Identification of novel biomarker as citrullinated inter-alpha-trypsin inhibitor heavy chain 4, specifically increased in sera with experimental and rheumatoid arthritis

Male DBA/1 mice were purchased from Charles River Japan (Tokyo, Japan), and used at 6–10 weeks of age. All mice were maintained under specific pathogen-free conditions at the University of Tsukuba. All experimental protocols were approved by the Institutional Animal Care and Use Committee of the University of Tsukuba, and all animal experiments were conducted in accordance with institutional ethics guidelines. All surgeries were performed under isoflurane anesthesia, and utmost care was taken to minimize suffering.

Serum samples were collected from Japanese patients with RA (n = 60, mean age 52.2 years, range 20–73 years, females 80%) diagnosed by rheumatologists according to the 1987 American College of Rheumatology (ACR) classification criteria [14] or the 2010 ACR/European League Against Rheumatism classification criteria [15]. Serum samples of the disease control subjects were collected from Japanese patients with primary Sjögren’s syndrome (SS) (n = 27, mean age 58.8 years, range 26–80 years, females 96%), systemic lupus erythematosus (SLE) (n = 15, mean age 33.9 years, range 16–55 years, females 80%), or osteoarthritis (OA) (n = 12, mean age 56.3 years, range 37–80 years, females 67%). All the patients with SS were diagnosed by rheumatologists according to the 1999 Japanese Ministry of Health criteria for diagnosis of SS [16]. All the patients with SLE fulfilled the 1997 ACR classification criteria [17]. None of the patients with SS or SLE had overlapping RA. Serum samples were collected from healthy subjects (HS) (n = 30, mean age 49.0 years, range 34–65 years, females 80%). Serum samples were also collected from 17 patients with RA before and 24 weeks after treatment with biologic drugs (infliximab, n = 9; abatacept, n = 8). All samples were collected at the University of Tsukuba Hospital after informed consent was obtained. This study was reviewed and approved by the ethics committee of the University of Tsukuba.

DBA/1 mice were immunized with 25 μg of pGPI (Invitrogen/Thermo Fisher Scientific, Carlsbad, CA, USA) in complete Freund’s adjuvant (CFA) (BD Biosciences, San Jose, CA, USA). pGPI was emulsified with CFA at a 1:1 ratio (vol/vol), or PBS + CFA was prepared as a vehicle control. For induction of arthritis, 150 μl of the emulsion was injected intradermally at the base of the tails of the mice. Each mouse was also given an injection of 200 ng of pertussis toxin (Sigma-Aldrich, St. Louis, MO, USA) intraperitoneally on days 0 and 2 after immunization to induce arthritis. Arthritis was assessed every other day and evaluated using a scale of 0–3 for swelling and redness of each paw. The clinical score was the sum of the scores for four paws, as described previously [11].

Sera were obtained on days 0–28 every week from mice immunized with pGPI or control, and antibodies were measured by enzyme-linked immunosorbent assay (ELISA). Sera were diluted 1:25 in dilution buffer and added to the 96-well plate (Immunoscan CCPlus test kit; Euro Diagnostica, Malmö, Sweden) for 1 h at room temperature. After a washing step, horseradish peroxidase (HRP)-conjugated polyclonal rabbit antimouse immunoglobulin (Dako, Carpinteria, CA, USA) diluted 1:1000 was added for 30 minutes at room temperature. After another washing step, color was developed with 3,3′,5,5′-tetramethylbenzidine (TMB) microwell peroxidase substrate (KPL/SeraCare, Milford, MA, USA). The optical density (OD) was measured at 450 nm by using a microplate reader.

A standard pool was obtained by mixing sera obtained from several mice on day 28. The concentrations of antibodies in this pool were considered 100 U/ml. A standard curve was obtained using serum dilutions, and the Michaelis-Menten equation was used to convert OD values into units, as described previously [18].

Native peptide ITIH4_428–447 and R438 citrullinated peptide ITIH4_428–447 were synthesized (serum, purity 95%) and used in ELISA. Ninety-six-well plates (Nunc MaxiSorp; Thermo Fisher Scientific) were coated with 10 μg/ml peptides for 12 h at 4 °C. After washing and blocking steps, sera from patients with RA (n = 60) and HS (n = 30) were diluted 1:200 in 1% bovine serum albumin (BSA) in PBS, then added for 2 h at room temperature. After a washing step, HRP-conjugated goat antihuman immunoglobulin G (IgG) (heavy and light chains [H + L]) (Abcam, Cambridge, MA, USA) diluted 1:10,000 was added for 1 h at room temperature. After washing, color was developed with TMB microwell peroxidase substrate (KPL/SeraCare). The OD was measured at 450 nm by using a microplate reader. The cutoff value was determined as the mean + 2 SD of HS.

Ankle joint tissue samples were harvested at days 0 and 14 from mice immunized with pGPI or control. They were fixed in neutralized 10% formalin, embedded in paraffin, and sectioned. To detect citrullinated proteins, we prepared modification buffer by mixing Reagent A (20% H₂SO₄, 25% H₃PO₄, and 0.025% FeCl₃) and Reagent B (1% diacetyl monoxime, 0.5% antipyrine, 1 M acetic acid) at a 2:1 ratio (vol/vol). The sections were covered with the modification buffer and incubated in a light-proof container at 37 °C for 2.5 h to modify citrulline residues. Then, the sections were incubated overnight at room temperature with rabbit anti-modified citrulline (AMC) polyclonal antibodies diluted 1:3200 in 2% BSA in PBS to detect modified citrulline residues [19]. The sections were also incubated with HRP-conjugated goat antirabbit IgG (H + L) (Bio-Rad Laboratories, Hercules, CA, USA) for 30 minutes at room temperature. The sections were also stained with 3,3′-diaminobenzidine (DAB) (Nichirei Biosciences, Tokyo, Japan) and hematoxylin. To detect ITIH4 or macrophages, IHC analysis of the joint or liver (as a positive control of ITIH4 staining) sections was performed as described above, using rabbit antihuman ITIH4 antibodies (Abcam) diluted 1:1000 and rat antimouse F4/80 antibodies (BioLegend, San Diego, CA, USA) diluted 1:200 as primary antibodies and using rabbit-specific IHC polymer detection kit HRP/DAB (Abcam) and HRP-conjugated polyclonal rabbit antirat immunoglobulin (Dako) diluted 1:200 as secondary antibodies, respectively.

We extracted total RNA from the ankle joints of pGIA or control mice using the ISOGEN (Wako Pure Chemical Industries, Tokyo, Japan) extraction method according to the instructions provided by the manufacturer. The extracted RNA was reverse-transcribed to complementary DNA with random primers. We performed real-time qPCR using a TaqMan gene expression assay (Applied Biosystems/Thermo Fisher Scientific, Foster City, CA, USA), and the Padi4 (Mm01341658_m1) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (NM_002046) primers. Real-time qPCR was carried out using an ABI 7500 analyzer (Applied Biosystems). The expression of GAPDH was used as the control.

Serum samples were obtained at days 0, 7, 14, and 28 from mice immunized with pGPI or control, as well as from patients with RA, SS, SLE, or OA and from HS. For analysis of citrullinated protein expression, serum samples were loaded into each well (50 μg for SDS-PAGE or 100 μg for two-dimensional PAGE [2D-PAGE]), separated by SDS-PAGE or 2D-PAGE, and transferred to polyvinylidene difluoride membranes. The modification buffer was added to the blots before incubation in a light-proof container at 37 °C for 2.5 h to modify citrulline residues, as described above. Blots were washed with 0.05% Tween 20 in Tris-buffered saline (TBST) and blocked with 5% milk in TBST, then incubated overnight at 4 °C with AMC antibodies diluted 1:3200 in 5% milk in TBST [19]. After a washing step, the blots were incubated with secondary antibody HRP-conjugated goat antirabbit IgG (H + L) (Bio-Rad Laboratories) diluted 1:5000 in 5% milk in TBST for 1 h at room temperature. Densitometric analysis was carried out using an ImageQuant LAS 4000 densitometer (GE Healthcare Life Sciences, Marlborough, MA, USA). The band intensity was determined with ImageQuant TL software (GE Healthcare Life Sciences) and normalized using the value of one sample (pGIA or RA) as 10. The difference of citrullinated ITIH4 levels from baseline to 24 weeks after treatment was normalized using each value at baseline. For analysis of ITIH4 expression, serum samples were loaded 10 μg/well and Western blot analysis was performed as described above, using rabbit antihuman ITIH4 antibodies (Abcam) diluted 1:3000.

Sera were obtained at days 0, 7, 14, and 28 from mice immunized with pGPI or control, and citrulline contents were measured by using the color development reagent (COLDER) assay, as described previously [20]. Sera were diluted 1:100 in COLDER buffer (50 mM NaCl, 10 mM CaCl₂, 2 mM dithiothreitol, and 100 mM Tris, pH 7.4), and 60 μl of samples were added to the 96-well plates (Nunc MaxiSorp). Then, 200 μl of COLDER were added to detect citrullinated proteins. The samples were incubated for 30 minutes at 95 °C. The OD was measured at 570 nm by using a microplate reader and compared with a standard curve of known citrulline concentrations. Data were normalized relative to the protein concentration in each sample.

All data were expressed as mean ± SEM. Differences between groups were evaluated for statistical significance using Student’s t test. The Kruskal-Wallis test was used for evaluation of band intensity among the five subject groups. P values less than 0.05 were considered significant. Statistical analyses were performed using IBM SPSS Statistics software (IBM, Armonk, NY, USA).

First, we investigated the production of ACPA in pGIA. DBA/1 mice were immunized with pGPI, and we induced symmetrical polyarthritis characterized by severe swelling of the limb joints (Additional file 1: Figure S1a). ACPA titers were measured in pGIA sera by ELISA, and they were significantly higher than those of control mice (Additional file 1: Figure S1b).

Then, we evaluated the expression of citrullinated proteins in pGIA because high titers of autoantibodies against citrullinated proteins were detected. Ankle joint sections were stained with AMC antibodies by IHC. Citrullinated proteins were detected in joints at day 14 (the peak arthritic phase) and expressed in areas of synovial hyperplasia in pGIA but not in controls (Fig. 1a). Citrullinated proteins were not expressed in several other tissues examined (e.g., lung, spleen, lymph node) (data not shown). We analyzed the gene expression of Padi4, which is an enzyme that catalyzes protein citrullination, in articular tissue samples from pGIA by qPCR. At day 14, Padi4 levels tended to be higher in the arthritic joints of pGIA mice than in control mice, albeit insignificantly (Additional file 2: Figure S2a).

In our previous study, PAD4 levels were increased in sera of patients with RA [21], so we next explored the expression of citrullinated proteins in sera. Western blot analysis of serum samples using AMC antibodies identified citrullinated proteins in pGIA mice but in none of the control mice (Fig. 1b). Interestingly, the band of citrullinated protein was detected as a common band at ~ 120 kDa. The band was detected from the prearthritic phase and with significantly stronger intensity in the peak arthritic phase, then it decreased in line with the self-limited arthritis (Fig. 1c). Whereas citrullinated proteins were detected in serum samples (Additional file 3: Figure S3a, right), they were not detected when Western blot analysis was performed without modification of citrulline residues (Additional file 3: Figure S3a, left), suggesting that the detected band was really specific for citrulline.

To further confirm the serum levels of citrullinated proteins, we measured citrulline contents in pGIA sera using the COLDER assay. Serum citrulline content tended to be higher in pGIA at day 14 than in the controls (Additional file 3: Figure S3b). These results suggest that citrullinated proteins appeared in joints and sera in association with arthritis in pGIA.

To identify the 120 kDa citrullinated protein in sera, we analyzed the protein by MS. First, we separated the serum samples of pGIA by 2D-PAGE and confirmed the spot of citrullinated proteins at ~ 120 kDa by Western blot analysis (Fig. 2a, right). Second, the spot in the gels was sliced after Coomassie brilliant blue staining (Fig. 2a, left), digested with trypsin, and analyzed by nanoUPLC-MS to identify the protein. ITIH4 was identified with high coverage (Table 1), and we examined citrulline modifications by nanoUPLC-MS^E. Citrullination of several sites containing R438 were identified in ITIH4 from pGIA (Table 1 and Fig. 2b). Analysis of the samples by MS after SDS-PAGE (Additional file 4: Figure S4a) also confirmed the presence of ITIH4 with high coverage, as well as citrullination at R438 in ITIH4 (Additional file 4: Figure S4b).

To further investigate the association between sera and arthritic joints, we explored the expression of ITIH4 in joints in pGIA. Ankle joint sections were stained with anti-ITIH4 antibodies by IHC, and liver sections were stained as positive controls to detect ITIH4 (Additional file 5: Figure S5a). In naïve mice, ITIH4 was not detected in joints, but ITIH4 was clearly expressed in inflamed synovium at day 14 in pGIA (Fig. 2c). Macrophages detected by anti-F4/80 antibodies in serial sections showed colocalization with ITIH4 in some areas (Fig. 2d), suggesting that synovial macrophages may express ITIH4 in joints of pGIA. Moreover, ITIH4 depositions colocalized with citrullinated proteins in inflamed synovium in comparative serial sections (Fig. 2e), suggesting the presence of citrullinated ITIH4 in joints of pGIA. These results indicate the specific expression of ITIH4 and citrullinated ITIH4 in arthritic joints in pGIA.

We also investigated the expression of citrullinated proteins in sera of 60 patients with RA and 30 HS, as well as in 12 patients with osteoarthritis (OA), 15 patients with SLE, and 27 patients with SS (disease controls). As well as in pGIA mice, the 120 kDa band of citrullinated protein was detected in 82% of patients with RA but in none of the control groups, except for being faintly detected in 2 patients with SS (7%) and 1 patient with SLE (7%) (Fig. 3a and Additional file 6: Figure S6a). The levels of this citrullinated protein were significantly higher in patients with RA than in controls (Fig. 3b).

We analyzed the 120 kDa citrullinated protein in patients with RA by 2D-PAGE and MS in the same way as pGIA mice. Using 2D-PAGE and Western blot analysis, we first confirmed the spot of citrullinated protein at ~ 120 kDa, similar to pGIA (Fig. 3c). Second, MS of this spot identified ITIH4 with high coverage (Table 2). Analysis of citrulline modifications identified citrullination of several sites in ITIH4 from patients with RA. Citrullinated sites varied among individual samples, but the citrullination of R438 was common among the examined samples (n = 3) (Table 2 and Fig. 3d).

To confirm whether the 120 kDa citrullinated protein was ITIH4, we performed Western blot analysis of serum samples using anti-ITIH4 antibodies. The bands were detected at 120 kDa by anti-ITIH4 antibodies in the same position as the bands detected by AMC antibodies (Fig. 3e). In addition, the levels of total ITIH4 were similar in patients with RA and controls, whereas the levels of citrullinated ITIH4 were specifically increased in patients with RA relative to controls (Fig. 3e). Using ELISA, we also measured total ITIH4 concentrations in sera from 60 patients with RA and 30 HS who were the same individuals used in Western blot analysis. There was no significant difference in total ITIH4 levels between patients with RA and HS (data not shown). These results indicate that the levels of citrullinated ITIH4 rather than total ITIH4 were specifically increased in association with arthritis in sera of patients with RA.

Citrullinated ITIH4 was previously identified as one of various citrullinated proteins in RA synovium [6]; however, its function and association with RA are not clear. We compared the clinical features of patients with RA with or without serum citrullinated ITIH4 as confirmed by Western blot analysis, using a band intensity cutoff value of 0.46, representing the mean + 3 SD of HS. C-reactive protein (CRP) and rheumatoid factor (RF) levels and Disease Activity Score in 28 joints measured by CRP (DAS28-CRP) scores were significantly higher in citrullinated ITIH4-positive than in ITIH4-negative patients (p = 0.047, p = 0.011, and p = 0.040, respectively), whereas there was no association between citrullinated ITIH4 and anti-CCP antibody titer or anti-CCP antibody-positive rate (Table 3). These results suggest that high levels of citrullinated ITIH4 were associated with inflammatory markers and disease activity of patients with RA.

In addition, we analyzed the relationship between serum citrullinated ITIH4 and RF/ACPA levels in patients with RA. The levels of citrullinated ITIH4 were significantly higher in RF⁺ ACPA⁺ patients compared with RF⁻ ACPA⁺, RF⁻ ACPA⁻ patients (p = 0.011, p = 0.025, respectively) (Table 4), suggesting that citrullinated ITIH4 levels were associated with RF levels in patients with RA. Interestingly, citrullinated ITIH4 were detected in 88% of RF⁻ ACPA⁻ patients, although those levels were low (Table 4). It was suggested that positivity of citrullinated ITIH4 could be a useful marker in seronegative patients with RA as well.

To further clarify whether serum citrullinated ITIH4 levels were correlated with the disease activity of patients with RA, we recruited patients who continued biologic treatment for 24 weeks (infliximab, n = 9; abatacept, n = 8) among citrullinated ITIH4-positive patients. Because treatment with tocilizumab downregulates inflammatory markers sometimes irrespective of arthritis activity, we chose infliximab and abatacept in this experiment. At baseline, the levels of citrullinated ITIH4 were positively correlated with DAS28-CRP, DAS28-ESR, Simplified Disease Activity Index, and Clinical Disease Activity Index (Fig. 4a–d). In our comparison of the differences from baseline to 24 weeks after treatment, we found that citrullinated ITIH4 levels were decreased after treatment in responders (Fig. 4e). The difference of citrullinated ITIH4 levels was significantly correlated with the difference of DAS28-CRP and DAS28-ESR before and after treatment (Fig. 4f, g). These results suggest that serum citrullinated ITIH4 levels were positively correlated with disease severity and that citrullinated ITIH4 could be a serum marker reflecting disease activity in patients with RA.

Next, we investigated the autoantigenicity of citrullinated ITIH4. We synthesized a citrulline-containing ITIH4_428–447 peptide (pITIH4) and the corresponding arginine-containing pITIH4 because R438 was identified as a dominant citrullination site. We assessed antibody reactivities against native or citrullinated pITIH4 by ELISA. On one hand, the levels of anticitrullinated pITIH4 antibody in patients with RA were significantly higher than those of HS (Fig. 5a, right). On the other hand, there was no significant difference in the levels of antinative pITIH4 antibody between patients with RA and HS (Fig. 5a, left). When the cutoff value was determined using sera from HS, 21.7% of RA sera was shown to be reactive with citrullinated pITIH4. The sensitivity and specificity were 21.7% and 96.7%, respectively. These results suggest the possibility of an autoimmune reaction to citrullinated ITIH4 in patients with RA.