SM03, a non-depleting anti-CD22 antibody, modulates B cell activation and immune crosstalk to attenuate autoimmunity Authors: Kin Lok Wong, Lung Hei Cheng, Cheuk Ling Chung, Pui Lam Ng, Shaik Abdullah Nawabjan, Muthu Iswarya Gandhi Sethuraman, Ka Kiu Luk, Wing Tung Tsoi, Billy Kwok Chong Chow, Shui On Leung Published: 2026 ABSTRACT AbstractObjectiveTo evaluate the disease-modifying potential of SM03, a novel humanized anti-CD22 monoclonal antibody, for B cell-mediated autoimmune diseases by investigating its mechanism for suppressing B cell dysregulation in autoimmune milieu.MethodsSM03's mechanism was assessed in-vitro using functional assays on stimulated human PBMC from healthy donors and patients with Systemic Lupus Erythematosus (SLE)/Sjögren's Syndrome (SS). The efficacy of SM03 to attenuate autoimmunity was then evaluated in-vivo in a humanized pristane-induced SLE mouse model as well as a preventive collagen-induced Rheumatoid Arthritis (RA) model in cynomolgus monkeys. Disease-specific biomarkers, histopathology, and immune cell phenotypes were analyzed.ResultsSM03 attenuated T cell–dependent B cell activation by reducing class-switched B cells, plasmablast differentiation, and pro-inflammatory cytokine production in B cell lines, healthy and disease PBMCs, without inducing B cell depletion. In the SLE model, SM03 suppressed key disease manifestations (splenomegaly, anti-dsDNA, proteinuria, glomerular deposits) and reduced activated T cells without broad B cell depletion. In the RA model, SM03 dose-dependently suppressed joint scores, cartilage damage, synovial hyperplasia, anti-collagen II antibodies, and IL-6.ConclusionBy enhancing CD22's inhibitory signalling to disrupt autoreactive B-T cell interactions, SM03 functions as a disease-modifying therapy that attenuates dysregulation of lymphocytes in autoimmunity. This non-depleting mechanism supports its translational potential for SLE and RA and implicates its broader utility for other B cell-driven autoimmune diseases.Keywords: Rheumatoid arthritis, Systemic lupus erythematosus, CD22, Antibody treatment, Humanized murine model, Non-human primate collagen induced arthritis model, B cell therapy Introduction Autoimmune diseases such as systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) are driven by dysregulated interactions between B cells and other immune cells, with autoreactive B cells serving as pivotal mediators of pathogenesis [1,2]. B cells contribute to disease progression through autoantibody production, antigen presentation, and cytokine-mediated recruitment of inflammatory cells, including T cells, monocytes, and neutrophils [[3], [4], [5]]. Multiple tolerance checkpoints normally limit pathogenic self-reactive B cells population, yet these controls can be perturbed in systemic autoimmune diseases such as SLE, RA, and Sjögren's syndrome. Impaired tolerance and altered B cell activation states have been linked to enrichment of autoreactive clones, although the relative contribution of specific checkpoints and diversification pathways likely varies by disease context [1]. CD22 is an inhibitory co-receptor that helps tune BCR signalling and has been implicated in maintaining B cell tolerance in multiple experimental systems [[6], [7], [8]]. CD22 is a sialic acid-binding immunoglobulin-like lectin (Siglec) that regulates BCR signalling upon interaction with α2,6 sialic acid glycans (CD22L) expressed in trans on adjacent cells, interaction with CD22-sialic acid ligand induces recruitment of tyrosine phosphatase SHP-1 through immunoreceptor tyrosine-based inhibitory motifs (ITIMs), suppressing calcium flux, this leads to regulation of B cell hyperactivation via direct modulation of BCR signalling [6,8]. Furthermore, studies using CD22 knockout mouse models have shown that CD22's regulatory function is essential for preventing spontaneous development of severe lupus-like autoimmune phenotypes [9]. Crosstalk between autoreactive B cells and T cells is a key driver of autoimmune disease. Through antigen presentation and co-stimulatory signalling, autoreactive B cells sustain activation of pathogenic lymphocytes and amplification of inflammatory cascades [2]. Self-antigen recognition initiates these processes, promoting activation of naïve T cells and chronic immune responses. As autoimmunity progresses, self-antigens targeted by the immune system often broaden and worsen tissue injury [2,10]. In SLE, autoreactive B cells targeting nuclear antigens such as SmD or Ro/SSA expand their specificity to additional spliceosome proteins (e.g., SmB, A-RNP), exacerbating pathology [11,12]. In SS, peptides derived from Ro/SSA and La/SSB complexes enhance T cell activation, contributing to glandular dysfunction and lymphocytic infiltration. In RA, responses to citrullinated neo-antigens broaden to additional post-translationally modified proteins, driving diversification of anti-citrullinated protein antibodies (ACPA) [13]. Similarly, in bullous pemphigoid (BP) [14], autoantibody expansion correlates with disease progression and severity [1,2]. These observations highlight how B–T cell cooperation and antigen diversification sustain autoimmune pathology. The regulation of epitope spreading is closely linked to B cell inhibitory checkpoints. CD22 is a critical checkpoint that maintains tolerance for self-antigens by negatively regulating B cell receptor (BCR) signalling upon recognition of self-ligands containing the α2,6 sialic acid molecule [7]. Indeed, engagement of CD22 using high-affinity ligands (such as STALs) can modulate B cell activation towards autoantigen stimulation [7]. However, when tolerogenic constraints are circumvented, autoreactive B cells not only escape tolerance but also become central players in orchestrating broader immune dysregulation. Autoreactive B cells interact with other immune cells—such as T cells, monocytes, and macrophages—via bystander activation and cytokine signalling, this is thought to contribute to the amplification of inflammatory cascades and perpetuate autoimmune pathology [12,15]. Current therapies for systemic autoimmunity are evolving from broad cytokine blockade (e.g., TNFα, IL-6, JAK inhibitors) toward cell-specific approaches. B cell–targeted strategies such as anti-CD20 (rituximab), BAFF/APRIL blockade (belimumab), and CD19 CAR-T highlight the central role of B cells [16,17]. However, these therapies often fail to fully resolve autoreactivity and may deplete protective subsets. CD22 has emerged as a promising target: agents like epratuzumab modulate BCR signalling without depletion, while synthetic ligands (e.g., STALs) co-engage CD22 and antigen to induce antigen-specific tolerance and modulate autoreactive B cell activity [18]. Thus, targeting autoreactive B cells and their interactions within the immune milieu is an area of active investigation that may offer a novel and potentially transformative direction in immunotherapy for autoimmune diseases. SM03, an anti-CD22 monoclonal antibody, offers a novel strategy in the treatment of autoimmune diseases by exploiting CD22's immune checkpoint role. We have previously demonstrated that SM03 preserves B cells and modulates their activity by enhancing CD22's inhibitory signalling through “trans” ligand binding rather than “cis” binding to autologous ‘self’ cells [19]. This shift reinforces CD22-mediated suppression of autoreactive B cell activation within our experimental models, suggesting a mechanism for enhanced checkpoint regulation in autoreactive contexts [19], while avoiding broad B cell depletion. In this paper, we provide evidence that SM03 downregulates gene signatures associated with BCR-mediated B cell activation and appears to modulate other immune cells by upregulating immune checkpoint genes, as suggested by exploratory RNA-sequencing analysis of SM03 treatment on anti-IgM stimulated healthy volunteer PBMC. Using B cell lines, healthy volunteer PBMCs and autoimmune disease patient PBMCs, we observed that SM03 effectively reduced autoantigen-driven B cell activation and differentiation. Furthermore, SM03 attenuated hallmark features of autoimmune disease in both a humanized mouse model of SLE and a collagen-induced arthritis (CIA) model in cynomolgus monkeys. These in-vitro and in-vivo studies suggest that by uniquely modulating B cells and interrupting their crosstalk with T cells, SM03 may offer a dual immunomodulatory effect on B cell activation and immune crosstalk without B cell depletion, providing a mechanistic rationale for its translational evaluation in autoimmune disease. Materials and methods RNA-seq At least 1 × 106 B cells were isolated from anti-IgM and SM03 or IgG1 isotype control (IgG1) treated PBMC cells. RNA was purified using Qiagen RNeasy Mini Kit with genomic DNA removal, and sequencing was performed by Novogene. After libraries were constructed and quantified, they were pooled and sequenced on Illumina platforms. Data quality check was performed, followed by mapping to reference genome. Results were obtained through differential gene expression analysis, functional analysis, and structure analysis. In-vitro studies DAUDI cells were stimulated with 50 ng/mL CD40L and 0.5 μg/mL α-2,6-sialic acid ligand, treated with 20 μg/mL SM03 or IgG1 Isotype control. After 16 h, cells were harvested and stained with CD69-FITC antibodies. DAUDI and primary B cells (1 × 106 cells/mL) isolated from healthy donors with naïve B cell isolation kit, were stimulated with 50 ng/mL IL-21, 50 ng/mL CD40L, and 1 μM ODN 2006, then treated with 10 μg/mL SM03 or IgG1. After 3 days incubation, the cells were stained with CD69-FITC, HLA-DR-BV605, and HLA-ABC-APC-Fire750 antibodies. For B and T cells co-culture, healthy peripheral blood mononuclear cells (PBMC)-isolated CD4+ T cells were pre-treated with 30 μg/mL Mitomycin C for 30 min. Autologous isolated Naïve B cells were co-incubated with the T cells in the presence of 10 μg/mL anti-human CD3 Antibody pre-coated well and co-treated with IgG1 Isotype control or SM03. After a 7-day treatment, the cells were stained with fluorochrome-conjugated antibodies (Supplementary Table 1) and analyzed by flow cytometry. PBMC from healthy volunteers (HV) were stimulated with Human CD3/CD28 T Cell Activator (αCD3/CD28) then 10 μg/mL Ro60 peptide (sequence: LQEMPLTALLRNLGKMT) or 20 μg/mL DEWYs peptide (sequence: DWEYSVWLSN) and treated with 10 μg/mL SM03 or IgG1 for 6 days. Supernatant was harvested to conduct ELISA analysis on expression of IL-6 and IFN-γ, or IL-6 and CXCL13 respectively. Similarly, SS patients and SLE patients derived PBMCs (SS PBMC and SLE PBMC respectively) were stimulated with Ro60 or DWEYS respectively and treated with 20 μg/mL SM03 or IgG1. After 6 days, cells were stained using the same antibody panel and analyzed by flow cytometry. PBMCs from healthy donors were purchased from iXcells Biotechnologies. According to the supplier, all donors were screened using a health questionnaire and provided written informed consent prior to sample collection. PBMCs from patients with systemic lupus erythematosus (SLE) and Sjögren's syndrome and CD34+ cord blood derived stem cells were purchased from HumanCells Biosciences. According to the vendor's documentation, all human samples were obtained under Institutional Review Board (IRB)–approved protocols with written informed consent from donors. Samples were de-identified prior to distribution. No human subjects were recruited or directly involved in this study, and no additional ethical approval was required for the use of these commercially sourced biospecimens. SLE humanized mice model Pregnant NOD. Cg-PrkdcscidIl2rgtm1Wjl/SzJ (NSG) mice were purchased from the Laboratory Animal Unit, the University of Hong Kong (accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International). All animals were housed in the Centre for Comparative Medicine Research facility at the University of Hong Kong. All mice were bred and maintained under specific pathogen-free conditions approved by the Department of Health in Hong Kong and the Committee on the Use of Live Animals in Teaching and Research (CULATR) of the University # 24-092. A total of 40 NSG pups subjected to sub-lethal irradiation dose (102 cGy) were injected with 1 × 105 cells of human cord blood CD34+ stem/progenitor cell to generate humanized mice (hu-mice) [20]. Mice Tail vein hydrodynamic injection (HDI) with human IL3, IL4, and GMCSF plasmids was performed to sustain growth and differentiation of human cells [21] and were randomly assigned to negative control (-ve), induction and Isotype control (IgG1), Rituximab (RTX) and SM03 treatment groups. Only hu-mice with more than 10% humanization rate (Supplementary Materials) then proceeded with pristane induction. Four weeks post-induction, drugs were administered weekly at 20 mg/kg in 100ul PBS through intravenous injection for a total of five times. Hu-mice were sacrificed for subsequent analysis a week after the final dose. Evaluation of outcomes were done by flow cytometry analysis, histological scoring and colorimetric assays. In brief, cells extracted from whole blood and spleen were fixed with 4% PFA and stained with two separate sets of fluorescent conjugated antibodies to investigate B or T cell population respectively. Histological analysis includes scoring for kidney, liver, and lungs using hematoxylin-eosin staining (H&E) [[22], [23], [24]], periodic acid-Schiff (PAS) staining, immunohistochemistry staining, and immunofluorescence imaging on FFPE sections. ELISA on different cytokines in serum and BCA assay on urine were performed. Collagen induced arthritis in Macaca fascicularis preclinical model Ten male crab-eating macaque (Macaca fascicularis), aged 4 to 7 years, 4-7 kg, were acquired from Guangxi Fangcheng Gang Spring Biological Technology Development Corporation Ltd. (Guangxi, China). Laboratory Animal Use License Number: SYXK Gui2023-0006 was issued by the Guangxi Zhuang Autonomous Region Science and Technology Department. The standard of care and management of the animals followed the standard operating procedure of Hkeybio (Shanghai, China) and was in accordance with Guide for care and use of laboratory animal, 2010 and US Animal Welfare Act 99-198. The experimental design, study protocols, and experimental procedures were approved by Institutional Animal Care and Use Committee (IACUC) #230038. Monkeys were injected with equal parts of Bovine type II collagen and Complete Freund's Adjuvant at day 0 and 21. All drug treatment groups (placebo n = 3; prophylactic high n = 3/low SM03 n = 4) had their first dose administered intravenously on the induction day, followed by weekly injection, for a total of ten doses. Clinical scoring was performed bi-weekly on observation of joint appearances (Supplementary Materials). Blood samples from all animals were collected before induction (day 0) and on day 28, 56 and 70, serum was also collected at all the above days and day 42. Cytokine and autoantibody levels were measured using ELISA and Cytometric Bead Array. PBMC was extracted for immune profiling through flow cytometry. To mimic synovial fluid collected from RA patients, intra-articular lavage of the knee was collected from all animals at day 37 and 70, 1 ml of physiological saline was injected to knee joint and aspirated for subsequent procedures. At experimental endpoint, one animal per group exhibiting clinical joint scores closest to group mean was euthanized. Histopathological findings are reported descriptively based on joint-level assessments of these representative animals. Eight joints (proximal interphalangeal joints of forelimbs and hindlimbs, elbows, and knees, bilaterally) were harvested and fixed in 4% PFA. All digit joints underwent EDTA decalcification, while longitudinally bisected knee and elbow joints were divided for parallel decalcification with either EDTA or formic acid. Decalcified tissues were processed into paraffin-embedded sections for histopathological evaluation (H&E) and immunofluorescence staining targeting CD3+ T cells and CD19+ B cells following an established protocol (Supplementary Materials). Clinical scores and blood collection derived parameters including cytokine and autoantibody levels as well as PBMC immunophenotyping were collected for all animals (placebo n = 3; LD n = 4; HD n = 3), whereas histopathology and joint IF were performed on one representative animal per group (n = 1/group; 8 joints/animal). Statistical analysis Data from flow cytometry were quantified by FlowJo (version 10, BD, Ashland, OR, USA). Statistical analyses were conducted by Prism (version 10, GraphPad Software, Boston, MA, USA). Statistical analysis: One-way ANOVA with Dunnett's post-hoc test for comparisons between control or IgG1 Isotype control, p < 0.05 is statistically significant between groups. For longitudinal clinical scoring in the cynomolgus monkey collagen-induced arthritis model, repeated-measures two-way analysis of variance (two-way ANOVA of time × treatment), followed by Dunnett's multiple comparisons test (each treatment vs placebo) A p value < 0.05 was considered statistically significant. Results Sequencing results from samples of SM03 treated PBMC activated by IgM implicated T cell related modulation parameters We previously proposed that SM03 facilitates CD22-mediated ‘self-recognition’ of autologous cells, thereby attenuating autoreactive BCR signalling [6,8,9,19]. To explore downstream mechanisms, HV PBMC were stimulated with anti-IgM to mimic B cell activation and co-treated with SM03 or IgG, followed by B cell isolation. Exploratory RNA-seq analysis was then conducted to generate hypotheses for subsequent functional validation (Supplementary Fig. 1A). SM03 treatment resulted in 988 differentially expressed genes (DEGs) with a nominal p-value <0.05 in their log2 fold change when compared with IgG1 (Supplementary Fig. 1B). This included DEGs that were associated with immunomodulation, encompassing B cell functions, immune tolerance, and checkpoint regulation (Supplementary Fig. 1C). For example, VPREB3 (log2FC = −1.587, p = 0.000) and CEACAM1 (log2FC = −1.053, p = 0.009) were downregulated, reflecting reduced B cell proliferative signalling [25,26]. Whereas CD276 (log2FC = 2.014, p = 0.015) and CD200 (log2FC = 1.047, p = 0.000) were upregulated, implicating enhanced T cell checkpoint inhibition and immunosuppressive function of B cells towards other immune cells [27,28]. Gene Set Enrichment Analysis (GSEA) using Gene Ontology (GO), Kyoto Encyclopaedia of Genes and Genomes (KEGG), Reactome and Hallmark gene set collections from MSigDB revealed significant downregulation of pathways related to BCR signalling and SLE pathogenesis (FDR <0.25) (Supplementary Fig. 1D and E). Collectively, these transcriptomic findings implicated that SM03 could potentially dampen BCR-driven activation and modulate immune-crosstalk, prompting further in-depth in-vitro investigation into the mechanistic action of SM03. SM03 modulated T cell dependent B cell activation and decreased autoantigen driven B cell differentiation in diseased patient PBMC Given that CD22 negatively regulates B cell receptor signaling upon engagement with sialic acid–containing self-ligands, we hypothesized that B cells activated by autoantigens may be susceptible to CD22-mediated inhibitory modulation by SM03 [19]. To investigate whether SM03 can intervene T cell dependent activation of B cells, B cell line DAUDI was stimulated with trans-ligand 2,6-sialic acid ligand and CD40L to mimic co-stimulatory signal from T cells, where SM03 significantly modulated B cell activation marker-CD69 expression (Fig. 1A, Supplementary Fig. 2A and B). TLR-9 stimulation induces autoreactive B cell activation in SLE patient PBMC [29]. It was found that SM03 treatment could suppress TLR-9 agonist, CD40L and IL-21 induced B cell activation (CD69) as well as decrease elevated MHC class I and II expression (HLA-ABC and HLA-DR respectively) levels in DAUDI cells and HV PBMC derived B cells (Fig. 1B and C and Supplementary Fig. 2C–F). Collectively, these results point to a role for SM03 in dampening B cell responses to T cell-like signals and downregulating surface proteins required for B cell antigen presentation. (A-J) SM03 Modulates B–T Cell Interaction and Autoantigen-Specific B Cell Responses In-Vitro. (A) Flow cytometry analysis of CD69 expression on B cell line Daudi stimulated with CD40L (50 ng/mL), 2,6-sialic acid and treatment antibody (20 μg/mL). (B-C) Flow cytometry analysis of Daudi cells (B) and PBMC derived naïve B cells (C) subjected to ODN (1 μM), CD40L (50 ng/mL) & IL21 (50 ng/mL) stimulation and treatment antibody (20 μg/mL). CD69 expression measured by percentage of parental cell, HLA-ABC and HLA-DR quantified by MFI. (D) Plasma cell population from naïve B cells after co-culturing with autologous isolated CD4+ T cells, pre-treated with mitomycin C (30 μg/mL), in the presence of antibody treatment for 7 days. (E-F) Cytokine levels measured by ELISA of healthy PBMCs pre-conditioned with T cell activator for 24h, then stimulated with respective autoantigen for 6 days (E) IL-6 and IFN-γ; stimulated with DWEYS (F) CXCL13 and IL-6; stimulated with Ro60. (G) IL-6 levels in PBMCs from Sjögren's donors, stimulated with CD40L (50 ng/mL), IL-21(50 ng/mL) and BAFF (400 ng/mL) plus R848(2.5 μg/mL) or IL4(50 ng/mL), measured by ELISA. (H) Plamsablast population in Sjögren's PBMCs stimulated with Ro60 and treated with SM03 or IgG1. (I-J) SLE PBMCs stimulated with DWEYS peptide (20 μg/mL), treated with antibody. (I) Total B cells and plasmablasts were analyzed by flow cytometry, and (J) anti-DWEYS were measures by ELISA. Error bars denote SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001 MFI: mean fluorescence intensity. To directly address whether SM03 can suppress T cell driven B cell differentiation to antibody producing populations, HV PBMC isolated B cells was first co-cultured and stimulated by activated autologous T cells, followed by SM03 treatment [30]. We were unable to observe any suppression of B cell differentiation into plasmablasts (data not shown), and only marginal modulation of plasma cell (Fig. 1D, Supplementary Fig. 3A). The absence of significant modulation in B cell differentiation by SM03 in healthy donor indicates SM03 does not affect canonical T cell stimulation directed pathways. To further assess whether this effect extends beyond B cell–intrinsic responses and influences broader immune interactions, we established an in-vitro autoantigen-driven PBMC model to evaluate SM03 activity in a more physiologically relevant setting (Supplementary Fig. 3B). To mimic conditions in autoimmune disease PBMCs, HVs were pre-conditioned with T cell activation and exposed to autoantigens Ro60 (SS) [31] or DWEYS (SLE) — peptides known to cross-react with pathogenic autoantibodies in approximately 30% of patients [32] to promote PBMC activation (Supplementary Fig. 3B) [33,34]. SM03 significantly reduced IL-6 and IFN-γ in DWEYS-stimulated HV PBMCs (Fig. 1E). SM03 also attenuated Ro60-induced IL-6 and CXCL13 secretion (Fig. 1F). The concept was also validated with diseased patient samples, where HV and SS PBMCs were stimulated with CD40L, IL-21, BAFF plus R848 or IL-4 (Fig. 1G and Supplementary Figure 3C) [[35], [36], [37], [38], [39]]. In another experiment, using combination of CD40L, IL-21, BAFF, and either R848 (to mimic TLR7/9 innate activation) or IL-4 to induce an autoreactive flare response in PBMCs, SM03 significantly reduced IL-6 levels in the R848 combination but not in those treated with IL-4 combination, for HVs (Supplementary Fig. 3C). Interestingly, stimulated SS PBMCs had higher secreted IL-6 levels than HVs, which SM03 significantly attenuated (Fig. 1G), supporting that SM03 exerts modulatory effects under autoantigen driven stimulation conditions. To further assess SM03's effect on autoantigen-specific responses, patient derived SS and SLE PBMCs were stimulated with Ro60 or DWEYS respectively (Supplementary Fig. 4A). Autoantigen alone increased plasmablasts in both diseases (SS: Fig. 1H and Supplementary Fig. 4B and 4D; SLE: Fig. 1I (Right) and Supplementary Fig. 4B and 4C). SM03 significantly suppressed DWEYS-induced plasmablast expansion in both sample groups without reducing total B cells (Fig. 1I Left). SM03 also significantly modulated anti-DWEYS IgG secretion in disease PBMC (Fig. 1J). This indicates SM03 effectively modulates autoantigen-driven B cell responses across multiple autoimmune contexts, including SLE and SS, without broadly suppressing normal B cell populations. SM03 modulates autoimmune related phenotypes in SLE pre-clinical studies A humanized mice model of systemic lupus erythematosus (SLE) was conducted to evaluate the therapeutic potential of SM03 in SLE (Fig. 2A). This model was selected due to the specific binding of SM03 to human CD22 , which does not bind to its murine homolog, precluding the use of standard murine SLE models [40]. Rituximab (RTX) - an anti-CD20 B cell depleting antibody, was used as positive control. SM03 modulates immune manifestations in humanized pristane induced SLE model (A) Illustration of experimental timeline covering the generation of SLE hu-mice and in-vivo validation of SM03 in comparison to IgG1 and RTX. (B) Anti-dsDNA antibody serum levels (IU/mL) upon sacrifice detected by ELISA kit (C) Comparison of spleen size, including length and weight, across different treatment groups (D) Urine protein levels (mg/mL) upon sacrifice detected by BCA Protein Assay Kit (E) Human B cells isolated from mice spleen were identified by surface markers CD3− CD19+ and were quantified as percentage of human CD45+ cells. Plasmablasts were further gated with CD27hi and CD38hi expression. (F) Human helper T cells isolated from mice spleen were identified by surface markers CD3+ CD4+ CD8− and were quantified as percentage of human CD3+ cells. Activated helper T cells were further distinguished with HLA-DRhi expression. NSG mice: NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ mice SLE: systemic lupus erythematosus; RTX: Rituximab; Error bars denote SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; -ve n = 6; IgG1 n = 8; SM03 n = 6; RTX n = 7. A hallmark of SLE is the presence of anti-double-stranded DNA (anti-dsDNA) antibodies, which form immune complexes driving inflammation and tissue injury [29,34,40]. SM03 treatment reduced elevated serum levels of anti-dsDNA autoantibodies from pristane-induced hu-mice (IgG1-control mice), comparable to RTX (Fig. 2B). Additionally, SM03 treatment induced a trend towards reduction in elevated serum cytokine levels of IL-18, IFN-γ and IL6 (Supplementary Fig. 5A), which were upregulated in IgG1-control mice. Furthermore, SM03 significantly modulated IL-6 and CXCL10 mRNA expression in liver tissues (Supplementary Fig. 5B) [40]. This provides a strong indication that SM03 can exert immunomodulation of autoreactivity through suppressing excessive autoantibody and inflammatory cytokine production. Consistent with clinical and preclinical SLE phenotypes, IgG1-control mice developed splenomegaly, indicative of systemic immune activation (Fig. 2C). Germinal center formation showed a moderate increase in IgG1-control mice, albeit not statistically significant, meanwhile SM03 and RTX trended toward modulation (Supplementary Fig. 6A). Renal involvement—a major SLE complication—was evident through immune complex deposition and inflammation. In line with previous studies, IgG1-control mice developed proteinuria (Fig. 2D) [40]. Notably, SM03 significantly reduced proteinuria to basal levels (Fig. 2D). Immunophenotyping revealed that SM03, comparable to RTX, modulated pathogenic CD27+CD38+ plasmablast populations; in contrast to the marked CD19+ B cell depletion induced by RTX, SM03 maintained total B cell counts (Fig. 2E, Supplementary Fig. 5C). IgG1-control mice displayed elevated CD4+ T helper cells and CD4+HLA-DR+ activated T helper cells of splenic origin. Both SM03 and RTX significantly reduced these populations to basal levels (Fig. 2F, Supplementary Fig. 5D), demonstrating the selective modulation of antibody-producing B cell subsets by SM03 may also indirectly downregulate helper T cell activation in SLE. Histopathological analysis of kidneys in IgG1-control mice revealed severe glomerular injury—immune-complex deposition, necrosis, and Bowman's capsule swelling (Fig. 3A). SM03 treatment significantly reduced these pathological features matching that of RTX (Fig. 3A). CD45+ leukocyte infiltration in glomeruli was elevated in IgG1-controls and was reduced to near-basal levels by both SM03 and RTX (Fig. 3B). Furthermore, immunofluorescence analysis demonstrated significantly elevated anti-IgG, anti-IgM (Fig. 3C) and complement deposits (Fig. 3D) within the glomeruli in IgG1-control mice. Both SM03 and RTX effectively reduced these deposits (Fig. 3C and D), consistent with suppression of pathogenic autoantibodies and mitigation of renal inflammation. SM03 attenuates lupus nephritis and pulmonary inflammation in NSG SLE model. (A) (Left) Kidney sections from SLE were PAS stained, images are representative of 4 treatment groups. (Right) Average pathological scoring of individual glomeruli from PAS-stained kidney sections. All glomerulus from at least 6 random regions per mice renal cortex was accounted for in the data. (B) Human CD45+ immune cell infiltration was assessed by IHC and quantified as percentage of DAB-stained area. Red arrows point to representative positive cells. (C-D) Glomerular immune complex and complement deposition. IF images of (C) human IgG and IgM, (D) C3 deposition on kidney sections from all groups, quantified by measuring the MFI within glomerular tufts. (E) (Left) Pathological scoring of lungs sections stained with H&E. (Right) Representative lung sections stained with H&E, highlighting pulmonary histopathological differences across groups (F) Percentage area of human CD45+ cells, denoted by DAB staining intensity, in lung IHC sections. PAS: periodic acid-Schiff IF: immunofluorescence MFI: mean fluorescent intensity IHC: immunohistochemistry. Error bars denote SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001; -ve n = 6; IgG1 n = 8; SM03 n = 7; RTX n = 7. Given the prevalence of pulmonary complications in SLE [41], the pulmonary involvement in this pristane induced SLE mouse model was also examined. IgG1-control mice showed perivascular and interstitial immune infiltrations (Supplementary Fig. 6B), serositis, and fibrosis (Fig. 3E and F), whereas SM03 significantly reduced these inflammatory and fibrotic changes. RTX did not confer similar lung protection, suggesting a tissue-specific advantage of SM03's non-depleting mechanism (Fig. 3E and F and Supplementary Fig. 6B). Beyond renal and pulmonary manifestations, IgG1 control mice exhibited liver damage, which neither SM03 nor RTX reversed (Supplementary Fig. 6C). SM03 exerts strong immunomodulatory effects as well as amelioration of SLE-specific disease pathogenesis, matching RTX in several key efficacy measures without full B cell depletion. These findings support SM03 as a potential disease-modifying therapy that suppresses dysregulated autoreactive B cells, attenuating tissue inflammation at multiple sites in SLE. SM03 ameliorated clinical symptoms and in a collagen induced arthritis in Macaca fascicularis preclinical model SM03 had previously demonstrated efficacy in a phase II clinical study for moderately-to-severely active RA [42]. Evidence suggests a pre-clinical RA phase marked by arthralgia, immune dysregulation, and autoantibody production without overt arthritis [43]. Since SM03 lacks murine CD22 binding, and robust transgenic or humanized RA models are unavailable — unlike established non-human primate models [[44], [45], [46]], a CIA model in Macaca fascicularis (cynomolgus macaque) was used to assess in-vivo efficacy and MOA of SM03, with groups receiving placebo, high-dose SM03 at 20 mg/kg (HD), or low-dose SM03 at 10 mg/kg (LD) (Fig. 4A). Reduction in clinical scoring of collagen induced arthritis (CIA) Macaca fascicularis model by SM03. (A) Experiment setup and timeline of in vivo validation of SM03 in CIA model in Macaca fascicularis. (B) Representative clinical photographs of joint presentation at five positions, red arrows denote typical swelling in RA (C) Progressive clinical scoring of sixty-four joints of treatment groups, evaluated twice weekly (D) (Left) Mean histopathological scoring of eight digit-joints per animal (Right) Decalcified FFPE tissues were stained with H&E, histopathological data represented by i) inflammatory cell infiltration (blue arrow) ii) vascular pannus formation (green arrow), iii) cartilage damage (black arrow) and iv) bone resorption (purple arow). (E) Immune infiltration of CD3 and CD19 cells in different IF joint sections, depicted by green fluorescent dye. (B-C) Statistical analysis for longitudinal clinical scores (C) were analyzed by repeated-measures two-way ANOVA (time × treatment), followed by Dunnett's multiple comparisons test (each treatment vs placebo). Error bars denote SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001. n = 4 in SM03 low dose, n = 3 in placebo and SM03 high dose. Histopathology and joint immunofluorescence (D–E) were performed on one representative animal per group (selected as closest to the group mean clinical score); eight joints were analyzed per animal (n = 8 joints per group). SM03 demonstrated a dose-dependent inhibition of arthritic joint damage progression. Compared to placebo, HD exhibiting near-complete repression of joint pathology (Fig. 4B and C and Supplementary Fig. 7C). A single representative animal with clinical score closest to the mean from each sample group was euthanized and joint sections processed for histopathological evaluation. H&E staining revealed severe cartilage destruction, subchondral bone resorption, synovial hyperplasia, disorganized synovial architecture, fibrous proliferation, and inflammatory infiltrates in placebo-treated joints (Fig. 4D and Supplementary Fig. 7A and B). HD and partially LD treatment attenuated these pathological features, restoring synovial integrity and joint architecture (Fig. 4D and Supplementary Fig. 7A and B). B and T cell infiltration to joints were observed in placebo group, and correlated to the pathological features observed, HD and LD progressively modulated the levels of lymphocyte infiltration to the joints (Fig. 4E and Supplementary Fig. 8), suggesting a dual immunomodulatory effect on lymphocyte trafficking. To investigate SM03's MOA in reducing synovitis and bone/cartilage damage, synovial cytokines from all animals were measured via saline joint lavage flushing due to challenges in fluid extraction from small primate joints. Inflammatory chemokines and cytokines CXCL5, MCP-1 and IL-6 as well as autoantibody IgG levels were modulated by SM03 HD (Fig. 5A). Albeit encouraging results, levels of conventional RA biomarkers alkaline phosphatase (ALP), C-reactive protein (CRP) and Erythrocyte Sedimentation Rate (ESR) were marginally moderated by SM03 at HD (Fig. 5B), which indicates SM03 enacts suppression of RA symptoms through unique pathway. Immune cell profiling and serum colorimetric measurements of CIA model. All experimented animals were analyzed for Joint lavage (A) and serum (B,E) ELISA as well as peripheral immunophenotyping (D). (A) Detection of CXCL5, MCP-1, IL-6 and IgG levels in joint lavage, obtained through flushing saline into joint cavities, using ELISA and cytometric bead array.(B) Selected laboratory tests results (CRP, ALP, ESR) conducted at four different time points (C-D) Quantification of different B cell (C) T cell and macrophage (D) populations through flow cytometry, blood samples were collected five times in total. (E) Serum levels of anti-collagen IgG, anti-collagen IgM, IL-6 and MCP-1, sample collected on day 70. Results acquired with ELISA (antibody) and cytometric bead array (cytokine). Error bars denote SEM. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001 n = 4 in SM03 low dose, n = 3 in placebo and SM03 high dose. Peripheral immunophenotypic analysis showed no significant alterations by LD or HD in total B cells, plasma cells, total T cells, or cytotoxic T cells (Fig. 5C and D and Supplementary Figs. 9 and 10). However, HD significantly modulated circulating plasmablasts and CD4+ T helper cells, while SM03 treatment showed a trend toward reducing class-switched B cells, T cells and macrophages (Fig. 5C and D). Furthermore, dose-dependent reductions in serum anti-collagen II IgG and IgM levels were also observed with SM03 treatment (Fig. 5E), correlating to RA autoantibody burden and aligning with prior literature [47]. HD also induced significant reduction in serum IL-6 and a trend in lowering MCP-1 levels (Fig. 5E) [45]. These findings collectively highlight SM03's efficacy in mitigating joint pathology and modulating key immune mediators in a primate CIA model, supporting its potential as a therapeutic agent for pre-RA. Discussion In preclinical SLE and RA models, SM03, an anti-CD22 monoclonal antibody specifically binding to B cells, consistently attenuated hallmark disease manifestations, including autoantibody production, tissue inflammation, and organ pathology. SM03 achieved efficacy in both SLE and RA without extensive unspecific B cell depletion and showed a better histopathological profile compared to RTX under the conditions tested. Although CD22 engagement may, in some contexts, contribute to inhibitory or apoptotic signalling in B cells, we did not observe evidence of broad B cell depletion following SM03 treatment in our in vitro or in vivo models; however, apoptosis in specific disease contributing B cell subsets was not directly assessed in the present study. These findings highlight SM03 as a non-depleting B cell modulation therapy, supporting the possibility that modulation of B cell activation and immune crosstalk contributes to reduced tissue pathology in these models. In-vitro studies using B cell lines, healthy donor PBMCs, and PBMCs from SLE or SS patients revealed that SM03 exerts broad immunomodulatory effects extending beyond direct B cell modulation. By reducing B cell–T cell crosstalk, SM03 may influence pathways linked to T cell activation and myeloid differentiation or activation. Exploratory transcriptomic profiling of anti-IgM–stimulated human PBMCs suggested that SM03 not only downregulates B cell activation signatures, including BCR signalling pathways, but could also induce upregulation of checkpoint regulators such as CD276 and CD200 (Supplementary Fig. 1A). These changes implicate that SM03 may indirectly modulate T cell responses to reshape the autoreactive immune milieu through its immunomodulatory properties on B cells [27,28]. Consistent with this, SM03 decreased activation marker on CD40L and TLR agonists stimulated DAUDI and naïve B cells (Fig. 1A–C), implicating SM03 could modulate T cell dependent CD40L-CD40 crosslinking induced B cell activation [37]. SM03 also suppressed MHC-I and MHC-II upregulation in B cells suggesting SM03 could reduce dysregulated antigen presentation, thus potentially dampen B and T cell crosstalk, and subsequent T cell activation (Fig. 1B and C). Beyond B cells, SM03 has previously been demonstrated to reduce IL-23 and IL-12 secretion from PBMC that were activated with TLR-4 agonist and IFN-γ through downregulating TLR-4 expression on monocytes [48], suggesting additional suppression in Th-1 and Th-17 activation [48]. Furthermore, in PBMC co-culture assays with autoantigens (e.g., Ro60, DWEYS peptide), SM03 reduced plasmablast differentiation, autoantibody production, and cytokine secretion (IL-6, IFN-γ) in SLE and SS patient PBMCs without affecting total B cells (Fig. 1E, F, H-J) [2,29]. Notably, SS PBMCs produced nearly twice as much IL-6 as healthy PBMCs under the same stimulation, and SM03 exerted a proportionally stronger modulatory effect (Fig. 1G and Supplementary Fig. 3C). Collectively, these findings highlight the therapeutic mechanism of action for SM03: acting through multiple immune pathways—reforming B cell phenotype, reducing antigen presentation, and dampening pro-inflammatory cytokine networks—thereby influencing T cell and macrophage activity in the autoreactive environment. In a humanized pristane-induced SLE mouse model engrafted with human CD34+ hematopoietic stem cells to enable human-specific CD22 targeting, SM03 effectively attenuated key disease hallmarks without broad B cell depletion (Fig. 2, Fig. 3) [19,40,42]. Treatment reduced splenomegaly, anti-dsDNA autoantibody levels, and glomerular immune deposits and damage, while modulating plasmablasts and activated T cells (Fig. 2, Fig. 3). Importantly, kidney pathology was comprehensively assessed beyond structural damage. IgG1-control mice exhibited severe glomerular injury, including immune complex deposition, necrosis, and Bowman's capsule swelling, accompanied by marked CD45+ leukocyte infiltration (Fig. 3A and B). SM03 significantly mitigated these pathological features, reducing immune infiltration to near basal levels to a degree comparable to RTX. Immunofluorescence analysis further revealed that SM03 markedly decreased pathogenic IgG, IgM, and complement deposits within glomeruli (Fig. 3C and D), indicating suppression of autoantibody-driven renal inflammation. Consistent with these histological improvements, SM03 normalized proteinuria to basal levels, underscoring its functional benefit in preserving renal structural integrity. Beyond nephritis, pulmonary involvement—a frequent complication in SLE—was also evaluated. IgG1-control mice developed extensive perivascular and interstitial immune infiltrates, serositis, and fibrosis (Fig. 3E and F), reflecting severe lung pathology. Interestingly, SM03 and RTX exhibited divergent effects on lung histopathology under the conditions tested. SM03 treatment yielded a more pronounced reduction in these inflammatory and fibrotic features, while RTX demonstrated a more modest impact on these parameters. These findings suggest that non-depleting CD22 modulation may differentially impact pulmonary versus renal manifestations in this humanized SLE model and warrant further investigation of tissue-specific mechanisms. Together with the renal improvements observed, the observations from lung tissues support that SM03's ability to exert immunomodulatory effects across multiple target organs in this pre-clinical model. Collectively, these findings provide a strong rationale for the translational evaluation of SM03 in SLE at both serological and tissue levels. Consistent with improvements of histopathological features, SM03 reduced activated CD4+ T helper cells and HLA-DR+ T cells in pristane-induced SLE hu-mice (Fig. 2F), and in a preventive CIA cynomolgus monkey model, decreased circulating plasmablasts, T helper cells, and macrophages (Fig. 5C–D). These changes correlated with reduced joint lymphocyte infiltration and lower pro-inflammatory cytokine levels, underscoring SM03's dual modulation of B–T cell crosstalk and myeloid interactions in the CIA model (Fig. 4, Fig. 5). This unique mechanism positions SM03 as a more comprehensive approach to mitigating autoreactive repertoire and cascades compared to conventional B cell-depleting strategies [12,15,16,30]. In the preventive CIA model, SM03 dose-dependently suppressed joint scores, cartilage destruction, synovial hyperplasia, and B and T cell infiltration, alongside reductions in serum and synovial anti-collagen II IgG and IgM autoantibodies and serum IL-6 (Fig. 4B–E, 5A&E) [19,42,45,47]. These outcomes highlight SM03's efficacy over traditional B cell depletion therapies, driven by its unique “cis-trans” binding MOA, which enhances CD22 inhibitory signalling via SHP-1 recruitment to dampen BCR hyperactivation — contrasting with depleting therapies that risk impairing protective immunity [6,7,19]. Furthermore, these findings indicate that SM03's mechanism of action is translatable to disease-related organ and tissue damage. Collectively, the preclinical evaluation of SM03 underscores its therapeutic potential in autoimmune diseases characterized by dysregulated B cell responses, including SLE and RA. While epitope spreading is a key driver of autoimmune progression [2,10], it was not directly assessed in this study. Nonetheless, SM03 reduced plasmablast differentiation, decreased autoantibody secretion (e.g., anti-DWEYS IgG), and reduced antigen-presentation markers, these cellular processes are commonly implicated in the propagation of autoreactive responses [1,11,13]. In SLE and RA, broadening of autoreactive specificities is associated with disease severity [1,11,13]. Our observation that SM03 suppresses autoantigen-stimulated plasmablast responses without reducing total B cell numbers is consistent with a modulatory effect on autoreactive B cell activation [6,7,19]. Whether CD22 modulation can shape the breadth of autoreactive B cell specificities, potentially through downregulating repertoire diversification or antigen-specific T-cell help, remains to be verified through longitudinal antigen-specific assays and BCR repertoire analyses [6,12]. By modulating B cell activation and antigen presentation markers (e.g., reduced HLA-ABC MFI), SM03 may indirectly dampen downstream T cell activation pathways in autoreactive settings. In turn, this could reduce the amplification of inflammatory networks that drive tissue pathology in organs such as the kidney and joints [15,16,49]. Thus, rather than depleting B cells, SM03 may represent non-depleting approach to modulating B cell–T cell crosstalk while preserving overall B cell numbers. Collectively, these preclinical findings support further evaluation of non-depleting CD22 modulation as an alternative to conventional B cell depletion strategies. Looking ahead, future investigations should directly evaluate whether SM03 influences autoantibody breadth and shapes the BCR repertoire over time. This includes assessing epitope spreading in diseases where this phenomenon is central, such as bullous pemphigoid (BP), in which humoral spreading targets multiple skin antigens, driving blistering [14]. A phase III randomized, double-blind, placebo-controlled study in moderate-to-severe RA patients on methotrexate background has been completed, data analysis is ongoing and results have yet been publicly disclosed [42]. Intriguingly, in addition, anecdotal observations from two refractory BP cases treated off label with SM03 showed clinical improvement with reduced anti-BP180/230 autoantibodies and normalization of skin immune infiltrates, without reported adverse events [unpublished case reports]. While these observations are hypothesis-generating, they motivate prospective studies to determine whether CD22 modulation can dampen autoreactive B cell activity and associated inflammatory networks in BP and other autoantibody-driven diseases [2,14,16]. Informed consent statement Human PBMCs and stem cells were obtained as commercially available, de-identified research materials from iXcells Biotechnologies and HumanCells Biosciences. Informed consent was obtained by the suppliers prior to sample collection. Institutional review board statement The study used only commercially available, de-identified human biospecimens; therefore, no additional institutional review board approval was required for the authors’ institutions. Generative artificial intelligence The authors used generative artificial intelligence to improve the readability and language of the manuscript. Funding The Study was funded by SinoMab BioScience Limited CRediT authorship contribution statement Kin Lok Wong: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. Lung Hei Cheng: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. Cheuk Ling Chung: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Pui Lam Ng: Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Shaik Abdullah Nawabjan: Resources, Supervision, Writing – review & editing. Muthu Iswarya Gandhi Sethuraman: Data curation, Investigation, Methodology, Writing – review & editing. Ka Kiu Luk: Data curation, Formal analysis, Investigation, Methodology. Wing Tung Tsoi: Data curation, Formal analysis, Investigation, Methodology. Billy Kwok Chong Chow: Resources, Supervision, Writing – review & editing. Shui On Leung: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing. Declaration of competing interest The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:Wong Kin Lok reports financial support was provided by SinoMab BioScience Limited. Lung Hei Cheng reports financial support was provided by SinoMab BioScience Limited. Cheuk Ling Chung reports financial support was provided by SinoMab BioScience Limited. Pui Lam Ng reports financial support was provided by SinoMab BioScience Limited. Ka Kiu Luk reports financial support was provided by SinoMab BioScience Limited. Wing Tung Tsoi reports financial support was provided by SinoMab BioScience Limited. Shui On Leung reports financial support was provided by SinoMab BioScience Limited. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgement We thank Mr. Oscar Chow from the University of Hong Kong for ordering animals and reagents in the present study. Footnotes Supplementary data to this article can be found online at https://doi.org/10.1016/j.jtauto.2026.100380. Supplementary data The following are the Supplementary data to this article: Multimedia component 1 RNA sequencing B cell isolated from anti-IgM stimulated PBMC. RNA-seq analyses were exploratory and intended to generate hypotheses for subsequent functional validation.(A) Schematic for RNA sequencing (B) Heat map of differential expressed genes (DEGs)with significant log2 fold changes (p < 0.05) SM03 versus IgG, left three columns are SM03, right three columns are IgG1. (C) DEGs selected from those with nominal p-value <0.05 and ranked among the greatest absolute log2 fold change, SM03 versus IgG. Most down regulated genes are associated with B cell receptor signaling, migration and activation, while up-regulated genes involve in inhibitory checkpoint and immunomodulation. (D) Gene set enrichment analysis (GSEA) plots SM03 versus IgG.(i) Enrichment plot: GOBP_MATURE_B_CELL_DIFFERENTIATION (ii) Enrichment plot: GOBP_B_CELL_RECEPTOR_SIGNALING_PATHWAY (iii) REACTOME_ANTIGEN_ACTIVATES_B_CELL_RECEPTOR_BCR_LEADING_TO_GENERATION_OF_SECOND_MESSENGERS (iv) KEGG_SYSTEMIC_LUPUS_ERYTHEMATOSUS (E) Volcano plot of differentially expressed gene SM03 vs IgG. SM03 modulated B cell activation and repressed MHC expression. (A) Gating strategy for Daudi subject to 50ng/ mL CD40L stimulation. Singlets (FSC-A vs. FCS-H) were gated for CD69+ population and (B) representative CD69+ populations across treatment groups for Daudi stimulated. (C-D) Gating strategy for (C) Daudi and (D) Naïve B cell isolated B cell from healthy volunteer subjected to ODN (1 μM), CD40L (50 ng/mL) & IL21 (50 ng/mL) stimulation and treatment antibody. (E) Flow cytometry analysis of Daudi cells subjected to ODN, CD40L & IL21 stimulation and treatment antibody (20 μg/mL). (Left) Dot plot demonstrates CD69 populations. (Right) histogram showing representative staining of HLA-ABC and HLA-DR (quantified by MFI) respectively. (F) Representative flow data of Naïve B cell from healthy volunteer subjected to ODN, CD40L & IL21 stimulation, CD69+ populations across -ve, IgG1 and SM03 group. SM03 modulated T cell dependent B cell differentiation and activation. (A) (Left) Illustration of the experimental setup of primary pan-T cells isolated from healthy human donors were conditioned with Mitomycin C and followed by 7-days of co-culture with autologous naïve B cells and treatment antibody at 20 μg/mL (Right) Plasma cell population was identified and characterized by flow cytometry, as demonstrated in the representative flow cytometry plot. (B) Illustration of the experimental setup of healthy PBMCs pre-conditioned with (aCD3/CD28) T cell activator (24 h), then stimulated with autoantigen DWEYS and Ro60 respectively for 6 days. (C) IL-6 levels in PBMCs from healthy volunteer stimulated with CD40L (50 ng/mL), IL-21(50 ng/mL) and BAFF (400 ng/mL) plus R848(2.5 μg/mL) or IL4(50 ng/mL), measured by ELISA. (A) & (B) illustrated with Biorender. SM03 modulated B cell differentiation in disease patient PBMC. (A-C) SLE PBMC stimulated with 20μg/ mL DWEYS, n = 3. (A) Illustration of the experimental setup. (B) Gating strategy: Singlets (FSC-A vs. FSC-H) were further identified as B cells (CD3− and CD19+), plasmablast (CD27+ CD38+) and plasma cell (CD3− CD38+ CD138+). (C) Representative flow data of B cells (CD3− and CD19+), plasmablast (CD27+ CD38+) and plasma cell (CD3− CD38+ CD138+) population across treatment groups. (D) Plasmablast (CD3− CD19+ CD38+ CD27+) analysis in Sjögren's PBMCs stimulated with Ro60 and treated with SM03 or IgG1. (A) Illustrated with Biorender. Additional characterization of inflammatory response in NSG SLE model. (A) Whole bloods were collected at euthanasia; serum was isolated via centrifugation at 2000 rpm for 10 min. Cytokines including IL-6, IL-18 and IFN-γ were detected with sandwich ELISA kit (B) Gene expression levels for IL-6 and CXCL10. RNA was extracted with RNAiso plus from homogenized mice liver tissue, RT-qPCR was performed to obtain 2−ΔΔCt value upon normalizing to –ve group and reference gene (GAPDH). (C-D) Gating strategy for PBMC isolated from NSG SLE mice spleen through Percoll fractionation. Singlets (FSC-A vs. FSC-H) gated with human CD45+ (hCD45 vs. muCD45/ SSC-A) were further identified as (C) B cells (CD3− and CD19+), plasmablast (CD27+ CD38+), plasma cell (CD3− CD38+ CD138+), and (D) T cells (CD3+), helper T (CD3+ CD4+ CD8−), cytotoxic T (CD3+ CD4− CD8+). Error bars denote SEM. ∗p < 0.05. Different histopathological presentation and immune infiltration across multiple organs in NSG SLE model. (A) GC formation in mice spleen investigated by IHC staining, PNA was employed as a marker for identifying GC formation. Red arrows depict areas that have positively stained with DAB. (B) Representative IHC images of lung tissue, DAB stained human CD45+ immune cells suggest immune infiltration (red arrows) (C) (Left) Representative H&E-stained liver sections, IgG1 group showing extensive hepatocyte ballooning and apoptosis, indicate IgG1 group has a trend towards more liver manifestations. (Right) Scoring covering criteria including hepatocyte inflammation, ballooning, apoptosis, and bile deposition. GC: germinal center. Histopathological scoring of CIA Macaca fascicularis model. (A) Pathological H& E score indexes include 4 separate criteria including i) inflammatory cell infiltration (blue arrow), ii) vascular pannus formation (green arrow), iii) cartilage damage (black arrow) and iv) bone resorption (purple arow). These parameters were scored according to the severity of each index: 0 = normal, 1 = mild; 2 = mild; 3 = Moderate; 4 = Serious. One monkey from each treatment group was euthanized and 8 joints (1 on each left and right fingers, 1 on each left and right toes, left and right elbows, and left and right knees) were taken and stored in paraformaldehyde. All joints were decalcified with either EDTA or formic acid, before sectioning and stained for H&E. The placebo group exhibited severe arthritic destruction, including extensive joint surface erosion, cartilage loss, significant bone resorption, and high inflammatory scores. The SM03 LD group showed a variable but intermediate phenotype, with some joints displaying similarly severe damage while others, particularly the knees, demonstrated clear but partial improvement with reduced pathology scores. In stark contrast, the HD treatment was consistently highly effective, presenting joints with a smooth surface covered by healthy hyaline cartilage, minimal inflammation, and very low scores across all damage categories. (B) A table of the scoring of each joint area is shown above. (C) Clinical scoring of sixty-four joints of individual animals. Blue: high dose; Red: low dose; Black: Placebo. Immune infiltration in CIA Macaca fascicularis model. (A) Each joint was decalcified by either EDTA or formic acid and stained with DAPI (blue fluorescence) and anti-CD3 or anti-CD19, corresponding secondary antibodies with eGFP conjugate were further probed (green fluorescence), respectively. Data was analyzed using Image-Pro Plus. (B) The average fluorescence intensity of CD3 and CD19 in placebo group was the highest, and the sum of the optical densities (OD) of each joint was 119.27 and 66.21, respectively; a decrease in OD for both CD3 and CD19 infiltrate was observed with SM03, 10 mg/kg at 85.37 and 50.25, respectively. CD3 and CD19 levels in SM03, 20 mg/kg were the lowest where the sum of the optical densities of each joint was 57.87 and 39.51, respectively. Gating strategy of CIA Macaca fascicularis model. (A) Gating strategy of non-human primate B cell and sub-populations. Live cells gated (Viability dye-) were further gated as CD45+ (CD45 vs. SSC-A) cells. B cells were selected through CD3− then CD19+ population. Plasma cell (CD138+) was gated directly from CD3− population. Plasmablast (CD38 vs. CD27) and class-switch memory B cells (IgD vs. CD27) were gated from parental B cells. (B) Representative data of population from different B cell type across placebo, and SM03 low/ high dose group. Gating strategy of CIA Macaca fascicularis model for T cells and macrophage. (A) Gating strategy of non-human primate T cell and macrophage, live cells gated (Viability dye-) were further gated as CD45+ (CD45 vs. SSC-A) cells. Within the population, macrophage (CD11b vs. SSC-A) and T cells (CD3 vs. SSC-A) were identified separately. Helper and cytotoxic T cells were distinguished by CD4+ and CD8+ (CD4 vs. CD8) respectively. (B) Representative flow data of different cell populations across placebo, and SM03 low/ high dose group. Data availability The RNA-sequencing data generated in this study are publicly available in the Mendeley Data repository at https://data.mendeley.com/datasets/5gv25jswsv/1. References Associated Data Supplementary Materials Multimedia component 1 Data Availability Statement The RNA-sequencing data generated in this study are publicly available in the Mendeley Data repository at https://data.mendeley.com/datasets/5gv25jswsv/1.