Lower serum interleukin‐22 and interleukin‐35 levels are associated with disease status in neuromyelitis optica spectrum disorders

Abstract Aims The exact pathogenesis of neuromyelitis optica spectrum disorder (NMOSD) remains unclear. A variety of cytokines are involved, but few studies have been performed to explore the novel roles of interleukin‐22 (IL‐22) and interleukin‐35 (IL‐35) in NMOSD. Therefore, this study was designed to investigate serum levels of IL‐22 and IL‐35, and their correlations with clinical and laboratory characteristics in NMOSD. Methods We performed a cross‐section study, 18 patients with acute NMOSD, 23 patients with remission NMOSD, and 36 healthy controls were consecutively enrolled. Serum levels of IL‐22 and IL‐35 were measured by enzyme‐linked immunosorbent assay (ELISA). The correlations between serum IL‐22 and IL‐35 levels and clinical and laboratory characteristics were evaluated by Spearman's rank or Pearson's correlation coefficient. Results The serum levels of IL‐22 and IL‐35 were significantly lower in patients with acute NMOSD and remission NMOSD than in healthy controls (IL‐22: 76.96 ± 13.62 pg/mL, 87.30 ± 12.79 pg/mL, and 94.02 ± 8.52 pg/mL, respectively, P < .0001; IL‐35: 45.52 ± 7.04 pg/mL, 57.07 ± 7.68 pg/mL, and 60.05 ± 20.181 pg/mL, respectively, P < .0001). Serum levels of IL‐35 were negatively correlated with EDSS scores and cerebrospinal fluid protein levels (r = −.5438, P = .0002 and r = −.3523, P = .0258, respectively) in all patients. Conclusions Lower serum levels of IL‐22 and IL‐35 are associated with disease status in NMOSD. Additionally, lower serum levels of IL‐35 are associated with disease severity in NMOSD.


| INTRODUC TI ON
Neuromyelitis optica spectrum disorder (NMOSD) has become one of the most rapidly diagnosed and progressing diseases in the field of neurology in the past 10 years. The median age of onset is 39 years old, and it is much more common among women than men (reported ratios range from 3:1 to 9:1). 1,2 It was formerly known as optic neuromyelitis or Devic disease. A specific antibody, NMO-IgG, binds to aquaporin-4 (AQP4) at the synaptic ends of astrocytes located on the blood-brain barrier (BBB), resulting in the loss of myelin. 3  Interleukin-22 belongs to the IL-10 cytokine family but has unique traits that make it different from IL-10. A broad variety of lymphocytes secrete IL-22, including Th17, Th22, and NK cells, neutrophils, and innate lymphoid cells. 6 The biological effects of  are mediated by the IL-22-IL-22R pathway. 7 Its involvement has been reported in many autoimmune diseases, including protective, pathogenic, and dual effects. In protective effects, a decrease in IL-22 may be considered a risk factor for type 2 diabetes and could worsen hepatitis. 8,9 However, in other diseases, IL-22 could be pathogenic.
For example, in rheumatoid arthritis, patients who responded to treatment showed reductions in plasma IL-22 levels, indicating that plasma IL-22 might play a detrimental role in this disease. 10 In psoriasis, IL-22 downregulates Cx43 expression, which could lead to keratinocyte hyperproliferation. 11 Interestingly, in some diseases, IL-22 plays a dual role. In systemic lupus erythaematosus (SLE), decreased plasma IL-22 levels are correlated with SLE disease activity and could result in the rs2227513 polymorphism, which might contribute to SLE susceptibility. 12,13 In contrast, the Yang 14 group showed that IL-22 levels in both the serum and the kidneys were significantly higher in lupus nephritis (LN) patients than in healthy controls. Considering the complexity of IL-22, we determined that identifying its role in NMOSD might be slightly challenging.
Interleukin-35 is the newest identified anti-inflammatory cytokine, which belongs to the interleukin-12 cytokine family, a unique group of heterodimeric cytokines that include IL-12, IL-23, IL-27, and IL-35. 15 Intracellular signaling is mediated by the activation of the JAK-STAT pathway. 16 In patients with conditions such as ulcerative colitis, Crohn's disease, primary Sjogren's syndrome, and chronic obstructive pulmonary disease (COPD), the serum levels of IL-35 are low. [17][18][19][20] Mesenchymal stem cells with IL-35 overexpression had stronger immunosuppressive effects than were exerted by nontransfected mesenchymal stem cells. 21 Wang 22 found that treating experimental autoimmune uveitis (EAU) with IL-35 suppressed uveitis by inhibiting Th17 and Th1 differentiation and inducing the production of regulatory Breg/IL-35 + Breg cells, as well as Tregs. Data from the above studies demonstrate that IL-35 has an immunosuppressive function in inflammation and that it might be a new target therapy for autoimmune diseases. As NMOSD has very high disability and relapse rates, some patients do not respond to current therapies. It is therefore urgent for us to find new therapies. However, few studies have explored the exact role of IL-35 in NMOSD.
Therefore, in this study, to improve our understanding of IL-22 and IL-35 interrelationships and immunopathologic roles in NMOSD, we investigated the serum levels of IL-22 and IL-35 and their correlations with clinical and laboratory characteristics.

| Study population
This was a cross-section study, which was carried out at Shanghai Renji Hospital affiliated with Shanghai Jiao Tong University. In all, 41 patients (18 with acute NMOSD and 23 with remission NMOSD) were consecutively recruited from the clinical centre from January 2015 to June 2016. The patients who were included in this study needed to meet the international consensus diagnostic criteria for NMOSDs established by the Wingerchuk 4 group in 2015. For acute NMOSD, they were required to be drug-naïve before sampling. Considering that other autoimmune diseases might influence the results, we had already excluded them before they were recruited in our research.
In total, 36 healthy individuals from the Health Care Centre of Renji Hospital (10 men and 26 women) were also enrolled in the study as a control group. There was no acute or chronic sickness in the control subjects.
The Ethics Committee of Renji Hospital approved all of the research procedures, and we obtained informed consent from all of the participants.

| Clinical assessment
Using the hospital's electronic medical record review system, we collected clinical data related to NMOSD patients, including gender, age, the total duration of the disease, the current duration of the disease, the first episode, the first episode diagnosis, the annualized relapse rate (ARR), the serum autoimmune antibodies, the expanded disability status scale (EDSS) score, the presentation of optic neuritis protein levels, immunoglobulin G (IgG) levels, the IgG index, the 24hour intrathecal IgG synthesis rate and other biochemical indicators as well as head MRI, cervical MRI, and lumbar MRI results. Regarding these data, the period from the time of the current attack to the time of sampling was regarded as the current disease duration. If the current disease duration was <30 days, the patient was defined as being in the acute phase of the attack. If the current disease duration was more than 30 days, the patient was defining as being in the remission phase of the attack. The period from the time of the first attack to the time of sampling was regarded as the total disease duration. EDSS scores were assessed at the time of sampling. In addition, considering the effects of drugs, patients in the acute phase were required to have not used immunomodulatory drugs before sampling in this study.

| Statistical analysis
Statistical analyses were performed using SPSS (version 23.0) and the GraphPad Prism (version 7.0) statistical program. The Kolmogorov-Smirnov Z test was used to detect the normality of the data. The count data are expressed as a percentage (%), and the chi-square test or Fisher's exact test was used for comparisons between two groups. Measurement data with a normal distribution are expressed as the mean ± standard deviation (SD). Student's t test was used for comparisons between two groups. One-way ANOVA was used for comparisons among the three groups (total NMOSD, acute NMOSD, remission NMOSD), and LSD multiple comparisons test was used for comparisons within a group. Measurements of non-normal distribution were expressed as the median and interquartile range (IQR). The Mann-Whitney U test was used for comparisons between groups. Correlations between continuous variables were analyzed by Spearman's rank correlation coefficient or the Pearson correlation coefficient. A two-tailed probability value <.05 was considered statistically significant.

| Demographic and clinical features of subjects with NMOSD
A total of 41 patients were enrolled. There were 18 patients in the acute phase and drug-naïve phase and 23 patients in the remission phase. The mean age was 50.00 ± 14.075 years old in the acute group and 40.65 ± 11.089 years old in the remission group (P < .05). The  Table 1).
We enrolled 36 healthy subjects as controls (mean age, 44.58 ± 13.137 years old, female/male ratio: 26/10), and there were no significant differences between the controls and the total number of patients (mean age, 44.76 ± 13.187 years old; female/male ratio: 29/12) in age and the sex ratio (P = .954 and P = .887, respectively; Table 1).

| Serum levels of IL-22 and IL-35 in NMOSD patients and healthy controls
The serum levels of IL-22 were significantly lower in patients with acute NMOSD and remission NMOSD than in the healthy controls (HC; 76.96 ± 13.62 pg/mL, 87.30 ± 12.79 pg/mL, and 94.02 ± 8.52 pg/ mL, respectively; P < .0001). Additionally, there were significant differences between the acute NNOSD and HC groups, between the acute NMOSD and remission NMOSD groups, and between the remission NMOSD and HC groups (P 1 < .0001, P 2 < .01, and P 3 < .05, respectively; Table 2, Figure 1).

| D ISCUSS I ON
F I G U R E 1 A, Serum interleukin-22 (IL-22) levels in patients with neuromyelitis optica spectrum disorder (NMOSD) and healthy controls. Using one-way ANOVA, we found there were significant differences among the three groups (P < .0001). Using LSD multiple comparisons test for acute NNOSD and HC, acute NMOSD and remission NMOSD, and remission NMOSD and HC, we found there were significant differences (P 1 < .0001, P 2 < .01, and P 3 < .05, respectively). B, Serum interleukin-35 (IL-35) levels in patients with NMOSD and healthy controls. Using one-way ANOVA, we show there were significant differences among the three groups (P < .0001). Using LSD multiple comparisons test for acute NMOSD and HC, acute NMOSD and remission NMOSD, and remission NMOSD and HC, we show there were significant differences (P 1 < .0001, P 2 < .01, and P 3 < .01, respectively; *P < .05; **P < .01; ***P < .001; ****P < .0001) Treg cells is an important factor underlying NMOSD. 33 In addition, because of findings related to AQP4 antibodies, it is clear that B cell-mediated humoral immunity plays an important role in the pathogenesis of NMOSD. [38][39][40] In some human autoimmune diseases, there is a lack of equivalent Breg cells with regard to quantity or function. 41 An interesting finding was that IL-35 was also correlated with the number of Breg cells. 22  and IL-10R2 mRNAs.

| CON CLUS ION
In conclusion, we found that serum levels of IL-22 and IL-35 were lower in the NMOSD group than in HCs and that serum levels of IL-35 were negatively associated with EDSS scores and CSF pro-