Multiple sclerosis (MS) is a chronic disease of the central nervous system (CNS), with inflammation, demyelination, and neurodegeneration. The pathogenic process is immune-mediated, and the etiology is probably multifactorial, with interaction of heritable and non-heritable factors (1).

Among the other factors, microbiota and gut function are increasingly recognized as relevant in this immune-mediated disorder (2). Several studies have recently shown that the microbiota, as a part of the intestine–brain axis, plays a role in the etiopathogenesis of MS (3, 4). However, a crucial component of this axis, the intestinal barrier, has received much less attention. The question of whether or not intestinal permeability (IP) is affected during the disease course is at least as important as the changes in the microbiota balance (5, 6).

IP changes may underlie gastro-intestinal or even far-from-gut autoimmune disorders. In fact, increased gut permeability allows the passage of macromolecules, toxins, and bacterial species that may trigger immune-mediated diseases in different systems, even distant from the gastrointestinal tract, such as the CNS (7, 8). On the other hand, CNS inflammation can increase gut permeability and alter mucosal structure in the small intestine (9).

In a previous work, we investigated the gut permeability in relapsing–remitting MS (RRMS) patients and healthy donors, finding that alteration of IP represents a relatively frequent event in patients with MS (10). This study and a previous one, showing that CD161^highCD8+ T cells, encompassing the mucosal associated invariant T (MAIT) cell subset, play a role in MS pathogenesis (11), prompted us to focus on the gut triggers that may lower the threshold for disease development in susceptible individuals.

Dimethylfumarate (DMF) has both neuroprotective and anti-inflammatory effects, and it is currently used as an oral, first-line, disease-modifying therapy (DMT) in MS. Some of the mechanisms responsible for its efficacy have been clarified, while others remain unexplored. Gastrointestinal tract irritation is one of the most frequent side effects of DMF (12). On the other hand, studies on experimental models of inflammatory bowel diseases showed that DMF might beneficially affect IP (13).

In this study, we investigated IP changes, the circulating CD161+CD8+ T-cell subset, and clinical/neuroradiological data in a cohort of RRMS patients before and after 9 months of DMF therapy, with a longitudinal design aimed at analyzing data at three time points: baseline (before therapy) and after 1 (T1) and 9 months (T2) of treatment.

The demographic, clinical, and neuroradiological characteristics of 25 patients at baseline are summarized in Table 1. Sixteen patients (64%) showed an altered IP, while 14 cases (56%) showed active MRI (4 of them were also in clinical relapses). Moreover, we investigated the frequencies of CD161+ CCR6+ CD8+ T lymphocytes in PBMCs obtained from 23 MS patients.

During DMF therapy, two significant changes emerged. At first, we could confirm the decrease of disease activity as evaluated by MRI (6/25 at T1, p = 0.035 for the comparison between T1 and T0; 3/25 at T2, p < 0.001 for the comparison between T2 and T0); consistent with this result, the Kaplan–Meier analysis showed that the proportion of patients with normal MRI signal is higher at T1 and T2 compared to T0 (Figure 1). Then, we showed that the frequency of circulating CD161+CCR6+CD8+ T cells in MS patients is reduced after 9 months of DMF treatment. In Figure 2A, representative plots depict the progressive drop of frequencies of subpopulations of CD8 T cells (CD161hi CCR6+) in an MS patient during DMF treatment, while Figure 2B shows the cumulative data with the significant drop of the T-cell subset at T2 (p < 0.001).

We found that the decrease in disease activity evaluated radiologically was 12 times higher in subjects showing reduced frequencies of CD161+CCR6+CD8+ T cells in the peripheral blood, and 15 times higher in cases with IP changes at T1 (Table 2). Furthermore, a logistic regression model showed a relationship between the drop of CD161+CD8+ T cells in the peripheral blood and IP changes at T2, considering as covariates both EDSS and MRI activity (p = 0.023). Consistent with this result, the Cox analysis showed that the decline of the T-cell subset was more evident in patients with persistent IP changes (Hazard Ratio, HR = 4.19; p = 0.03; Figure 3).

Concerning the possible effects of DMF treatment on gut barrier alterations, no significant difference emerged: some cases improved, while others worsened during follow-up, without a clear longitudinal pattern (Supplementary Table); the proportion of patients having IP changes at T1 and T2 was 16/25 (64%) and 15/25 (60%), respectively (figures that were quite comparable to the baseline data). A minority of patients (3/25) had mild gastro-intestinal side effects during treatment with DMF; other mild side effects were within the known safety profile of the drug (not shown). One patient presented a relapse during the follow-up, and no significant change occurred in patients' EDSS at T2 compared to baseline.

This work, together with our pilot studies on IP and mucosal immunity in RRMS (10, 11), provides evidences that the gut barrier is frequently altered in these patients and that the CD161+CD8+ T-cell subset shows dynamics compatible with disease course and therapy. No other studies on IP changes in MS have been reported so far, since the other works are mainly focused on microbiota changes (3). However, we deem IP alterations and dysbiosis as two faces of the same coin (16), and further studies correlating IP and gut microbiota changes in MS will certainly be informative on disease etiopathogenesis.

The CD161+CD8+ T-cell subset encompasses the MAIT cells, which were the object of several investigations after our first study on MS in 2011 (11), all largely confirming the involvement of MAIT cells in MS pathogenesis. Among the evidences repeatedly reported were an IL18-driven activation and consequent CNS infiltration of MAIT cells in the diseased brain, and an increased type-17 differentiation and oligoclonality of circulating MAIT cells in MS patients compared to controls (17–21). The IL18-driven activation, and the consequent CNS infiltration of CD8+ MAIT cells in MS, may cause reduced frequency in blood, helping to reconcile, at least in part, the conflicting results on the frequency of circulating MAIT cells in MS. A recent work showed indeed that MAIT cell subtype, smoking habit, and disease onset (primary progressive vs. relapsing–remitting) affect the number of circulating MAIT cells (22). Smokers with primary progressive MS showed low frequency of circulating MAIT cells, suggesting a tendency to reside in the inflamed organ, in apparent contrast to what was observed in most studies on patient with RRMS.

Concerning the effects of DMF on the variables under study, we found the expected decline of disease activity, which was in keeping with the initial pivotal trials (23–25). The parallel drop in the fraction of circulating CD161+CD8+ T cells is in accord with two previous works on the effects of DMF therapy in MS patients (22, 26). The action of DMF on all the proinflammatory T-cell subsets, including the CD161+ IL17-producing T cells, is mediated by a dose-dependent induction of apoptosis and decrease of proliferation (27). Other works, showing a decrease of proinflammatory MAIT cells after hematopoietic stem cell transplant or alemtuzumab for treatment-refractory forms (28, 29), support results obtained after DMF treatment, and indirectly confirm the pathogenic role of MAIT cells in MS.

No clear DMF effects were evident on IP changes, and the gastrointestinal side effects in our group of patients were relatively rare and apparently unrelated to IP changes. The meaning of this finding requires further studies (such as those based on novel multi-sugar assay for site-specific gastrointestinal permeability analysis) and suggests that the alterations of the gut barrier in MS are complex: the decreased disease activity at MRI and the reduction of the percentage of circulating MAIT cells during treatment with DMF seem to occur more frequently in patients with IP changes. These relationships raise the possibility that the gut barrier alteration may represent a predictor of pathophysiological transitions, besides its possible role in disease pathogenesis. Our study adds evidences to the potential role of mucosal immunity in MS pathogenesis, and yet suggests questions that remain unanswered. Among those are whether IP changes somehow drive demyelinating process [as seen in experimental models of MS; (9)] or simply contribute to the organ-specific immune dysfunction. Also, it is unclear through which mechanisms MAIT cells [or subsets of them; (22)] become activated and pathogenic at the CNS level in apparently sterile conditions. Answering these questions may provide new fruitful lines of attack against neuroinflammation, such as IP enhancers or stabilizers, already under scrutiny in gastro-intestinal conditions, as well as compounds coming from reworking the increasingly growing data coming from microbiota studies in experimental and human autoimmune diseases.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Azienda Ospedaliero-Universitaria Sant'Andrea, Università Sapienza. The patients/participants provided their written informed consent to participate in this study.

MSa, GR, and LB conceived the study. MB, GB, and MSi coordinated the project. FG, MC, and LL performed data analysis. CF carried out statistical analysis. All authors contributed to the project, writing of the manuscript, and approved its final version.

MSa receives research support and has received fees as speaker from Sanofi, Biogen, Roche, Novartis, Bayer Schering, and Merck Serono. LB received funds for research from TEVA, Baxter, Roche, and Merck Serono and speaker honoraria by Genzyme, Teva, Novartis, Roche, and Merck Serono. MB has received fees as speaker from Sanofi, Biogen, Roche, Merck Serono, and Novartis. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.