Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Nov 1.
Published in final edited form as: Arthritis Rheumatol. 2015 Nov;67(11):2877–2887. doi: 10.1002/art.39283

Local joint inflammation and histone citrullination provides a murine model for the transition from preclinical autoimmunity to inflammatory arthritis

Dong Hyun Sohn 1,2, Christopher Rhodes 1,2, Kazuhiro Onuma 1,2, Xiaoyan Zhao 1,2, Orr Sharpe 1,2, Tal Gazitt 1,2, Rani Shiao 1,2, Justyna Fert-Bober 3,*, Danye Cheng 1,2, Lauren J Lahey 1,2, Heidi H Wong 1,2, Jennifer Van Eyk 3,*, William H Robinson 1,2, Jeremy Sokolove 1,2
PMCID: PMC4626401  NIHMSID: NIHMS707993  PMID: 26227989

Abstract

Objective

Anti-citrullinated protein antibodies (ACPAs) are characteristic of rheumatoid arthritis. However, their presence years before onset of clinical RA is perplexing. Although multiple putative citrullinated antigens have been identified, no studies have demonstrated the specific capacity of these antigens to initiate inflammatory arthritis. We sought to recapitulate the transition from preclinical to clinical RA and to demonstrate the capacity of local citrullination to facilitate this transition.

Methods

We performed proteomic analysis of activated human neutrophils to identify citrullinated proteins including those targeted as part of the RA immune response. Using ELISA we compared RA and OA synovial fluid for levels of citrullinated histone 2B (cH2B) and its immune complex. Using macrophage activation assays we assessed the effect of histone citrullination on immunostimulatory capacity and evaluated the stimulatory capacity of native and citrullinated H2B immune complexes. Finally, we assessed the potential for anti-cH2B antibodies to mediate arthritis in vivo.

Results

We identified robust targeting of neutrophil-derived citrullinated histones by the ACPA immune response. Over 90% of RA patients have anti-cH2B antibodies. Histone citrullination increases innate immunostimulatory capacity and immune complexes containing citrullinated histones activate macrophage cytokine production and propagate neutrophil activation. Finally, we demonstrate that immunization with H2B is arthritogenic, but only in the setting of underlying articular inflammation.

Conclusion

We identify citrullinated histones, specifically cH2B as an antigenic target of the ACPA immune response. Furthermore, local generation of citrullinated antigen during low grade articular inflammation provides a mechanistic model for the conversion from preclinical autoimmunity to inflammatory arthritis.

Rheumatoid arthritis (RA) is associated with antibodies targeting proteins which have undergone post-translational modification of arginine to citrulline by a family of enzymes known as peptidyl-arginine deiminases (1-3). Although several lines of evidence implicate these antibodies in RA pathogenesis, the presence of these anti-citrullinated protein antibodies (ACPA) up to several years before the onset of disease development (4-6) has called into question their direct role in the mediation of synovial inflammation. Similarly, there are multiple citrullinated protein targets of the ACPA immune response with no dominant immunopathogenic antigen identified. Recent reports suggest that proteins generated in multiple pathways of neutrophil activation may serve as antigenic targets of the ACPA immune response (7-10). However, no studies to date have demonstrated the capacity of these citrullinated products to initiate inflammatory arthritis, nor has a mechanisms by which these citrullinated antigens might contribute to the initiation and pathogenesis of RA been elucidated. The presence of neutrophils and their activation products are ubiquitous to sites of inflammation. Thus, we hypothesized that generation of citrullinated products of neutrophil activation could, in the setting of circulating ACPA, provide the nidus for immune complex generation and the transition from preclinical autoimmunity to clinical RA.

Materials and Methods

Sample collection

Serum, plasma, or synovial fluid was obtained from patients with RA, Psoriatic arthritis (PsA), or osteoarthritis (OA). All RA patients met the American College of Rheumatology criteria for the disease (11) and all samples were obtained under IRB approved protocols at Stanford University. RA serum was obtained from the VA Palo Alto Health Care system (n = 62) or the ABCoN cohort of the North American Rheumatoid Arthritis Consortium (n = 123) (12). RA, OA, and psoriatic arthritis (PsA) synovial fluid specimens for quantitation of H2B-IC were obtained at the VA Palo Alto by the investigators (JS, WHR) or by a generous gift from Dr. David Lee (Brigham and Women's Hospital) while RA and OA synovial fluid specimens for measurement of H2B levels were obtained as above with additional samples purchased from Bioreclamation LLC (Hicksburg, NY).

Generation and proteomic interrogation of products of neutrophil activation

Human neutrophils were isolated as below. Products of neutrophil activation were generated by incubating 3 × 107 neutrophils with 10 μM ionomycin, 30 nM PMA, or 200 ng/ml TNFα for 4 hours at 37 °C. After removing supernatants, each dish was washed and products of neutrophil activation were digested with 10 U/ml micrococcal nuclease. Samples were centrifuged at 300 × g to remove intact cells, then at 4,000 × g to remove debris. Neutrophil activation and generation of neutrophil extracellular traps was visualized by staining with DAPI, anti-neutrophil elastase, or anti-citrullinated H3 (Abcam). Alternatively, neutrophil activation was quantitated by measurement of DNA content in the supernatant or by incubation with Sytox Green (Invitrogen) followed by measurement of fluorescence at 485 nm (excitation) / 520 nm (emission).

Products of neutrophil activation induced by ionomycin were separated on parallel SDS-PAGE gels and stained with Coomassie blue or transferred to PVDF membranes followed by probing with ACPA-positive RA serum IgG (RA-IgG); anti-modified citrulline antibody (Millipore); or rabbit anti-H2B polyclonal antibody (Abcam).

Coomassie-stained bands corresponding to bands identified by RA-IgG and/or anti-modified citrulline were cut from gel, digested with trypsin, and subjected to mass spectrometry analysis as previously described (13, 14).

To confidently identify citrullinated residues, proteomic analysis of products from ionomycin activated neutrophils was performed using FASP protocol (15) followed by cation exchange chromatography and mass spectroscopy as previously described (7).

Immunoprecipitation and immunoblot analysis

Products of neutrophil activation were denatured at 95°C in the presence of 0.1% SDS, 0.5 mM EDTA, and 1 mM DTT, incubated with anti-H2B antibody or human RA-IgG, then with Protein G beads. Beads were eluted by boiling and proteins were separated by SDS-PAGE and transferred to PVDF membrane with identification of citrullinated proteins using an anti-modified citrulline antibody (Millipore). Anti-modified citrulline blot was stripped (and re-exposed to confirm stripping) and re-blocked with 5% milk and re-probed with rabbit anti-H2B antibody directly conjugated to HRP (Abcam).

Detection of antibodies to nH2B, cH2B

Detection of antibodies to native H2B (nH2B) or citrullinated H2B (cH2B), as well as a panel of 21 additional citrullinated epitopes (or arginine containing controls) was performed using a bead-based direct immunoassay as previously described (6).

Measurement of H2B, cH2B, and H2B immune complex in RA serum or RA synovial fluid

Levels of cH2B protein were measured using a novel ELISA developed in our lab. Plates were coated with 2 μg/ml of rabbit anti-H2B capture antibody (Abcam), washed and blocked with 2% BSA, washed again, and incubated with RA or OA synovial fluid diluted 1:20 in PBS containing 0.1% BSA and 0.05% Tween 20. Plates were washed and incubated with 2 μg/ml of mouse anti-citrulline antibody (clone F95, Millipore) followed by incubation with an HRP conjugated anti-mouse IgM. Levels of H2B immune complex (H2B-IC) were measured using a C1q capture assay as previously described (13).

H2B immunization studies

All animal studies were performed in adherence to the NIH Guide for the Care and Use of Laboratory Animals and under protocols approved by the VA Palo Alto IACUC. DBA/1J mice (Jackson Laboratory) were immunized with adjuvant alone; 50 μg bovine type II collagen (low dose CIA); native or citrullinated H2B (125 μg of native or in vitro citrullinated H2B); or low dose collagen and native or citrullinated H2B (ldCIA + cH2B). Briefly, DBA/1 J mice were immunized intradermally on the proximal tail with the above immunogens emulsified in complete Freund's adjuvant (CFA). Twenty-one days after immunization, mice were boosted by subcutaneous injection at the base of the tail with the listed immunogens in incomplete Freund's adjuvant (IFA). Each limb was graded with a score of 0-4 (maximum possible score of 16 for each mouse). Hindpaw thickness was measured with microcalipers. At experimental termination, scoring of pathology was performed by two investigators blinded to immunization condition as previously described (16). Results presented represent outcomes of experiments performed at least twice.

Serum transfer studies

DBA/1J mice were immunized with 125 μg cH2B or adjuvant alone followed by booster injection at day 21 as above. Mice were sacrificed and blood was collected on experimental day 50 and serum was confirmed to have antibody reactivity to cH2B. Twenty DBA/1J mice were immunized with 50 μg bovine type II collagen and boosted at day 21 as described above (low dose CIA). Mice with a score of > 4 on day 28 were excluded leaving 8 mice per group. To assure randomization, ldCIA mice were ranked by level of disease at day 28 and matched to a mouse with similar arthritis score before receiving cH2B or control serum. On days 28 and 30, ldCIA mice were received 300 μl of cH2B or control serum via i.p. injection. Mice were scored 3 times weekly until termination (day 46). Mice which failed to manifest any arthritis during the course of the study were excluded leaving 6 mice analyzed per group.

Cell isolation and culture

Bone marrow derived macrophages (BMMs) were isolated from wild-type C57BL/6, TLR4 deficient B6.B10ScN-Tlr4lps-del/JthJ mice, TLR2 deficient mice (The Jackson Laboratory), or TLR9 deficient mice (a gift from Dr. Lawrence Steinman, Stanford University) and cultured as previously described (17).

Generation of human monocyte-derived macrophages was performed as previously described (18). Neutrophils were isolated from fresh whole blood or buffy coats using the dextran/ficoll method as previously described (19).

Antibodies and reagents

LPS was from Sigma-Aldrich, CpG, Pam3CSK4, and TLR4 inhibitor CLI-095 were from InvivoGen. Anti-human CD32 (FcγRIIa) antibody (clone IV.3) from Stem Cell Technologies. Total histone, as well as H1, H2A, H2B, and H3, all purified from calf thymus were purchased from Immunovision. Histones were used in either unmodified native form or citrullinated form. In vitro citrullination was performed as previously described (18) and confirmed by mobility shift on SDS-PAGE and by western blot using human ACPA-positive RA sera as well as anti-modified citrulline antibody (Millipore). Native histones underwent sham citrullination in an identical manner to citrullinated proteins but without the addition of the PAD enzyme (Sigma-Aldrich). PAD enzyme incubated with citrullination buffer and DTT but without histones served as a control to assure no contribution or contamination from the small amount of PAD enzyme remaining in histone citrullination reaction mixtures.

Macrophage stimulation

Human macrophages (5 × 104/well) or murine macrophages (8 × 104/well) were incubated with native or citrullinated H1, H2A, H2B, or H3 for 16 hours and TNFα levels in supernatants determined by ELISA (PeproTech). The TLR4 ligand LPS (1 ng/ml) and TLR2 ligand Pam3CSK4 (10 ng/ml) served as controls for TNFα induction by TLR4 and TLR2, respectively. Immune complexes were generated in vitro by incubation of native or citrullinated histone (total histones, H2A, H2B, or H3) with a polyclonal rabbit antibody against H2A (Abcam), H2B (eBioscience), H3 (Abcam) or, as a control, with normal polyclonal rabbit IgG (DakoCytomation) at 37°C for 45 minutes. Cross-titration of antibody and antigen yielded an optimal ratio for formation of IC; final concentrations of 20 μg/ml of histones and 10 μg/ml of polyclonal antibody were used for IC stimulation of monocyte-derived macrophages. Given a molar excess of histone antigen, this likely generates relatively small plate-bound complexes. At final dilutions, all reagents used in the stimulation assays were tested for endotoxin contamination by application of the Limulus amebocyte assay (Associates of Cape Cod), and were shown to possess endotoxin levels below the detectable range (< 0.03 endotoxin units/ml). To further exclude endotoxin contamination in select experiments, histones were treated with endotoxin-removal resin (Detoxi-Gel) or stimulations were performed in the presence of 15 μg/ml of polymyxin B (Sigma-Aldrich) or after treatment of histones with proteinase K followed by boiling.

For IC stimulation of human macrophages, human IgG derived from patients with ACPA-positive RA was used to generate plate-bound human histones immune complexes (total histones, cH2A, cH2B, or cH3). IgG from 3 pooled plasma samples shown by ELISA to contain high levels of anti-cH2A and cH2B antibodies was purified on protein G columns (Pierce). Eluted IgG fractions were concentrated and buffer exchanged to PBS (Amicon Ultra) and depleted of endotoxin over an endotoxin-removal column (Detoxi-Gel). IgG concentrations were estimated by optical density at 280 nm; IgG was aliquoted and stored at −80°C. For generation of histone immune complexes, flat-bottomed 96-well culture plates were coated with 50 μl of total histones, cH2A, cH2B, or cH3 (20 μg/ml), washed, and blocked with 2% low endotoxin BSA, washed and incubated with 100 μl of ACPA-positive IgG (1.0 mg/ml) in PBS containing 0.05% Tween 20 or, as a control, IgG isolated from ACPA-negative patients with OA. Wells were again washed and human macrophages (5 × 104/well) in 200 μl RPMI containing 10% FBS were then added to the wells. To block TLR4 activation, macrophages were pre-incubated for 1 hour with the small-molecule TLR4 inhibitor CLI-095 (InvivoGen). As prior studies have demonstrated ACPA-immune complex induced macrophage activation to be FcγRIIa dependent (20), an anti-human CD32 (FcγRIIa) antibody (clone IV.3; Stem Cell Technologies) was used to block Fcγ receptor activation.

Quantification of products of neutrophil activation

Freshly isolated neutrophils in RPMI without phenol red were seeded into 96-well plates (5 × 104 cells/well) and incubated with PBS, PMA, Ionomycin, LPS, nH2B, cH2B, nH2B-IC, or cH2B-IC for 4-8 hours in the presence of 1 μM Sytox Green (Invitrogen), a non cell-permeant DNA binding dye, to detect extracellular DNA. The plates were read in a fluorescence microplate reader at 485 nm (excitation) / 520 nm (emission).

Statistical analysis

For in vitro cell stimulation assays, an unpaired t-test was used to compare cytokine production between groups. Kruskal-Wallis one-way ANOVA with Dunns post-test was used to evaluate the association of anti-cH2B antibodies with circulating TNFα levels. Differences in arthritis time points as well as arthritis scores and histology scores were compared by the Mann-Whitney U test (two tailed). These analyses were performed using GraphPad Prism version 5.

Results

Products of neutrophil activation are targets of the autoantibody response in RA

There are multiple pathway which can result in neutrophil activation and histone citrullination including apoptosis, NETosis, and necroptosis/autophagy, as well as immune-mediated membranolytic pathways, mediated by perforin and the membrane attack complex (7). We tested different stimuli to activate neutrophils including PMA, ionomycin, LPS, and TNF. We have found that ionomycin activated neutrophils generated the most robust pattern of citrullinated proteins compared to PMA while we observed negligible production of citrullinated proteins upon activation by LPS or TNF. Due to its robust and reproducible pattern of citrullination, and to assure the broadest profile of neutrophil activation, further studies utilized ionomycin-induced neutrophil activation.

We demonstrate the generation of multiple proteins upon neutrophil activation (Figure 1A, Supplementary Tables 1 and 4), including several proteins targeted by ACPA-positive RA-IgG (Figure 1B), the majority of which are identified to be citrullinated as assessed by staining with an anti-modified citrulline antibody (Figure 1B). The citrullinated proteins targeted by ACPA-positive RA-IgG include the most prominent bands (as defined by densitometry) at 14 and 13 kD, subsequently identified by mass spectrometry to contain histones, predominantly H2B and H2A, respectively (Supplementary Table 1). The identity of H2B was confirmed using a parallel immunoblot probed with a polyclonal anti-H2B antibody (Figure 1B). Furthermore, immunoprecipitation with either purified human IgG derived from a pool of RA patients or a rabbit polyclonal anti-H2B antibody generated bands at 14 kD which were recognized by both anti-modified citrulline antibody and anti-H2B antibody (Figure 1C and D). Though it is possible that anti-modified citrulline antibodies as well as human RA IgG could recognize other modifications including carbamylation, use of unbiased mass spectroscopy neutrophil activation products identified citrullination of neutrophil-derived H2B, and to a lesser extent H2A and H3, but not carbamylated histones (Supplementary Table 2).

Figure 1.

Figure 1

Open in a new tab

Proteomic analysis of the products of neutrophil activation for targeting by the RA immune response. A and B, neutrophil activation was induced by 10 μM ionomycin, and protein products separated by SDS-PAGE and (A) stained with Coomassie blue, (B) immunoblotted with healthy IgG, isolated RA-IgG, citrullinated residues were acid modified followed by immunoblotting with anti-modified citrulline antibody, or immunoblotted with anti-H2B antibody. Mass spectroscopic identification of proteins in (A) provided in Supplementary Table 1. C and D, Products of neutrophil activation were immunoprecipitated with human RA-IgG (C) or anti-H2B antibody (D) and immunoblotted with anti-modified citrulline antibody or anti-H2B antibody. NB: Band #5 with prominent staining by RA-IgG on (B) was confirmed to be micrococcal nuclease (thermonuclease; S. aureus) used in the digestion of neutrophil activation products.

Citrullinated histone 2B is a target of the ACPA immune response in vitro and in vivo

Guided by proteomic studies of activated human neutrophils, we evaluated for the presence of autoantibodies to native or citrullinated H2B in plasma and synovial fluid derived from RA patients. The vast majority of ACPA-positive RA patients possess antibodies targeting citrullinated H2B (cH2B), while no such antibodies were found in patients with OA or ACPA-negative RA (P < 0.0001) (Figure 2A). There was no significant targeting of native H2B in any group (Figure 2A). Consistent with an influx of neutrophils and ongoing neutrophil activation we observed significantly elevated levels of cH2B (P < 0.0001) (Figure 2B) in RA compared with OA synovial fluids as measured by a citrulline-specific H2B ELISA.

Figure 2.

Figure 2

Open in a new tab

Prominent targeting of citrullinated H2B by RA-associated autoantibodies. A, Bead-based immunoassay analysis of anti-citrullinated and anti-native H2B autoantibodies in anti-CCP2+ RA (n = 81), anti-CCP2- RA (n= 85), and OA (n = 36) patient plasma. B, Citrullinated H2B ELISA analysis of RA (n = 19) and OA (n = 13) synovial fluids. C and D, Analysis of H2B immune complexes in anti-CCP2+ RA (n = 20), anti-CCP2- RA (n = 10), and PsA (n = 14) plasma (C), or RA (n = 14), OA (n = 10), and PsA (n = 4) synovial fluid (D).

We then measured levels of H2B-containing immune complexes (ICs) in plasma and synovial fluid from patients with RA, OA, or PsA. Since only citrulline modified H2B is recognized by RA sera, it was inferred that these ICs contain cH2B (as demonstrated in Figure 1B, C, and D). Levels of such circulating H2B immune complex (H2B-IC) were detectable in the plasma of over half of ACPA-positive RA patients tested but not in the plasma of ACPA-negative RA or PsA subjects (Figure 2C). Similarly, H2B-ICs were elevated in the majority of synovial fluid specimens derived from ACPA-positive RA but not from patients with OA or PsA (P = 0.007) (Figure 2D).

Immunization with native or citrullinated H2B is arthritogenic, but only in the setting of low grade joint inflammation

Having identified cH2B as a possible target of the RA immune response, we sought to demonstrate that autoimmunity against cH2B can directly mediate inflammatory arthritis. We immunized DBA/1J mice with resultant induction of a robust anti-cH2B antibody response (Figure 4A, B, and C), but without visible induction of arthritis (Figure 3). We hypothesized that a lack of cH2B expression in the intact joint would not allow generation of cH2B immune complexes and that the conversion from asymptomatic autoimmunity to clinical RA could be the result of a low grade articular inflammatory response resulting in generation of intra-articular cH2B. Supporting this hypothesis, when mice were co-immunized with low dose type II collagen (ldCIA) and cH2B, we observed development of robust arthritis in the co-immunized mice compared to those immunized with cH2B or low dose collagen alone (P < 0.01) (Figure 3). This effect was observed despite generation of similar levels anti-cH2B antibodies and anti-CII antibodies, respectively (Figure 4A).

Figure 4.

Figure 4

Open in a new tab

Assessment of immunoreactivity developed after immunization with cH2B and/or low dose type II collagen. A, Similar levels of anti-cH2B and anti-CII antibodies on day 17 and at experimental termination (day 50) in mice receiving single or co-immunization. B and C, Multiplex immunoassay was performed to measure a panel of citrullinated antigens and their non-citrullinated controls. Significance analysis of microarray comparison between groups identified ACPA subtypes elevated before boost (day 17) (B) and at experimental termination (day 50) (C) (n = 9 ∼ 10 per group). D, Serum transferred from cH2B immunized mice exacerbates arthritis compared with serum from mice immunized with adjuvant alone (n = 6 per group).

Figure 3.

Figure 3

Open in a new tab

Immunization with cH2B induces inflammatory arthritis in the presence of low-grade joint inflammation. A and B, Arthritis score (A) and paw swelling (B) over time in mice immunized with adjuvant alone (CFA), cH2B alone (cH2B), low dose type II collagen (ldCIA), or a combination of low dose collagen and cH2B (ldCIA + cH2B) (n = 5 per group). C, Representative images of H&E-stained sections of ankle joints from mice described in A and B. Scale bar, 500 μm. D, Histological scores of inflammation, pannus formation, and bone or cartilage erosions of ankle joints from mice described above. * = P < 0.05; *** = P < 0.001 compared with CFA control.

Using a multiplex ACPA antigen array, we observed a dominant antibody response targeting cH2B and the closely related citrullinated H2A. We also observed a process of epitope spreading in both cH2B and ldCIA/cH2B co-immunized mice, with a pattern similar to that observed in studies of preclinical human RA (6), yet again, without development of arthritis in mice immunized with cH2B alone (Figure 4B and C, and Supplementary Figure 1). We observed very little cross targeting of native H2B (Figure 4B and C) thus confirming generation of a predominantly citrulline-specific immune response closely recapitulating the human phenotype.

Notably, immunization with native H2B resulted in generation of robust immunoreactivity to both native and citrullinated H2B and was associated with similar augmentation of low grade arthritis as was observed with cH2B immunization (data not shown). This is consistent given generation of both native and citrullinated histones in the setting of low grade inflammation. Also possible is that native H2B immunization generates antibodies cross-reactive with both native and citrullinated H2B. Though our data demonstrates minimal targeting of native H2B upon cH2B immunization (Figure 4) and we did observe relatively more cross reactivity with cH2B upon native H2B immunization (Supplementary Figure 1C). However, as demonstrated above, human RA is not associated with anti-native histone reactivity. Thus although native histone immunization does not well recapitulate the human phenotype of anti-citrulline autoimmunity, it does further suggests the capacity of low grade inflammation to generate local antigen-immune complexes in the setting of circulating autoantibody. Finally, we observed that transfer of ACPA-containing serum from mice immunized with cH2B (Figure 4C) was able to significantly enhance low grade arthritis compared to control serum (Figure 4D), but again, only in the setting of an underlying low grade joint inflammation and generation of intra-articular protein citrullination.

Histone citrullination enhances innate immunostimulatory capacity

Given the plethora of potential citrullinated antigens, we sought to identify mechanisms by which anti-citrullinated histone autoreactivity might result in chronic arthritis. Recent studies identified the ability of extracellular histones to act as mediators of innate immunity (21) including stimulation of cytokine production via TLRs (21-24). Thus, we assessed the stimulatory capacity of extracellular histones and specifically, the effect of citrullination on stimulatory potency. We observed dose dependent stimulation of TNFα production from human monocyte-derived macrophages stimulated with total histones as well as individually purified mammalian histones with increased stimulation by citrullinated histones, most prominently by citrullinated H2B compared to unmodified proteins (Figure 5A). Using macrophages derived from mice deficient in TLR2, TLR4, or TLR9, we observed that H2B-induced TNFα production is TLR4-dependent (Figure 5B), an effect not abrogated by the addition of the endotoxin-chelator polymyxin B (Figure 5C) but significantly abrogated after treatment with proteinase K and boiling (P = 0.003; Figure 5D). Interestingly, H2B (Type 1-M) contains 126 amino acids including 8 arginine residues and thus has the potential for citrullination at multiple sites. By mass spectroscopy, only 2 of these arginines were determined to be citrullinated during the process of neutrophil activation and this pattern of citrullination was closely reproduced by our in vitro citrullination (Supplementary Tables 2 and 3).

Figure 5.

Figure 5

Open in a new tab

Effect of histone citrullination on innate immunostimulatory capacity. A-D, ELISA quantification of TNFα production from (A) human macrophages stimulated with native or citrullinated histones, (B) murine macrophage deficient in TLR2, TLR4, or TLR9, (C) murine macrophage stimulated with H2B in the presence of polymyxin B or (D) with H2B after treatment with proteinase K and boiling.

Immune complexes containing citrullinated histones activate macrophage via co-stimulation of TLR4 and Fcγ receptor

To examine whether incorporation of citrullinated histones into immune complexes further enhances their ability to stimulate cytokine production, we stimulated human macrophages with in vitro-generated citrullinated histone or cH2B immune complexes with or without inhibitors of TLR4 and/or the FcγRIIa. We demonstrate significantly increased induction of macrophage TNFα production in response to histone-IC generated from polyclonal anti-H2B antibody and either native or citrullinated total histones (Figure 6A). Despite similar specificity of this polyclonal antibody for native and citrullinated H2B, we observed significantly increased stimulation by citrulline-containing immune complex generated in this model system. To more closely mimic disease physiology, citrullinated H2B immune complex (cH2B-IC) was generated using human IgG derived from a pool of ACPA-positive RA patients (RA-IgG) demonstrated to have autoantibody reactivity to citrullinated H2B. Incubation of RA-IgG with plate-bound citrullinated H2B isolated only specific ACPA without the inclusion of non-specific IgG. Citrullinated H2B immune complex was powerful inducer of TNFα from human monocyte-derived macrophages (Figure 6B). Similar effects were observed for highly homologous cH2A-IC but not cH3-IC (data not shown) demonstrating specific reactivity to select citrullinated histones by the RA immune response.

Figure 6.

Figure 6

Open in a new tab

Citrullinated histone-containing immune complexes stimulate macrophage activation and propagate neutrophil activation. A, ELISA quantification of human macrophage TNFα production in response to native or citrullinated histone immune complex generated with a polyclonal anti-H2B antibody. B, ELISA quantification of human macrophage TNFα production in response to citrullinated H2B immune complex (cH2B-IC) generated with human RA patient derived IgG (RA-IgG) with blockade of FcγRIIa (FcB), TLR4 (TLR4B), or both. C, ELISA quantitation of serum TNFα levels from RA patients across tertiles of anti-cH2B antibody levels. D, Sytox green quantitation of neutrophil activation in response to cH2B-IC generated with human RA-IgG, or native and citrullinated H2B immune complexes generated with polyclonal anti-H2B antibody.

To dissect the requirement of TLR4 and/or the Fcγ receptor (FcγR) in the stimulation of macrophage TNFα production, we evaluated the effect of the small-molecule TLR4 inhibitor CLI-095 and/or inhibitory antibodies targeting the FcγRIIa. Inhibition of either TLR4 or FcγRIIa resulted in significant abrogation of the macrophage response to citrullinated H2B immune complex with further abrogation upon combined inhibition of TLR4 and FcγR pathways (Figure 6B). These results suggest that immune complexes containing specific citrullinated histones serve as potent inducers of macrophage TNFα production by co-stimulation of TLR4 and the FcγR.

To determine whether anti-cH2B immunoreactivity is associated with increased inflammation in human RA, we evaluated whether levels of anti-cH2B antibodies were associated with levels of circulating cytokines and/or RA disease activity. Increasing tertiles of anti-cH2B titers were associated with increasing levels of circulating TNFα (Figure 6C), further supporting the immunostimulatory capacity of cH2B-IC. A similar observation is made for IL-6 (comparison of tertiles of anti-cH2B antibodies, P = 0.030 by ANOVA) as well as disease activity (DAS28 score; comparison of tertiles of anti-cH2B antibodies, P ≤ 0.001 by ANOVA).

Immune complexes containing citrullinated histones propagate neutrophil activation

Given the ability of cH2B-IC to co-stimulate macrophage activation, we hypothesized a similar capacity to induce neutrophil activation. cH2B-IC generated with human RA-IgG stimulates neutrophil activation (Figure 6D). Furthermore, cH2B-IC formed with a rabbit polyclonal anti-H2B antibody which equally recognizes native and citrullinated H2B induces greater neutrophil activation from cH2B-IC as compared to native H2B-IC (P = 0.008; Figure 6D) again supporting the additional stimulatory capacity imparted by histone citrullination. Thus, the generation of citrullinated cH2B-IC may result in additional generation of citrullinated antigens by further induction of neutrophil activation.

Discussion

RA is immunologically characterized by the production of ACPAs targeting a variety of citrullinated proteins and these antibodies are associated with increased disease activity, disease severity, and disease damage (25). Converging lines of evidence suggest that these antibodies may be directly involved in RA pathogenesis. Emerging literature suggests a common pathogenic mechanism by which induction of local and/or systemic inflammation is produced as a result of autoantibodies with the capacity to bind antigens which themselves possess innate immunostimulatory capacity (18, 26-28) including at least some citrullinated antigens such as citrullinated fibrinogen (18, 29) and now citrullinated histones. Although others may exist, we have observed that not all citrullinated proteins (including several recognized by ACPA) possess immunostimulatory capacity after citrullination.

The current study provides critical evidence on how ACPAs, and specifically antibodies targeting citrullinated histones, might participate in the initiation and propagation of synovial inflammation in RA.

Specifically, we identify citrullinated histones including cH2A and cH2B as novel autoantigens with innate immunostimulatory capacity. We further demonstrate the arthritogenic potential of these antibodies, but only in the setting of low grade articular inflammation. This observation builds on a prior study which observed that administration of an anti-citrullinated fibrinogen antibody alone was unable to induce murine arthritis but was able to significantly enhance submaximal arthritis induced by administration of anti-type II collagen antibodies (30). Thus, the emerging concept brought to light by our data identifies the requisite role for not only break in tolerance and autoantibody production, but a critical role for antigen generation to facilitate onset of clinically apparent autoimmunity and joint inflammation. Our models support local generation of citrullinated antigen as the penultimate event leading to generation of ACPA-immune complexes and induction of clinical inflammation.

A limitation of this study is the use of adjuvant to induce a forced immune response against citrullinated antigens, in this case bovine cH2B. In human RA, the events initiating the immune response against citrullinated antigens are still unclear environmental factors (including the association of smoking with generation of citrullinated antigen) (31) as well as genetics (especially the ability of certain HLA haplotypes to preferentially present citrullinated antigen to T cells) (32). As such, this study does not address the origins of anti-citrulline autoimmunity in human RA. However, the generation of relatively citrulline-specific autoantibodies supports utility of this model to study the effector phase of RA including the transition from preclinical autoimmunity to clinical arthritis. Another limitation of this study is the ability of mass spectroscopy to identify all products of neutrophil activation, especially citrullinated proteins, and others have recently reported an overlapping spectrum of protein citrullination during NETosis (7, 8). Similarly, it is possible that our analysis may have lacked sensitivity to identify all citrullinated residues including those which are not a result of neutrophil activation and the intracellular process of NETosis, but rather due to the potential process of extracellular citrullination. Additionally, the in vitro stimulation assays performed do not identify a direct physical interaction between TLR4 and cH2B, and although extensive steps were taken to exclude the possibility, there remains the potential for low levels of residual endotoxin. Finally, in our murine models it was necessary to generate low grade articular inflammation to observe the arthritogenic effects of anti-cH2B antibodies. Though potentially advancing out understanding of the events involved in the transition from preclinical autoimmunity, such a window of initiation has not been clearly identified in development of human RA.

In conclusion, we demonstrate that citrullinated products of neutrophil activation, especially cH2B, can serve as autoantigens with the capacity to link innate and adaptive immunity potentially driving the initiation and propagation of RA-associated inflammation. We propose that local histone citrullination generates antigenic fuel, which initiates a feed-forward synergy between innate and adaptive immune activation, resulting in the transition from preclinical ACPA immunity to clinical rheumatoid arthritis.

Supplementary Material

Supp FigureS1

Supplementary Figure 1. ACPA repertoire induced by citrullinated and native H2B immunization.

Supp TableS1-S4

Supplementary Table 1. Identification of predominant products of ionomycin activated neutrophils.

Supplementary Table 2. Proteomic identification of citrullination sites from ionomycin activated neutrophils.

Supplementary Table 3. Proteomic identification of in vitro histone citrullination sites.

Supplementary Table 4. Proteomic interrogation of all products of ionomycin activated neutrophils.

Acknowledgments

These studies were supported by grants from the Department of Veterans Affairs (IK2 BX001301), an Arthritis Foundation Innovative Research Award, and a Rheumatology Research Foundation Disease Targeted Research Grant (JS); NIH R01 AI085268, R01 AR063676, and NHLBI Proteomics Center contract N01 HV 28183 (W.H.R.). D.H.S. was supported by Arthritis Foundation Postdoctoral Fellowship (Award # 5119).

Footnotes

Author contributions: D.H.S., X.Z., W.H.R., and J.S. initiated the investigation of histone citrullination in RA. D.H.S., K.O., and H.H.W. performed animal studies. D.H.S., C.R., T.G., R.S., D.C, O.S., and J.S. performed in vitro cellular stimulation and western analysis of products of neutrophil activation. O.S. performed proteomic identification of products of neutrophil activation. J.F.B. and J.V.E. conducted or contributed to the proteomic analysis of protein citrullination. L.J.L. performed antigen and cytokine immunoassays. All authors analyzed the data. D.H.S., W.H.R., J.S. wrote and edited the manuscript; all authors approved the manuscript.

References

  • 1.Vossenaar ER, van Venrooij WJ. Citrullinated proteins: sparks that may ignite the fire in rheumatoid arthritis. Arthritis Res Ther. 2004;6(3):107–11. doi: 10.1186/ar1184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Firestein GS. Evolving concepts of rheumatoid arthritis. Nature. 2003;423(6937):356–61. doi: 10.1038/nature01661. [DOI] [PubMed] [Google Scholar]
  • 3.McInnes IB, Schett G. The pathogenesis of rheumatoid arthritis. N Engl J Med. 2011;365(23):2205–19. doi: 10.1056/NEJMra1004965. [DOI] [PubMed] [Google Scholar]
  • 4.Rantapaa-Dahlqvist S, de Jong BA, Berglin E, Hallmans G, Wadell G, Stenlund H, et al. Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis Rheum. 2003;48(10):2741–9. doi: 10.1002/art.11223. [DOI] [PubMed] [Google Scholar]
  • 5.Nielen MM, van Schaardenburg D, Reesink HW, van de Stadt RJ, van der Horst-Bruinsma IE, de Koning MH, et al. Specific autoantibodies precede the symptoms of rheumatoid arthritis: a study of serial measurements in blood donors. Arthritis Rheum. 2004;50(2):380–6. doi: 10.1002/art.20018. [DOI] [PubMed] [Google Scholar]
  • 6.Sokolove J, Bromberg R, Deane KD, Lahey LJ, Derber LA, Chandra PE, et al. Autoantibody epitope spreading in the pre-clinical phase predicts progression to rheumatoid arthritis. Plos ONE. 2012;7(5):e35296. doi: 10.1371/journal.pone.0035296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Romero V, Fert-Bober J, Nigrovic PA, Darrah E, Haque UJ, Lee DM, et al. Immune-mediated pore-forming pathways induce cellular hypercitrullination and generate citrullinated autoantigens in rheumatoid arthritis. Science translational medicine. 2013;5(209):209ra150. doi: 10.1126/scitranslmed.3006869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Khandpur R, Carmona-Rivera C, Vivekanandan-Giri A, Gizinski A, Yalavarthi S, Knight JS, et al. NETs are a source of citrullinated autoantigens and stimulate inflammatory responses in rheumatoid arthritis. Science translational medicine. 2013;5(178):178ra40. doi: 10.1126/scitranslmed.3005580. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Dwivedi N, Upadhyay J, Neeli I, Khan S, Pattanaik D, Myers L, et al. Felty's syndrome autoantibodies bind to deiminated histones and neutrophil extracellular chromatin traps. Arthritis Rheum. 2012;64(4):982–92. doi: 10.1002/art.33432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Pratesi F, Dioni I, Tommasi C, Alcaro MC, Paolini I, Barbetti F, et al. Antibodies from patients with rheumatoid arthritis target citrullinated histone 4 contained in neutrophils extracellular traps. Ann Rheum Dis. 2013 doi: 10.1136/annrheumdis-2012-202765. [DOI] [PubMed] [Google Scholar]
  • 11.Arnett F, Edworthy S, Bloch D, McShane D, Fries J, Cooper N, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1988;31:315–24. doi: 10.1002/art.1780310302. [DOI] [PubMed] [Google Scholar]
  • 12.Liu C, Batliwalla F, Li W, Lee A, Roubenoff R, Beckman E, et al. Genome-wide association scan identifies candidate polymorphisms associated with differential response to anti-TNF treatment in rheumatoid arthritis. Molecular medicine. 2008;14(9-10):575–81. doi: 10.2119/2008-00056.Liu. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Zhao X, Okeke NL, Sharpe O, Batliwalla FM, Lee AT, Ho PP, et al. Circulating immune complexes contain citrullinated fibrinogen in rheumatoid arthritis. Arthritis Res Ther. 2008;10(4):R94. doi: 10.1186/ar2478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sokolove J, Brennan MJ, Sharpe O, Lahey LJ, Kao AH, Krishnan E, et al. Brief report: citrullination within the atherosclerotic plaque: a potential target for the anti-citrullinated protein antibody response in rheumatoid arthritis. Arthritis Rheum. 2013;65(7):1719–24. doi: 10.1002/art.37961. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nature methods. 2009;6(5):359–62. doi: 10.1038/nmeth.1322. [DOI] [PubMed] [Google Scholar]
  • 16.Paniagua RT, Chang A, Mariano MM, Stein EA, Wang Q, Lindstrom TM, et al. c-Fms-mediated differentiation and priming of monocyte lineage cells play a central role in autoimmune arthritis. Arthritis Res Ther. 2010;12(1):R32. doi: 10.1186/ar2940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Sohn DH, Sokolove J, Sharpe O, Erhart JC, Chandra PE, Lahey LJ, et al. Plasma proteins present in osteoarthritic synovial fluid can stimulate cytokine production via Toll-like receptor 4. Arthritis Res Ther. 2012;14(1):R7. doi: 10.1186/ar3555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Sokolove J, Zhao X, Chandra PE, Robinson WH. Immune complexes containing citrullinated fibrinogen costimulate macrophages via Toll-like receptor 4 and Fcgamma receptor. Arthritis Rheum. 2011;63(1):53–62. doi: 10.1002/art.30081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Clark RA, Nauseef WM. Isolation and functional analysis of neutrophils. Current protocols in immunology / edited by John E Coligan [et al] 2001;Chapter 7:23. doi: 10.1002/0471142735.im0723s19. Unit 7. [DOI] [PubMed] [Google Scholar]
  • 20.Clavel C, Nogueira L, Laurent L, Iobagiu C, Vincent C, Sebbag M, et al. Induction of macrophage secretion of tumor necrosis factor alpha through Fcgamma receptor IIa engagement by rheumatoid arthritis-specific autoantibodies to citrullinated proteins complexed with fibrinogen. Arthritis Rheum. 2008;58(3):678–88. doi: 10.1002/art.23284. [DOI] [PubMed] [Google Scholar]
  • 21.Xu J, Zhang X, Pelayo R, Monestier M, Ammollo CT, Semeraro F, et al. Extracellular histones are major mediators of death in sepsis. Nat Med. 2009;15(11):1318–21. doi: 10.1038/nm.2053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Huang H, Evankovich J, Yan W, Nace G, Zhang L, Ross M, et al. Endogenous histones function as alarmins in sterile inflammatory liver injury through Toll-like receptor 9 in mice. Hepatology. 2011;54(3):999–1008. doi: 10.1002/hep.24501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Allam R, Scherbaum CR, Darisipudi MN, Mulay SR, Hagele H, Lichtnekert J, et al. Histones from dying renal cells aggravate kidney injury via TLR2 and TLR4. J Am Soc Nephrol. 2012;23(8):1375–88. doi: 10.1681/ASN.2011111077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Semeraro F, Ammollo CT, Morrissey JH, Dale GL, Friese P, Esmon NL, et al. Extracellular histones promote thrombin generation through platelet-dependent mechanisms: involvement of platelet TLR2 and TLR4. Blood. 2011;118(7):1952–61. doi: 10.1182/blood-2011-03-343061. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Zendman AJ, van Venrooij WJ, Pruijn GJ. Use and significance of anti-CCP autoantibodies in rheumatoid arthritis. Rheumatology (Oxford) 2006;45(1):20–5. doi: 10.1093/rheumatology/kei111. [DOI] [PubMed] [Google Scholar]
  • 26.Leadbetter EA, Rifkin IR, Hohlbaum AM, Beaudette BC, Shlomchik MJ, Marshak-Rothstein A. Chromatin-IgG complexes activate B cells by dual engagement of IgM and Tolllike receptors. Nature. 2002;416(6881):603–7. doi: 10.1038/416603a. [DOI] [PubMed] [Google Scholar]
  • 27.Boule MW, Broughton C, Mackay F, Akira S, Marshak-Rothstein A, Rifkin IR. Toll-like receptor 9-dependent and -independent dendritic cell activation by chromatin-immunoglobulin G complexes. J Exp Med. 2004;199(12):1631–40. doi: 10.1084/jem.20031942. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Marshak-Rothstein A, Rifkin IR. Immunologically active autoantigens: the role of tolllike receptors in the development of chronic inflammatory disease. Annual review of immunology. 2007;25:419–41. doi: 10.1146/annurev.immunol.22.012703.104514. [DOI] [PubMed] [Google Scholar]
  • 29.Sanchez-Pernaute O, Filkova M, Gabucio A, Klein M, Maciejewska-Rodrigues H, Ospelt C, et al. Citrullination enhances the pro-inflammatory response to fibrin in rheumatoid arthritis synovial fibroblasts. Ann Rheum Dis. 2013;72(8):1400–6. doi: 10.1136/annrheumdis-2012-201906. [DOI] [PubMed] [Google Scholar]
  • 30.Kuhn KA, Kulik L, Tomooka B, Braschler KJ, Arend WP, Robinson WH, et al. Antibodies against citrullinated proteins enhance tissue injury in experimental autoimmune arthritis. J Clin Invest. 2006;116(4):961–73. doi: 10.1172/JCI25422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Padyukov L, Silva C, Stolt P, Alfredsson L, Klareskog L. A gene-environment interaction between smoking and shared epitope genes in HLA-DR provides a high risk of seropositive rheumatoid arthritis. Arthritis Rheum. 2004;50(10):3085–92. doi: 10.1002/art.20553. [DOI] [PubMed] [Google Scholar]
  • 32.Hill JA, Southwood S, Sette A, Jevnikar AM, Bell DA, Cairns E. Cutting edge: the conversion of arginine to citrulline allows for a high-affinity peptide interaction with the rheumatoid arthritis-associated HLA-DRB1*0401 MHC class II molecule. J Immunol. 2003;171(2):538–41. doi: 10.4049/jimmunol.171.2.538. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp FigureS1

Supplementary Figure 1. ACPA repertoire induced by citrullinated and native H2B immunization.

NIHMS707993-supplement-Supp_FigureS1.tif (2.6MB, tif)
Supp TableS1-S4

Supplementary Table 1. Identification of predominant products of ionomycin activated neutrophils.

Supplementary Table 2. Proteomic identification of citrullination sites from ionomycin activated neutrophils.

Supplementary Table 3. Proteomic identification of in vitro histone citrullination sites.

Supplementary Table 4. Proteomic interrogation of all products of ionomycin activated neutrophils.

NIHMS707993-supplement-Supp_TableS1-S4.docx (43.6KB, docx)

RESOURCES