Anti-Ro52 Autoantibodies Are Related to Chronic Graft-vs.-Host Disease After Allogeneic Hematopoietic Stem Cell Transplantation

Chronic graft-vs.-host disease (cGVHD) remains a major cause of morbidity and mortality after allogeneic hematopoietic stem cell transplantation (allo-HSCT). Previous studies have shown that autoantibodies play an important role in the development of cGVHD. Anti-nuclear autoantibodies (ANA) is the most frequently detected autoantibodies in patients with cGVHD, but the role of anti-Ro52 autoantibodies (anti-Ro52) in cGVHD remains largely unknown. In this study, we analyzed autoantibodies from 84 patients after allo-HSCT, including 42 with active cGVHD and 42 without cGVHD. Autoantibodies were found in 36 (42.9%) patients. Among these autoantibody-positive patients, 28 (77.8%) patients had active cGVHD. The most frequent autoantibodies in patients with active cGVHD were ANA (50.0%), anti-Ro52 (28.6%) and anti-mitochondrial autoantibodies type 2 (4.8%). We further explored the association between anti-Ro52 and cGVHD. Patients with active cGVHD had higher anti-Ro52 levels than patients without cGVHD (P < 0.05). The increases of anti-Ro52 levels were more significant in patients with moderate/severe cGVHD compared to those of patients without cGVHD (P < 0.05). Stratified and multivariable logistic regression analysis demonstrated that moderate/severe cGVHD was an independent risk factor for the levels of anti-Ro52 (P < 0.01). ROC analysis confirmed anti-Ro52 as a risk factor for progression of skin cGVHD. Moreover, the anti-Ro52 levels were highly correlated with the levels of B cell-activating factor (BAFF) and IgG1 antibodies. Our study demonstrates that anti-Ro52 is associated with cGVHD. The increased levels of anti-Ro52 were associated with higher levels of BAFF and IgG1 antibodies, suggesting a mechanistic link between elevated anti-Ro52 levels and aberrant B cell homeostasis.


Study Design and Patient Eligibility
Patients with hematological malignancy undergoing allo-HSCT were enrolled in this study. This study included 42 patients with active cGVHD. Eligibility criteria were as follows: (1) >3 months from time of allo-HSCT; (2) not received prednisone (≥0.5 mg/kg per day) 2 weeks before sample collection; and (3) never received rituximab (anti-CD20 mAb) or ibrutinib (inhibitor of Bruton's tyrosine kinase). Forty-two patients without cGVHD were matched to 42 patients with active cGVHD according to age, gender, primary disease, time after transplantation, conditioning regimen, HLA typing, source of graft, and grade of acute GVHD. This study was performed in accordance with the Declaration of Helsinki and was approved by the institutional review board of Nanfang Hospital. All patients and donors gave written informed consent to participate in the study.

GVHD Prophylaxis and Treatment
Generally, all HLA-haploidentical donor (HID) patients were transplanted with a combination of bone marrow (BM) and peripheral blood stem cell (PBSC) grafts, whereas most HLAmatched sibling donor (MSD) patients received PBSC grafts (25,26). Cyclosporine A (CsA), methotrexate (MTX), and mycophenolate mofetil (MMF) were administered to most patients undergoing MSD transplant for GVHD prophylaxis. CsA + MTX + MMF + antithymocyte globulin (ATG) was administered to patients undergoing HID transplants for GVHD prophylaxis (25)(26)(27). Patients received CsA, MMF and steroids for acute GVHD treatment as detailed in a previous report (28). Anti-CD25 monoclonal antibody and other immunosuppressive drugs were used to treat steroid-resistant acute GVHD. Steroids and CsA were used initially to treat cGVHD and were used in combination with various immunosuppressive agents to treat cGVHD that was unresponsive to initial therapy (29).

GVHD Assessment
The diagnosis and grade of cGVHD on the day of sample collection, not at first diagnosis, were documented by clinical   (30). Patients with active cGVHD were defined as requiring the addition of highdose prednisone (≥2 mg/kg per day) or continued multiagent immunosuppression after sample collection (11,31). Patients without cGVHD were defined as patients who had not developed cGVHD by the time of sample collection. Patients with previous cGVHD that had resolved or who became asymptomatic by the time of sample collection were not included (11,31).

Detection of Serum Autoantibodies
The enrolled patients were screened for the presence of the following autoantibodies: anti-Ro52 autoantibodies (anti-Ro52), anti-nuclear autoantibodies (ANA), anti-histone autoantibodies (AHA), anti-ribosomal P protein autoantibodies (anti-Rib-P), anti-polymyositis/scleroderma autoantibodies (anti-PM/Scl), anti-histidyl tRNA synthetase autoantibodies (anti-Jo-1), anti-mitochondrial autoantibodies type 2 (AMA-M2), and anti-centromere-B autoantibodies (anti-CENP-B) (Euroimmun, Lubeck, Germany). The detection of ANA was performed by indirect immunofluorescence assay (IFA) using HEp-2 cells (AESKU ANA-IFA reagent kit). Patient' s serum was diluted 1:80 and allocated into the appropriate cells and was incubated slides 30 min. After the incubation, rinsed off the serum with washing buffer in a slide staining dish and following covered with FITC labeled anti-human IgG for 30 min. Slides were washed with washing buffer and sealed with mounting medium for automatic interpretation by the HELIOS system (AESKU

Enzyme-Linked Immunosorbent Assay
The levels of soluble B cell-activating factor (BAFF) and IgG1 in patient plasma samples were measured by commercially available enzyme-linked immunosorbent assay (ELISA) (DBLYS0B R&D Systems, Minneapolis, USA and 88-50560-22, Invitrogen, CA, USA, respectively). The plates were read using the CLARIO star system following the manufacturer's recommended procedures (BMG Labtech, Cary, NC, USA).

Statistical Analysis
The descriptive analysis of patient characteristics included median, minimum and maximum values for continuous variables and numbers and frequencies for categorical variables.
Fisher's exact test was performed in comparison of categorical variables. For continuous variables, Student's t-test was performed for comparisons between two groups. Univariable logistic regression analysis was performed for the factors listed in Table 1 to identify variables that were associated with the presence of autoantibodies. Factors that were significant at the 0.1 level from the univariable logistic regression were included in the multivariable logistic regression. Correlation studies were performed using Pearson's correlation test. Anti-Ro52 levels, a highly skewed variable, was transformed to logarithm with base 10 for meeting the normality assumption. Receiver operating characteristic (ROC) curves analysis and area under the curve (AUC) estimation were also performed in order to discriminate our interests and the optimum cut-off value was according to the Youden's index. All statistics were analyzed in GraphPad Software (Prism Version 6.0; GraphPad Software, San Diego, CA) or SPSS version 22.0 (SPSS, Chicago, USA). Tests for significance were 2-sided, with a significance P level of 0.05 or less.

Patient Characteristics
There There were no significant differences in age, gender, primary disease, time after transplantation, conditioning regimen, HLA typing, source of graft, and grade of acute GVHD between patients with and without cGVHD in our study ( Table 1).

Association Between Autoantibodies and cGVHD
There were no statistically significant differences in age, gender, primary disease, conditioning regimen, and acute GVHD grade between patients who developed autoantibodies and patients who did not develop autoantibodies. Compared with patients who did not develop autoantibodies, patients who developed autoantibodies have several characteristics, including HLA-matched transplant, PBSC graft, non-ATG based GVHD prophylaxis, and moderate/severe cGVHD ( Table 4).
Further stratified and multivariable logistic regression analysis demonstrated that moderate/severe cGVHD was an independent risk factor for the levels of autoantibodies (P < 0.001) ( Figure 1C and Table 5).

Association Between Anti-Ro52 and cGVHD
In our study, patients with active cGVHD had higher anti-Ro52 levels than patients without cGVHD (P < 0.05) ( Figure 3A). These increases of anti-Ro52 levels were more significant in patients with moderate/severe cGVHD compared to those of patients without cGVHD (median, 7.0 vs. 5.3 AU/mL; P < 0.05) (Figure 3B). Further stratified and multivariable logistic regression analysis demonstrated that moderate/severe cGVHD was an independent risk factor for the levels of anti-Ro52 (P < 0.01) ( Figure 3C and Table 7).

Correlation Between Anti-Ro52 and cGVHD Target Organ
We further explored the correlation between anti-Ro52 and cGVHD target organ by receiver operating characteristic (ROC) analyses. ROC analysis confirmed anti-Ro52 as a risk factor for progression of skin cGVHD (Figure 4A, cut-off = 8.60 at 85.7% sensitivity and 61.9% specificity, P < 0.05) but showed no correlation with other cGVHD target organs (Figures 4B-H).

Anti-Ro52 Correlated With the Generation of B-Cell Activating Factor (BAFF) and IgG1
It has been widely demonstrated that B cell homeostasis altered and BAFF and IgG1 levels increased in cGVHD patients (11,(33)(34)(35). We further examined whether anti-Ro52 was correlated with BAFF and IgG1 levels in these patients. Patients with anti-Ro52 positive had significantly higher BAFF levels than patients with anti-Ro52 negative (median, 7.0 vs. 5.1 pg/mL; P < 0.05) ( Figure 5A). Importantly, the anti-Ro52 levels were strongly correlated with the levels of BAFF (r = 0.64, P < 0.01) (Figure 5B). A higher level of IgG1 was observed in patients with anti-Ro52 positive when compared to patients with anti-Ro52 negative (median, 3.8 vs. 3.1 µg/mL; P < 0.05) (Figure 5C). The levels of anti-Ro52 were also strongly correlated with the levels of IgG1 (r = 0.47, P < 0.05) ( Figure 5D).

DISCUSSION
Recently, antibodies have been reported to play an important role in the development of cGVHD (7,33,36). Srinivasan et al.
showed that donor B cell-derived antibodies augmented the development of bronchiolitis obliterans in a murine model of cGVHD (9). Immunoglobulin G (IgG) deposition in the skin has been observed in murine cGVHD models (7,37). We previously  reported that donor B cell antibodies augment cutaneous cGVHD in mice by damaging the thymus and increasing tissue infiltration of pathogenic Th17 cells (7). In humans, Miklos et al. reported that alloantibodies to Y chromosome-encoded proteins correlated significantly with clinical cGVHD development (21,22). Our previous study showed that the levels of IgG1 correlated significantly with clinical cGVHD severity (11). It has also been demonstrated that circulating autoantibodies are associated with the development of clinical cGVHD (20,23,38). In this study, autoantibodies were detected in 36 (42.9%) patients: 28 (77.8%) patients had active cGVHD, and 8 (22.2%) patients had no cGVHD. The most common autoantibodies in patients with active cGVHD were ANA and anti-Ro52. ANA and anti-Ro52 were found in 21 (50.0%) and 12 (28.6%) active cGVHD  (23), which is consistent with our findings. In our study, patients with moderate/severe cGVHD had a trend toward higher ANA titers than patients without cGVHD (≥1:160: 41.9 vs. 7.1%, P < 0.01). Among 42 patients without cGVHD, two patients subsequently developed cGVHD 3.5 and 8.9 months later. These results indicate that autoantibodies are not initiated but augmented the development of cGVHD. These findings are consistent with our previous findings that antibodies from donor B cells perpetuate cutaneous cGVHD in mice (7). Ro52 is a RING finger protein that belongs to the tripartite motif family (TRIM) (24,39). Ro52 was identified as a major autoantigen in autoimmune disease, including rheumatoid arthritis, SLE, and Sjögren's syndrome (40)(41)(42). Like several other TRIM proteins, Ro52 acts in the process of ubiquitination and regulates immune responses by targeting key molecules involved in cell proliferation, survival or death (43)(44)(45). Several studies demonstrated that increased expression of the Ro52 autoantigen might be directly involved in the reduced cellular proliferation and increased apoptotic cell death observed in Sjögren's syndrome and SLE patients and might contribute to the autoantigenic load and induction of autoimmune B and T cell responses observed in rheumatic patients (45,46). Therefore, anti-Ro52 can be detected in patients with several different autoimmune diseases (47)(48)(49). In SLE as well as systemic sclerosis and autoimmune myositis patients, anti-Ro52 is detected in approximately one-third of the patients (50,51). Anti-Ro52 is also the most common specificity in patients with primary Sjögren's syndrome (66.7%) (52). The presence of anti-Ro52, either as a single specificity or in a combination with other specificities, is a factor associated with interstitial lung disease (53,54). However, the presence of anti-Ro52 in the cGVHD patients is rarely reported (55,56). Sarantopoulos et al. reported that the levels of anti-Ro52 in patients with unresponsive cGVHD after rituximab treatment increased (56). In our study, we found that patients with active cGVHD had higher anti-Ro52 levels than patients without cGVHD (P < 0.05). These increases of anti-Ro52 levels were more significant in patients with moderate/severe cGVHD compared to those of patients without cGVHD (median, 7.0 vs. 5.3 AU/mL; P < 0.05). Further stratified and multivariable logistic regression analysis demonstrated that moderate/severe cGVHD was an independent risk factor for the levels of anti-Ro52 (P < 0.01). ROC analysis confirmed anti-Ro52 as a risk factor for progression of skin cGVHD.
The presence of autoantibodies emphasizes the importance of B cells in the development of cGVHD (7,9,23,33,(36)(37)(38). The important role of B cells has also been confirmed by the successful treatment of some subgroups of cGVHD patients with the B cell-depleting agent rituximab (57)(58)(59)(60). It has been reported that Ro52 can bind to almost all B cells due to its interaction with the Fc domain of IgM and IgG. By binding directly to the B cell receptor, Ro52 might be capable of activating B cells in the absence of conventional immune receptor interactions (61,62). It has been widely demonstrated that B cell homeostasis altered and BAFF increased in cGVHD patients (33)(34)(35). BAFF expression might be indirectly regulated by Ro52 (63,64). We further examined whether anti-Ro52 was correlated with the levels of BAFF in these patients. Patients with anti-Ro52 positive had significantly higher BAFF levels than patients with anti-Ro52 negative (median, 7.0 vs. 5.1 pg/mL; P < 0.05). Importantly, the levels of anti-Ro52 were strongly correlated with the levels of BAFF (r = 0.64, P < 0.01). Several investigators have demonstrated that Ro52 might bind the Fc part of IgG molecules via the B30.2/PRYSPRY domain with unexpectedly high affinity. Ro52 functionally regulates quality control of IgG1 in B cells or plasma cells through the endoplasmic reticulum-associated degradation (ERAD) system (65)(66)(67). It has also been reported that the levels of IgG, especially IgG1, increased in Ro52-null mice with dermatitis (68). Our previous study showed that the levels of IgG1 correlated significantly with clinical cGVHD severity (11). We further examined the correlation between anti-Ro52 and IgG1 levels. A higher level of IgG1 was observed in patients with anti-Ro52 positive when compared to patients with anti-Ro52 negative (median, 3.8 vs. 3.1 µg/mL; P < 0.05). The levels of anti-Ro52 were also strongly correlated with the levels of IgG1 (r = 0.47, P < 0.05). Espinosa et al. observed that loss of the lupus autoantigen Ro52 induced tissue inflammation and systemic autoimmunity by dysregulating the IL-23-Th17 pathway (68). The development of cGVHD is mediated by pathogenic Th17 cells (7,69). Further studies are needed to explore whether anti-Ro52 are associated with Th17 cell development in cGVHD patients.
One limitation of this study was the limited sample size of patients. A kinetic study of anti-Ro52 prevalence was absent.
Kinetic studies of more patients will be conducted to explore the effect of anti-Ro52 on cGVHD development.

CONCLUSION
Our study demonstrates that the anti-Ro52 is associated with cGVHD. ROC analysis confirmed anti-Ro52 as a risk factor for progression of skin cGVHD. The levels of anti-Ro52 correlated with the severity of cGVHD and the levels of BAFF and IgG1 antibodies. Therefore, our findings support a mechanistic link between elevated anti-Ro52 levels and aberrant B cell homeostasis. Further studies will be needed to investigate the exact mechanisms of anti-Ro52 in cGVHD.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Requests to access the datasets should be directed to Hua Jin, echohua1124@163.com.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Nanfang Hospital. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
KY analyzed the data and wrote the manuscript. YC and HQ collected and analyzed the data. YY, ZF, FH, HZ, and YS assisted in the research. HJ and QL designed the study, supervised the research, and critically revised the manuscript. All authors contributed to the article and approved the submitted version.