BTK inhibition limits B-cell-T-cell interaction through modulation of B-cell metabolism: implications for multiple sclerosis therapy

Abstract

Inhibition of Bruton's Tyrosine Kinase (BTKi) is now viewed as a promising next-generation B-cell-targeting therapy for autoimmune diseases including multiple sclerosis (MS). Surprisingly little is known; however, about how BTKi influences MS disease-implicated functions of B cells. Here, we demonstrate that in addition to its expected impact on B-cell activation, BTKi attenuates B-cell:T-cell interactions via a novel mechanism involving modulation of B-cell metabolic pathways which, in turn, mediates an anti-inflammatory modulation of the B cells. In vitro, BTKi, as well as direct inhibition of B-cell mitochondrial respiration (but not glycolysis), limit the B-cell capacity to serve as APC to T cells. The role of metabolism in the regulation of human B-cell responses is confirmed when examining B cells of rare patients with mitochondrial respiratory chain mutations. We further demonstrate that both BTKi and metabolic modulation ex vivo can abrogate the aberrant activation and costimulatory molecule expression of B cells of untreated MS patients. Finally, as proof-of-principle in a Phase 1 study of healthy volunteers, we confirm that in vivo BTKi treatment reduces circulating B-cell mitochondrial respiration, diminishes their activation-induced expression of costimulatory molecules, and mediates an anti-inflammatory shift in the B-cell responses which is associated with an attenuation of T-cell pro-inflammatory responses. These data collectively elucidate a novel non-depleting mechanism by which BTKi mediates its effects on disease-implicated B-cell responses and reveals that modulating B-cell metabolism may be a viable therapeutic approach to target pro-inflammatory B cells.

Keywords: Autoimmune diseases; B cells; Bruton’s tyrosine kinase (BTK); Co-stimulatory molecules; Cytokines; Immunometabolism.

Fig. 1

BTK inhibition decreases B-cell activation, co-stimulatory molecule expression and proliferation. Peripheral CD19⁺ B cells were isolated from venous blood of healthy donors (purity confirmation by flow cytometry routinely > 98%) and pre-treated with BTKi (1uM) or vehicle for 1 h, then left without stimulation (Nil) or stimulated under various conditions including activation through the BCR alone using BCR cross-linking antibody (αBCR), combined activation through the BCR together with CD40-ligand (CD40L) stimulation and IL-4 (40X4), or activation with CpG alone. CD69, CD80 and CD86 surface expression was measured on day 2 using flow cytometry and CFSE dilution was used to measure B-cell proliferation on day 5. BTKi treatment strongly decreases activation induced CD69 expression by B cells across the different stimulation conditions (n > 8) (a–c). CD80 expression is decreased with BTKi treatment across stimulation conditions (d and e), while CD86 expression is only reduced by BTKi under αBCR and CD40L + αBCR + IL-4 stimulation conditions (f and g, n = 12/13). h and i BTKi decreases B-cell proliferation, in particular upon CpG stimulation (n = 3/5). Repeated measure Two-way ANOVA, each line represents paired results from individual donors. ns not significant, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001

Fig. 2

Inhibition of BTK diminishes allogenic and antigen-specific B-cell:T-cell interactions. BTKi or vehicle pre-treated human B cells were co-cultured with cell trace violet (CTV)-labelled allogenic CD4⁺ T cells for 7 days. BTKi treatment substantially diminishes allogenic T-cell responses (n = 5/6) (a–d). BTKi or vehicle pre-treated CFSE labeled human B cells were co-cultured with Cell trace violet (CTV) labelled CD4⁺ T cells in presence of Tetanus Toxoid (TT) for 12 days. Proliferation of B cells and CD4⁺ T cells was assessed as dilution of CFSE or CTV, respectively. T-cell cytokines (IFNγ and TNFα) were measured by FACS intracellular staining. BTK inhibition decreases both B-cell (e and f) and T-cell (g and h) proliferative responses to tetanus toxoid (n = 7) in the same co-culture system. In addition, BTKi decreases both IFNγ (i and j) and TNFα (k and l) expression by T cells (n = 6). a–d Repeated measure one way ANOVA; e–l paired t test. Repeated measure Two-way ANOVA, each line represents paired results from individual donors. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001

Fig. 3

Transcriptomic changes induced by BTKi across different modes of B-cell activation. Purified healthy donor human B cells, pre-treated with either vehicle or BTKi for 30 min, were stimulated with either CD40L + αBCR + IL-4 or CpG for 18 h followed by bulk RNA-sequencing to test the impact of BTKi on activation-induced B-cell transcriptomic changes. Heatmaps together with hierarchical clustering (one minus spearman rank correlation) and principal component analyses were used to visualize the differentially expressed genes (a–d) with adjusted p value (Adj p) < 0.05; Fold change (FC) > 2 across different conditions (n = 3). Gene-set enrichment analysis (GSEA) against KEGG- and Hallmark-gene sets for those genes that were similarly impacted by BTKi across the stimulation conditions (e). As indicated by the negative NES (normalized enrichment score) values, 25 shared gene sets are inhibited in the BTKi condition relative to the vehicle control, among which nearly 50% (12/25) are implicated in either metabolic pathways (9/25, red arrows) or important signaling pathways that control metabolic processes (3/25, blue arrows)

Fig. 4

BTK inhibition down-modulates B-cell mitochondrial respiration and mitochondrial respiratory-chain mutations result in diminished B-cell activation and APC potential. Purified human peripheral B cells, pre-treated with either vehicle or BTKi for 30 min, were then stimulated with either CD40L + αBCR + IL-4 or CpG for 36 h. Seahorse mitochondrial stress testing was used detect mitochondrial-related metabolic changes of B cells. As illustrated in a representative oxygen consumption rate (OCR) curve of the seahorse mitochondrial stress assay (a), inhibition of BTK decreases both basal respiration (b) and maximal respiration (c) of B cells (n = 3). To assess the impact of mitochondrial respiratory chain mutations on human B cells, flow cytometry was applied to B cells purified ex vivo from patients with known mitochondrial complex I (ND3 or ND6) or V (USMG5) respiratory chain mutations (n = 11; see Table S1) and healthy controls (n = 48). Compared to healthy controls (HC), B cells of patients with mitochondrial respiratory chain mutations (Mito Dis) exhibited decreased activation profiles assessed as reduced frequencies of CD71⁺ and GITR⁺ B cells (d), and reduced expression of the co-stimulatory molecules CD80 and CD86 (e and f). b and c Repeated measure two-way ANOVA; Each line represents paired results from individual donors. d–f Non-parametric t test. *p < 0.05, **p < 0.01, and ****p < 0.0001

Fig. 5

In vitro inhibition of mitochondrial respiration recapitulates the effects of BTKi on normal human B-cell activation and both inhibition of mitochondrial respiration and BTKi abrogate aberrant activation and costimulatory molecule expression of MS patient B cells. Purified healthy donor B cells were pre-treated with either vehicle, Rotenone (to inhibit mitochondrial respiration), 2-DG (to inhibit glycolysis), or the combination of Rotenone and 2-DG for 30 min, and then left unstimulated or stimulated with CD40L + αBCR + IL-4 for 2 days to assess expression of the activation marker CD69 and co-stimulatory molecules CD80, CD86; or for 5 days to assess proliferation by CFSE dilution, using flow cytometry. In the absence of stimulation (data not shown), average mean fluorescence intensities (MFIs) of B cell CD69, CD80 and CD86 expression in the ‘vehicle’ condition were approximately 1000, 75 and 125, respectively. Following stimulation which induced the expected proliferation and upregulation of activation and costimulatory molecules, inhibition of glycolysis was found to modestly limit B-cell proliferation (b) while having limited or no appreciable effect on B-cell activation (a) or activation-induced upregulation of costimulatory molecules (c and d, n = 10). In contrast, inhibiting mitochondrial respiration significantly decreased B-cell activation and proliferation, and limited the activation-induced upregulation of costimulatory molecules (a–d). To assess the effects of inhibiting mitochondrial respiration or BTK signaling on B-cell costimulatory molecule expression of MS patients, peripheral B cells were purified from either newly diagnosed (never-treated) MS patients, or healthy controls (HC) balanced for age and sex (Table S2 for participant demographics). The B cells were then treated with vehicle, BTKi or Rotenone and stimulated with CD40L + αBCR + IL-4 for 2 days. Expression of the costimulatory molecules CD80 (e) and CD86 (f) was measured by flow cytometry. Compared to HC, activated B cells of MS patients express abnormally higher levels of CD80 and CD86, and inhibition of BTK or inhibition of mitochondrial respiration with Rotenone, decrease the activation-induced B-cell expression of CD80 and CD86 by both HC and MS patient B cells, such that levels in MS patients are similar to those observed in HC (n = 13–16). a–d Repeated measure two-way ANOVA; a–d Mix-effects two-way ANOVA; Each line represents paired results from individual donors. ns not significant, *p < 0.05, **p < 0.01 ***p < 0.001 and ****p < 0.0001

Fig. 6

BTK inhibition alters B-cell metabolism and modulates B-cell and T-cell functional phenotypes in vivo. Healthy volunteers (n = 6) in a Phase I study were treated with 300 mg oral BTKi twice a day for 7 days (a). PBMC samples were collected and cryopreserved prior to treatment (D0) or day 7 on treatment (Post). B cells were purified from the PBMC and stimulated with CD40L + αBCR + IL-4 for 36 h, and seahorse mitostress assays were used to determine B-cell mitochondrial respiration. As illustrated in a representative oxygen consumption rate (OCR) curve (b), in vivo BTKi treatment significantly decreases both basal (c) and maximal (d) B-cell mitochondrial respiration. Multi-color flow cytometry panels were applied to PBMC from the same phase I trial participants to interrogate the phenotypic and functional profiles of both B cells and T cells. BTKi treatment substantially limits the frequency of proliferating Ki-67⁺ circulating B cells (e) and diminishes their activation-induced CD80 and CD86 expression (f and g). TriMAP dimensionality reduction was applied to the B-cell immunophenotyping data to visualize B-cell subset composition (h and i). In vivo BTKi treatment increases transitional and naïve B-cell frequencies while decreasing frequencies of circulating class-switched memory B cells (j), which results in an increased ratio of antigen inexperienced B-cell subsets (transitional + naïve B cells) to memory B cells (k). TriMAP dimensionality reduction applied to T-cell immunophenotyping (l and m) reveals that effector memory T cells are preferentially decreased by BTKi treatment (n), resulting in decreased ratios of effector memory (TEM) to naïve (nT) T cells (o). In contrast, the frequency of phenotypically defined regulatory T cells (Treg) increases (p), which together results in an anti-inflammatory shift of the balance of Treg and TEM T-cell subsets (q). In keeping with these phenotypic changes, in vivo BTKi treatment is found to diminish activation-induced pro-inflammatory cytokine-expressing T cells (r–t). e–l Paired t test. Each line represents paired results from individual donors. *p < 0.05