Accepted Manuscript Macrophages are the dominant effector cells responsible for IgG-mediated passive systemic anaphylaxis challenged by natural protein antigen in BALB/c and C57BL/6 mice Delong Jiao, Yuan Liu, Xiao Lu, Beiyi Liu, Qingjun Pan, Yanjun Liu, Yang Liu, Ping Zhu, Ning Fu PII: DOI: Reference:
S0008-8749(14)00063-X http://dx.doi.org/10.1016/j.cellimm.2014.03.018 YCIMM 3310
To appear in:
Cellular Immunology
Received Date: Accepted Date:
17 February 2014 31 March 2014
Please cite this article as: D. Jiao, Y. Liu, X. Lu, B. Liu, Q. Pan, Y. Liu, Y. Liu, P. Zhu, N. Fu, Macrophages are the dominant effector cells responsible for IgG-mediated passive systemic anaphylaxis challenged by natural protein antigen in BALB/c and C57BL/6 mice, Cellular Immunology (2014), doi: http://dx.doi.org/10.1016/j.cellimm. 2014.03.018
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Macrophages are the dominant effector cells responsible for IgG-mediated passive systemic anaphylaxis challenged by natural protein antigen in BALB/c and C57BL/6 mice Authors: Delong Jiao a, Yuan Liua, b, Xiao Lu a, Beiyi Liu a, Qingjun Pana, c, Yanjun Liu a, Yang Liud, Ping Zhu a, Ning Fua,* Affiliations: a
Department of Immunology, School of Basic Medical Sciences, Southern Medical University,
Guangzhou, China; b
Present address: Medical genetic center, Guangdong Women and Children Hospital, Guangzhou,
China; c
Present address: Institute of Nephrology, Guangdong Medical College, Zhanjiang, China;
d
Department of Immunology, Chengdu Medical College, Chengdu, China
* Corresponding author Ning Fu, Department of Immunology, Southern Medical University, 510515 Guangzhou, Guangdong, P. R. China. Tel: 86-20-61640405. E-mail: [email protected]
1
Abstract IgG-induced passive systemic anaphylaxis (PSA), a serious adverse effect of passive immune therapy using therapeutic monoclonal antibodies, has been greatly emphasized. However, controversy exists regarding the type of effector cells involved in IgG-induced anaphylaxis as a result of the induction of PSA by different IgG subtypes or the use of mice with varying genetic backgrounds. To clarify the effector cells for PSA, the PSA model with serious hypothermia was established by IgG monoclonal antibody (mAb) against natural protein or complete antigen, not hapten conjugate, in BALB/c and C57BL/6 mice. The results indicated that PSA could be remarkably inhibited by the depletion of macrophages but not by the depletion of whole leukocytes, basophils, neutrophils or monocytes. We further confirmed that macrophages are indispensable for the PSA induced by all six IgG-natural antigen complexes in both strains of mice. Additionally, platelet-activating factor (PAF) was found to be the major effector mediator for IgG-induced anaphylaxis. Moreover, gene knock-out of the complete third component (C3) did not affect PSA-related hypothermia in C57BL/6 mice. These results indicate that macrophages and PAF act as dominant effector cells and mediator molecules, respectively, and are indispensable components in the induction of IgG-mediated PSA induced by IgG mAb and natural protein antigen. Based on the above results, we hypothesize that inconsistencies in effector cells for PSA may be associated with different features of the mAb-antigen system that might affect the magnitude of FcγRs cross-linking on effector cells. Keywords: IgG; passive systemic anaphylaxis; mouse model; macrophage; platelet activating factor
2
1. Introduction Both active systemic anaphylaxis (ASA), which occurs in hosts immunized with antigen, and passive systemic anaphylaxis (PSA), which occurs in hosts following the infusion of therapeutic antibodies or immunoglobulins, are life-threatening allergic reactions [1-4]. Recently, IgG-induced PSA has attracted significant attention, as allergic reactions are severe potential adverse effects of therapy with monoclonal antibodies (mAbs) and immunoglobulins [2, 3, 5, 6], both of which are practiced extensively in the clinic. Moreover, over 30 therapeutic mAbs for clinical administration have been approved by the FDA, and more than 300 therapeutic mAbs and immunoglobulin reagents have been tested in clinical trials [7]. The frequency of allergic reaction, including PSA, ranges from 0.09 to 77% depending on the type of antibody administered [8]. Therefore, a better understanding of PSA is necessary to elucidate the underlying mechanisms of anaphylaxis and to create a new strategy to evaluate the safety of therapeutic antibody administration to prevent these common allergic reactions observed clinically. In mouse models, anaphylaxis can be mediated through a classical pathway that involves IgE, Fcε receptors (FcεRs), mast cells, histamine and leukotrienes, as well as through an alternative pathway that involves IgG, Fcγ receptors (FcγRs), macrophages or other cells and platelet activating factor (PAF) [9, 10]. Similarly, IgG, FcγRs, macrophages and PAF are thought to play
PSA: passive systemic anaphylaxis; ASA: active systemic anaphylaxis; mAbs: monoclonal antibodies; FcεRs: Fcε receptors; FcγRs: Fcγ receptors; PAF: platelet-activating factor; Ai-ICs: anaphylaxis-inducing immune complexes; C3: complement third component; C3-/-: complement third component knock-out; WT: wild type; OVM: ovomucoid; HSA: human serum albumin; OVA: ovalbumin; BSA: bovine serum albumin; TNP: 2,4,6-trinitrophenol; DNP: 2,4-dinitrophenol; PenV: penicillin V; GaMD: goat anti-mouse IgD; GPI: rabbit glucose-6-phosphate isomerase; CPM: cyclophosphamide; PBS-lipo: PBS encapsulated liposomes; Clo-lipo: Clodronate encapsulated liposomes; WBCs: white blood cells; RBCs: red blood cells; BPO, benzylpenicilloyl; GSA, guinea pig serum albumin. 3
pivotal roles in IgG-induced anaphylaxis in humans, suggesting that activation of the alternative pathway occurs in both mice and humans [11]. However, the mechanisms underlying IgG-induced PSA remain elusive, particularly because there are conflicting viewpoints regarding the type of effector cells involved. Mukai et al. proposed that basophils are the main effector cells in alternative pathway 1 and that macrophages are the main effector cells in alternative pathway 2 [12]. Tsujimura et al. demonstrated that basophils are the effector cells for PSA [13], whereas Jönsson et al. indicated that neutrophils and monocytes/macrophages are the most important effector cells for ASA and PSA in mouse models [14, 15]. The latest publication of Khodoun et al., covering all three known effector cell types, concluded that all macrophages, basophils and neutrophils participated in IgG-induced anaphylaxis [16]. The general consensus is that the divergence regarding effector cell types in IgG-mediated PSA may be due to IgG isotypes and variations in the genetic background of inbred mice. In the present work, we focused on the effector cells responsible for IgG-induced PSA by IgG mAb against the natural protein antigen that is more common clinically, compared with hapten conjugates such as DNP and TNP used in murine PSA models. In our previous work, we raised and proved that only a fraction of IgG immune complexes, defined as anaphylaxis-inducing immune complexes (Ai-ICs), are responsible for mediating murine IgG-induced PSA, which was then confirmed using 36 combinatory reactions with 8 natural proteins or complete antigens and 27 strains of corresponding mAbs [17]. In 14 ICs capable of inducing PSA, a remarkable temperature drop occurred, but seldom death, although 2~3-folds mAbs and 2~10-folds allergen were used in the preliminary experiments, which were considered non-lethal PSA. Here, by using 6 kinds of Ai-ICs containing various natural complete antigens, not hapten conjugates, and IgG mAbs or mAb cocktails of IgG1, IgG2a and IgG2b isotypes, we demonstrate that the dominant effector cells 4
mediating PSA in BALB/c and C57BL/6 mice are macrophages rather than basophils, neutrophils or monocytes. Consequently, we hypothesize that different features and dosages of the mAb-antigen system might be associated with the effector cell type by affecting the magnitude of FcγR cross-linking.
5
2. Materials and methods 2.1. Mice Female BALB/c and C57BL/6 mice (6~10-week-old) were purchased from the Laboratory Animal Center of Sun Yat-Sen University (Guangzhou, Guangdong, China). The wild type (WT) and complement third component knock-out (C3-/-) C57BL/6 mice used for complement third component (C3) dependent study were provided by Dr. Qiang Zou (Chengdu Medical College, Chengdu, Sichuan, China). The animal experiments were approved by the Ethics Committee for Experimental Animals at Southern Medical University and were performed according to the national guidelines for animal welfare. 2.2, Reagents and antibodies Ovomucoid (OVM), human serum albumin (HSA), Triprolidine, CV-3988, gadolinium chloride (GdCl3) and PAF were purchased from Sigma-Aldrich (St. Louis, MO, USA), and cyclophosphamide (CPM) was purchased from Lingnan Pharmaceutical Ltd (Guangzhou, Guangdong, China). Clodronate and PBS encapsulated liposomes (Clo-lipo and PBS-lipo) were purchased from ClodronateLiposome.org (Amsterdam, The Netherlands). The antibodies used for flow cytometry analysis against mouse Gr-1, IgE, CD49b and F4/80 were obtained from BioLegend (San Diego, CA, USA), and anti-mouse CD11b was purchased from eBioscience (San Diego, CA, USA). Anti-Gr-1 (Rat IgG2b) used for neutrophil depletion was purchased from BioLegend, and MAR-1 (hamster IgG) used for basophil depletion was purchased from eBioscience. Rat IgG2b and hamster IgG isotype control antibodies were also purchased from eBioscience. All mAbs and antigens for the induction of PSA have been used in previously published work [17]. 2.3. PSA model 6
PSA models using 6 Ai-ICs have been previously reported. Briefly, mice were primed with intravenous (i.v.) mAbs or mAb cocktails and challenged i.v. with antigens 2 hours later. Rectal temperature was monitored from 0 to 90 min post-challenge using a digital thermometer (Model BAT-12; Physitemp Instruments, Inc., Clifton, NJ, USA). 2.4. In vivo depletion or inhibition of cells Macrophages and monocytes were depleted via intraperitoneal (i.p.) injection of 0.5 ml/mouse Clo-lipo followed by i.v. injection of 0.2 ml/mouse Clo-lipo 24 hours later. Macrophages and monocytes were inhibited by i.v. injection of 1 mg/mouse GdCl3 in 200 µl normal saline. Basophils were depleted by i.p. injection of 5 µg/mouse MAR-1 in 200 µl PBS twice a day for 3 days. Neutrophils were depleted by i.v. injection of 300 µg/mouse anti-Gr1 in 300 µl PBS. White blood cells (WBCs) in the peripheral blood were depleted by i.p. injection of 4 mg/mouse CPM in 200 µl PBS. PSA was induced at 1 day following cell depletion or at 4 days after CPM injection for WBC depletion. 2.5. In vivo cell receptor blockage PAF receptors were blocked by i.v. injection of 200 µg/mouse CV-3988 in 200 µl PBS containing 5% ethanol 30 minutes before PSA challenge. Histamine receptors were blocked by i.p. injection of 100 µg/mouse triprolidine in 200 µl PBS 30 minutes before PSA challenge. 2.6. Flow cytometry To detect specific cell populations, cells from the blood, bone marrow, spleen and peritoneal cavity of mice were incubated with the indicated fluorescently labelled antibodies at 4°C for 30 minutes, followed by the removal of red blood cells (RBC) using RBC lysis buffer (BioLegend) and analysis using a flow cytometer (FACS Calibur, BD Biosciences, San Jose, CA, USA). Data analyses were 7
performed with FCS Express software (De Novo Software, Los Angeles, CA USA). The phenotypes of cells were identified using specific markers, such as Gr1highCD11b + for neutrophils, CD115+CD11b+ for monocytes, F4/80highCD11b+ for macrophages in the peritoneal cavity, F4/80 +CD11blow for macrophages in the spleen, IgEhighSSClow for basophils in the blood and IgEhighCD49b+ for basophils in the spleen and bone marrow. 2.7. Statistical analysis All analyses were performed with SPSS 13.0 software (IBM Corporation, Armonk, NY, USA). Changes in rectal temperature for mice in different groups were assessed by repeated measurements. The correlation between PAF dose and maximum change in rectal temperature (Figure 6A) was assessed by bivariate correlation analysis. P values of less than 0.05 were considered significant.
8
3. Results 3.1. Depletion of basophils did not affect 2C2-OVM induced PSA To determine the role of basophils in IgG-induced PSA, we depleted basophils with MAR-1, a mAb specific to FcεRIα, and we observed the effects on 2C2 (IgG1)-OVM induced PSA using BALB/c and C57BL/6 mice. We found that the treatment with mAb MAR-1 efficiently depleted basophils in peripheral blood, spleen and bone marrow in BALB/c and C57BL/6 mice (Fig 1, upper panel), yet symptoms of 2C2-OVM induced PSA in both mice strains were unaffected (Fig 1, lower panel). 3.2. Depletion of neutrophils did not affect 2C2-OVM-induced PSA Similarly, we investigated the role of neutrophils by specific lineage depletion using anti-Gr1 mAb. As illustrated in the upper panel of Fig 2, neutrophils were efficiently depleted in the blood, spleen and bone marrow of BALB/c and C57BL/6 mice after anti-Gr1 mAb treatment, whereas symptoms of 2C2-OVM-induced PSA in both mouse strains were not significantly affected (Fig 2, lower panel). 3.3. Depletion of WBCs did not affect 2C2-OVM induced PSA To further clarify the role of basophils, neutrophils and other white blood cells in IgG-induced PSA, we induced leucopenia in mice using CPM, an agent that prevents cell division by cross-linking and breaking DNA strands [18]. We found that the WBC counts, including basophils, neutrophils and monocytes, decreased dramatically following CPM treatment in BALB/c and C57BL/6 mice (Fig 3, left panel). Intriguingly, symptoms of 2C2-OVM-induced PSA were almost unchanged by CPM pretreatment (Fig 3, right panel). Conversely, combined pretreatment using CPM plus Clo-lipo, a depletion agent for monocytes/macrophages, significantly protected both mice strains from 2C2-OVM-induced PSA (Fig 3, right panel). 9
3.4. Depletion and functional blockage of macrophages inhibited 2C2-OVM-induced PSA The use of specific depletion or blocking agents also demonstrated a role for macrophages in IgG-induced PSA. As shown in the upper panel of Fig 4, monocytes in the peripheral blood and macrophages in the spleen and peritoneum were efficiently depleted by Clo-lipo in BALB/c and C57BL/6
mice.
Clo-lipo
pretreatment
alone
drastically attenuated
the
symptoms of
2C2-OVM-induced PSA (Fig 4A, lower panel), consistent with the results of CPM plus Clo-lipo pretreatment shown in the right panel of Fig 3. Moreover, we found that GdCl3, a functional blocker of monocytes/macrophages, could also significantly protect both mouse strains from 2C2-OVM induced PSA (Fig 4B). The results shown in Fig 3 and Fig 4 indicate that macrophages are indispensable for non-lethal PSA induced by 2C2-OVM, while WBCs, including basophils, monocytes and granulocytes, do not play pivotal roles. 3.5. Depletion of macrophages, rather than WBCs, inhibited PSA induced by 5 other Ai-ICs As macrophages are indispensable for PSA induced by 2C2-OVM, an IgG1 mAb-antigen immune complex, we attempted to confirm an important role for macrophages in PSA induced by other Ai-ICs. We selected 5 additional Ai-ICs composed of different complete antigens and corresponding mAb or mAb cocktails for PSA induction as follows: mAb 6H4-OVM (IgG2a-Ag), mAb
1G3&4F12-HSA
(IgG2a&IgG1-Ag),
mAb
H1A6&4A2-HSA
(IgG2b&IgG1-Ag),
M4&M12-MP1P (recombinant mannoprotein 1 of penicillium) (IgG1-Ag-IgG2b) and 3H9&4D11rHb zeta (recombinant human hemoglobin ζ) (IgG1&IgG1-Ag), in addition to mAb 2C2-OVM (IgG1-Ag). PSA initiated by these Ai-ICs could be significantly inhibited by pretreatment with Clo-lipo but not with CPM (Fig 5), similar to 2C2-OVM, as shown in the right panel of Fig 3. 3.6. PAF is a major biological mediator responsible for PSA-related hypothermia PAF, even at very low concentrations, is considered to be an important mediator of shock-related 10
hypotension and hypothermia in both ASA and PSA induced by IgE and IgG antibodies. Here, PAF was used as a positive control for inducing typical hypothermia (Fig 6A). The use of CV-3988, an antagonist of the PAF receptor (Fig 6B), and triprolidine, an antagonist of the histamine receptor (Fig 6C), showed that the major biologic mediator for PSA induced by Ai-IC is PAF, not histamine. 3.7. C3 knock-out did not affect IgG-induced PSA Although IgG immune complexes were reported to act on FcγRs during PSA induction, there is consideration that they might perform their biological activities, including PSA induction, by activating complements. Therefore, C3 -/- mice on C57BL/6 background were used. As illustrated in Fig 7, PSA symptoms remained unchanged in C3 -/- mice compared to WT C57BL/6 mice, indicating that IgG type PSA is not complement dependent.
11
4. Discussion In this study, we ascertained that macrophages are the dominant effector cells in IgG-induced striking but non-lethal PSA challenged with natural protein antigens. Nevertheless, the participation of neutrophils and basophils cannot be completely excluded, as the results in Fig 1 and Fig 2 appeared to show a very slight but not significant protection following anti-Gr1 and MAR-1 treatment. Based on relevant data acquired by our group and four others, as summarized in Table 1, we hypothesize that the conflicting opinion regarding effector cells in PSA may be attributed the use of different antibody-antigen systems and the severity of PSA, which causes FcγR cross-linking of various magnitudes, despite the previous assumption that the disagreement was associated with the IgG isotypes and genetic background of the animals. Furthermore, the nature, number or location of epitopes on antigen, and antibody features including specificity, affinity, isotype, and sialylation might affect the participation of specific effector cells. It is generally accepted that IgG-inducing PSA is provoked by antigen-IgG ICs containing non-monomer IgG to form cross-linking of FcγRs on effector cells in both mice and humans. Administration of polyclonal antibodies capable of binding multiple epitopes on antigen molecules can result in a higher degree of FcγRs cross-linking [19]. A mAb against hapten-carriers that contains multiple identical epitopes, such as DNP or TNP, on carrier proteins, such as BSA or OVA, behaves similarly to a polyclonal antibody. If there is only one epitope on a natural protein recognized by a mAb, then cross-linking of FcγR cannot be initiated, except in the case where multiple similar or repeated epitopes are present on one antigen molecule. Moreover, the mAb cocktails consisting of at least two strains, as listed in Table 1, can initiate cross-liking of FcγR by binding with different epitopes on antigen molecules [20], which could not be as high in magnitude as that induced by mAb-hapten or polyclonal antibody against natural antigen. As Table 1 12
demonstrates, high doses of mAbs or ascites and high doses of hapten conjugates (BPO-GSA, PenV-BSA, DNP-BSA and TNP-OVA) as antigen, as well as natural complete antigen (GPI, with polyclonal antiserum), were used to trigger PSA. These studies consistently find that neutrophils, basophils or monocytes/macrophages were defined as effector cells responsible for PSA by depleting neutrophils, mast cells, basophils, NK and monocytes/macrophages. In the above murine PSA, the high doses of polyclonal antiserum or mAbs against hapten-conjugates with high doses of antigens for challenge may form a great amount of ICs to bind FcγRs expressed on monocyte/macrophages, neutrophils and basophils, inducing a high magnitude FcγRs cross-linking and serious hypothermia or death. Distinctively, in our work, 6 types of ICs consisting of natural protein as complete antigens and corresponding mAb or mAb cocktails in medium to low doses were used to induce PSA, resulting in hypothermia of 2~7°C temperature drop. No deaths occurred in the BALB/c and C57BL/6 mice, in which macrophages are proven to be dominant effector cells. Based on all the above data listed in Table 1, we presume that a great amount of ICs or FcγRs cross-liking in high magnitude may lead to the participation of several types of effector cells in occurrence with PSA, while a low to medium amount of ICs may cause cross-liking of FcγRs at a moderate level, activating only one or two effector cells, particularly macrophages. Moreover, in Khodoun’s report, macrophage depletion showed a greater reduction of hypothermia compared with depleting basophils and neutrophils in C57BL/6 mice, but not in BALB/c mice. Jönsson et al. concluded that both neutrophils and monocytes/macrophages are effector cells mediating PSA; however, their data showed that the reduction of hypothermia by depleting monocytes/macrophages is more significant than that caused by depleting neutrophils or basophils [15]. Additionally, in Jönsson’s work, they demonstrated neutrophils were responsible for inducing anaphylaxis [14]. Depleting neutrophils was also shown to effectively inhibit ASA and PSA, whereas depleting 13
basophils could not alter PSA, and depleting monocytes/macrophages could not alter ASA, and no data were reported regarding the effects of macrophage depletion on PSA [14]. Aside from the degree of FcγR cross-linking, the affinity, isotype or sialylation of mAb can also impact the occurrence and severity of PSA. For instance, murine IgG2a with a high activation/inhibition (A/I) ratio can bind with FcγRIV, and sialylation of IgG may function as anti-inflammation treatment [21]. Notably, neutrophils and basophils may participate in serious hypothermia and lower survival in the PSA models by the antibody-hapten conjugates [13, 14, 16], whereas macrophages as dominant effector cells only induced remarkable hypothermia but seldom death in PSA models by low to medium doses of mAb-natural protein antigens. Concerning the role of the IgG isotype, it appears to have no significant impact on the type of effector cells based on the results of our work performed with 6 ICs consisting of mAb IgG1, IgG2a and IgG2b, as well as the work from other teams summarized in Table 1. The association of the genetic backgrounds of inbred mice with the type of effector cells involved in IgG-induced PSA was also considered. Aside from some gene knockout and transgenic mice used for special purposes, most studies only used a single strain of mice to verify the major effector cells [9, 13-15]. Specifically, BALB/c or C57BL/6 mice were typically used. Khodoun established PSA models that showed some differences in survival and temperature drop after cell depletion in BALB/c or C57BL/6 mice, but no remarkable influence on the type of effector cells. The results in our present work suggest no differences of IgG-induced PSA between BALB/c and C57BL/6 mice. The same result was also observed in Swiss hybrid mice (data not shown). Only the work reported by Kimura showed the type of effector cells in PSA associated with genetic background as listed in Table 1, in which anti-Gr mAb could inhibit the PSA in WBB6F1-W:Wv mice, but not in CTS and 14
DS mice [22]. In addition to effector cells, we also confirmed that PAF, rather than histamine, was the major mediator molecule in IgG-induced PSA. These results suggest functional suppression or blockade of PAF as a potential strategy for the therapy of IgG-induced PSA. Moreover, we excluded the involvement of an important complement component C3 in IgG-induced PSA, indicating that IgG-induced anaphylaxis is independent of complement activation and distinguishes IgG-induced PSA from systemic type III hypersensitivities such as serum sickness. In conclusion, we demonstrated that macrophages were the major effector cells for IgG-induced PSA and that PAF is the major effector molecule. Importantly, our work indicated that the occurrence of PSA is not associated with the type of antigen, the isotype of the monoclonal antibody in immune complexes, or genetic variations of mouse strains. These data provide integrated evidence of the PSA mechanism distinguishing from relevant publications in existence.
15
Acknowledgements This work is supported by a research grant from the National Natural Science Foundation of China (30671970 to Ning Fu). We also acknowledge Prof. Zou Qiang for providing C3-/- mice.
Conflict of interest The authors declare no conflict of interest.
16
References [1] D.A. Moneret-Vautrin, M. Morisset, J. Flabbee, E. Beaudouin, and G. Kanny, Epidemiology of life-threatening and lethal anaphylaxis: a review. Allergy 60 (2005) 443-51. [2] T.T. Hansel, H. Kropshofer, T. Singer, J.A. Mitchell, and A.J. George, The safety and side effects of monoclonal antibodies. Nat Rev Drug Discov 9 (2010) 325-38. [3] A. Daguet, and H. Watier, 2nd Charles Richet et Jules Hericourt Workshop: Therapeutic antibodies and anaphylaxis; May 31-June 1, 2011, Tours, France. MAbs 3 (2011) 417-21. [4] D.R. McCluskey, and N.A. Boyd, Anaphylaxis with intravenous gammaglobulin. Lancet 336 (1990) 874. [5] K.L. Kolho, T. Ruuska, and E. Savilahti, Severe adverse reactions to Infliximab therapy are common in young children with inflammatory bowel disease. Acta Paediatr 96 (2007) 128-30. [6] A. Vultaggio, A. Matucci, F. Nencini, S. Pratesi, P. Parronchi, O. Rossi, S. Romagnani, and E. Maggi, Anti-infliximab IgE and non-IgE antibodies and induction of infusion-related severe anaphylactic reactions. Allergy 65 (2010) 657-61. [7] http://www.phrma.org/sites/default/files/1776/biotech2011.pdf [8] E. Maggi, A. Vultaggio, and A. Matucci, Acute infusion reactions induced by monoclonal antibody therapy. Expert Rev Clin Immunol 7 (2011) 55-63. [9] R.T. Strait, S.C. Morris, M. Yang, X.W. Qu, and F.D. Finkelman, Pathways of anaphylaxis in the mouse. J Allergy Clin Immunol 109 (2002) 658-68. [10] F.D. Finkelman, M.E. Rothenberg, E.B. Brandt, S.C. Morris, and R.T. Strait, Molecular mechanisms of anaphylaxis: Lessons from studies with murine models. Journal of Allergy and Clinical Immunology 115 (2005) 449-457. [11] F.D. Finkelman, Anaphylaxis: lessons from mouse models. J Allergy Clin Immunol 120 (2007) 17
506-15; quiz 516-7. [12] K. Mukai, K. Obata, Y. Tsujimura, and H. Karasuyama, New insights into the roles for basophils in acute and chronic allergy. Allergol Int 58 (2009) 11-9. [13] Y. Tsujimura, K. Obata, K. Mukai, H. Shindou, M. Yoshida, H. Nishikado, Y. Kawano, Y. Minegishi,
T.
Shimizu,
and
H.
Karasuyama,
Basophils
play
a
pivotal
role
in
immunoglobulin-G-mediated but not immunoglobulin-E-mediated systemic anaphylaxis. Immunity 28 (2008) 581-9. [14] F. Jonsson, D.A. Mancardi, Y. Kita, H. Karasuyama, B. Iannascoli, N. Van Rooijen, T. Shimizu, M. Daeron, and P. Bruhns, Mouse and human neutrophils induce anaphylaxis. J Clin Invest 121 (2011) 1484-96. [15] F. Jonsson, D.A. Mancardi, W. Zhao, Y. Kita, B. Iannascoli, H. Khun, N. van Rooijen, T. Shimizu, L.B. Schwartz, M. Daeron, and P. Bruhns, Human FcgammaRIIA induces anaphylactic and allergic reactions. Blood 119 (2012) 2533-44. [16] M.V. Khodoun, Z.Y. Kucuk, R.T. Strait, D. Krishnamurthy, K. Janek, C.D. Clay, S.C. Morris, and F.D. Finkelman, Rapid desensitization of mice with anti-FcgammaRIIb/FcgammaRIII mAb safely prevents IgG-mediated anaphylaxis. J Allergy Clin Immunol 132 (2013) 1375-87. [17] D. Jiao, Y. Liu, X. Lu, Q. Pan, J. Zheng, B. Liu, Y. Liu, Y. Wang, and N. Fu, Characteristics of anaphylaxis-inducing IgG immune complexes triggering murine passive systemic anaphylaxis. Allergy 68 (2013) 236-45. [18] J.Y. Wang, G. Prorok, and W.P. Vaughan, Cytotoxicity, DNA cross-linking, and DNA single-strand breaks induced by cyclophosphamide in a rat leukemia in vivo. Cancer Chemother Pharmacol 31 (1993) 381-6. [19] F. Nimmerjahn, and J.V. Ravetch, Analyzing antibody-Fc-receptor interactions. Methods Mol 18
Biol 415 (2008) 151-62. [20] R. Ishikawa, Y. Tsujimura, K. Obata, Y. Kawano, Y. Minegishi, and H. Karasuyama, IgG-mediated systemic anaphylaxis to protein antigen can be induced even under conditions of limited amounts of antibody and antigen. Biochem Biophys Res Commun 402 (2010) 742-6. [21] Y. Kaneko, F. Nimmerjahn, and J.V. Ravetch, Anti-inflammatory activity of immunoglobulin G resulting from Fc sialylation. Science 313 (2006) 670-3. [22] S. Kimura, M. Nagata, M. Takeuchi, K. Takano, and M. Harada, Anti-granulocyte antibody suppression of active and passive anaphylactic shock in WBB6F1-W/Wv mice. Cell Mol Life Sci 53 (1997) 663-6.
19
Figure legends Fig 1. Effects of basophils depletion on 2C2-OVM-induced PSA. BALB/c and C57BL/6 mice (n=3 each unless specified otherwise) were subject to MAR-1 or isotype control (ISO) treatment, and the basophils in blood, spleen and bone marrow were then evaluated by flow cytometry analysis. Basophil depletion using MAR-1 was also established before the induction of PSA by 2C2-OVM while changes in rectal temperature were monitored and shock symptoms evaluated. Data are representative of 3 independent experiments. Fig 2. Influences of neutrophil depletion on 2C2-OVM induced PSA. BALB/c and C57BL/6 mice were subject to anti-Gr1 or its ISO treatment, and then neutrophils in the blood, spleen and bone marrow were evaluated by flow cytometry analysis. Neutrophil depletion using anti-Gr1 was also established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments. Fig 3. Impacts of WBC depletion on 2C2-OVM induced PSA. BALB/c and C57BL/6 mice were subject to CPM or control (PBS) treatment. WBCs were then counted using a hemocytometer, and neutrophils, monocytes and basophils in blood were analyzed by flow cytometry. Depletion of WBCs using CPM with or without Clo-lipo was established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 3 independent experiments. Fig 4. Effects of monocyte/macrophage depletion or inhibition on 2C2-OVM-induced PSA. BALB/c and C57BL/6 mice were subject to Clo-lipo or PBS-lipo treatment, and then monocytes in blood and macrophages in the spleen and peritoneum were evaluated by flow cytometry (A, upper panel). Monocyte/macrophage depletion using Clo-lipo (A, lower panel) and inhibition using GdCl3 20
(B) were established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of at least 3 independent experiments. Fig 5. Impacts of monocyte/macrophage depletion and WBC depletion on PSA induced by five additional IgG-antigen complexes. Monocytes/macrophages depletion using Clo-lipo and WBC depletion using CPM were established before the induction of PSA by five other IgG-antigen complexes in BALB/c mice, while hypothermia was monitored as previously described. Superscripts on mAb names indicate the IgG isotypes. Fig 6. Role of PAF and histamine in 2C2-OVM-induced PSA. (A) BALB/c mice were i.v. injected with PAF at different doses, and hypothermia was measured. (B and C). CV-3988 and Triprolidine pretreatment was conducted before 2C2-OVM-induced PSA in BALB/c mice, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments. Fig 7. Role of C3 in 2C2-OVM-induced PSA. PSA was established in C3-/- (n=3 each) and WT (n=4 each) C57BL/6 mice by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments.
21
Table 1. Reports concerning the effector cells in IgG-induced murine PSAa) Publicatio n
–
Reference
In their/our work Effector
cell
type proved
Mice
PSA Protocol b)
strain
Ab for priming
Methods for effector cell research
Antigen for challenge
no. Kimura S
Neutrophil
1997 [22]
Tsujimura
v
1
WBB6F1-W:W
anti-BPO mAb ascites
mice
(200µl)
Not
CTS
neutrophil
mice
and
DS
Basophil
C57BL/6 mice
2008 [13]
Jönsson
c)
BSA antiserum (200µl)
BPO-GSA (1mg)
PenV-BSA (500µg)
(1000µg)
Neutrophil
2011 [14]
WT and 5KO
mAb
C57BL/6 mice
(200µg),
anti-DNP
2b
K/BxN serum anti-GPI
DNP-HSA (500µg)
GPI (40µg)
(20µl) anti-OVA1
Jönsson
Macrophage
WT,
3KOIIA
mAb
2012 [15]
and
and
5KOIIA
(1000µg),
neutrophil
C57BL/6 mice
K/BxN serum anti-GPI
OVA (1000µg)
Monocyte/ma
BALB/c
and
2013 [16]
crophage,
C57BL/6 mice
anti-TNP2a
mAb
methods
Neutrophil
anti-Gr1 mAb
Mast cell
WBB6F1-W:Wv strain mice
Basophil
anti-CD200R3 mAb
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo
NK cell
anti-NK1.1 mAb
Neutrophil
anti-Gr1 mAb
Monocyte/
Clo-lipo; GdCl3
Neutrophil
anti-Gr1 mAb
Basophil
anti-CD200R3 mAb
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo; GdCl3
Neutrophil
anti-Gr1 and Anti-Ly6G
GPI (50ug)
(20µl)
Khodoun
Depletion/inhibition
depleted/inhibited
BSA (1mg)
anti-e)PenV1
mAb
d)
Lineage
TNP-OVA (500µg)
(100µg)
mAb Basophil
anti-CD200R3
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo; GdCl3
Neutrophil
anti-Gr1
basophil, and
Macrophage
work
mAb
hydroxyurea
neutrophil Our present
mAb
BALB/c
and
C57BL/6 mice
Basophil
anti-CD200R3 mAb
OVM (10µg)
Basophil
anti-FcεR1α mAb
6H4 (100µg)
OVM (10µg)
Neutrophil
anti-Gr1 mAb
1G32a (50µg) & 4F121
HSA (100µg)
Monocyte/Macrophage
Clo-lipo; GdCl3
White blood cell
CPM
1
2a
2C2 (100µg)
(50µg) 2b
1
HSA (100µg)
M41 (50µg) & M122b
MP1P (50µg)
H1A6
(50µg) & 4A2
(50µg)
(50µg) 3H91 (50µg) & 4D111
rHbζ (100µg)
(50µg) a)
Only PSA data summarized in table, though there are some data on ASA in above papers. Superscripts on mAb and mAb ascites indicate the IgG isotypes. 1: IgG1; 2a: IgG2a, 2b:IgG2b. c) BPO, benzylpenicilloy; d) GSA, guinea pig serum albumin; e)PenV, penicillin V. b)
22
and
24
25
26
27
28
30
31
1. Macrophages are the dominant effector cells responsible for IgG-mediated anaphylaxis. 2. The type of effector cell may be associated with different antigen-antibody systems. 3. The type of effector cell is not affected by IgG isotype and genetic background of mice. 4. PAF is the major mediator responsible for IgG-induced anaphylaxis. 5. IgG-mediated anaphylaxis is independent of complete third component.
32
S0008-8749(14)00063-X http://dx.doi.org/10.1016/j.cellimm.2014.03.018 YCIMM 3310
To appear in:
Cellular Immunology
Received Date: Accepted Date:
17 February 2014 31 March 2014
Please cite this article as: D. Jiao, Y. Liu, X. Lu, B. Liu, Q. Pan, Y. Liu, Y. Liu, P. Zhu, N. Fu, Macrophages are the dominant effector cells responsible for IgG-mediated passive systemic anaphylaxis challenged by natural protein antigen in BALB/c and C57BL/6 mice, Cellular Immunology (2014), doi: http://dx.doi.org/10.1016/j.cellimm. 2014.03.018
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Macrophages are the dominant effector cells responsible for IgG-mediated passive systemic anaphylaxis challenged by natural protein antigen in BALB/c and C57BL/6 mice Authors: Delong Jiao a, Yuan Liua, b, Xiao Lu a, Beiyi Liu a, Qingjun Pana, c, Yanjun Liu a, Yang Liud, Ping Zhu a, Ning Fua,* Affiliations: a
Department of Immunology, School of Basic Medical Sciences, Southern Medical University,
Guangzhou, China; b
Present address: Medical genetic center, Guangdong Women and Children Hospital, Guangzhou,
China; c
Present address: Institute of Nephrology, Guangdong Medical College, Zhanjiang, China;
d
Department of Immunology, Chengdu Medical College, Chengdu, China
* Corresponding author Ning Fu, Department of Immunology, Southern Medical University, 510515 Guangzhou, Guangdong, P. R. China. Tel: 86-20-61640405. E-mail: [email protected]
1
Abstract IgG-induced passive systemic anaphylaxis (PSA), a serious adverse effect of passive immune therapy using therapeutic monoclonal antibodies, has been greatly emphasized. However, controversy exists regarding the type of effector cells involved in IgG-induced anaphylaxis as a result of the induction of PSA by different IgG subtypes or the use of mice with varying genetic backgrounds. To clarify the effector cells for PSA, the PSA model with serious hypothermia was established by IgG monoclonal antibody (mAb) against natural protein or complete antigen, not hapten conjugate, in BALB/c and C57BL/6 mice. The results indicated that PSA could be remarkably inhibited by the depletion of macrophages but not by the depletion of whole leukocytes, basophils, neutrophils or monocytes. We further confirmed that macrophages are indispensable for the PSA induced by all six IgG-natural antigen complexes in both strains of mice. Additionally, platelet-activating factor (PAF) was found to be the major effector mediator for IgG-induced anaphylaxis. Moreover, gene knock-out of the complete third component (C3) did not affect PSA-related hypothermia in C57BL/6 mice. These results indicate that macrophages and PAF act as dominant effector cells and mediator molecules, respectively, and are indispensable components in the induction of IgG-mediated PSA induced by IgG mAb and natural protein antigen. Based on the above results, we hypothesize that inconsistencies in effector cells for PSA may be associated with different features of the mAb-antigen system that might affect the magnitude of FcγRs cross-linking on effector cells. Keywords: IgG; passive systemic anaphylaxis; mouse model; macrophage; platelet activating factor
2
1. Introduction Both active systemic anaphylaxis (ASA), which occurs in hosts immunized with antigen, and passive systemic anaphylaxis (PSA), which occurs in hosts following the infusion of therapeutic antibodies or immunoglobulins, are life-threatening allergic reactions [1-4]. Recently, IgG-induced PSA has attracted significant attention, as allergic reactions are severe potential adverse effects of therapy with monoclonal antibodies (mAbs) and immunoglobulins [2, 3, 5, 6], both of which are practiced extensively in the clinic. Moreover, over 30 therapeutic mAbs for clinical administration have been approved by the FDA, and more than 300 therapeutic mAbs and immunoglobulin reagents have been tested in clinical trials [7]. The frequency of allergic reaction, including PSA, ranges from 0.09 to 77% depending on the type of antibody administered [8]. Therefore, a better understanding of PSA is necessary to elucidate the underlying mechanisms of anaphylaxis and to create a new strategy to evaluate the safety of therapeutic antibody administration to prevent these common allergic reactions observed clinically. In mouse models, anaphylaxis can be mediated through a classical pathway that involves IgE, Fcε receptors (FcεRs), mast cells, histamine and leukotrienes, as well as through an alternative pathway that involves IgG, Fcγ receptors (FcγRs), macrophages or other cells and platelet activating factor (PAF) [9, 10]. Similarly, IgG, FcγRs, macrophages and PAF are thought to play
PSA: passive systemic anaphylaxis; ASA: active systemic anaphylaxis; mAbs: monoclonal antibodies; FcεRs: Fcε receptors; FcγRs: Fcγ receptors; PAF: platelet-activating factor; Ai-ICs: anaphylaxis-inducing immune complexes; C3: complement third component; C3-/-: complement third component knock-out; WT: wild type; OVM: ovomucoid; HSA: human serum albumin; OVA: ovalbumin; BSA: bovine serum albumin; TNP: 2,4,6-trinitrophenol; DNP: 2,4-dinitrophenol; PenV: penicillin V; GaMD: goat anti-mouse IgD; GPI: rabbit glucose-6-phosphate isomerase; CPM: cyclophosphamide; PBS-lipo: PBS encapsulated liposomes; Clo-lipo: Clodronate encapsulated liposomes; WBCs: white blood cells; RBCs: red blood cells; BPO, benzylpenicilloyl; GSA, guinea pig serum albumin. 3
pivotal roles in IgG-induced anaphylaxis in humans, suggesting that activation of the alternative pathway occurs in both mice and humans [11]. However, the mechanisms underlying IgG-induced PSA remain elusive, particularly because there are conflicting viewpoints regarding the type of effector cells involved. Mukai et al. proposed that basophils are the main effector cells in alternative pathway 1 and that macrophages are the main effector cells in alternative pathway 2 [12]. Tsujimura et al. demonstrated that basophils are the effector cells for PSA [13], whereas Jönsson et al. indicated that neutrophils and monocytes/macrophages are the most important effector cells for ASA and PSA in mouse models [14, 15]. The latest publication of Khodoun et al., covering all three known effector cell types, concluded that all macrophages, basophils and neutrophils participated in IgG-induced anaphylaxis [16]. The general consensus is that the divergence regarding effector cell types in IgG-mediated PSA may be due to IgG isotypes and variations in the genetic background of inbred mice. In the present work, we focused on the effector cells responsible for IgG-induced PSA by IgG mAb against the natural protein antigen that is more common clinically, compared with hapten conjugates such as DNP and TNP used in murine PSA models. In our previous work, we raised and proved that only a fraction of IgG immune complexes, defined as anaphylaxis-inducing immune complexes (Ai-ICs), are responsible for mediating murine IgG-induced PSA, which was then confirmed using 36 combinatory reactions with 8 natural proteins or complete antigens and 27 strains of corresponding mAbs [17]. In 14 ICs capable of inducing PSA, a remarkable temperature drop occurred, but seldom death, although 2~3-folds mAbs and 2~10-folds allergen were used in the preliminary experiments, which were considered non-lethal PSA. Here, by using 6 kinds of Ai-ICs containing various natural complete antigens, not hapten conjugates, and IgG mAbs or mAb cocktails of IgG1, IgG2a and IgG2b isotypes, we demonstrate that the dominant effector cells 4
mediating PSA in BALB/c and C57BL/6 mice are macrophages rather than basophils, neutrophils or monocytes. Consequently, we hypothesize that different features and dosages of the mAb-antigen system might be associated with the effector cell type by affecting the magnitude of FcγR cross-linking.
5
2. Materials and methods 2.1. Mice Female BALB/c and C57BL/6 mice (6~10-week-old) were purchased from the Laboratory Animal Center of Sun Yat-Sen University (Guangzhou, Guangdong, China). The wild type (WT) and complement third component knock-out (C3-/-) C57BL/6 mice used for complement third component (C3) dependent study were provided by Dr. Qiang Zou (Chengdu Medical College, Chengdu, Sichuan, China). The animal experiments were approved by the Ethics Committee for Experimental Animals at Southern Medical University and were performed according to the national guidelines for animal welfare. 2.2, Reagents and antibodies Ovomucoid (OVM), human serum albumin (HSA), Triprolidine, CV-3988, gadolinium chloride (GdCl3) and PAF were purchased from Sigma-Aldrich (St. Louis, MO, USA), and cyclophosphamide (CPM) was purchased from Lingnan Pharmaceutical Ltd (Guangzhou, Guangdong, China). Clodronate and PBS encapsulated liposomes (Clo-lipo and PBS-lipo) were purchased from ClodronateLiposome.org (Amsterdam, The Netherlands). The antibodies used for flow cytometry analysis against mouse Gr-1, IgE, CD49b and F4/80 were obtained from BioLegend (San Diego, CA, USA), and anti-mouse CD11b was purchased from eBioscience (San Diego, CA, USA). Anti-Gr-1 (Rat IgG2b) used for neutrophil depletion was purchased from BioLegend, and MAR-1 (hamster IgG) used for basophil depletion was purchased from eBioscience. Rat IgG2b and hamster IgG isotype control antibodies were also purchased from eBioscience. All mAbs and antigens for the induction of PSA have been used in previously published work [17]. 2.3. PSA model 6
PSA models using 6 Ai-ICs have been previously reported. Briefly, mice were primed with intravenous (i.v.) mAbs or mAb cocktails and challenged i.v. with antigens 2 hours later. Rectal temperature was monitored from 0 to 90 min post-challenge using a digital thermometer (Model BAT-12; Physitemp Instruments, Inc., Clifton, NJ, USA). 2.4. In vivo depletion or inhibition of cells Macrophages and monocytes were depleted via intraperitoneal (i.p.) injection of 0.5 ml/mouse Clo-lipo followed by i.v. injection of 0.2 ml/mouse Clo-lipo 24 hours later. Macrophages and monocytes were inhibited by i.v. injection of 1 mg/mouse GdCl3 in 200 µl normal saline. Basophils were depleted by i.p. injection of 5 µg/mouse MAR-1 in 200 µl PBS twice a day for 3 days. Neutrophils were depleted by i.v. injection of 300 µg/mouse anti-Gr1 in 300 µl PBS. White blood cells (WBCs) in the peripheral blood were depleted by i.p. injection of 4 mg/mouse CPM in 200 µl PBS. PSA was induced at 1 day following cell depletion or at 4 days after CPM injection for WBC depletion. 2.5. In vivo cell receptor blockage PAF receptors were blocked by i.v. injection of 200 µg/mouse CV-3988 in 200 µl PBS containing 5% ethanol 30 minutes before PSA challenge. Histamine receptors were blocked by i.p. injection of 100 µg/mouse triprolidine in 200 µl PBS 30 minutes before PSA challenge. 2.6. Flow cytometry To detect specific cell populations, cells from the blood, bone marrow, spleen and peritoneal cavity of mice were incubated with the indicated fluorescently labelled antibodies at 4°C for 30 minutes, followed by the removal of red blood cells (RBC) using RBC lysis buffer (BioLegend) and analysis using a flow cytometer (FACS Calibur, BD Biosciences, San Jose, CA, USA). Data analyses were 7
performed with FCS Express software (De Novo Software, Los Angeles, CA USA). The phenotypes of cells were identified using specific markers, such as Gr1highCD11b + for neutrophils, CD115+CD11b+ for monocytes, F4/80highCD11b+ for macrophages in the peritoneal cavity, F4/80 +CD11blow for macrophages in the spleen, IgEhighSSClow for basophils in the blood and IgEhighCD49b+ for basophils in the spleen and bone marrow. 2.7. Statistical analysis All analyses were performed with SPSS 13.0 software (IBM Corporation, Armonk, NY, USA). Changes in rectal temperature for mice in different groups were assessed by repeated measurements. The correlation between PAF dose and maximum change in rectal temperature (Figure 6A) was assessed by bivariate correlation analysis. P values of less than 0.05 were considered significant.
8
3. Results 3.1. Depletion of basophils did not affect 2C2-OVM induced PSA To determine the role of basophils in IgG-induced PSA, we depleted basophils with MAR-1, a mAb specific to FcεRIα, and we observed the effects on 2C2 (IgG1)-OVM induced PSA using BALB/c and C57BL/6 mice. We found that the treatment with mAb MAR-1 efficiently depleted basophils in peripheral blood, spleen and bone marrow in BALB/c and C57BL/6 mice (Fig 1, upper panel), yet symptoms of 2C2-OVM induced PSA in both mice strains were unaffected (Fig 1, lower panel). 3.2. Depletion of neutrophils did not affect 2C2-OVM-induced PSA Similarly, we investigated the role of neutrophils by specific lineage depletion using anti-Gr1 mAb. As illustrated in the upper panel of Fig 2, neutrophils were efficiently depleted in the blood, spleen and bone marrow of BALB/c and C57BL/6 mice after anti-Gr1 mAb treatment, whereas symptoms of 2C2-OVM-induced PSA in both mouse strains were not significantly affected (Fig 2, lower panel). 3.3. Depletion of WBCs did not affect 2C2-OVM induced PSA To further clarify the role of basophils, neutrophils and other white blood cells in IgG-induced PSA, we induced leucopenia in mice using CPM, an agent that prevents cell division by cross-linking and breaking DNA strands [18]. We found that the WBC counts, including basophils, neutrophils and monocytes, decreased dramatically following CPM treatment in BALB/c and C57BL/6 mice (Fig 3, left panel). Intriguingly, symptoms of 2C2-OVM-induced PSA were almost unchanged by CPM pretreatment (Fig 3, right panel). Conversely, combined pretreatment using CPM plus Clo-lipo, a depletion agent for monocytes/macrophages, significantly protected both mice strains from 2C2-OVM-induced PSA (Fig 3, right panel). 9
3.4. Depletion and functional blockage of macrophages inhibited 2C2-OVM-induced PSA The use of specific depletion or blocking agents also demonstrated a role for macrophages in IgG-induced PSA. As shown in the upper panel of Fig 4, monocytes in the peripheral blood and macrophages in the spleen and peritoneum were efficiently depleted by Clo-lipo in BALB/c and C57BL/6
mice.
Clo-lipo
pretreatment
alone
drastically attenuated
the
symptoms of
2C2-OVM-induced PSA (Fig 4A, lower panel), consistent with the results of CPM plus Clo-lipo pretreatment shown in the right panel of Fig 3. Moreover, we found that GdCl3, a functional blocker of monocytes/macrophages, could also significantly protect both mouse strains from 2C2-OVM induced PSA (Fig 4B). The results shown in Fig 3 and Fig 4 indicate that macrophages are indispensable for non-lethal PSA induced by 2C2-OVM, while WBCs, including basophils, monocytes and granulocytes, do not play pivotal roles. 3.5. Depletion of macrophages, rather than WBCs, inhibited PSA induced by 5 other Ai-ICs As macrophages are indispensable for PSA induced by 2C2-OVM, an IgG1 mAb-antigen immune complex, we attempted to confirm an important role for macrophages in PSA induced by other Ai-ICs. We selected 5 additional Ai-ICs composed of different complete antigens and corresponding mAb or mAb cocktails for PSA induction as follows: mAb 6H4-OVM (IgG2a-Ag), mAb
1G3&4F12-HSA
(IgG2a&IgG1-Ag),
mAb
H1A6&4A2-HSA
(IgG2b&IgG1-Ag),
M4&M12-MP1P (recombinant mannoprotein 1 of penicillium) (IgG1-Ag-IgG2b) and 3H9&4D11rHb zeta (recombinant human hemoglobin ζ) (IgG1&IgG1-Ag), in addition to mAb 2C2-OVM (IgG1-Ag). PSA initiated by these Ai-ICs could be significantly inhibited by pretreatment with Clo-lipo but not with CPM (Fig 5), similar to 2C2-OVM, as shown in the right panel of Fig 3. 3.6. PAF is a major biological mediator responsible for PSA-related hypothermia PAF, even at very low concentrations, is considered to be an important mediator of shock-related 10
hypotension and hypothermia in both ASA and PSA induced by IgE and IgG antibodies. Here, PAF was used as a positive control for inducing typical hypothermia (Fig 6A). The use of CV-3988, an antagonist of the PAF receptor (Fig 6B), and triprolidine, an antagonist of the histamine receptor (Fig 6C), showed that the major biologic mediator for PSA induced by Ai-IC is PAF, not histamine. 3.7. C3 knock-out did not affect IgG-induced PSA Although IgG immune complexes were reported to act on FcγRs during PSA induction, there is consideration that they might perform their biological activities, including PSA induction, by activating complements. Therefore, C3 -/- mice on C57BL/6 background were used. As illustrated in Fig 7, PSA symptoms remained unchanged in C3 -/- mice compared to WT C57BL/6 mice, indicating that IgG type PSA is not complement dependent.
11
4. Discussion In this study, we ascertained that macrophages are the dominant effector cells in IgG-induced striking but non-lethal PSA challenged with natural protein antigens. Nevertheless, the participation of neutrophils and basophils cannot be completely excluded, as the results in Fig 1 and Fig 2 appeared to show a very slight but not significant protection following anti-Gr1 and MAR-1 treatment. Based on relevant data acquired by our group and four others, as summarized in Table 1, we hypothesize that the conflicting opinion regarding effector cells in PSA may be attributed the use of different antibody-antigen systems and the severity of PSA, which causes FcγR cross-linking of various magnitudes, despite the previous assumption that the disagreement was associated with the IgG isotypes and genetic background of the animals. Furthermore, the nature, number or location of epitopes on antigen, and antibody features including specificity, affinity, isotype, and sialylation might affect the participation of specific effector cells. It is generally accepted that IgG-inducing PSA is provoked by antigen-IgG ICs containing non-monomer IgG to form cross-linking of FcγRs on effector cells in both mice and humans. Administration of polyclonal antibodies capable of binding multiple epitopes on antigen molecules can result in a higher degree of FcγRs cross-linking [19]. A mAb against hapten-carriers that contains multiple identical epitopes, such as DNP or TNP, on carrier proteins, such as BSA or OVA, behaves similarly to a polyclonal antibody. If there is only one epitope on a natural protein recognized by a mAb, then cross-linking of FcγR cannot be initiated, except in the case where multiple similar or repeated epitopes are present on one antigen molecule. Moreover, the mAb cocktails consisting of at least two strains, as listed in Table 1, can initiate cross-liking of FcγR by binding with different epitopes on antigen molecules [20], which could not be as high in magnitude as that induced by mAb-hapten or polyclonal antibody against natural antigen. As Table 1 12
demonstrates, high doses of mAbs or ascites and high doses of hapten conjugates (BPO-GSA, PenV-BSA, DNP-BSA and TNP-OVA) as antigen, as well as natural complete antigen (GPI, with polyclonal antiserum), were used to trigger PSA. These studies consistently find that neutrophils, basophils or monocytes/macrophages were defined as effector cells responsible for PSA by depleting neutrophils, mast cells, basophils, NK and monocytes/macrophages. In the above murine PSA, the high doses of polyclonal antiserum or mAbs against hapten-conjugates with high doses of antigens for challenge may form a great amount of ICs to bind FcγRs expressed on monocyte/macrophages, neutrophils and basophils, inducing a high magnitude FcγRs cross-linking and serious hypothermia or death. Distinctively, in our work, 6 types of ICs consisting of natural protein as complete antigens and corresponding mAb or mAb cocktails in medium to low doses were used to induce PSA, resulting in hypothermia of 2~7°C temperature drop. No deaths occurred in the BALB/c and C57BL/6 mice, in which macrophages are proven to be dominant effector cells. Based on all the above data listed in Table 1, we presume that a great amount of ICs or FcγRs cross-liking in high magnitude may lead to the participation of several types of effector cells in occurrence with PSA, while a low to medium amount of ICs may cause cross-liking of FcγRs at a moderate level, activating only one or two effector cells, particularly macrophages. Moreover, in Khodoun’s report, macrophage depletion showed a greater reduction of hypothermia compared with depleting basophils and neutrophils in C57BL/6 mice, but not in BALB/c mice. Jönsson et al. concluded that both neutrophils and monocytes/macrophages are effector cells mediating PSA; however, their data showed that the reduction of hypothermia by depleting monocytes/macrophages is more significant than that caused by depleting neutrophils or basophils [15]. Additionally, in Jönsson’s work, they demonstrated neutrophils were responsible for inducing anaphylaxis [14]. Depleting neutrophils was also shown to effectively inhibit ASA and PSA, whereas depleting 13
basophils could not alter PSA, and depleting monocytes/macrophages could not alter ASA, and no data were reported regarding the effects of macrophage depletion on PSA [14]. Aside from the degree of FcγR cross-linking, the affinity, isotype or sialylation of mAb can also impact the occurrence and severity of PSA. For instance, murine IgG2a with a high activation/inhibition (A/I) ratio can bind with FcγRIV, and sialylation of IgG may function as anti-inflammation treatment [21]. Notably, neutrophils and basophils may participate in serious hypothermia and lower survival in the PSA models by the antibody-hapten conjugates [13, 14, 16], whereas macrophages as dominant effector cells only induced remarkable hypothermia but seldom death in PSA models by low to medium doses of mAb-natural protein antigens. Concerning the role of the IgG isotype, it appears to have no significant impact on the type of effector cells based on the results of our work performed with 6 ICs consisting of mAb IgG1, IgG2a and IgG2b, as well as the work from other teams summarized in Table 1. The association of the genetic backgrounds of inbred mice with the type of effector cells involved in IgG-induced PSA was also considered. Aside from some gene knockout and transgenic mice used for special purposes, most studies only used a single strain of mice to verify the major effector cells [9, 13-15]. Specifically, BALB/c or C57BL/6 mice were typically used. Khodoun established PSA models that showed some differences in survival and temperature drop after cell depletion in BALB/c or C57BL/6 mice, but no remarkable influence on the type of effector cells. The results in our present work suggest no differences of IgG-induced PSA between BALB/c and C57BL/6 mice. The same result was also observed in Swiss hybrid mice (data not shown). Only the work reported by Kimura showed the type of effector cells in PSA associated with genetic background as listed in Table 1, in which anti-Gr mAb could inhibit the PSA in WBB6F1-W:Wv mice, but not in CTS and 14
DS mice [22]. In addition to effector cells, we also confirmed that PAF, rather than histamine, was the major mediator molecule in IgG-induced PSA. These results suggest functional suppression or blockade of PAF as a potential strategy for the therapy of IgG-induced PSA. Moreover, we excluded the involvement of an important complement component C3 in IgG-induced PSA, indicating that IgG-induced anaphylaxis is independent of complement activation and distinguishes IgG-induced PSA from systemic type III hypersensitivities such as serum sickness. In conclusion, we demonstrated that macrophages were the major effector cells for IgG-induced PSA and that PAF is the major effector molecule. Importantly, our work indicated that the occurrence of PSA is not associated with the type of antigen, the isotype of the monoclonal antibody in immune complexes, or genetic variations of mouse strains. These data provide integrated evidence of the PSA mechanism distinguishing from relevant publications in existence.
15
Acknowledgements This work is supported by a research grant from the National Natural Science Foundation of China (30671970 to Ning Fu). We also acknowledge Prof. Zou Qiang for providing C3-/- mice.
Conflict of interest The authors declare no conflict of interest.
16
References [1] D.A. Moneret-Vautrin, M. Morisset, J. Flabbee, E. Beaudouin, and G. Kanny, Epidemiology of life-threatening and lethal anaphylaxis: a review. Allergy 60 (2005) 443-51. [2] T.T. Hansel, H. Kropshofer, T. Singer, J.A. Mitchell, and A.J. George, The safety and side effects of monoclonal antibodies. Nat Rev Drug Discov 9 (2010) 325-38. [3] A. Daguet, and H. Watier, 2nd Charles Richet et Jules Hericourt Workshop: Therapeutic antibodies and anaphylaxis; May 31-June 1, 2011, Tours, France. MAbs 3 (2011) 417-21. [4] D.R. McCluskey, and N.A. Boyd, Anaphylaxis with intravenous gammaglobulin. Lancet 336 (1990) 874. [5] K.L. Kolho, T. Ruuska, and E. Savilahti, Severe adverse reactions to Infliximab therapy are common in young children with inflammatory bowel disease. Acta Paediatr 96 (2007) 128-30. [6] A. Vultaggio, A. Matucci, F. Nencini, S. Pratesi, P. Parronchi, O. Rossi, S. Romagnani, and E. Maggi, Anti-infliximab IgE and non-IgE antibodies and induction of infusion-related severe anaphylactic reactions. Allergy 65 (2010) 657-61. [7] http://www.phrma.org/sites/default/files/1776/biotech2011.pdf [8] E. Maggi, A. Vultaggio, and A. Matucci, Acute infusion reactions induced by monoclonal antibody therapy. Expert Rev Clin Immunol 7 (2011) 55-63. [9] R.T. Strait, S.C. Morris, M. Yang, X.W. Qu, and F.D. Finkelman, Pathways of anaphylaxis in the mouse. J Allergy Clin Immunol 109 (2002) 658-68. [10] F.D. Finkelman, M.E. Rothenberg, E.B. Brandt, S.C. Morris, and R.T. Strait, Molecular mechanisms of anaphylaxis: Lessons from studies with murine models. Journal of Allergy and Clinical Immunology 115 (2005) 449-457. [11] F.D. Finkelman, Anaphylaxis: lessons from mouse models. J Allergy Clin Immunol 120 (2007) 17
506-15; quiz 516-7. [12] K. Mukai, K. Obata, Y. Tsujimura, and H. Karasuyama, New insights into the roles for basophils in acute and chronic allergy. Allergol Int 58 (2009) 11-9. [13] Y. Tsujimura, K. Obata, K. Mukai, H. Shindou, M. Yoshida, H. Nishikado, Y. Kawano, Y. Minegishi,
T.
Shimizu,
and
H.
Karasuyama,
Basophils
play
a
pivotal
role
in
immunoglobulin-G-mediated but not immunoglobulin-E-mediated systemic anaphylaxis. Immunity 28 (2008) 581-9. [14] F. Jonsson, D.A. Mancardi, Y. Kita, H. Karasuyama, B. Iannascoli, N. Van Rooijen, T. Shimizu, M. Daeron, and P. Bruhns, Mouse and human neutrophils induce anaphylaxis. J Clin Invest 121 (2011) 1484-96. [15] F. Jonsson, D.A. Mancardi, W. Zhao, Y. Kita, B. Iannascoli, H. Khun, N. van Rooijen, T. Shimizu, L.B. Schwartz, M. Daeron, and P. Bruhns, Human FcgammaRIIA induces anaphylactic and allergic reactions. Blood 119 (2012) 2533-44. [16] M.V. Khodoun, Z.Y. Kucuk, R.T. Strait, D. Krishnamurthy, K. Janek, C.D. Clay, S.C. Morris, and F.D. Finkelman, Rapid desensitization of mice with anti-FcgammaRIIb/FcgammaRIII mAb safely prevents IgG-mediated anaphylaxis. J Allergy Clin Immunol 132 (2013) 1375-87. [17] D. Jiao, Y. Liu, X. Lu, Q. Pan, J. Zheng, B. Liu, Y. Liu, Y. Wang, and N. Fu, Characteristics of anaphylaxis-inducing IgG immune complexes triggering murine passive systemic anaphylaxis. Allergy 68 (2013) 236-45. [18] J.Y. Wang, G. Prorok, and W.P. Vaughan, Cytotoxicity, DNA cross-linking, and DNA single-strand breaks induced by cyclophosphamide in a rat leukemia in vivo. Cancer Chemother Pharmacol 31 (1993) 381-6. [19] F. Nimmerjahn, and J.V. Ravetch, Analyzing antibody-Fc-receptor interactions. Methods Mol 18
Biol 415 (2008) 151-62. [20] R. Ishikawa, Y. Tsujimura, K. Obata, Y. Kawano, Y. Minegishi, and H. Karasuyama, IgG-mediated systemic anaphylaxis to protein antigen can be induced even under conditions of limited amounts of antibody and antigen. Biochem Biophys Res Commun 402 (2010) 742-6. [21] Y. Kaneko, F. Nimmerjahn, and J.V. Ravetch, Anti-inflammatory activity of immunoglobulin G resulting from Fc sialylation. Science 313 (2006) 670-3. [22] S. Kimura, M. Nagata, M. Takeuchi, K. Takano, and M. Harada, Anti-granulocyte antibody suppression of active and passive anaphylactic shock in WBB6F1-W/Wv mice. Cell Mol Life Sci 53 (1997) 663-6.
19
Figure legends Fig 1. Effects of basophils depletion on 2C2-OVM-induced PSA. BALB/c and C57BL/6 mice (n=3 each unless specified otherwise) were subject to MAR-1 or isotype control (ISO) treatment, and the basophils in blood, spleen and bone marrow were then evaluated by flow cytometry analysis. Basophil depletion using MAR-1 was also established before the induction of PSA by 2C2-OVM while changes in rectal temperature were monitored and shock symptoms evaluated. Data are representative of 3 independent experiments. Fig 2. Influences of neutrophil depletion on 2C2-OVM induced PSA. BALB/c and C57BL/6 mice were subject to anti-Gr1 or its ISO treatment, and then neutrophils in the blood, spleen and bone marrow were evaluated by flow cytometry analysis. Neutrophil depletion using anti-Gr1 was also established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments. Fig 3. Impacts of WBC depletion on 2C2-OVM induced PSA. BALB/c and C57BL/6 mice were subject to CPM or control (PBS) treatment. WBCs were then counted using a hemocytometer, and neutrophils, monocytes and basophils in blood were analyzed by flow cytometry. Depletion of WBCs using CPM with or without Clo-lipo was established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 3 independent experiments. Fig 4. Effects of monocyte/macrophage depletion or inhibition on 2C2-OVM-induced PSA. BALB/c and C57BL/6 mice were subject to Clo-lipo or PBS-lipo treatment, and then monocytes in blood and macrophages in the spleen and peritoneum were evaluated by flow cytometry (A, upper panel). Monocyte/macrophage depletion using Clo-lipo (A, lower panel) and inhibition using GdCl3 20
(B) were established before the induction of PSA by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of at least 3 independent experiments. Fig 5. Impacts of monocyte/macrophage depletion and WBC depletion on PSA induced by five additional IgG-antigen complexes. Monocytes/macrophages depletion using Clo-lipo and WBC depletion using CPM were established before the induction of PSA by five other IgG-antigen complexes in BALB/c mice, while hypothermia was monitored as previously described. Superscripts on mAb names indicate the IgG isotypes. Fig 6. Role of PAF and histamine in 2C2-OVM-induced PSA. (A) BALB/c mice were i.v. injected with PAF at different doses, and hypothermia was measured. (B and C). CV-3988 and Triprolidine pretreatment was conducted before 2C2-OVM-induced PSA in BALB/c mice, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments. Fig 7. Role of C3 in 2C2-OVM-induced PSA. PSA was established in C3-/- (n=3 each) and WT (n=4 each) C57BL/6 mice by 2C2-OVM, while hypothermia was monitored as previously described. Data are representative of 2 independent experiments.
21
Table 1. Reports concerning the effector cells in IgG-induced murine PSAa) Publicatio n
–
Reference
In their/our work Effector
cell
type proved
Mice
PSA Protocol b)
strain
Ab for priming
Methods for effector cell research
Antigen for challenge
no. Kimura S
Neutrophil
1997 [22]
Tsujimura
v
1
WBB6F1-W:W
anti-BPO mAb ascites
mice
(200µl)
Not
CTS
neutrophil
mice
and
DS
Basophil
C57BL/6 mice
2008 [13]
Jönsson
c)
BSA antiserum (200µl)
BPO-GSA (1mg)
PenV-BSA (500µg)
(1000µg)
Neutrophil
2011 [14]
WT and 5KO
mAb
C57BL/6 mice
(200µg),
anti-DNP
2b
K/BxN serum anti-GPI
DNP-HSA (500µg)
GPI (40µg)
(20µl) anti-OVA1
Jönsson
Macrophage
WT,
3KOIIA
mAb
2012 [15]
and
and
5KOIIA
(1000µg),
neutrophil
C57BL/6 mice
K/BxN serum anti-GPI
OVA (1000µg)
Monocyte/ma
BALB/c
and
2013 [16]
crophage,
C57BL/6 mice
anti-TNP2a
mAb
methods
Neutrophil
anti-Gr1 mAb
Mast cell
WBB6F1-W:Wv strain mice
Basophil
anti-CD200R3 mAb
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo
NK cell
anti-NK1.1 mAb
Neutrophil
anti-Gr1 mAb
Monocyte/
Clo-lipo; GdCl3
Neutrophil
anti-Gr1 mAb
Basophil
anti-CD200R3 mAb
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo; GdCl3
Neutrophil
anti-Gr1 and Anti-Ly6G
GPI (50ug)
(20µl)
Khodoun
Depletion/inhibition
depleted/inhibited
BSA (1mg)
anti-e)PenV1
mAb
d)
Lineage
TNP-OVA (500µg)
(100µg)
mAb Basophil
anti-CD200R3
Mast cell
KitW-sh/W-sh strain mice
Monocyte/Macrophage
Clo-lipo; GdCl3
Neutrophil
anti-Gr1
basophil, and
Macrophage
work
mAb
hydroxyurea
neutrophil Our present
mAb
BALB/c
and
C57BL/6 mice
Basophil
anti-CD200R3 mAb
OVM (10µg)
Basophil
anti-FcεR1α mAb
6H4 (100µg)
OVM (10µg)
Neutrophil
anti-Gr1 mAb
1G32a (50µg) & 4F121
HSA (100µg)
Monocyte/Macrophage
Clo-lipo; GdCl3
White blood cell
CPM
1
2a
2C2 (100µg)
(50µg) 2b
1
HSA (100µg)
M41 (50µg) & M122b
MP1P (50µg)
H1A6
(50µg) & 4A2
(50µg)
(50µg) 3H91 (50µg) & 4D111
rHbζ (100µg)
(50µg) a)
Only PSA data summarized in table, though there are some data on ASA in above papers. Superscripts on mAb and mAb ascites indicate the IgG isotypes. 1: IgG1; 2a: IgG2a, 2b:IgG2b. c) BPO, benzylpenicilloy; d) GSA, guinea pig serum albumin; e)PenV, penicillin V. b)
22
and
24
25
26
27
28
30
31
1. Macrophages are the dominant effector cells responsible for IgG-mediated anaphylaxis. 2. The type of effector cell may be associated with different antigen-antibody systems. 3. The type of effector cell is not affected by IgG isotype and genetic background of mice. 4. PAF is the major mediator responsible for IgG-induced anaphylaxis. 5. IgG-mediated anaphylaxis is independent of complete third component.
32
