Features of the intestinal microbiota composition in multiple sclerosis patients receiving oral disease-modifying therapy

Full Text

Abbreviations

EDSS — Expanded Disability Status Scale; DMT — disease modifying therapy; UT — untreated; FG — fingolimod; TF — teriflunomide; IFNs — Interferons; GA — Glatiramer Acetate; DMF — Dimethyl Fumarate; GIT — gastrointestinal tract; DD — disease duration; CFU — colony forming units.

Background

Multiple sclerosis (MS) is a chronic inflammatory autoimmune disease of the CNS, characterized by myelin loss and damage of nerve cells [1]. The causes of MS are unknown, and the mechanisms of the disease progression are not fully understood. The potential role of gut microbiota in the pathogenesis of MS has been actively discussed in recent years [2]. Intestinal microbiota affects the development and function of the immune and nervous systems by promoting differentiation of Th [3, 4] and Treg [5, 6] cell subpopulations, blood-brain and gut barrier integrity [7–9] and myelination [10].

Several studies show that gut microbiota in MS patients are altered compared to healthy subjects and is characterized by dysbiosis [11–14]. In addition, 70–90% of patients with MS have GIT dysfunction [14, 15], resulting from intestinal microbiota dysbiosis.

With the key role of immune auto-aggression in MS pathogenesis recognized, immunomodulators/immunosuppressors were developed to modify this disease. IFN-β based drugs and glatiramer acetate (GA) are the first and best-studied substances for disease-modifying therapy (DMT) [16, 17]. These drugs have a complex, multifaceted effect on the body; however, not all of these effects are fully understood. Several small studies have shown that GA and interferon therapy may influence gut microbiota composition [18–20]. At the same time, the efficacy of GA therapy may be influenced by the presence of some microorganisms [21].

Studying microbiota in patients with MS has become even more relevant with the development and widespread clinical use of oral DMTs [22, 23]. Upon entering the intestine, drugs may affect the gut microbiota causing or exacerbating dysbiosis, which is implicated in GIT disorders and reduces the patient’s quality of life. On the other hand, gut microbiota composition may influence treatment efficacy either by altering drug metabolism and bioavailability or by affecting the immunocytes targeted by these DMTs. A recent study has evaluated the potential effects of dimethyl fumarate on gut microbiota composition [24].

The purpose of our study was to compare gut microbiota composition in MS patients treated with two widely used DMTs – Fingolimod or Teriflunomide and in untreated patients.

Materials and methods

Ethics approval and patient consent. The study was approved by the ethics committees of both Institute of Experimental Medicine and the Institute of the human brain; written informed consent has been obtained from all participants.

Patients. 81 patients with MS were enrolled in the observational study. All patients were under observation in the Clinical Department of the Institute of Experimental Medicine and the Institute of the Human Brain from 10.01.2017 to 01.09.2019. They had relapsing-remitting MS (RR-MS), secondary progressive MS (SP-RS), or primary progressive MS (PP-MS). Patients received oral DMT, Fingolimod (FG, n = 31) or Teriflunomide (TF, n = 19), or were untreated (UT, n = 31). All patients were in the remission stage during an assessment. The characteristics of the studied cohort are shown in Table 1.

 

Table 1. The characteristics of enrolled multiple sclerosis patients

Characteristic	UT	DMT
		FG	TF
Age, years	42.9 ± 2.3	42.9 ± 2.0	36.0 ± 2.0
Duration of disease, years	10.5 ±1.7	12.3 ± 1.1	7.8 ± 1.6
EDSS score	3.2 ± 0.4	3.7 ± 0.4	3.2 ± 0.3
Age of MS onset, years	32.2 ± 1.9	30.9 ± 1.5	28.3 ± 1.6
Duration of therapy, years	–	5.0 ± 0.4	2.1 ± 0.1
Total patients (male/female)	31 (9/22)	31 (11/20)	21 (10/11)

Notes: UT — untreated group; DMT — disease-modifying therapy group; FG — Fingolimod; TF — Teriflunomide; EDSS — Expanded Disability Status Scale; MS — multiple sclerosis.

 

Methods. A modified Neurogenic Bowel Dysfunction Score [25] questionnaire was used to assess GIT functions.

Two methods were used for the analysis of microorganism’s levels in fecal samples: the culture method according to the algorithm described earlier [19] and real-time polymerase chain reaction (RT-PCR) with the “Colonoflor” Kit (Alphalab, St.-Petersburg, Russia). Fecal samples were delivered to the laboratory; the same sample was used for two methods. The sample were analyzed without a storage period; freezing of samples was not allowed.

Statistical analysis. Analysis of variance with post-hoc HSD test for unequal groups was performed in Statistica-8 for compare effects of DMTs. Fisher exact test was used to compare proportions. Differences at p < 0.05 were considered statistically significant.

Results

Assessment of gastrointestinal tract function. The questionnaire-based survey revealed that 64.5% of subjects not receiving DMT had GIT dysfunction. The percentage of subjects with GIT disorders was higher in groups receiving treatment: 81% for subjects treated with Teriflunomide (TF) and 100% for those treated with Fingolimod (FG). The most frequent complaints among patients are listed in Table 2. Subjects with GIT dysfunction had 2 to 6 symptoms in varying combinations. The distribution of patients depending on the number of complaints presented is shown in Figure 1.

 

Table 2. Incidence of various gastrointestinal tract symptoms and metabolic disorders in a cohort of study subjects with multiple sclerosis

Symptom		Percentage of patients with symptom, %
		Untreated (UT) (n = 31)		FG (n = 31)		TF (n = 21)
Defecation disorders, total: Diarrhea/constipation, %		45.2 19.4/25.8		74.2* 29.0/45.2		66.6 33.3/33.3
Bloating, %		22.6		35.5		23.8
Rumbling, %		16.1		29.0		52.6#
Stomach heaviness, %		22.6		29.0		21.1
Abdominal pain, %		22.6		54.8*		52.4*
Nausea/Vomiting, %		35.5		29.0		36.8
Heartburn, %		22.6		45.2		52.6*
Change in appetite, %		22.6		29.0		31.6
Body mass change (total %)	↑ (%)	45.2	35.5#	29.0#	12.9	42.1	0
	↓ (%)		9.7		16.1		42.1#

Notes: * differences from UT group, p < 0.05; ^# differences from other groups p < 0.05 (Fisher test). UT — untreated group; FG — Fingolimod; TF — Teriflunomide.

 

Fig. 1. Proportion of multiple sclerosis patients according to number of symptoms. UT — untreated; FG — Fingolomod-treated; TF — Teriflunomide-treated; 0 S — without symptoms; 2 S, 3 S, 4 S, >5 S — two, three, fore, and five and more symptoms GIT disorders from table 2

 

The functional defecation disorders (fecal incontinence or constipation) were 1.2 and 1.8 times less prevalent in the untreated patients than those receiving TF and FG, respectively. Half of the TF-treated patients experienced rumbling and heartburn, which is more common than patients from the other groups. Half of the patients in both groups receiving DMT experienced abdominal pain, twice the share in the untreated group.

In addition to GIT disorders, some patients in all groups reported changes in appetite and body mass. Thus, most untreated patients (35.5%) with a change in body mass reported an increase in body mass, whereas all patients receiving TF with a change in body mass reported a decrease in body mass. In addition, the patients receiving FG reported both an increase in body mass (13%) and a reduction in body mass (16%).

Generally, oral DMT exacerbated GIT disorders in MS patients, with specific changes depending on the drug used.

Analysis of the intestinal microbiota by the culture method. The symbiotic bacteria species (Lactobacillus spp., Bifidobacterium spp., Escherichia coli (N E. coli), Enterococcus spp.) and atypical (with reduced fermenting activity, hemolytic, lactose-negative) opportunistic bacteria species (E. coli (A E. coli)), Enterobacter spp., Citrobacter spp., Proteus spp., Klebsiella spр. Staphylococcus aureus, Clostridium spp.) and yeast of genus Candida spp. were detected by the cultural method in fecal samples. The levels of some microorganisms differed in treated and untreated patients. There were also differences in patients receiving different drugs (Fig. 2).

 

Fig. 2. The level of symbiotic (a) and opportunistic (b) bacterial species in intestinal microbiota of multiple sclerosis patients with Fingolimod or Teriflunomide therapy (culture method). Data represented as mean ± standard error. In Y axis – the bacterial level in lg colony forming units (CFU) on fecal g; in Х axis – bacterial species found; UT – untreated, FG – Fingolimod-treated, TF – Teriflunomide-treated. ANOVA with post-hoc HSD for unequal N, * difference between FG and TF groups, •differences from UT and TF groups, ^# differences from UT group, p < 0.05. N — normal, A — atypical

 

Figure 2 shows that mean levels of Lactobacillus spp., Bifidobacterium spp., and Enterococcus spp. for every group were within the reference range, with the lowest levels of these bacteria found in patients receiving TF. However, the analysis of individual values showed that not all patients had normal bacteria levels, with levels being both lower and higher than normal. Thus, reduced levels (<6.0 lg CFU/g) of Lactobacillus spp. were found in 68% of patients in the TF group vs 42% in UT and 32% in FG. Reduced levels (<5.0 lg CFU/g) of Enterococcus spp. were found in 26% of patients receiving FG or TF vs 6.5% in the untreated group. At the same time, 13% of patients in UT group and 19% in FG group had a high (8.0 lg CFU/g) level of enterococci. No patients with high levels of enterococci were observed in TF group.

Subjects in TF treated group had lower levels of Bifidobacterium spp. (Fig. 2). This, however was not due to a reduction in the absolute quantity of bacteria, but because the percentage of patients with a high level (9.0–10.0 lg CFU/g) of Bifidobacterium spp. in this group was lower than in other groups (37% vs 74% and 68%, in UT and FG, respectively). It is noteworthy that high levels of Bifidobacterium spp. were found mainly in patients with higher EDSS scores (4.0–8.0) in all groups.

Normal levels (7.0–8.0 lg CFU/g) of Escherichia coli were found in 68% of untreated patients, in 47% of TF treated patients, and only in 29% FG treated patients (φ = 3.13; p < 0.01; vs UT). That is, symbiotic E. coli were decreased in MS patients, with a further reduction when DMTs are used, especially FG.

At the same time, in this group, not only the level of E. coli decreased, but also their properties changed. The “normal” E. coli were replaced by its atypical forms or other Enterobacteriaceae, particularly Enterobacter spp. The proportion of such patients in FG group was 2.7 times more compared UT group (52% vs 19%; φ = 2.72, p < 0.01). Interestingly, TF-treated patients showed only a decrease (to 4.0–6.0 lg CFU/g) in the level of E. coli.

This data has shown that Fingolimod had the most substantial adverse effect on E. coli compared to the two other groups (Fig. 2).

Analysis of intestinal microbiota by Real-Time PCR. We also compared the quantity of anaerobic MS marker bacteria of two dominant phyla, Bacteroidetes (Bacteroides fragilis, Bacteroides thetaiotaomicron, Prevotella spp.), Firmicutes (Faecalibacterium prausnitzii, Roseburia spp., Ruminococcus spp.) and minor phyla Proteobacteria (Sutterella wadsworthensis) and Verrucomicrobia (Akkermansia muciniphila) by RT-PCR method.

We found no significant differences in the levels of B. fragilis, B. thetaiotaomicron or Prevotella spp. (Fig. 3), but the proportion of patients with high (>12.5 lg CFU/g) B. fragilis level was significantly lower in the FG group compared to the UT group (φ = 2.31; p < 0.01).

 

Fig. 3. Changes in bacterial levels in Fingolimod- and Teriflunomide-treated patients with multiple sclerosis (PCR method). Data represented as mean ± standard error. In Y axis — the bacterial level in lg colony forming units (CFU) on fecal g; in Х axis — bacterial species; UT — untreated, FG — Fingolimod-treated, TF — Teriflunomide-treated. ANOVA with post-hoc HSD for unequal N, * differences between FG and TF groups, • differences from UT and TF groups, ^# differences from UT group, p < 0.05

 

In addition, FG-treated patients had lower levels of butyrate-producing bacteria, F. prausnitzii, Roseburia spp., the proportion of patients with high (>10.5 lg CFU/g) level of F. prausnitzii was 5 times lower than in the untreated group.

Ruminococcus spp. was found in 59.1% UT patients and only in 23% of patients receiving DMT (φ = 2.30, p < 0.05 and φ = 2.15, p < 0.05 for FG and TF, respectively).

59. wadsworthensis was also found in 59.1% UT patients, but TF had less effect on the presence of these bacteria in patients (38.5%) than FG (17.6%; φ = 2.75, p < 0.01).

In contrast, A. muciniphila was found in a comparable number in UT- and FG-treated patients (44% and 41%, respectively), and only in 23% of patients receiving TF, although the differences did not reach statistical significance.

Thus, treatment with FG and TF affected the qualitative and quantitative composition of the gut microbiota, having an inhibiting effect on different groups of studied microorganisms.

Discussion

Published data demonstrate that MS patients have altered composition of intestinal microbiome compared to healthy subjects. Several studies have found an increased abundance of Bifidobacterium spp. (phylum Actinobacteria) [11, 20, 26], A. muciniphila (phylum Verrucomicrobia) [12, 27–31], methane-producing Euriarchaeota [12, 32] and decreased abundance of Bacteroides (phylum Bacteroidetes) and Clostridia (phylum Firmicutes) – producers of butyrate [12].

There are also limited data on the influence of DMTs on microbiota composition [18, 20, 24, 27, 28]. These studies show alterations in intestinal microbiome composition during GA, IFNs, or DMF treatment. The authors interpret these results as a positive effect of DMT on the intestinal microbiome.

Our study noted quantitative alterations in the same bacterial species described by other authors using sequencing methods [11–13]. At the same time, we note the negative impact of DMT on symbiotic bacterial species, which leads to an increased proportion of related opportunistic species.

Different drugs likely affect specific target bacteria. In particular, the antimicrobial effect of GA on the gram-negative bacteria E. coli and Pseudomonas aeruginosa has been shown in vitro [33]. Therefore it is not surprising that some GA-treated patients have reduced E. coli, as shown earlier [19].

FG, TF, and DMF inhibited in vitro growth of Clostridium perfringens [34]. Since these bacteria are present in a small number of patients (about 11%) [35, 36], we could not assess the anti-clostridial effects in our cohort of FG or TF treated patients. The drugs may be expected to have the same effect on other Clostridia. However, FG and TF had a different impact on F. prausnitzii and Roseburia spp. levels, with the first drug decreasing their quantity and the second not affecting their number compared to UT patients.

Interestingly, other authors observed an increase in the abundance of Faecalibacterium after 12 weeks of DMF treatment [24]. Such differences may be related to the different influences on the Clostridia class of used drugs or durations of drugs treatment. Therefore, it is possible that at the beginning of treatment (12 weeks with DF) there is an increase, but as the duration of treatment increases (2 years with TF) there is a decrease, first to the level of UT patients and then lower (5 years with FG).

The decrease in Lactobacillus spp. and Enterococcus spp. levels in TF-treated patients, observed in this study, suggests that the drug is active against gram-positive Bacilli class species (phylum Firmicutes), while FG has a more pronounced effect on Clostridia class, which belongs to the same phylum.

Storm-Larsen et al. have shown that after two weeks on DMF, Actinobacteria abundance was decreased mainly driven by a reduction of Bifidobacteria [24]. We registered a lower level of Bifidobacteria spp. in TF-treated patients. The level of these bacteria in FG-treated patients was comparable to the untreated group. Interestingly, regardless of receiving DMT, the level of Bifidobacteria was higher in patients with high EDSS scores. Considering that the level of Actinobacteria phylum, especially Вifidobacteria spp., significantly increased in MS patients, reducing the level of these bacteria associated with TF treatment can be seen as a positive effect of treatment.

In this study, we confirmed an earlier finding in a larger group of patients that the quantity of symbiotic E. coli decreases, and it is substituted with related opportunistic species in FG-treated patients [19]. Also, it is worthy of note that TF did not have this anti-coliform effect.

This study is the first to compare the effects of two oral DMTs (FG and TF) on the composition of the intestinal microbiota and the spectrum of GIT disorders. The impact of FG and TF on the levels of different classes of bacteria can cause differences in the GIT disorders’ range observed in patients.

The substitution of symbiotic species of Proteobacteria phylum (E. coli, S. wadsvorthensis) by opportunistic bacteria in FG-treated patients is consistent with the presence of diarrhea/constipation and pain characteristic of inflammatory bowel diseases.

Since Escherichia coli can synthesize antibiotic-like substances – colicins and compete for adhesion and metabolite sites, they actively participate in the development of colonization resistance, suppressing the growth and multiplication of related pathogenic and opportunistic microorganisms in the intestine [37]. Therefore, it is logical that as the levels of symbiotic E. coli decrease, the gastrointestinal tract of FG-treated MS patients is colonized by opportunistic species. These can be atypical forms of E. coli, Enterobacter spp., Citrobacter spp., Klebsiella spp., which persist in the gastrointestinal tract of MS patients, causing intestinal disorders and possibly affecting the MS course. It is noteworthy that clearance of these bacteria is considerably slower in mice infected with Citrobacter rodentium when FG is administered [38], which is associated with a decrease in the number of Th17 cells that control pathogens in the intestines. The weakening of the control function of Th17 cells in the intestine, especially in combination with reduced levels of symbiotic species, may lead to excessive growth of pathogenic species and their translocation to other niches, including CNS [39]. This is likely the cause of Listeria monocytogenes rhombencephalitiss and other infections described in FG treated MS patients [40–42].

Bacteroides fragilis, B. thetaiotaomicron, Prevotella spp. (Bacteroidetes phylum), were found to be more resistant to FG- or TF-treatment, as their levels did not experience significant reductions compared to the untreated group. Other studies have described an increase in the abundance of Bacteroides with DMF treatment [43], or Prevotella spp. with IFN-β or GA treatment [12].

Interestingly, only patients receiving TF experienced a decrease in body mass without any patients undergoing an increase. Thus, TF likely affects metabolic processes or the bacteria involved in them.

Thus, we have demonstrated in this study that DMTs alter the composition of the gut microbiota. Furthermore, according to our preliminary data (data not shown), these changes can increase with increasing duration of treatment, and as a result, dysbiosis becomes more pronounced in patients receiving DMTs for a long time.

However, to consider the effect of the therapy duration and other factors, for example, the severity of the disease, further studies involving larger cohorts of patients are needed. In addition, expanding the list of determined microorganisms may also be advisable since it will allow to more fully characterize the changes in the microbial composition caused by the DMTs.

Conclusion

A great deal of attention is given to the study of intestinal microbiota in various CNS diseases. Alterations in the intestinal microbiota composition can contribute to GIT dysfunction and modulate the immune functions of the macroorganism, contributing to the pathological process and aggravating the clinical course of the disease. We suggest that an increase in pro-inflammatory opportunistic species in patients receiving DMTs is an adverse side-effect of the drugs, negatively affecting the MS course. Moreover, the alterations in gut microbiota composition may reduce treatment efficacy. With this in mind, regular monitoring of the microbiota and its correction can be helpful in the management of patients with MS.

Additional information

Funding. The study was funded by Ministry of Education and Science of the Russian Federation (grant 0557-2019-001).

Conflict of interest. The authors have no conflicts of interest regarding the publication of this article.

Author contributions. E.A. Tarasova — investigation, writing — original draft. V.I. Lioudyno — investigation, writing — original draft. A.V. Matsulevich — investigation, formal analysis. I.G. Negoreeva, A.G. Ilves, E.V. Ivashkova, G.G. Shkilnyuk — data curation. I.N. Abdurasulova — conceptualization, formal analysis, writing – review and discussion, funding acquisition, supervision.

About the authors

Elena A. Tarasova

Institute of Experimental Medicine

Email: tarasovahellen@mail.ru
ORCID iD: 0000-0003-0160-9590
Scopus Author ID: 25937494300
ResearcherId: J-6990-2018

Research Associate, Pavlov Department of Physiology

Russian Federation, 12, Academician Pavlov Str., Saint Petersburg, 197376

Victoria I. Lioudyno

Institute of Experimental Medicine

Email: vlioudyno@mail.ru
ORCID iD: 0000-0002-1449-7754
SPIN-code: 8980-8497
Scopus Author ID: 6504455988
ResearcherId: E-3797-2014 H-6

Cand. Sci. (Biol.), Senior Research Associate, Pavlov Department of Physiology

Russian Federation, 12, Academician Pavlov Str., Saint Petersburg, 197376

Anna V. Matsulevich

Institute of Experimental Medicine

Email: cat_fly@bk.ru
ORCID iD: 0000-0002-0030-9548
SPIN-code: 8464-1814
Scopus Author ID: 57190964381
ResearcherId: J-8280-2018

Research Associate, Pavlov Department of Physiology

Russian Federation, 12, Academician Pavlov Str., Saint Petersburg, 197376

Irina G. Negoreeva

N.P. Bechtereva Institute of the Human Brain of the Russian Academy of Sciences

Email: nip@ihb.spb.ru
ORCID iD: 0000-0002-1497-7109
SPIN-code: 7742-7720
Scopus Author ID: 23498576100

MD, Cand. Sci. (Med.), Research Associate, Laboratory of Neuroimmunology

Russian Federation, Saint Petersburg

Aleksandr G. Ilves

N.P. Bechtereva Institute of the Human Brain of the Russian Academy of Sciences

Email: ailves@hotmail.com
ORCID iD: 0000-0002-9822-5982
SPIN-code: 1068-7281
Scopus Author ID: 36113684700
ResearcherId: AAO-7683-2021

MD, Cand. Sci. (Med.), Senior Research Associate, Laboratory of Neuroimmunology

Russian Federation, Saint Petersburg

Elena V. Ivashkova

N.P. Bechtereva Institute of the Human Brain of the Russian Academy of Sciences

Email: ivashkova@ihb.spb.ru
ORCID iD: 0000-0002-0201-0136
SPIN-code: 5861-9531
Scopus Author ID: 6507961979

MD, Cand. Sci. (Med.), Research Associate, Laboratory of Neuroimmunology

Russian Federation, Saint Petersburg

Galina G. Shkilnyuk

N.P. Bechtereva Institute of the Human Brain of the Russian Academy of Sciences

Email: galinakima@mail.ru
ORCID iD: 0000-0001-7175-668X
Scopus Author ID: 57193109310
ResearcherId: AAZ-3672-2020

MD, Cand. Sci. (Med.), Research Associate, Laboratory of Neuroimmunology

Russian Federation, Saint Petersburg

Irina N. Abdurasulova

Institute of Experimental Medicine

Author for correspondence.
Email: i_abdurasulova@mail.ru
ORCID iD: 0000-0003-1010-6768
Scopus Author ID: 22233604700
ResearcherId: J-6887-2018 H-3

Cand. Sci. (Biol.), Head of Laboratory, Pavlov Department of Physiology

Russian Federation, 12, Academician Pavlov Str., Saint Petersburg, 197376

References

Supplementary files