CD40 interacts with CD40L and plays an essential role in immune regulation and homeostasis. Recent research findings, however, support a pathogenic role of CD40 in a number of autoimmune diseases. We previously showed that memory B cells from relapsing-remitting multiple sclerosis (RRMS) patients exhibited enhanced proliferation with CD40 stimulation compared with healthy donors. In this study, we used a multiparameter phosflow approach to analyze the phosphorylation status of NF-κB and three major MAPKs (P38, ERK, and JNK), the essential components of signaling pathways downstream of CD40 engagement in B cells from MS patients. We found that memory and naive B cells from RRMS and secondary progressive MS patients exhibited a significantly elevated level of phosphorylated NF-κB (p-P65) following CD40 stimulation compared with healthy donor controls. Combination therapy with IFN-β-1a (Avonex) and mycophenolate mofetil (Cellcept) modulated the hyperphosphorylation of P65 in B cells of RRMS patients at levels similar to healthy donor controls. Lower disease activity after the combination therapy correlated with the reduced phosphorylation of P65 following CD40 stimulation in treated patients. Additionally, glatiramer acetate treatment also significantly reduced CD40-mediated P65 phosphorylation in RRMS patients, suggesting that reducing CD40-mediated p-P65 induction may be a general mechanism by which some current therapies modulate MS disease.
A member of the TNFR superfamily, CD40 is expressed constitutively on B cells, macrophages, microglia, and other APCs. CD40 interacts with CD40L, which is displayed on T cells and acts as a costimulatory molecule for B cells. CD40 interactions are essential for normal B cell responses (i.e., survival, proliferation, and differentiation), particularly in the context of germinal center reactions (1, 2). CD40 stimulation leads to activation of canonical NF-κB (3), noncanonical NF-κB signaling (4), as well as activation of MAPKs and PI3K (5).
Because CD40–CD40L interactions are a critical component of immune cell activation, it stands to reason that CD40 signaling perturbations in B cells are a common feature of autoimmune disorders (6–9). Similarly, in the absence of CD40 signaling, B cell activation is severely impaired (10). For example, the hallmark feature of X-linked hyper-IgM syndrome is the lack of B cell activation, which is caused by a mutation in the CD40L gene (11) and has a dramatic impact on the Ab genetics and function of B cells in these patients (12).
In the context of multiple sclerosis (MS), CD40–CD40L interactions represent an important, therapeutically relevant step in the activation of immune cells that mediate damage to the CNS. CD40-expressing cells, including macrophages, microglia, and B cells, are present in CNS tissues in close proximity to CD40L-expressing cells (13). Mutations in CD40 have been associated with MS in some studies (14–16), although others have not found significant associations. In mice, prophylactic treatment with a neutralizing Ab to CD40L prevented experimental autoimmune encephalomyelitis, a mouse model of MS (13). This mAb was tested in MS patients but clinical trials were halted due to side effects thought to be unrelated to the immunopathology of the disease (17). These results have intensified the pursuit of other biological agents that would potentially interfere with CD40–CD40L interactions (18), but more focus in this area requires a better understanding of the impact CD40–CD40L interactions have on the autoimmune process in MS.
We previously reported that memory B cells from treatment-naive relapsing-remitting MS (RRMS_TN) patients exhibited enhanced proliferation when stimulated with a low dosage of CD40L compared with memory B cells from healthy donor (HD) controls (19). In fact, B cells from glatiramer acetate (GA; Copaxone)–treated MS patients no longer display hyperresponses to low-dose CD40 stimulation (20). To understand why B cells from MS patients exhibit a hyperactive response to CD40, we asked whether key signaling intermediates downstream of CD40 displayed enhanced activity. The relative low frequency of memory B cells makes this a challenging pursuit; however, by using a sensitive phosflow technique, we were able to detect phosphorylation of essential components of the canonical NF-κB pathway and the MAPK pathway downstream of CD40 engagement in RRMS patients and HD controls. We found that memory and naive B cells from RRMS and secondary progressive MS (SPMS) patients exhibited a significantly elevated level of phosphorylated NF-κB (p-P65) following CD40 stimulation compared with HD controls. We also found that both GA therapy and IFN-β-1a/Cellcept combination therapy reduce the hyperresponsiveness of B cells from RRMS patients to CD40 stimulation. These results, based on analysis of key signaling proteins involved in CD40 signal transduction, demonstrate that NF-κB signaling downstream of CD40 engagement is altered in B cells of RRMS and SPMS patients and some therapy interventions modulate this key signaling pathway.
Materials and Methods
Patients were recruited to this study according to Institutional Review Board–approved criteria. Informed consent was received from all subjects before inclusion in the study. Twelve RRMS patients (11 female and 1 male; mean [SD] age, 37.1 [8.7] y), before (defined as T0 time point or treatment-naive) and 12 mo (defined as T1 time point) after receiving either monotherapy with IFN-β-1a (Avonex) or combination therapy with IFN-β-1a and mycophenolate mofetil (Avonex and Cellcept) were recruited at the University of Texas Southwestern Medical Center and cell samples from them were evaluated in the study (21). Seven treatment-naive SPMS patients (six female and one male; mean [SD] age, 56 [9.3] y) and five treatment-naive neuromyelitis optica (NMO) patients (four female and one male; mean [SD] age, 44.6 [7.2] y) recruited to University of Texas Southwestern Medical Center were also included in this study. Additionally, cohorts of treatment-naive (n = 12, 9 female and 3 male; mean [SD] age, 41.2 [8.2] y) and GA-treated RRMS (n = 8, 7 female and 1 male; mean [SD] age, 39.8 [10.7] y) patients were recruited at the University of Colorado Denver after institutional approval and informed consent. The length of GA treatment on RRMS patients ranges from 8 to 22 mo. All the treatment-naive patients included in the study had received no treatment with any disease-modifying therapy previously and no steroids in the past 60 d.
PBMCs from RRMS (n = 12, collected at T0 time point before treatment and T1 time point after treatment) recruited at the University of Texas Southwestern Medical Center were prepared from leukapheresis pack by Ficoll density gradient centrifugation and cryopreserved on the same day of leukapheresis (22). HD leukapheresis packs (n = 9) were purchased from HemaCare (Van Nuys, CA) and processed using the same protocol at the University of Texas Southwestern Medical Center. PBMCs from treatment-naive NMO patients (n = 5) and treatment-naive SPMS patients (n = 7) and their HD control cohort (n = 9, HD) were prepared by Ficoll density separation from peripheral blood collected in ACD tubes. PBMCs from the RRMS_TN (n = 12) and GA-treated RRMS (n = 8) cohorts recruited to University of Colorado Denver were also prepared by Ficoll density separation from peripheral blood collected in ACD tubes.
CD40 stimulation and phosflow
Cryopreserved PBMCs were thawed and immediately diluted with 20 ml of RPMI complete medium (RPMI 1640 medium supplemented with 10% FBS, 1% penicillin, streptomycin, l-glutamine, and HEPES). Cells were washed with 10 ml of complete medium and were resuspended in the same medium to a final concentration of 2.5 × 106 cells/ml. One million PBMCs were placed in sterile FACS tubes and rested for 2 h at 37°C in a humidified 5% CO2 incubator. Human CD40L (Cell Signaling Technology, Danvers, MA) or human TNF-α (R&D Systems, Minneapolis, MN) in complete medium was added at indicated concentrations. In some experiments, PBMCs were stimulated with CD40L in the presence or absence of the inhibitor TCPA-1 (Cayman Chemical, Ann Arbor, MI) at indicated concentrations. Five hundred microliters of 4% PFA in PBS was added to the cells to immediately stop the reaction at indicated time points and kept at 37°C for an additional 10 min. Rested cells with medium only served as the 0 min time point or unstimulated control. Cells were cooled on ice for 1 min and centrifuged at 1500 rpm for 5 min at 4°C. Cells were permeabilized in 1 ml of ice-cold 100% methanol and incubated on ice for 30 min before transferring the tubes to −80°C for storage for up to 1 wk. To minimize the phosflow staining variation, cells stimulated at different dates were recovered from −80°C storage at the same time and were washed with FACS buffer (1% BSA in PBS) three times, then stained with directly conjugated Abs against CD3 (UCHT1) (Tonbo Biosciences, San Diego, CA), CD4 (RPA-T4) (Tonbo Biosciences), CD20 (H1), CD27 (L128), and phosphorylated epitopes (BD Biosciences, San Jose, CA, unless otherwise specified). Intracellular stains consisted of phospho-specific Abs for p-P65 (pS529, clone K10-895.12.50), p-P38 (pT180/pY182, clone 36/p38), p-ERK (pT202/pY204, clone 20A), p-IKKγ (pS376, clone N19-39), and p-JNK (pT183/pY185, clone G9; Cell Signaling Technology). To detect p-IKKα/β, an unconjugated primary Ab against its phospho-epitopes (pS176/180, clone 16A6; Cell Signaling Technology) was incubated with the cells first and further probed with a fluorescence-conjugated secondary Ab. To detect TNFR-associated factor (TRAF)2 or TRAF6 protein, an unconjugated primary Ab against TRAF2 (C-20; Santa Cruz Biotechnology, Dallas, TX) or TRAF6 (EP592Y; Abcam, Cambridge, MA) was incubated with fixed and permeabilized cells and further probed with a fluorescence-conjugated secondary Ab. Cells were washed twice with FACS buffer and 100,000–300,000 events were acquired on a BD FACSCanto (BD Biosciences). Analysis was performed using online Cytobank software (Cytobank, Mountain View, CA).
Cell stimulation and Western blot
PBMCs (1 × 108) from each human subject were recovered from liquid nitrogen and washed with RPMI complete medium. CD19+ B cells were isolated by positive selection using human CD19 microbeads (Miltenyi Biotec, San Diego, CA). The purity of isolated CD19+ B cells was >90% by flow cytometry. CD19+ B cells (2 × 106) were resuspended in RPMI complete medium to a final concentration of 2.5 × 106 cells/ml and were stimulated with 2 ng/ml human CD40L for 15 min. A tube with the same number of CD19+ B cells was left unstimulated as a control. Stimulation was terminated by washing cells with 10 ml of ice-cold PBS three times. After the last wash, cells were resuspended in 1× RIPA buffer supplemented with 1× protease inhibitor mixture and 1× phosphatase inhibitor mixture (all from Thermo Fisher Scientific, Waltham, MA). After incubation on ice for 15 min, cell lysates were spun at 14,000 rpm for 10 min and the supernatants were collected. BCA protein assay (Thermo Fisher Scientific) was performed to determine the protein concentrations in the cell lysates. Ten micrograms of protein from each sample was resolved on SDS-PAGE and then electrophoretically transferred onto nitrocellulose membranes. The membranes were first probed with a polyclonal rabbit primary Ab recognizing the phosphorylated form of P65 (pS529, clone ab47395; Abcam). Following TBST washes, the membrane was incubated with an anti-rabbit IgG-HRP secondary Ab (Santa Cruz Biotechnology) and Ab-bound p-P65 was then detected using an ECL system (Thermo Fisher Scientific). The membranes were exposed to a LI-COR Imaging System (Lincoln, NE) to acquire the images, and the band densities were analyzed using the Image Studio software (Lincoln, NE). Next, the same membrane was reprobed with a polyclonal rabbit Ab recognizing p-P38 (pT180/pY182, clone ab4822; Abcam). After p-P38 detection, the membrane was stripped and reprobed with anti–β-actin (Santa Cruz Biotechnology).
All data were analyzed using GraphPad Software (La Jolla, CA). Statistical analysis was carried out with an unpaired, two-tailed t test using the Welch correction.
CD40 stimulation activates NF-κB and MAPK pathways in human B cells
To investigate whether CD40 signaling is abnormal in B cells from MS patients, we first performed experiments that would allow us to monitor the activation of NF-κB and MAPK pathways, the major signaling cascades downstream of CD40 engagement (1). To determine the optimal timing of phosphorylation of proximal signaling components downstream of CD40 (NF-κB [P65], P38, ERK and JNK), we stimulated total PBMCs with a relatively high concentration of CD40L (1 μg/ml) and monitored the phosphorylation status at different time points in different cell populations (gating strategy shown in Fig. 1A) after stimulation (Fig. 1B, Supplemental Fig. 1A). We found elevated phosphorylation of P65 in memory and naive B cells as early as 5 min poststimulation compared with the basal level (0 min time point). The peak phosphorylation level of P65 was achieved at 15 min and sustained until 30 min. P65 phosphorylation decreased significantly 45 min after stimulation and returned to basal levels 120 min after stimulation. A similar time course was observed for phosphorylation of MAPKs in B cells, although CD40-induced JNK phosphorylation was minimal at all time points tested (Supplemental Fig. 1A).
Next, we tested a series of CD40L concentrations (0.4–1000 ng/ml) and detected the peak response at 15 min (Fig. 1C, Supplemental Fig. 1B). We observed a dose-dependent response to CD40L in B cells for all the signaling molecules and found that a concentration of 2 ng/ml CD40L is optimal to induce an intermediate response. We hypothesized that MS patients would have higher CD40 signaling status compared with HD, and the intermediate response induced with such a low/suboptimal concentration would allow us to identify differences in responses to CD40 stimulation. Interestingly, we did not observe any obvious changes in the phosphorylation status of all tested signaling molecules upon CD40 stimulation in other immune cells, including CD4 (CD4+CD3+) and CD8 T cells (CD4−CD3+) or monocytes (defined as CD4dimCD3dim and confirmed by CD14 positive staining, data not shown), in any of the conditions we tested (Supplemental Fig. 1B), suggesting that in this assay system with this dose of CD40L, signaling activation only occurs in B cells. Based on these data, we used CD40L at 2 ng/ml and time points (0, 5, 15, and 45 min) as the standard stimulation conditions for the rest of the study.
B cells from RRMS and SPMS patients exhibited significantly high levels of NF-κB phosphorylation
To investigate whether MS patients have abnormal CD40 signaling in their B cells, the basal level (0 min time point) as well as the level of CD40-induced phospho-epitopes were measured and compared between PBMCs from HD and PBMCs from RRMS patients before any treatment. Although we found no differences in the basal level of p-P65 and the three MAPK molecules (p-P38, p-ERK, p-JNK) in B cells, we observed that both memory and naive B cells from RRMS patients exhibited significantly higher phosphorylation of P65 than did those from HD at 15 min after CD40L stimulation (Fig. 2A, p = 0.002 for 15 min in memory B cells [upper panel] and p < 0.0001 for 15 min in naive B cells [lower panel]). P65 phosphorylation was similar in the RRMS patients and HD controls at the signaling recovery phase (45 min time point). We also monitored the phosphorylation levels of P38, ERK, and JNK along the time course and found no difference between RRMS patients and HD at any time point tested (Fig. 2A).
Next, we applied Western blot to validate the phosflow data. We magnetically isolated CD19+ B cells from HD and RRMS_TN patients and stimulated the cells with 2 ng/ml CD40L for 15 min (Fig. 2B, 2C). These Western blot data confirmed that CD40 stimulation triggers the phosphorylation of P65 and P38 in isolated B cells of both RRMS_TN patients and HDs. However, B cells from RRMS_TN patients showed a higher fold increase of P65 phosphorylation upon CD40 stimulation compared with HD samples (Fig. 2B, 2C). To identify whether the abnormal NF-κB signaling we observed in MS patients is specific to B cells and specific to CD40 stimulation, we tested the effect of TNF-α stimulation on phosphorylation of P65 and P38 in both B cells and T cells in MS patients. Although TNF-α stimulation induced phosphorylation of P65 and P38 in both B and T cells, no significant difference in the fold induction of p-P65 and p-P38 after stimulation was observed in either B or T cells between treatment-naive RRMS patients and HD in response to TNF-α stimulation (data not shown). The result suggested that the abnormal NF-κB signaling is likely a phenomenon specific to B cells and CD40 stimulation.
More than 80% of untreated RRMS patients will eventually transition to SPMS after 20–25 y, a more aggressive form of MS (23, 24). In a cohort of treatment-naive SPMS patients, we observed a significantly elevated phosphorylation of P65 in their B cells compared with HD controls (Fig. 3A, left two panels, p = 0.042 for 15 min in memory B cells and p = 0.028 for 15 min in naive B cells). Additionally, naive B cells (but not memory B cells) from treatment-naive SPMS patients showed a significantly elevated phosphorylation of P38 compared with HD controls (Fig. 3A, lower right panel, p = 0.041 for 15 min in naive B cells).
We further investigated whether B cells from patients with NMO, a disease that frequently presents with similar clinical features to RRMS (25), also displayed dysregulation of CD40 signaling. B cells from NMO patients had similar levels of phosphorylation of tested signaling molecules (both the basal levels and 5, 15, and 45 min after CD40 stimulation) compared with HD (Fig. 3B, data not shown) with the exception of memory B cells from treatment-naive NMO patients, which showed a higher basal level of p-P65 and p-P38 compared with HD controls (Fig. 3B, left two panels, p = 0.01 for p-P65 at 0 min and p = 0.02 for p-P38 at 0 min), No correlation was observed between clinical parameters measuring disease status (expanded disability status scale [EDSS]) and phospho-activation level of P65 in B cells from RRMS patients that were either unstimulated or stimulated with CD40 (data not shown).
CD40-induced phosphorylation of the upstream kinases within the NF-κB pathway were different in B cells from RRMS patients compared with HD B cells
Next, we focused on expression and activation status of major upstream molecules involved in canonical NF-κB signaling. We focused on TRAF2 and TRAF6, because previous studies demonstrated that TRAF2 and TRAF6 bind directly to CD40 and mediate NF-κB activation (26, 27). We found that the expression of TRAF6 in B cells was similar between RRMS_TN patients and HD controls and was not affected by CD40 stimulation (Supplemental Fig. 2A). TRAF2 expression was lower after CD40 stimulation compared with the unstimulated control condition, but the level of TRAF2 in B cells was similar between RRMS_TN patients and HD controls in the unstimulated condition (p = 0.1752 for memory B cells and p = 0.2017 for naive B cells) and CD40 stimulated condition (p = 0.6899 for memory B cells and p = 0.7729 for naive B cells) (Supplemental Fig. 2B). Surface expression of CD40 on B cells from RRMS_TN patients is similar to HDs (Supplemental Fig. 4A, 4B).
We also examined the activation of upstream IκB kinases using Abs specific to phospho-epitopes present on IKKα/β and IKKγ (28). We found that the phosphorylation status of IKKγ did not change upon CD40 stimulation in either RRMS patients or HD controls (Supplemental Fig. 2C). However, memory B cells from RRMS patients showed a significantly higher level of CD40-induced IKKα/β phosphorylation at 15 min compared with HD controls (Fig. 4A, 4B).
Next, we asked whether inhibition of IKKα/β activity could dampen the aberrant CD40 signaling in B cells from RRMS patients. To do this, we used TCPA-1, an IKKα/β-specific inhibitor that has been used in clinical trials for treating rheumatoid arthritis (29). We showed that in RRMS patients, TCPA-1 inhibited phosphorylation of P65 in a dose-dependent manner (Fig. 4C, 4D). At 500 nM, when the effect on phosphorylation of P38 is minimal, TCPA-1 inhibited phosphorylation of P65 by ∼50% in both memory and naive B cells following CD40 stimulation (Fig. 4C–F). These data further confirmed the presence of aberrant canonical NF-κB activation upon CD40 stimulation in B cells from MS patients and that inhibition of upstream kinase activity, specifically IKKα/β, could dampen the response.
Effect of immunotherapy on CD40-induced NF-κB signaling in B cells from RRMS patients
Finally, we examined the effects of disease-modifying therapies on CD40 signaling in B cells from RRMS patients. The samples we used in this study were from a clinical trial that demonstrated that the combination of FN-β-1a therapy (Avonex) and mycophenolate mofetil (Cellcept) had greater therapeutic efficacy than did FN-β-1a therapy alone in a cohort of RRMS patients (21). We found that IFN-β-1a alone has no effect on the phosphorylation status of P65 and P38 in B cells from treated RRMS patients at either the basal level or following CD40 stimulation (Supplemental Fig. 3A–D, RRMS_TN versus RRMS_A). Next, we tested the effect of combination therapy of IFN-β-1a and mycophenolate mofetil on CD40-mediated B cell signaling (21). Prior to the combination therapy, B cells from RRMS patients exhibited a significantly higher activation level of P65 (T0) in comparison with HD controls (HD versus RRMS_TN [T0]), p = 0.0206 for memory B cells and p = 0.0037 for naive B cells, Fig. 5A–D). However, after 12 mo of combination treatment with IFN-β-1a and Cellcept, B cells from RRMS patients exhibited similar activation levels of p-P65 in response to CD40 stimulation compared with HD controls (RRMS_TN [T0] versus RRMS_A+C [T1], p = 0.065, and HD versus RRMS_A+C [T1], p = 0.4234, for memory B cells and RRMS_TN [T0] versus RRMS_A+C [T1], p = 0.065, and HD versus RRMS_A+C [T1], p = 0.7430 for naive B cells, Fig. 5A–D). Nevertheless, we did observe a significant correlation between the ΔEDSS and mean fluorescence intensity (MFI) values for p-P65 in both memory and naive B cell compartments following CD40 stimulation (rs = 0.5889, p = 0.0262 for memory B cells and rs = 0.5582, p = 0.0331 for naive B cells, Fig. 5E, 5F).
GA therapy is considered the first-line treatment for RRMS patients, and so we also tested the impact of GA therapy on CD40 signaling in RRMS patients. We found a significantly lower level of P65 phosphorylation in GA-treated RRMS patients compared with RRMS_TN patients (Fig. 6A, 6C, RRMS_TN versus RRMS_GA, p = 0.0314 for memory B cells and p = 0.0387 for naive B cells). However, GA therapy had no significant effect on CD40-induced P38 phosphorylation status (Fig. 6B, 6D).
We have recently demonstrated that B cells from RRMS patients, but not HD, are hyperresponsive to CD40 stimulation as measured by proliferation (19, 20, 22, 30). Based on these data, we hypothesized that dysregulation of CD40 signaling contributes to the hyperactivity of B cells from RRMS patients. In this study, we used both phosflow and Western blot to demonstrate that the CD40-induced canonical NF-κB pathway but not MAPK activation is aberrant in B cells from RRMS and SPMS patients compared with B cells from HD controls (Figs. 2, 3A). We measured CD40 expression levels on B cells of treatment-naive RRMS and SPMS patients and found that MS patients had similar CD40 expression on their B cells compared with HD (Supplemental Fig. 4). B cells and Ab production are central to the underpinnings of NMO (25), but we did not observe dysregulation of CD40-induced NF-κB signaling in B cells from NMO patients (Fig. 3B). We also noted a significant lower basal level of p-P65 and p-P38 in memory B cells from NMO patients compared with HD (Fig. 3B). These differences in signaling responses to CD40 stimulation may suggest that the role of B cells in these two diseases are distinct, despite similarities in clinical features that often lead to misdiagnosis of one or the other (31–33).
Resting B cells from RRMS patients and HD exhibited similar endogenous expression (basal level) of phosphorylated P65, P38, ERK, and JNK (Fig. 2A, HD versus RRMS_TN, 0 min). However, following signaling induction through CD40, the hyperphosphorylation status of P65 in B cells from the RRMS patients was evident (Fig. 2A, HD versus RRMS_TN, 15 min). This suggests that enhanced signaling rather than constitutive phosphorylation of P65 is altered in B cells from RRMS patients. Indeed, we detected an altered activation status of IKKα/β, the upstream kinases of the NF-κB pathway in memory B cells from RRMS_TN patients (Fig. 4A, 4B). Genome-wide association studies and targeted genomic studies have identified 97 variants associated with MS susceptibility, including 17 genes that are either within or proximal to NF-κB signaling genes (34–36). A study published recently confirmed that two of the genetic variants were associated with increased activation of NF-κB signaling after TNF-α stimulation in MS patients (37). It is quite possible that the aberrant NF-κB signaling following CD40 activation of B cells from RRMS patients identified in the present study could be due to the genetic variation and should be investigated further. We found that TNF-α stimulation induced phosphorylation of P65 and P38 in both B and T cells, but no significant difference of fold induction of p-P65 and p-P38 after stimulation in either B or T cells was observed between RRMS_TN patients and HD (data not shown). Thus, the abnormal NF-κB signaling we observed in MS patients is likely a phenomenon specific to B cells and CD40 stimulation. Further studies are needed to test genetic contributions to this phenotype.
We did not observe a relationship between EDSS (the measure of disease severity) and levels of phosphorylation of NF-κB in B cells from RRMS patients prior to immunotherapy (data not shown). This result is not surprising because MS pathogenesis involves many different cell types, including T cells (38). Further investigation is needed to determine whether T cells from RRMS patients exhibit a similar NF-κB signaling dysregulation profile in our cohorts upon relevant stimulation. Such studies should provide better insights into the relationship between the dynamic changes of signaling in multiple cell types and disease progression. Additionally, it is important to correlate the signaling changes with the downstream output of the NF-κB pathway because a delicate balance in signaling pathways is needed for proper cell function. Previously we showed that B cells from RRMS patients overproduce IL-6 compared with B cells from HD upon CD40 stimulation in vitro for at least 3 d (19, 20). However, the stimulation duration we used in this signaling study (<1 h) was too short to allow us to observe significant changes in cytokine production. A method to simultaneously detect cell signaling dynamics and cytokine changes following CD40 stimulation is needed.
Currently, molecular markers that are predictive of responsiveness to particular therapeutic regimens in MS remain unknown. IFN-β is the first disease-modifying drug approved for MS treatment and has remained an important treatment option (39, 40). Previous studies showed significant inhibitory effects of IFN-β on T cell activation in MS patients, but the effect of the drug on B cells is less clear (41, 42). In this study, we monitored the effect of IFN-β-1a (Avonex) on aberrant NF-κB signaling in MS, either alone or in combination with mycophenolate mofetil (Cellcept). We found that IFN-β-1a alone shows no modifying effect of aberrant NF-κB phosphorylation in B cells from RRMS patients (Supplemental Fig. 3A–D), but combination therapy with Cellcept showed some inhibitory effect on NF-κB signaling in RRMS patients (Fig. 5A–D, Supplemental Fig. 3E–H). Furthermore, we observed a significant correlation between the change in EDSS and MFI values of p-P65 following CD40 stimulation in patients treated with this combination therapy (Fig. 5E, 5F). Data from this small cohort of patients suggest that evaluation of p-P65 with CD40 stimulation may be reflective of drug responsiveness, at least in the context of IFN-β-1a and Cellcept combination therapy. Additionally, we tested the impact of GA treatment on CD40 signaling in a separate cohort of RRMS patients. We found that GA-treated RRMS patients also showed a significantly lower level of CD40-activated p-P65 than did RRMS_TN patients (Fig. 6).
These data demonstrate the utility of multiparameter phosflow analysis for monitoring the activation status of memory and naive B cells using a relatively small number of PBMCs. We can envision the use of this powerful tool to analyze signaling events in other cell populations known to be important in MS pathogenesis, such as regulatory B cells, plasmablasts, and many other cell populations. As proposed in our model (Fig. 7), the NF-κB cascade downstream of CD40 engagement could be the possible underlying dysregulated mechanism contributing to the hyperproliferation by B cells from RRMS patients upon CD40 stimulation (19). Further studies are warranted to determine the impact of dysregulation of the NF-κB cascade on hyperresponses by B cells from MS patients upon CD40 stimulation. Additionally, our data from a small cohort of patients suggested that GA therapy and IFN-β-1a/Cellcept combination therapy modulate the aberrant signaling that may contribute to their therapeutic mechanisms (Fig. 7). It is possible that reducing CD40-mediated p-P65 induction may be one mechanism by which current treatments modulate MS disease. Ongoing studies are directed to further investigate the impact of current therapies to correct this dysregulated pathway of CD40 stimulation in B cells from RRMS patients.
We thank the patients who participated in this study. We also thank Dr. Sean Morrison and his team in the Moody Foundation Flow Cytometry Facility at Children’s Research Institute for use of instruments (University of Texas Southwestern Medical Center). Acquisition of samples at the University of Colorado was supported by the Rocky Mountain MS Center and coordinated by Sean Selva (University of Colorado).
This work was supported by funding from the National Multiple Sclerosis Society. M.K.R. received grant support from the National Multiple Sclerosis Society and the National Institutes of Health. B.G. received grant support from Biogen, Chugai, MedImmune, Acorda Therapeutics, the National Institutes of Health, and the Patient-Centered Outcomes Research Institute. N.L.M. received grant funding from MedImmune and Teva Neuroscience.
The online version of this article contains supplemental material.
Abbreviations used in this article:
E.A. has received consulting fees from Biogen, Genentech, Genzyme, Novartis, and Teva Neuroscience. He has received research support from Acorda, Alkermes, Biogen, Genentech, Novartis, and the Rocky Mountain MS Center. M.K.R. has received consulting fees from Biogen, Roche, Genentech, Amarantus, and Novartis outside the submitted work. B.G. has received consulting fees from MedImmune, Novartis, and EMD Serono. The other authors have no financial conflicts of interest.