In-Depth Characterization of L1CAM+ Extracellular Vesicles as Potential Biomarkers for Anti-CD20 Therapy Response in Relapsing-Remitting Multiple Sclerosis

Abstract

The effective suppression of inflammation using disease-modifying therapies is essential in the treatment of multiple sclerosis (MS). Anti-CD20 monoclonal antibodies are commonly used long-term as maintenance therapies, largely due to the lack of reliable biomarkers to guide dosing and evaluate treatment response. However, prolonged use increases the risk of infections and other immune-mediated side effects. The unique ability of brain-derived blood extracellular vesicles (EVs) to cross the blood-brain barrier and reflect the central nervous system (CNS) immune status has sparked interest in their potential as biomarkers. This study aimed to assess whether blood-derived L1CAM⁺ EVs could serve as biomarkers of treatment response to rituximab (RTX) in patients with relapsing-remitting MS (RRMS). Serum samples (n = 25) from the baseline (month 0) and after 6 months were analyzed from the RTX arm of the ongoing randomized clinical trial OVERLORD-MS (comparing anti-CD20 therapies in RRMS patients) and were compared with serum samples from healthy controls (n = 15). Baseline cerebrospinal fluid (CSF) samples from the same study cohort were also included. EVs from both serum and CSF samples were characterized, considering morphology, size, and concentration, using transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA). The immunophenotyping of EV surface receptors was performed using flow cytometry with the MACSPlex exosome kit, while label-free quantitative proteomics of EV protein cargo was conducted using a proximity extension assay (PEA). TEM confirmed the presence of EVs with the expected round morphology with a diameter of 50-150 nm. NTA showed significantly higher concentrations of L1CAM⁺ EVs (p < 0.0001) in serum total EVs and EBNA1⁺ EVs (p < 0.01) in serum L1CAM⁺ EVs at baseline (untreated) compared to in healthy controls. After six months of RTX therapy, there was a significant reduction in L1CAM⁺ EV concentration (p < 0.0001) and the downregulation of TNFRSF13B (p = 0.0004; FC = -0.49) in serum total EVs. Additionally, non-significant changes were observed in CD79B and CCL2 levels in serum L1CAM⁺ EVs at baseline compared to in controls and after six months of RTX therapy. In conclusion, L1CAM⁺ EVs in serum showed distinct immunological profiles before and after rituximab treatment, underscoring their potential as dynamic biomarkers for individualized anti-CD20 therapy in MS.

Keywords: anti-CD20 therapy; brain-derived blood exosomes; characterization of L1CAM+ EVs; multiple sclerosis (MS); rituximab; treatment-response biomarkers.

Figure 1

Schematic overview of the study design, analysis set-up, and main findings: Significantly higher concentrations of L1CAM⁺ EVs (p < 0.0001) in serum total EVs and EBNA1⁺ EVs (p < 0.01) in serum L1CAM⁺ EVs at baseline (untreated) compared to in healthy controls were observed. After six months of RTX therapy, there was a significant reduction in L1CAM⁺ EV concentration (p < 0.0001) and the downregulation of TNFRSF13B (p = 0.0004; FC = −0.49) in serum total EVs.

Figure 2

Characteristics of serum and CSF EVs: (a) Morphology detected with TEM imaging (scale bar: 200 nm) and (b) tetraspanin (CD81, CD9, and CD63) (bead population numbers: 65, 53, and 56, respectively—flow cytometry analysis) expression profile of MS patients (month 0—m0 and month 6—m6) and healthy controls, with total EVs and L1CAM⁺ EVs expressed in mean fluorescence intensities (MFIs) as floating bar plots with min. and max. ranges. NTA: (c). Bubble plots show the relationship between mean EV particle size and the concentration of total EVs and L1CAM⁺ EVs. Each point represents an individual sample, with bubble size proportional to EV concentration. (d). L1CAM and EBNA concentration in serum total EVs and L1CAM⁺ EVs. Significantly higher concentrations of L1CAM⁺ EVs (p < 0.0001) in serum total EVs and EBNA1⁺ EVs (p < 0.01) in serum L1CAM⁺ EVs at baseline (untreated) compared to in healthy controls were observed. After six months of RTX therapy, there was a significant reduction in L1CAM⁺ EV concentration (p < 0.0001). Statistical significance between HC and m0 was assessed using the U Mann–Whitney test. Statistical significance between m0 and m6 was determined using the Wilcoxon signed-rank test.

Figure 3

(a,b). Surface immune profiling of total and L1CAM⁺ EVs shown as heatmaps (EV markers’ geometric mean fluorescence intensity (MFI) was normalized to the mean MFI for specific EV markers (CD9, CD63, and CD81), obtaining normalized MFI values) from CSF and serum samples comparing MS patients (m0, m6) and healthy controls. Statistical significance between HC and m0 was assessed using the U Mann–Whitney test and Wilcoxon signed-rank test (m0 and m6) with no significant findings. (c). Linear regression models adjusted for sex and age were used to estimate differences over time (MS = m0 and m6) and between sample groups (HC and MS) for the 21-flex panel proteins. Differentially expressed proteins with significant changes in both serum total and L1CAM⁺ EVs are presented. After six months of RTX therapy, there was a significant downregulation of TNFRSF13B (p = 0.0004; FC = −0.49) in serum total EVs. The models were adjusted for the covariates ‘age’ and ‘sex’, and inferential tests were two-tailed with a nominal alpha level of 0.05. Raw p-values were adjusted for multiple testing by controlling the false discovery rate with the Benjamini and Hochberg method, and the critical value (q-value) was set to ≤0.01.