The role of multiple sclerosis therapies on the dynamic of human gut microbiota

,


Introduction
The human gut microbiota, composed of a large diversity of microorganisms that reside inside our gastrointestinal tract, is taxonomically classified into species, genus and phyla, containing both beneficial and pathogenic microbes (Rinninella et al., 2019;Camara-Lemarroy et al., 2018). It is mostly shaped in early life of childhood depending on several factors, such as the type of delivery, milk feeding, weaning and medication. After the first three years of life, it reaches a partial stability, with continuous subtle changes caused by external factors: diet, exercise, body mass index, environmental and medication exposure (Rinninella et al., 2019). A microbiome is formed of the total genes, proteins, metabolites of all organisms (Maglione et al., 2021).
Multiple sclerosis (MS) is an autoimmune disease of the central nervous system, where genetic susceptibilities meet environmental factors. It includes impaired blood-brain barrier integrity, inflammation, demyelination, oligodendrocyte loss, gliosis and axonal degeneration. The disease is considered of three main stages, a pre-clinical stage where the combination of genetics and environmental factors could be a pathological trigger, an inflammatory stage where we encounter episodes of neurological dysfunction, such as optic neuritis, pyramidal, cerebellar or brainstem impairment, sensory symptoms, bladder dysfunction, and a final neurodegenerative stage of disease progression when we find a progressive disability, mostly with patient's gait ability. The disease has different clinical forms: clinically isolated syndrome (CIS), relapsing-remitting MS (RRMS), secondary progressive MS (SPMS) and primary progressive MS (PPMS). The modifiable risk factors connected to MS have been strongly analysed in order to find a specific role in the pathological process, including microbiome composition (Baecher-Allan et al., 2018). Recent studies outline that MS patients have a distinct microbiome composition compared to healthy controls (HC) (Mirza et al., 2020).
Due to the fact that the microbiome is unique for each individual and contains an enormous variety of microorganisms, it is difficult to describe a specific healthy microbiota (Rinninella et al., 2019). A healthy microbiota provides many essential functions, including gut permeability and motility, vitamins synthesis, absorption, but also the development of innate immune system (Chu et al., 2018). Based on this role in immunomodulation, recent studies indicate a possible connection between the microbiome and some neurological immune-mediated pathologies (Mirza et al., 2020).
The concept of gut microbiota-brain axis refers to different communication mechanisms between the gastrointestinal and nervous system (Strandwitz, 2018). Microbiota can influence the central nervous system (CNS) by different pathways. Firstly, it relates to the brain networks, it modulates different neurotransmitters (such as gamma aminobutyric acid, serotonin, dopamine, histamine) and it is connected with the sympathetic or parasympathetic nervous system, especially the vagus nerve. Secondly, it is involved into the endocrine pathway, as a stress response. The corticoids released by the hypothalamic-pituitaryadrenal axis can alter the microbiota composition and increase intestinal permeability. Another mechanism of which the microbes can alter the nervous system is immunoregulation. This includes impaired antigen presentation, cytokines and lymphocytes production, differentiating different types of T cells in the gut-associated lymphoid tissue (GALT) (Chu et al., 2018). Gut microorganisms can release metabolites (lipopolysaccharide, peptidoglycan) and the short chain fatty acids (SCFAs) metabolism is involved in microglial cells activation and the blood-brain barrier leakage (Strandwitz, 2018).
Complementary and alternative medicine, including homeopathy, is mostly associated with conventional forms of therapy, to relieve general symptoms, pain, spasticity, sensitivity problems, walking impairment, fatigue, memory loss. Of all MS patients, up to 30% use or have used a form of complementary and alternative medicine (Kim et al., 2018). It is estimated that 70-80% of MS patients presented an improvement in their general state (Olsen, 2009). Homeopathy's principle "let like be treated by like" is based on the idea that an illness can be treated by substances which trigger the same pathological symptoms in healthy individuals (Fisher and Ernst, 2015). Homeopathic remedies are highly diluted natural substances (Manzalini and Galeazzi, 2019), a more diluted solution is considered to have a higher potency (Chikramane et al., 2010). The potency choice is related to the health level: a lower health level implies a low potency, such as 30C, for a longer period of time, up to 3 months, while a 200CH treatment is limited for a few days. They are prescribed on a case-by-case basis, taking into account other aspects apart from specific medical signs, as well as particular psychological details or relevant daily habits of the patient (Whitmarsh, 2003). Some symptoms in MS are frequently ameliorated by some substances: Causticum for urinary disfunction, Phosphorus for optic neuritis, Cuprum metallicum and Nux vomica for spasms, Secale for sensory symptoms (Whitmarsh, 2003).
As presented further on in this manuscript, numerous studies outlined that MS patients have a distinct microbiota compared to HC. It was observed that MS treatments can modify the gut microbiome, but homeopathy has not been studied in this field.
The aims of this study are to investigate if MS patients' gut microbiota changes due to immunomodulatory drugs or complementary homeopathic treatments, including homeopathic treatment and to assess the differences in microbial taxonomy and diversity between MS and HC.

Subjects
The study was prospective, longitudinal, analytical, observational and case-control type. It included 50 patients diagnosed with MS, from the Neurology Department in The Emergency Hospital Cluj-Napoca, between January 2019 to May 2020. Twenty-one HC were recruited. The ethical authorization was obtained (number 2394/28.01.2020, Emergency Hospital Cluj-Napoca) and each subject signed an informed consent.
The MS group included adult patients with confirmed relapsingremitting MS (RRMS) or clinically isolated syndrome (CIS), according to the revised McDonald diagnosis criteria (Thompson et al., 2018), with an EDSS score of maximum 5 points, who haven't undergone a disease modifying therapy (DMT) a year prior to study enrolment. We excluded patients with progressive MS, unclear diagnostic criteria, active gastrointestinal pathologies, who were pregnant, breastfeeding or who underwent long term treatment with probiotics, antivirals, antibiotics, non-steroidal anti-inflammatory drugs or proton-pumps inhibitors in the past 6 months. Of all patients, 7 encountered transient mild gastric pathologies in the past, but with no active gastrointestinal symptoms at enrolment, 4 patients had abdominal surgeries in their medical history and 4 patients have endocrine pathologies (mostly thyroid related) under medical control. Other pathologies (hypertension, diabetes) were in isolated cases and clinically compensated at the study onset. The HC group was composed of healthy individuals without MS, matched for age and gender with the MS patients, not diagnosed with any pathology, and with no chronic treatment. All subjects declared that no major changes occurred in their diet or lifestyle during the study.
At the study onset, each subject provided a gut sample. After the first sample, they were prescribed a form of treatment depending on several factors, such as disease severity, patients' consent, comorbidities, drug safety and accessibility (Chikramane et al., 2010). Some refused a conventional therapy for the moment; they accepted however the homeopathic treatment. The second sample was obtained 2 months after the treatment initiation.
Regarding the homeopathic treatment distribution, each patient received a different individualized medication, according to their MS symptoms, including one of the following: Phosphorus, Lycopodium, Natrium muriaticum, Lac Caninum, Nux Vomica, Lachezis, Nitric acid, Rhus Toxicodendron, Tarentula Hispanica, Pulsatilla, Calcarea Carbonica, Sulphur, Ignatia, Aconitum or Causticum, with a concentration of either 20CH or 200CH. The substances were prescribed by a neurologist accredited in Homeopathy and they were taken daily, 7 granules sublingually, 3 consecutive days for the patients assigned with a concentration of 200CH and 1 to 3 months for those with the 30CH concentration.
Our MS patients were divided into separate groups, according to their treatment, as presented in Table 1. Group G-DMT received a disease modyfing treatment: intramuscular interferon beta 1a, 30 μg / 0.5 ml once a week (subgroup G-IFN) or oral teriflunomide 14 mg daily (subgroup G-TER). Group G-DMT + HOM received a DMT combined with homeopathy: G-IFN + HOM interferon beta1a and homeopathy, G-TER + HOM teriflunomide and homeopathy. Group G-HOM consisted in homeopathic treatment. Control group, G-HC was formed of HC, with no treatment.

Sample collection
All subjects provided fecal samples corresponding to the laboratory instructions: at any moment of the day, with no restrictions, using special stool containers. The provided samples were stored at minus 20 degrees Celsius and then shipped to the laboratory for the DNA extraction. Each subject collected two fecal samples: one before treatment (sample 1, S1) and one two months after starting the therapy or study enrollment (sample 2, S2).

The 16rRNA sequencing and statistical analysis
The metagenomic analysis of the universally present 16S subunit ribosomal RNA (rRNA) gene used specific hypervariable regions (V1-V3, V3-V4, ITS1&ITS2) throughout the Illumina MiSeq platform. The results are the OTUs (operational taxonomic units), microorganisms that have a DNA similarity of at least 97% to a laboratory database, classified into specific taxa, with their frequency and relative abundance.
Gut microbiota diversity is a mathematical measure of variability, characterized by richness (the number of different species) and evenness (the uniformity of different species). For alpha diversity, the diversity within the sample, the Chao index is an estimator of species richness. Shannon and Simpson are parameters for both richness and evenness, their values increase with the number of species and a more even distribution (Kim et al., 2017).
Using the frequency matrices, each combination of pairs was analysed using the Wilcoxon Rank-Sum Test. For beta diversity, the diversity between samples, we used the Bray-Curtis and the Jaccard indexes, where the graphical distance between them indicates the difference. Principal coordinate analysis (PCoA) was obtained from relative abundance matrices. Linear Discriminant Analysis Effect Size (LEfSe) figures use relative abundance matrices to calculate differential organism abundance between MS and HC, with an alpha value of 0.05 and a logarithmic linear discriminant analysis (LDA) score threshold of 2.0. All features have a score either lower than − 2.0 or larger than 2.0. The bacteria enriched for MS patients are represented in red, while those more abundant in HC are coloured in green. In the relative abundance stacked bars, each bar represents the average relative abundance of the top 25 species, selected according to the most abundant ones relative to both groups. Heatmaps were also generated.

Results
The MS group consisted of 31 females and 19 males, with a median age of 30.5±9.9 years, 44 of them being diagnosed with RRMS and 6 with CIS, a mean Expanded Disability Status Scale (EDSS) score of 1.8 points and a mean disease duration of 3.4 years. All the MS specific characteristics are presented in Table 2. The majority of them were naive (with no previous disease modifying treatment (DMT)), only 4 patients received treatment that was interrupted in the previous year.
The healthy control (HC) group was assembled considering people with similar demographic features as the MS group: 62% females and 38% males, a median age of 28±9.7 years.

Microbiome diversity
We studied overall differences in microbial composition by using alpha and beta diversity. There was no statistically significant difference in alpha diversity between MS patients and HC before (p = 0.85) and after the treatment (p = 0.95). Also, there was no difference between other group combinations for sample 2 (S2): G-DMT vs G-DMT + HOM (p = 0.89), G-DMT + DOM vs G-HOM (p = 0.98), G-DMT vs G-HOM (p = 0.64), and their subgroups, G = IFN + HOM vs G-TER + HOM (p = 0.23), G-IFN vs G-IFN + HOM (p = 0.06) between sample 1 (S1) and S2 for MS patients (p = 0.93), G-DMT + HOM (p = 0.38) or G-HOM (p = 0.79). Also, there was no difference in alpha diversity between the MS patients treated with DMT (G-DMT combined with G-DMT + HOM) and the patients not following DMT (G-HOM), p = 0.81. All p values are presented in Table 3.
As we analysed beta diversity, there was a difference for S2 between G-HOM and G-TER + HOM (p = 0.007) (Fig. 1) and between G-HOM and G-IFN (p = 0.012), but no statistically significant difference between the MS patients and HC or between the two samples for MS patients (all p > 0.05). Other results are outlined in Table 3

Taxonomic differences
The 16S analysis results include the frequency and relative abundance of the species.
Another point of interest in our research was to emphasize how the microbiome changes in time, from the moment the patients were prescribed a therapeutic scheme (S1) until two months after the treatment began (S2). For the MS cohort, the taxonomic changes are shown in Fig. 3 with the most abundant 25 organisms. The statistically significant differences between the two samples were more prevalent for Firmicutes phylum, as there were identified reduced levels for Ruminococcus

Discussion
Our data revealed no major changes in diversity between MS patients and HC, before or after any kind of treatment. Compared to HC, untreated patients presented an increase for Prevotella stercorea and decreased levels in Actinobacteria and Faecalibacterium prausnitzii. The taxonomic changes after two months of treatment were slightly different, with increased Gemella levels and decreased Ruminococcus ones for MS subjects. Comparisons between groups after treatment emphasize different taxonomic modifications. For example, patients treated with homeopathic treatment had an increase for Faecalibacterium prausnitzii, Akkermansia muciniphila and Bacteroides compared to combination therapy. Also, there were beta diversity significant results for treatment with homeopathy versus homeopathy combined with teriflunomide and also for homeopathy versus interferon beta1a. Analyzing changes that have occured in time, between the two samples, we noticed a decrease of Ruminococcus, Lachnospiraceae and Eubacterium oxidoreducens levels among MS treated patients compared to their baseline sample. Homeopathy induced in time the reduction of Eubacterium exidoreducens.
There is a consensus that a healthy microbiota is characterized by a balance between the microorganisms and the host, with a large diversity, resilience and stability, in order to maintain the host homeostasis and immune functions (Freedman et al., 2018). Although external factors can rapidly modify the microbiota, healthy stable bacteria return to their original composition and change only with persistent habits. A higher biodiversity means higher differences in species and thus, more biological functions they can exert, a better stability, the ability to resist to changes and to recover (Riccio and Rossano, 2018). Dysbiosis refers to an imbalance in bacterial composition, with an increase in harmful microorganisms and a decrease in the beneficial species with a change towards an inflammatory state (Freedman et al., 2018).
We notice in our MS cohort's microbiota profiles an increase in Firmicutes and Actinobacteria and a decrease in Bacteroidetes phylum. These Table 3 Overview of all p values for microbial diversity. changes are specific for a typical Western diet high in fats and sugar, where Firmicutes are more capable to extract energy from food, promoting weight gain. On the other hand, a diet based on complex carbohydrates, rich in fibers promotes the increase of Bacteroidetes and benefic metabolites (Magne et al., 2020). MS therapies could modulate the gut microbiota, but evidence is not well established in human studies so far. Interferon beta inhibits T cells and proinflammatory cytokines, stimulates Tregs and suppressive B cells, modulates the interaction between the microbes and epithelial cells and stabilize the intestinal barrier by upregulating tight junction proteins in endothelial cells. Teriflunomide inhibits dihydroorotate dehydrogenase, pyrimidine synthesis and proinflammatory cytokines and could influence the gut microbiome by suppressing the STAT-6 signaling pathway, increasing specific T reg cells (Camara-Lemarroy et al., 2018;Baecher-Allan et al., 2018).

Comparison between MS patients and HC at baseline
In our study, none of the alpha or beta diversity metrics differed significantly between the MS patients and the HC, before initiating treatment. This suggests that the microbiota of the MS patients doesn't show sign of unusual trends, corresponding with literature research (Mirza et al., 2020). At the taxonomy level, we observed several organisms which had a significant different relative abundance between the MS cases and the HC. Actinobacteria is a beneficial microorganism which has a role in the formation of the immune system (Adamczyk-Sowa et al., 2017) and it is lower among MS patients. Bifidobacterium is decreased among our MS cohort. The literature provides conflicting data regarding its role in immune diseases, but most studies have shown that Bifidobacterium induces an anti-inflammatory immune response (Budhram et al., 2017;Tankou et al., 2018). Prevotella genus is a well-studied bacteria in MS, implicated in the phytoestrogen metabolism, generally accepted as less prevalent in MS and more present for treated patients (Chen et al., 2016;Brown et al., 2021). Our MS patients presented higher levels of one species compared to HC, before or after treatment, Prevotella stercorea, but no relevant data on the Prevotella genus. This is probably due to the fact that this genus has several species with different roles (Mirza et al., 2020). Some bacterial metabolites can directly influence the central nervous system, such as short-chain fatty acid (SCFA) which has an immunosuppressive role in gut mucosa. Faecalibacterium prausnitzii and Bacteroides coprophilus participate in SCFA metabolism (Freedman et al., 2018). Bacteroides is a beneficial bacterium promoting IL-10 and is generally present in lower levels in MS cases. Faecalibacterium prauznitzii is established as a marker for a healthy microbiota and it is depleted among MS patients, as it can be observed in literature (Tremlett et al., 2016) and also in our cohort. All these differences in bacterial taxonomy compared to the HC suggest a dysbiosis for MS microbiota.

Comparison between groups after treatment
When analyzing if the treatment (of any kind) for MS significantly modifies the microbiome, we outlined that there is no major change in alpha or beta diversity for the MS patients compared to the HC, in the two months after beginning any form of therapy, as suggested by literature research (Mirza et al., 2020). Regarding the taxonomy changes, Ruminococcus is a beneficial bacterium and it is usually restored in MS cases after DMT , our data reflects a lower relative abundance among the MS patients compared to the HC for the second sample. Clostridium is a butyrate producing bacteria involved in the production of regulatory T cells and the IL-10 anti-inflammatory cytokine (Miyake et al., 2015) and it is reduced among MS patients after treatment.
The differences between groups after two months of therapy outline a possible intervention of confounders: DMT or homeopathic treatment on gut microbiome. Homeopathy might have an influence on gut microbiota, as our data suggests. Changes between patients who received any form of DMT (G-DMT) and a combination of DMT and homeopathy (G-DMT + HOM) are subtle, but we notice a large number of significant differences regarding the bacterial relative abundance between subgroups, with changes partially inconsistent and contradictory. The pair G-IFN and G-IFN + HOM outlines a possible influence of homeopathy on interferon beta1a treated patients, while the G-TER and G-TER + HOM pair suggests its influence on teriflunomide patients.
Compared to combination therapy, patients receiving homeopathy presented an increase in Akkermansia muciniphila, a controversial Bray-Curtis diversity value quantifies the difference between the group with homeopathy and teriflunomide and the group with homeopathy, after two months treatment. The figure on the right shows a 95% confidence ellipse for each group. PERMANOVA test results are included into the table. species, with an unknown role in immune maturation, elevated among untreated MS patients (Cekanaviciute et al., 2017). In vitro experiments outline its pro-inflammatory role, but in vivo studies have not reproduced these effects, but the opposite, it might have a benefic function for some metabolic disorders. Although the therapeutic choice did not influence alpha diversity for any group, we found isolated differences in beta diversity, similar to our literature research (Mirza et al., 2020). Our findings were related to homeopathy and interferon beta1A as treatments that could modify the microbiome diversity. The beta diversity of the group who received homeopathic treatment was different than the  group treated with a combination of homeopathy and teriflunomide. Also, the homeopathy-treated group had a different beta diversity than the group treated with interferon beta1A, all p values are presented in Table 3. With a statistically unsignificant p value of 0.06, close to 0.05, the group with interferon treatment had a minor change in beta diversity compared to the interferon and homeopathy group. These results suggest that homeopathy might have a role in the beta diversity of MS patients, but further studies are required.
Analyzing relative abundance in the selected phyla and species, we notice a decrease for Firmicutes for the group receiving the combination therapy compared to other groups, but not statistically significant. Firmicutes are related to shorter relapse times in MS (Sand and Baranzini, 2018), thus combining these two therapies could be beneficial in delaying the disease progress.
The two beneficial bacteria involved in the SCFAs metabolism, Bacteroides and Faecalibacterium prauznitzii, have similar relative abundances for the HC and the homeopathy group, of approximately 12% for Bacteroides and respectively 7% for Faecalibacterium. This suggests that homeopathic treatment might have a role in maintaining the balance of these healthy microorganisms.

Comparison in time (between the second and the first sample)
Microbiome changes in time were analysed for the MS cohort, where the results outline the effects of all treatments applied in our study: DMT, homeopathy or a combination of both. MS treated patients presented more diverse bacterial changes, from multiple phyla, which suggests that treatment may have a positive influence on microbiota in general. Patients had a lower relative abundance of Lachnospiraceae, Ruminococcus, Eubacterium oxidoreducens after treatment, as expected by literature research (Mirza et al., 2020). Lachnospiraceae is involved in the reduction of mucosal permeability through SCFAs production and increased expression of tight junction proteins in epithelial cells (Rinninella et al., 2019).
For the two groups including homeopathy, G-DMT + HOM and G-HOM, the bacteria more prevalent before treatment were attenuated in the second sample. This suggests that homeopathic treatment or its combination with DMT had an impact on gut microbiota after two months of treatment.
Homeopathic treatment has not been studied on gut microbiota among MS patients and no relevant studies were found to compare our findings with. Our patients who received only a homeopathic treatment presented a relative stability after 2 months of treatment, with the only statistically significant modification of Eubacterium oxidoreducens, less abundant in the second sample. Also, in the group treated with a combination of therapies, there are a few bacterial species that were less abundant in the second sample. These results suggest that homeopathic treated patients might have a more stable microbiome.

Strengths and limitations
Our study has a contribution in this medical field because it analyses gut samples in dynamics, comparing not only treatment exposure but also the changes that occur in time. Also, it is unique by also introducing homeopathic treatment as a possible con founder.
As possible limitation factors, all environmental and dietary compounds that can interfere and modify the microbiota in two months cannot be excluded or quantified. Further studies are needed.

Conclusions
MS microbiota is characterized by dysbiosis, with various taxonomic changes compared to HC. In untreated MS patients we find a higher abundance of Prevotella stercorea and a reduction of Actinobacteria, Bifidobacterium and Faecalibacterium prauznitzii. Treatment with interferon beta1a, teriflunomide or homeopathy implied several taxonomic changes. MS treated patients had a different microbiota compared to HC, with reduced Bifidobacterium, Ruminococcus and Clostridiales and a higher prevalence of Gemella, Megaspherae and Prevotella stercorea.
Comparing the second with the first sample, we outlined the modifications in MS microbiota in time, such as decreased Lachnospiraceae and Ruminococcus and increased Enterococcus faecalis. These dynamic changes may be due to MS treatment.
Overall, treatment exposure had a small influence on microbiome diversity. The only statistically significant diversity modifications we found implied the effect of homeopathic treatment on beta diversity indexes, compared to the effect of immunomodulatory drugs. Eubacterium oxidoreducens was also reduced after homeopathic treatment. Our results suggest that homeopathic treatment has also a role in influencing the gut microbiota.
This study outlines the important role of gut microbiota in MS and brings up further pathways to be studied. The microbiome is a vast ecosystem, influenced by numerous factors which cannot be completely quantified. Our project focused on dynamic changes in microbiota and the possible influence of some specific DMTs, as well as homeopathy as a complementary and alternative medicine used in clinical practice.

Funding
This research received no external funding.

Ethical approval
The study was conducted in accordance with the Declaration of Helsinki, and approved by Cluj Emergency County Hospital Ethics Committee, code 2394/28.01.2020.

Informed consent statement
Informed consent was obtained from all subjects involved in the study.

Declaration of Competing Interest
The authors declare no conflict of interest.

Data availability
The data and other information of this study are available from the corresponding author upon request, due to privacy issues.