.