Early restoration of immune and vascular phenotypes in systemic lupus erythematosus and rheumatoid arthritis patients after B cell depletion

Abstract This translational multi‐centre study explored early changes in serologic variables following B lymphocyte depletion by rituximab (RTX) treatment in systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) patients and investigated in vitro effects on the activity of other immune cells and the vascular endothelium. Eighty‐five SLE patients, seventy‐five RA patients and ninety healthy donors were enrolled. Two additional cohorts of selected SLE and RA patients were treated with RTX for 3 months. Changes in circulating levels of inflammatory mediators, oxidative stress markers and NETosis‐derived bioproducts were evaluated. Serum miRNomes were identified by next‐generation sequencing, and RTX‐induced changes were delineated. Mechanistic in vitro studies were performed to assess activity profiles. Altered inflammatory, oxidative and NETosis‐derived biomolecules were found in SLE and RA patients, closely interconnected and associated to specific miRNA profiles. RTX treatment reduced SLE and RA patients' disease activity, linked to a prominent alteration in those biomolecules and the reversal of altered regulating miRNAs. In vitro studies showed inhibition of NETosis and decline of pro‐inflammatory profiles of leucocytes and human umbilical vein endothelial cells (HUVECs) after B cell depletion. This study provides evidence supporting an early RTX‐induced re‐setting of the pro‐inflammatory status in SLE and RA, involving a re‐establishment of the homeostatic equilibrium in immune system and the vascular wall.


| INTRODUC TI ON
The presence of autoantibodies, and therefore the role of B cells, is a fundamental characteristic of both rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). 1,2 One of the main traits of both diseases, apart from a "vicious cycle" of inflammation, is determined by the interaction between B-and T cells. 1 In fact, in addition to antibody-dependent mechanisms, B cells act as antigen-presenting cells and co-stimulate T lymphocytes and other immune cells, including monocytes and neutrophils. This ultimately results in the increase of oxidative stress, neutrophils extracellular traps activation and release (NETosis), and the delivery of cytokines and cell-surface ligands, which in turn provides "help" to other B cells. [3][4][5] Recent studies have investigated the epigenetic processes implicated in inflammation, including an altered expression of microRNAs (miRNAs) that might play a key role in regulating inflammation. [6][7][8] Based on the evidence supporting the crucial role of B cells in the pathophysiology of these two autoimmune diseases, 1 While the role of RTX in controlling the clinical manifestations and disease progression of RA and SLE has been widely documented 9 to date, the exact mechanisms underlying the complex interplay between B cell depletion and pro-inflammatory status leading to clinical response need further elucidation.
In this study, we aimed to explore changes in serologic parameters related to inflammation, oxidative stress, NETosis and regulating miRNAs, following B lymphocyte depletion in RA and SLE patients, along with their effects on the activity of other immune cells and the vascular endothelium.

| Patients
Seventy-five RA patients, eighty-five SLE patients and ninety healthy donors (HD) were included in the study (during a 24-month period).
Additionally, two independent cohorts of sixteen consecutive RA patients and sixteen SLE patients treated with RTX were studied.
Both cohorts attended the Reina Sofia University Hospital of Cordoba (Spain) and the St. Giovanni Bosco Hospital of Turin (Italy). All patients fulfilled the American College of Rheumatology revised criteria for SLE 6 or RA. 7 Approval from the ethics committees was obtained, and subjects provided written informed consent. Clinical/laboratory parameters of patients are displayed in Table S1. Clinical/laboratory profiles of the patients treated with RTX are displayed in Table 1.
Patients were treated with RTX as rescue therapy because they were intolerant/refractory to standard immunosuppressants. The therapy with RTX was based on treating physicians' judgement and agreed case-by-case by multidisciplinary evaluation. RTX regimens included two infusions of RTX 1 g 2 weeks apart (in RA patients), or 375 mg/m 2 weekly for 4 weeks (in SLE patients).

| Blood collection and assessment of inflammatory, oxidative stress and NETosis parameters
Plasma and serum samples and purified leucocytes were obtained as previously reported. 10,11 Inflammatory mediators, oxidative stress parameters and NETosis were analysed as previously described. 8,12 For further details, see Appendix S1.

| RNA and miRNA isolation, profiling and quantitative real-time PCR
Total RNA from lymphocytes, monocytes, neutrophils and human umbilical vein endothelial cells (HUVECs) was extracted using TRI Reagent (Sigma). Inflammatory genes were evaluated as described elsewhere. 10 List of primers used is displayed in Table S2. Total RNA, including the miRNA fraction, was extracted from serum by using the QIAzol miRNeasy kit (Qiagen), as previously reported. 13 The protocols and primers used for both, miRNA profiling and RT-PCR, are described in the Appendix S1.

| In vitro study with RTX on lymphocyte population
Dose-and time-response experiments were carried out on lymphocytes purified from RA patients. Two doses of RTX (1 and 10 μg/mL) at 2× (24 and 48 hours) were evaluated in parallel, based on previous studies. [14][15][16] Lymphocytes purified from eight SLE and eight RA patients were incubated in vitro with 1 µg/mL of RTX for 24 hour. SLE and RA pro-inflammatory status in SLE and RA, involving a re-establishment of the homeostatic equilibrium in immune system and the vascular wall.

K E Y W O R D S
endothelial dysfunction, inflammation, NETosis, rheumatoid arthritis, rituximab, systemic lupus erythematosus patients selected had high activity scores (mean SLEDAI-2K of eight and mean DAS28 of 4.7, respectively).
The percentages of B cells before and after RTX treatment were analysed by flow cytometry on a flow cytometer FACSCalibur after incubation with PE human anti-CD19 and FITC human anti-CD3 (Miltenyi Biotech) as previously described. 17 Changes in gene expression of selected pro-inflammatory molecules (IL-1β, IL-6, IFN-γ and TNF-α) were evaluated using RT-PCR. brain extract, 100 U/mL penicillin, 100 mg/mL streptomycin and 250 pg/L fungizone (BioWhittaker) at 37°C in a humidified 5% CO 2 atmosphere.

| In vitro treatment of
Monocytes and HUVECs were incubated for 24 hours with serum (10%) obtained from SLE and RA patients at baseline and after 3-month RTX therapy; neutrophils were treated only for 6 hours to avoid the effect of the naturally occurring apoptotic process.

| Statistical analysis
All data were expressed as mean ± SD or median

| Early clinical response to RTX therapy
Clinical and analytical assessments after 3 months of RTX therapy revealed that 81.25% of SLE patients were early responders to the treatment (mean SLEDAI-2K after treatment = 1.5). The mean levels of acute-phase reactants (CRP and ESR), complement factors C3 and C4, and anti-dsDNA levels were also significantly reversed (Table 1).
Among RA patients, 62.5% showed early positive response to RTX therapy, as demonstrated by the reduction of disease activity (mean DAS28 of 3.2). Mean levels of RF and acute-phase reactants were also significantly reduced (Table 1).  Figure 2D).

| Serum inflammatory, oxidative stress and
In RA patients, a direct relationship among high disease activity (DAS28 > 3.2) and some inflammatory parameters, as well as with cell-free DNA levels, was demonstrated ( Figure 2G). Accordingly, positivity for ACPAs was associated with plasma levels of IL-17, TNF-α and IFN-γ ( Figure 2H). were also confirmed ( Figure 2J).

| Changes in the inflammatory parameters, oxidative stress and NETosis in the serum of SLE and RA patients after RTX therapy
After 3 months of RTX treatment, the levels of pro-inflammatory molecules were significantly reduced ( Figure 3A and 3). This is in line with the observed reduction of disease activities in both SLE and RA patients ( Table 1). 85% of RA patients and 90% of SLE patients were positive responders to therapy. Therefore, no comparative molecular analyses among responders and non-responders could be performed.
Levels of RF, anti-dsDNA and acute-phase reactants were also reduced. RTX further reduced the serum levels of LPO ( Figure 3B and G).
NETosis-derived products showed a trend of reduction in SLE patients ( Figure 3C-D) and reached statistical significance in RA ( Figure 3H-I).

| Modulation of altered circulating miRNAs in SLE and RA patients after RTX therapy
The  Figure S1B). The in silico study allowed to select four miRNAs, including miR-146a-3p, miR-125a-5p, miR-16-5p and miR-23a-3p, as potential modulators of mRNA targets involved in the pathogenesis of RA ( Figure   S2B). Besides, the increased levels of miR-146a-3p and the reduced levels of miR-125a-5p, miR-16-5p and miR-23a-3p found in the array were validated by RT-PCR in the whole cohort of RA patients. After RTX therapy, the levels of those miRNAs were significantly reversed ( Figure 3J).

| Correlation and association studies
In the cohorts of SLE and RA patients treated with RTX, association studies showed that the basal levels of deregulated circulating parameters related to inflammation, oxidative stress and NETosis were interrelated and linked to a number of clinical features ( Figure S3).
Circulating levels of microRNAs validated in SLE patients were found linked to levels of inflammatory parameters, as well as to clinical features, including high activity of the disease or hypocomplementemia ( Figure S4A-B).
In RA patients, similar profiles of correlation were demonstrated, involving a direct relationship between levels of the circulating miR-NAs validated, inflammatory molecules and NETosis-derived products, as well as with clinical and autoimmune parameters such as activity of the disease and positivity for RF and ACPAs ( Figure S4C-D).  Table S4).

| In vitro changes in the activation state of endothelial cells and monocytes induced by serum from SLE and RA patients treated with RTX
The in vitro treatment of monocytes with SLE and RA serum before RTX treatment promoted increased expression of pro-thrombotic and pro-inflammatory molecules (TF, VEGF, MCP-1, IL-8, IL-1ß). Besides, the treatment with the serum of those patients after RTX therapy abridged this activation status on healthy monocytes ( Figure S6A-B).
Similarly, the in vitro treatment of HUVECs with serum from SLE and RA patients before RTX therapy promoted the increased expression of inflammatory mediators such as VEGF, IL-8, ICAM-1 and endothelial nitric oxide synthase. The incubation with SLE and RA serum after RTX therapy significantly reduced the expression of the markers with HUVECs' activation ( Figure S6C-D). In parallel, the serum from RA and SLE patients before RTX therapy promoted a pro-inflammatory status in neutrophils, demonstrated by the increased expression of TF, IL-8 and MCP-1. The incubation with SLE and RA serum after RTX therapy mitigated this activation process ( Figure 5E and J).

| D ISCUSS I ON
In the present study, we demonstrated that RTX induced an early re-setting of the immune and vascular system status involving a  This effect, probably related to the re-establishment of a homeostatic equilibrium in the immune system, and consequently on the vascular wall, might further contribute to prevent CVD in these autoimmune disorders.
A set of serum miRNAs, acting as regulators of potential key targets involved in the physiopathology of both disorders, were significantly reversed after RTX therapy. Moreover, that reversion paralleled the improvement of disease activity, as well as the changes occurred in the immune and inflammatory profiles, as well as the down-regulation in the expression of NETs extrusion products.
The miRNAs validated by RT-PCR in the whole cohort of RA patients have been previously reported to act as relevant regulators of immune cell development, playing a crucial role in the inflammatory response, and acting as key players in the pathogenesis of chronic and autoimmune disorders, including RA. 33,34 Consistent with our results, a previous study by our team found that after a 6-month anti-TNF-α/DMARDs combination therapy, similar changes in three of the miRNAs found in the present RA cohort significantly increased in response to RTX (miR16-5p, mi-R125a-5p, and miR23a-3p). 13 The up-regulation of those miRNAs was also associated with the improvement in the disease activity and the reduction in the levels of acute-phase reactants after therapy.
On the other hand, this is the first study that identifies the changes promoted in the serum miRNA profile of SLE patients after treatment with RTX. The selected miRNAs were validated as up-regulated after RTX treatment included main regulators of inflammation, atherosclerosis, CV disease and nephropathy (ie miR-28-5p, miR-106-3p and miR-148b-3p), [35][36][37][38] as well as inhibitors of IL-6 production (ie miR-151a-3p). 39 Similar in the case of RA, the changes promoted by RTX therapy in that miRNAs were found parallel to the changes promoted in both disease activity and the inflammatory profile of these patients. Overall, our data suggest that differentially expressed miRNAs in the serum of SLE and RA patients before and after 3 months of RTX therapy have potential to serve as novel biomarkers for monitoring therapy responsiveness. Intriguingly, RTX was able to induce an inflammatory status re-assessment after only 3 months since induction therapy. One might speculate that changes in miRNAs profile might represent an early biomarker of clinical response when other conventional parameters lack adequate accuracy. In the future, these observations might guide strategic therapeutic options, especially when assessing treatment failure and/or poor response in order to identify tailored strategies.
In our series of patients, clinical effectiveness of RTX was demonstrated in a real-life setting, with rapid improvement in SLE and RA signs and symptoms, only after 12 weeks of therapy.
Although it is well recognized that complete remission is reached only after 6-12 months of treatment, and therefore it is not possible to conclude the precise timescale of improvement in the serological indices measured, this study allowed us to demonstrate, as previously reported, 40,41 that a significant reduction of disease activity in both autoimmune conditions is already visible only after 3 months of treatment. This study also allowed us to recognize a number of cellular and molecular mechanisms underlying this early response.
Because of the heterogeneity of clinical manifestations, add-on therapies and applied protocols, we could not investigate the impact of different RTX regimens on immune and vascular phenotypes. In future studies, larger study populations and longer-term time-points may identify additional important patient-centred outcome and molecular targets.
Taken together, our data demonstrated that: • SLE and RA patients display altered inflammatory, netotic and epigenetic profiles, which are interconnected and further related to the specific autoimmune profile and the activity of each disease.
• B lymphocytes play key roles in shaping pathological immune responses of SLE and RA.
• Associated with clinical response, RTX induced early restoration of homeostasis in immune and vascular systems.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.