Fingolimod is an effective treatment for relapsing-remitting multiple sclerosis. It is well established that fingolimod, a modulator of the sphingosine-1-phosphate pathway, restrains the egress of CCR7+ lymphocytes from lymphatic tissues into the blood, thus resulting in reduced lymphocyte counts in peripheral blood. CXCR5+ T follicular helper (Tfh) cells provide help to B cells, are essential for the generation of potent Ab responses, and have been shown to be critically involved in the pathogenesis of several autoimmune diseases. Besides lymphoid tissue-resident Tfh cells, CXCR5+ circulating Tfh (cTfh) cells have been described in the blood, their numbers correlating with the magnitude of Tfh cells in lymphoid tissues. Although the effect of fingolimod on circulating lymphocyte subsets has been established, its effect on cTfh cells remains poorly understood. In this study, we found that although fingolimod strongly and disproportionally reduced cTfh cell frequencies, frequencies of activated cTfh cells were increased, and the composition of the cTfh cell pool was skewed toward a cTfh1 cell phenotype. The circulating T follicular regulatory cell subset and CXCR5+ CD8+ T cell frequencies were also strongly and disproportionally decreased after fingolimod treatment. In contrast, relative frequencies of CXCR5– memory Th cells as well as regulatory T and B cells were increased. In summary, these data provide new insights into fingolimod-induced compositional changes of lymphocyte populations in the blood, in particular cTfh cells, and thus contribute to a better understanding of the mechanism of action of fingolimod in multiple sclerosis patients.
Multiple sclerosis (MS) is an autoimmune disease characterized by immune responses against components of the CNS (1). Inflammatory lesions in MS are believed to be induced in a way that immune cells from the blood, such as T cells, enter the CNS and cause tissue damage. This results in neurologic damage and disability affecting more than 2 million people worldwide (2). The success of different immunotherapies targeting T and/or B cells argues that both cell types are involved in the pathogenesis of MS (1, 3, 4). For example, the success of B cell depletion therapies in MS patients highlights a role for B cells in MS pathogenesis (5–7), although the precise mechanisms that lead to the observed improvements in these patients remain under investigation (8). One immune cell type that links B cells and T cells is the T follicular helper (Tfh) cell. Tfh cells interact with B cells and are critical for efficient Ab responses. Altered numbers or dysfunctional Tfh cells have been described in various disease settings, including autoimmunity (9–12). Tfh cells as well as mature B cells highly express the chemokine receptor CXCR5, which allows localization of these cells to CXCL13-rich (chemokine ligand of CXCR5) B cell aggregates. Tfh cells in the germinal center express the master transcription factor BCL6, whereas their counterparts in the blood, so-called circulating Tfh (cTfh) cells, do not (13–16). cTfh cells can be divided into cTfh1, cTfh2, and cTfh17 cells, analogous to classical Th cell populations according to their expression of the chemokine receptors CXCR3 and CCR6 (13). In contrast to Tfh cells, which promote Ab responses, T follicular regulatory (Tfr) cells have been described as negative regulators of Ab responses (17, 18). These cells share characteristics of Tfh and regulatory T (Treg) cells (e.g., coexpression of the Tfh cell–characteristic molecules CXCR5 and BCL6 as well as the Treg cell–characteristic transcription factor FOXP3). Similar to cTfh cells, Tfr cells are also present in peripheral blood.
There are several indications that Tfh cells are critically involved in MS (11, 12): CXCR5 is a genetic risk locus for MS (19, 20), ∼20% of T cells in cerebrospinal fluid are CXCR5+ (21), and frequencies of activated cTfh cells correlate with disease progression in secondary progressive MS patients (22). Mice that lack CXCL13 develop only mild experimental autoimmune encephalomyelitis (EAE), an established mouse model of MS, and anti-CXCL13–blocking Ab treatment reduces EAE symptoms (23). Besides B cells, which express high levels of CXCR5, Tfh cells may also be affected in these settings either directly because of their high CXCR5 expression levels or indirectly through impaired cross-talk with B cells. Tfh cells play important roles in the functioning and in the organization of follicles in lymphatic tissue (10), and follicle-like aggregates have been detected in the meninges (24) and linked to cortical pathological conditions in MS (25). Finally, Tfh cell gene signatures are enriched in CD4+ T cells isolated from the cerebrospinal fluid of MS patients, and deficiency in Tfh cells resulted in reduced disease in the EAE model (26).
In this study, we found that, in contrast to natalizumab and rituximab (which block the cell adhesion molecule α4-integrin and deplete CD20-positive B cells, respectively), fingolimod profoundly reduced the frequencies of cTfh cells in the blood of MS patients. Fingolimod is a clinically approved sphingosine-1-phosphate receptor (S1PR) modulator that prevents the egress of lymphocytes from lymph nodes to the blood and has effects on multiple immune cell subsets (27–29). Fingolimod is highly effective in preventing new relapses in relapsing-remitting MS (RRMS) patients. It is well established that fingolimod retains CCR7+ immune cells in lymphatic tissue resulting in a strong decrease of CCR7+ T cells in blood (30, 31), but effects of fingolimod on Tfh cells are poorly understood. We thus set out to analyze the effects of fingolimod on cTfh and related immune cells in more detail.
Materials and Methods
Patients and ethics statement
We analyzed blood of patients with the diagnosis of MS according to the McDonald criteria (32). Inclusion criteria were treatment with natalizumab, rituximab, or fingolimod. In addition, 10 untreated MS patients (including one patient with clinically isolated syndrome) and 15 blood samples from healthy controls (HC) were included. Gender, age, type of MS, and Expanded Disability Status Scale scores of all blood donors are included in Supplemental Table I. Patients in the fingolimod group had received treatment for at least four consecutive months at the time of blood draw. Male and female blood donors were analyzed together as we did not observe major immune cell composition differences between genders (data not shown). All patient and control samples were collected following written informed consent, and the study was approved by the institutional ethics review board of LMU Munich.
Human blood cell preparation
PBMCs were isolated from human whole blood collected in EDTA-coated tubes (Sarstedt) by Biocoll (Biochrom) gradient in SepMate tubes (STEMCELL Technologies) according to standard procedures within 24 h from blood draw. Cells were further analyzed directly after isolation.
Phenotypic analyses of lymphocytes by flow cytometry
PBMCs were first incubated with FcR Blocking Reagent (Miltenyi Biotec) and then stained at 4°C with the following fluorescently labeled Abs: for the T cell panel, CD4-FITC (clone RPA-T4), CD3-PE-Cy7 (OKT3), CD19-PerCP-Cy5.5 (SJ25C1), CD8a-PerCP (RPA-T8), CD56-PerCP (HCD56), CD14-PerCp-Cy5.5 (HCD14), CXCR5-allophycocyanin (J252D4), CXCR3-BV605 (G025H7), ICOS-BV510 (C398.4A), PD-1-BV785 (EH12.2H7), CD25-PE/Dazzle 594 (M-A251), CD38-AF700 (HIT2), CCR7-BV650 (G034H7), CD45RO-BV421 (UCHL1; BioLegend), and CCR6-PE (11A9; BD Biosciences); for the B cell panel, CD19-BV605 (SJ25C1), CD27-PE (O323), CD38-AF700 (HIT2), CD138-FITC (MI15), IgD-PE/Dazzle 594 (IA6-2), IgM-PE-Cy7 (MHM-88), CD8a-PerCP-Cy5.5 (RPA-T8), CD56-PerCP (HCD56), and CD14-PerCP-Cy5.5 (HCD14; BioLegend). Dead cells were excluded with the fixable viability dye eFluor 780 (eBioscience/Thermo Fisher Scientific). PBMCs were washed and then fixed with 1% paraformaldehyde prior to acquisition. Events were recorded on a BD LSRFortessa, and data were analyzed with FlowJo 10.
Determination of absolute cell numbers
RBC lysis of 100 μl of whole blood collected in EDTA-coated tubes was performed with 1× RBC Lysis Buffer (BioLegend). Cells were incubated with FcR Blocking Reagent (Miltenyi Biotec) and then stained at 4°C with the following fluorescently labeled Abs: CD4-allophycocyanin (RPA-T4), CD8a-PerCP-Cy5.5 (RPA-T8), CD45-BV421 (2D1), CD3-BV785 (OKT3), CD45RO-AF700 (UCHL1), and CD19-BV605 (SJ25C1; BioLegend). After washing and fixation with 1% paraformaldehyde, 30 μl of 123count eBeads (eBioscience/Thermo Fisher Scientific) were added prior to acquisition.
Visualization of flow cytometry data using dimensionality reduction (t-distributed stochastic neighbor embedding)
Dimensionality reduction and visualization of multiparameter flow cytometry data were performed with Barnes-Hut t-distributed stochastic neighbor embedding (tSNE) in FlowJo 10. Manually pregated and down-sampled cell populations from different treatment groups and donors were processed together in one tSNE plot (perplexity = 40, iterations = 1000, θ = 0.5) to allow direct comparison.
Analysis of previously published RNA-sequencing data
Previously published RNA-sequencing data (33) was converted to human gene nomenclature using the biomaRt R package. Relevant reads per kb per million mapped reads values were extracted in R programming language, and further analysis was performed with Prism 8 (GraphPad).
Statistical analyses were performed with Prism 8 (GraphPad) and are specified in each corresponding figure legend.
Fingolimod treatment reduces the frequencies and numbers of cTfh and cTfr cells
To analyze the changes imposed by fingolimod on T and B cells in MS patients, we used multiparameter flow cytometry to systematically profile various T and B cell populations (Supplemental Fig. 1). Compared with HC and untreated MS patients, we observed a relative increase in the CD45RO+ memory compartment of CD4+ T cells in fingolimod-treated MS patients (MS + fingolimod) and a reciprocal decrease in the frequency and number of CD45RO– naive CD4+ T cells (Fig. 1A), which is in accordance with previous studies (31), whereas treatment with natalizumab or rituximab did not affect the frequency of the memory CD4+ T cell compartment (Supplemental Fig. 2A). Despite the relative increase in the frequency of CD45RO+ memory cells, absolute cell numbers of memory CD4+ T cells were strongly decreased in fingolimod-treated MS patients compared with HC and untreated MS patients (Fig. 1A), which was expected as patients undergo lymphopenia during fingolimod treatment (31, 34). No differences were observed between HC and untreated MS groups in frequency or absolute cell number of memory CD4+ T cells (Fig. 1A). Interestingly, CXCR5+ cTfh cell frequencies among memory CD4+ T cells were strongly decreased in fingolimod-treated MS patients, with a reciprocal increase in the frequency of CXCR5− circulating memory Th (cTmem) cells (defined as CD4+CD45RO+CXCR5– and comprised of effector and central memory T cells) compared with HC and MS groups (Fig. 1B). Yet, the absolute cell numbers of both cTfh and CXCR5– cTmem cell populations were decreased in fingolimod-treated MS patients compared with HC and MS groups, with cTfh cell numbers being more affected than CXCR5– cTmem cell numbers (Fig. 1B). Furthermore, we did not observe a significant change in the frequency or number of cTfh cells in untreated MS patients compared with HC (Fig. 1B), which is in line with a recent study (35). We next assessed the frequencies of circulating Tfr (cTfr) and circulating Treg (cTreg) cells, which were defined as CXCR5+CD25hiCD4+ and CXCR5–CD25hiCD4+ T cells, respectively (Supplemental Fig. 1A). The identity of cTfr and cTreg cells was further supported by their lack of CD127 expression, as 93.5% of the CD25hiCD4+ T cells did not express CD127 (data not shown). CXCR5-expressing cTfr cell frequencies were reduced during fingolimod treatment compared with untreated MS patients, whereas the frequencies of CXCR5– cTreg cells were increased compared with HC and MS groups (Fig. 1C). Total cell numbers of both cTfr and cTreg cells were lower in fingolimod-treated MS patients compared with the control groups (Fig. 1C). In summary, MS patients treated with fingolimod exhibited a different blood CD4+ T cell composition than HC and MS groups with decreased frequencies of CXCR5-expressing cTfh and cTfr cells and of naive CD4+ T cells (Fig. 1D). In contrast, natalizumab and rituximab did not impact the frequency of all CXCR5-expressing CD45RO+ memory cells (including cTfh and cTfr cells) or the frequency of CXCR5–CD45RO+ memory cells (including CXCR5– cTmem and cTreg cells) as compared with both control groups (Supplemental Fig. 2B).
Frequencies of activated cTfh cells, but not CXCR5– cTmem cells, are elevated upon fingolimod treatment, and both cell types change their subset composition
We next performed more detailed phenotypical analyses of CD4+ T cells from the HC, MS, and MS + fingolimod groups by measuring the expression of the activation markers CD38, ICOS, and PD-1 as well as the chemokine receptors CXCR3 and CCR6. The frequency of activated CD38+ cells among the cTfh cell population was increased in fingolimod-treated MS patients as compared with HC and untreated MS patients (Fig. 2A), whereas there were no changes in the frequency of activated cells among the CXCR5– cTmem cell population (Fig. 2B). ICOS and PD-1 along with CD38 expression by cTfh cells have been reported to correlate with specific Ab responses (36–38). In line with this, various combinations of the activation markers also revealed a relative increase in activated cTfh cells in fingolimod-treated MS patients (Supplemental Fig. 3). cTfh and CXCR5– cTmem cells can be divided into further subsets according to their CXCR3 and CCR6 expression (13, 39). We observed an increase in the frequency of cTfh1 (CXCR3+CCR6–) cells in fingolimod-treated MS patients compared with HC and untreated MS patients (Fig. 2C). This was at the expense of cTfh1-17 (CXCR3+CCR6+) and cTfh17 (CXCR3–CCR6+) cell frequencies (Fig. 2C). The same analyses were performed on CXCR5– cTmem cells where a significant decrease of circulating Th1-17 (cTh) cells was observed in the fingolimod-treated MS patient group (Fig. 2D). Activated CD38+ cTfh and CXCR5– cTmem cells were also analyzed regarding their subset composition, with activated CD38+ cTfh cells containing a higher fraction of cTfh1-type cells in fingolimod-treated MS patients as compared with HC and untreated MS patients (Fig. 2E). Similar to the subset distribution within total cTfh cells, the fraction of cTfh1-17 and cTfh17 cells was decreased in fingolimod-treated MS patients compared with untreated MS patients and both control groups, respectively (Fig. 2E). Although increased circulating Th (cTh)1 cell frequencies were not observed within total CXCR5– cTmem cells in untreated MS patients (Fig. 2D), pregating on activated CD38+ CXCR5– cTmem cells revealed a higher frequency of cTh1 cells and a lower frequency of cTh2 cells (Fig. 2F). In summary, fingolimod treatment increased the frequencies of activated cTfh cells and skewed the composition of the cTfh cell pool toward a cTfh1 cell phenotype.
Dimensionality reduction of multiparameter flow cytometry data reveals changes in the composition of blood CD4+ T cells after fingolimod treatment
tSNE analysis of CD4+ T cells from seven untreated and 15 fingolimod-treated MS patients and the overlay with manually gated CD4+ T cell populations (compare gating strategies in Supplemental Fig. 1A) revealed and confirmed compositional changes of blood CD4+ T cells after fingolimod treatment (Fig. 3A). A comparison of the expression patterns of surface molecules that were used to generate the tSNE analysis revealed that naive CD4+ T cells expressing CD38 (40) were largely absent among naive CD4+CD45RO– T cells in the blood of fingolimod-treated patients compared with untreated MS patients. Furthermore, the increase of cTreg cells and the reduction of naive CD4+ T cells, cTfh cells, and cTfr cells were clearly apparent in this side-by-side comparison (Figs. 1A–C, 3A). Interestingly, in fingolimod-treated MS patients, a large fraction of cTreg cells expressed CCR6 (Fig. 3B). Finally, expression of the activation markers CD38, ICOS, and PD-1 distinguished untreated MS cTfh cells from cTfh cells derived from fingolimod-treated MS patients (Figs. 2A, 3B, Supplemental Fig. 3). In summary, dimensionality reduction of the multiparameter flow cytometry data set confirmed the reduction of naive and CXCR5-expressing CD4+ T cells in the blood of MS patients upon fingolimod treatment and, additionally, provided insights into the differential distribution of surface markers after S1PR modulation.
CD8+ T cells show similar compositional changes in the blood as CD4+ T cells upon fingolimod treatment
In addition to CD4+ T cells, CD8+ T cell numbers were also decreased in fingolimod-treated MS patients compared with the control groups (Fig. 4A), although not as strongly as CD4+ T cell numbers. In fact, more CD8+ T cells than CD4+ T cells were present in the blood of most fingolimod-treated MS patients, as reflected by a drop of the CD4+/CD8+ T cell ratio under 1 (Fig. 4B), consistent with other studies (31, 34). Similar to the observations made for CD4+ T cells, CD45RO+ memory CD8+ T cell frequencies were increased in fingolimod-treated MS patients compared with untreated MS patients (Fig. 4C, Supplemental Fig. 1B), and CXCR5-expressing CD8+ T cell frequencies were decreased compared with both control groups (Fig. 4D). As with CXCR5+CD4+ T cells, a higher percentage of CXCR5+CD8+ T cells showed expression of the activation marker CD38 (Fig. 4E). In summary, circulating CXCR5+CD8+ T cells were similarly affected by fingolimod treatment as their CD4+ counterparts.
B cell numbers in the blood decrease upon fingolimod treatment, and B cell subset frequencies change
Because of the close functional connection between Tfh and B cells, we next analyzed B cells in the blood of fingolimod-treated MS patients. Similar to T cells, the absolute number of B cells was decreased upon fingolimod treatment compared with HC and untreated MS patients (Fig. 5A), which is in line with previous studies (34, 41, 42). As some B cell subsets, especially memory B cells, were reported to act as pathogenic or proinflammatory B cells in MS and some as regulatory B (Breg) cells (43, 44), we next assessed the frequency of different B cell subsets (compare gating strategies in Supplemental Fig. 1C). CD27– nonmemory B cells were increased whereas CD27-expressing memory B cells were decreased in frequency in fingolimod-treated patients compared with untreated MS patients (Fig. 5B). CD27– nonmemory B cells were divided into CD38-expressing transitional B cells and CD38– naive B cells. There was a strong increase of regulatory transitional B cells in fingolimod-treated MS patients compared with HC and untreated MS patients (Fig. 5C). Additionally, the memory B cell compartment was divided into different memory B cell subsets. In this study, non–class-switched memory B cells (CD27+IgD+) were the only memory B cell subset that decreased in fingolimod-treated MS patients compared with HC and untreated MS patients (Fig. 5D), which is in line with a previous report (42). Significant changes were not observed in either the frequencies of circulating Ab-secreting cells (CD27+IgD–CD138+) or in IgM+ and class-switched memory B cells between all three groups (data not shown). In general, fingolimod treatment of MS patients decreased the frequency of memory B cells and increased the frequency of transitional Breg cells in their blood (Fig. 5E).
Naive CD4+ T cells and cTfh cells share characteristics of lymph node homing, whereas CXCR5– cTmem cells do not
S1PR1 expression overrides CCR7 signaling and enables the egress of immune cells from lymph nodes along a gradient of sphingosine-1-phosphate, the ligand of S1PR1, into the bloodstream (45). We, therefore, hypothesized that high expression of CCR7 in CD4+ T cells would retain them in lymph nodes during fingolimod treatment (31). The frequency of CCR7-expressing cells among CXCR5− cTmem cells and cTreg cells was significantly lower than in cTfh, cTfr, and circulating CD45RO– naive CD4+ T cells of healthy donors (Fig. 6A). High CCR7 expression was associated with lower frequencies of these Th cell subsets in the blood of fingolimod-treated patients (Fig. 1A–C). As expected, fingolimod treatment resulted in a severe reduction of CCR7-expressing total CD4+ T cells as well as CD8+ T cells in the blood of MS patients (Fig. 6B). These findings suggested that cTfh and naive CD4+ T cells not only expressed CCR7 at high levels but that they may also preferentially express other secondary lymphoid organ (SLO)–homing receptors, whereas CXCR5– cTmem cells may preferentially express receptors associated with homing to sites of inflammation. To test this, we made use of a previously published RNA-sequencing dataset (33) to check the expression of various genes in cTfh (CD4+CXCR5+) cells known to be important for homing to SLOs or sites of inflammation compared with circulating naive CD4+ T cells (CD45RA+CCR7+) and terminal effector CD4+ T cells (CD45RA–CCR7–) (healthy donors). The mRNAs of molecules associated with homing to and retention in lymph nodes (such as CD62L [SELL], CD69, and CCR7) were expressed at higher levels in naive and Tfh cells compared with terminal effector CD4+ T cells, whereas the mRNAs of molecules associated with migration toward peripheral tissues or sites of inflammation (such as integrin αL, integrin ß2, and CX3CR1) were higher expressed in terminal effector CD4+ T cells (Fig. 6C). Interestingly, S1PR1 mRNA was not differentially expressed in the different T cell types (Fig. 6C). In summary, fingolimod treatment particularly affected cTfh, cTfr, and circulating naive T cells that all share high expression of lymph node–homing receptors, such as CCR7, as compared with CXCR5– cTmem and cTreg cells.
cTfh cells reflect ongoing Tfh cell responses in SLOs (46, 47). In this study, we investigated how fingolimod, which is known to differentially regulate the egress of T cells and B cells from lymph nodes, affects cTfh cells in MS patients. Fingolimod profoundly reduced the frequencies of CXCR5+ cTfh cells among total memory CD4+ T cells. This effect of fingolimod on cTfh cells was similar to known effects of S1PR blockade on naive CD4+ T cells. Interestingly, fingolimod also reduced the frequencies of other CXCR5-expressing T cell subsets (i.e., Tfr and CXCR5+CD8+ T cells). A previous study reported that the frequency of CD25–CD127+CXCR5+PD-1+ cells remained stable within the CD4+ population upon fingolimod treatment and that only naive CD4+ T cells decreased (42). However, the gating strategy used in that study omitted many CXCR5+ cells because only a few CXCR5+ T cells in the blood are PD-1+ (16). Nevertheless, the observed increase in PD-1 mean fluorescence intensity in that study (42) is compatible with our finding that the frequency of activated cTfh cells, as measured by various combinations of CD38, ICOS, and PD-1, was increased in fingolimod-treated patients. Another study showed that the frequency of CXCR5+ cells among all CD4+ T cells did not differ between untreated and fingolimod-treated RRMS patients (34). This observation could be explained by the strong decrease of naive CD4+ T cells and the reciprocal overrepresentation of CD45RO+ memory T cells in the blood of fingolimod-treated MS patients. Similar to our findings, that study also showed that the number of CXCR5-expressing CD4+ T cells was strongly decreased upon fingolimod treatment, whereas patients with lymphopenia that received dimethyl fumarate treatment had similar numbers of CXCR5+CD4+ T cells as untreated RRMS patients (34), suggesting a preferential effect of fingolimod on Tfh cells. Our findings are also in line with another recent report that investigated the origin of cTfh cells and found a reduction in the frequency of CXCR5+ Tfh cells in the blood of MS patients upon fingolimod treatment (48).
The disproportional reduction of cTfh cells is most likely connected to their high CCR7 expression, which supports the view that CCR7 is a critical determinant of how likely a cell is to enter SLOs and then be retained by S1PR blocking (49). Furthermore, in addition to CCR7, cTfh cells also contained high mRNA levels of other SLO–homing/retention factors. This homing potential is in line with the function of naive CD4+ T cells to recognize Ag and differentiate within lymph nodes, although it also supports the theory that cTfr cells (which also express high levels of CCR7 in the blood) and especially cTfh cells patrol SLOs to quickly establish germinal centers upon reinfection (50). In mice, it was shown that cTfh and cTfr cells represent memory-like cells that recirculate to SLOs, and FTY720 treatment effectively reduced their frequencies in the circulation (51). These data are in line with a model in which recruited cTfh cells would downregulate CCR7 again to gain access to B cell follicles and germinal centers comparable to Tfh cells that differentiate during a primary immune response (52, 53). It should be noted that Tfh cells may not only be important for providing help to B cells for Ab responses during CNS inflammation. The Ag-presenting function of B cells, but not their ability to secrete Abs, has been shown to be critical for disease pathogenesis in a B cell–dependent EAE model (54). Given the requirement for continued interactions of lymphoid tissue-resident bona fide Tfh cells with B cells for the maintenance of the Tfh cell phenotype (55, 56), it is likely that the B-Tfh cell axis is also important for MS pathogenesis. This would be in line with a recent report that showed that deletion of the Tfh-associated transcription factor Bcl6 specifically in T cells ameliorated disease in the EAE model (26). Nevertheless, the contributions and relationships of lymphoid tissue-resident bona fide Tfh cells and blood-resident cTfh cells remain to be elucidated in more detail.
Similar to T cells, B cells are also reduced in blood by S1PR blocking. Our study confirms that memory B cells are preferentially reduced in blood during fingolimod treatment (42, 57, 58) and among those especially non–class-switched memory B cells. Effects on memory B cells might be relevant because memory B cells are recruited to the cerebrospinal fluid (4, 59, 60) and are considered to be of particular pathogenic relevance (59). We found that transitional B cells are relatively increased during fingolimod treatment, which confirms previous reports (59, 61). Transitional B cells have been described as regulatory IL-10–secreting Breg cells, and therefore, their strong relative increase upon fingolimod treatment as well as the relative increase of cTreg cells might suppress pathological conditions (41, 62, 63). Interestingly, the frequency of CCR6-expressing cTreg cells was particularly increased in the blood after fingolimod treatment. These cells might be more prone than CCR6-negative Treg cells to home to the CNS because of their responsiveness to a gradient of CCL20, the ligand of CCR6. Thus, these CCR6-positive Treg cells might control inflammatory immune responses directly in the tissue (64).
The prominent reduction of cTfh cells we describe in this article might also result in reduced development of ectopic lymphoid follicle–like structures (ELFs). ELFs, which are composed of B cell aggregates, T cells, and accessory cells, have been observed in the meninges of MS patients and have been linked to cortical pathological conditions (24, 25). Additionally, Tfh cell subsets have different B cell helper capabilities with Tfh2 and Tfh17 cells exhibiting good B cell helper functions, whereas Tfh1 cells do not (13). We observed that relative frequencies of cTfh17 cells were reduced, whereas frequencies of cTfh1 cells were correspondingly increased in the blood of MS patients during fingolimod treatment. As cTfh1 cells are believed to be less suited for providing B cell help than cTfh2 or cTfh17 cells (9), this might impair ELF formation and function. Because cTfh cells are considered to be important for potent recall responses (50), the reduction in total cTfh cell numbers as well as our observed change in the quality of cTfh cells (i.e., increase in cTfh1 cells that are not good B helper cells) could explain the observations of previous studies that showed that vaccination recall responses are impaired in fingolimod-treated vaccine recipients (65, 66).
In summary, we showed that S1PR blocking not only reduced total T and B cell numbers in the blood but also decreased the frequency of CXCR5-expressing Tfh, Tfr, and CD8+ T cells as well as memory B cells while increasing the frequency of Treg and Breg cells. These quantitative and qualitative changes in circulating lymphocyte populations add to our understanding of S1PR blockade and its benefits in MS patients.
We thank Martin Kerschensteiner and Reinhard Hohlfeld for helpful discussions; Anne Krug and Angelika Schmidt for critical reading of the manuscript; Miriam Schlüter, Sabine Pitter, Anida Muhic, and Angelika Bamberger for providing clinical samples; and the Biomedical Center Munich Core Facility Flow Cytometry of LMU Munich for providing equipment.
This work was supported in part by Novartis Pharma GmbH Germany, the Deutsche Forschungsgemeinschaft (Emmy Noether Programme BA 5132/1-1, SFB 1054 Teilprojekt B12, and SFB TR128 Teilprojekt B08), the Clinical Competence Network for Multiple Sclerosis, and the Verein zur Therapieforschung für Multiple Sklerose Kranke. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
The online version of this article contains supplemental material.
Abbreviations used in this article:
circulating memory Th
experimental autoimmune encephalomyelitis
ectopic lymphoid follicle–like structure
secondary lymphoid organ
T follicular helper
T follicular regulatory
t-distributed stochastic neighbor embedding.
The authors have no financial conflicts of interest.