Research ArticleHEALTH AND MEDICINE

High frequency of intestinal TH17 cells correlates with microbiota alterations and disease activity in multiple sclerosis

See allHide authors and affiliations

Science Advances  12 Jul 2017:
Vol. 3, no. 7, e1700492
DOI: 10.1126/sciadv.1700492
  • Fig. 1 Study design.

    Tissue samples from the duodenal mucosa were collected during EGD from RRMS patients (n = 19) and HCs (n = 17). Simultaneously, we collected 6 to 7 ml of peripheral blood from the same individuals. Total lymphocytes were isolated from one to two fragments of intestinal tissue by collagenase digestion and from peripheral blood by Ficoll gradient. Mucosal immune cells and peripheral blood mononuclear cells (PBMCs) were FACS-analyzed for immunological profiling of the following TH cell subsets: IL-17A+CD4+ TH17 cells, IFN-γ+CD4+ TH1 cells, and IL-22+ TH22 cells. Microbial DNA was extracted from frozen intestinal tissue samples (seven RRMS and seven HC samples), and 16S rRNA sequencing was performed on the Roche 454 platform. At the 2-year end point, we stratified RRMS patients in two cohorts on the basis of evidence of disease activity by clinical and MRI criteria. RRMS patients with no evidence of disease activity in the 2-year follow-up were classified as RRMS/NEDA (n = 9), whereas RRMS patients with evidence of disease activity were classified as RRMS/EDA (n = 10). SSCA, side scatter FACS parameter.

  • Fig. 2 Intestinal T cell subsets in RRMS patients and HCs.

    (A) Relative percentages of CD3+, CD4+, and CD8+ cells out of total lymphocytes isolated from the small intestinal mucosa of individuals with RRMS and of HCs. (B) Frequencies of different TH cell subsets (TH17, TH1, and TH22) in the intestinal mucosa (GUT) and PBMCs of HCs and RRMS patients. Blood samples were collected from the same individuals at the time of intestinal tissue sample collection. Data are means ± SEM. (C) RRMS patients were divided into two cohorts on the basis of disease activity. RRMS patients with evidence of disease activity at 2-year follow-up were classified as EDA, whereas patients with no disease activity were classified as NEDA. Percentages of intestinal TH17 cells out of total CD4+ T cells in HCs and RRMS/NEDA and RRMS/EDA patients are shown. (A to C) **P < 0.01 and *P < 0.05 by unpaired t test. (D) Logistic regression analysis showing correlation between TH17 cell percentages in the gut mucosa and predicted probability of showing disease activity at 2-year follow-up (NEDA score: RRMS/NEDA, 1; RRMS/EDA, 0). Odds ratio, 0.96; 95% CI, 0.02 to 0.98; P = 0.01.

  • Fig. 3 Differences of gut mucosa-associated microbiota composition between RRMS patients and HCs.

    (A) α-Diversity of gut microbiota profiles assessed by 16S amplicon sequencing of DNA purified from human small intestinal tissue of HCs and RRMS patients with active (RRMS/EDA) and nonactive (RRMS/NEDA) disease. (B) Average phylum-level composition of microbiota isolated from small intestinal samples of HC and total RRMS, RRMS/NEDA, and RRMS/EDA patients. (C) Means ± SEM of the relative abundance of different phyla. *P < 0.05 by unpaired t test. (D) Average genus-level composition in four cohorts of individuals. (E) Left: Means ± SEM of the relative abundance of Prevotella and Streptococcus strains. *P < 0.05 by unpaired t test. Right: Regression linear plot in all individuals (RRMS + HC) showing inverse relation between percentages of intestinal TH17 cells and relative abundance of Prevotella strains.

  • Table 1 Demographics of study population.

    All individuals recruited for the study underwent EGD for diagnostic purposes. Exclusion criteria for both HCs and RRMS patients were treatment with antibiotics or corticosteroids in the 3 months before EGD and history of gastroenteritis, gastric ulcer, irritable bowel disease, celiac disease, inflammatory bowel disease, and gastric and colorectal cancers. All patients recruited in this study have a diagnosis of RRMS defined according to Poser’s diagnostic criteria. None of the RRMS patients received corticosteroids in the 3 months before the EGD. The mean age of healthy individuals was slightly higher than that of RRMS patients, but our correlative analysis did not reveal any association between age and TH17 cell percentages and Prevotella/Streptococcus relative abundance. All samples analyzed were donated for research purposes, and each subject signed an informed consent. Data are means ± SD, unless otherwise indicated. RR-NEDA, RRMS/NEDA; RR-EDA, RRMS/EDA.

    CohortnAge (years)Age (mean)Sex (women/men)Other autoimmune diseases
    Healthy1727–7448 ± 310/7NA
    RRMS1925–5741 ± 211/8Autoimmune thyroiditis (1/19)
    RR-NEDA926–5640 ± 47/2
    RR-EDA1025–5741 ± 34/6
  • Table 2 Clinical data.

    All patients recruited in this study have a diagnosis of RRMS defined according to Poser’s diagnostic criteria. None of the RRMS patients received corticosteroids in the 3 months before the EGDs (time of intestinal biopsy collection). RRMS patients were divided into two subgroups on the basis of evidence of disease activity at 2-year follow-up: RRMS patients with no evidence of disease activity in the 2-year follow-up were classified as NEDA (n = 9), whereas RRMS patients with evidence of disease activity were classified as EDA (n = 10). All RRMS patients recruited in this study were undergoing different therapeutic immunomodulatory regimens (IFN-β, GA, and fingolimod) at the time of EGD and intestinal tissue sample analysis.

    MS patient IDSexAge (years)Disease activity (2-year follow-up)EDSSCurrent therapy
    MS001F38NEDA3Fingolimod
    MS002M49EDA3IFN-β
    MS003F27EDA2IFN-β
    MS004F49NEDA5.5GA
    MS005F38EDA1.5IFN-β
    MS006M42NEDA2GA
    MS007M43EDA1.5GA
    MS008M37EDA2Fingolimod
    MS009M35EDA4GA
    MS010M57EDA4GA
    MS011F55NEDA1.5GA
    MS012F26NEDA5GA
    MS013F29NEDA1IFN-β
    MS016M27NEDA1.5IFN-β
    MS017F42NEDA1.5IFN-β
    MS018F57EDA3.5Fingolimod
    MS020F25EDA1GA
    MS021F56NEDA2IFN-β
    MS022M40EDA1GA

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/7/e1700492/DC1

    fig. S1. Percentages of double-positive IL-17+IFN-γ+ and IL-17+IL-22+ T cells in the intestinal mucosa of HCs and RRMS patients with active disease (RRMS/EDA) or inactive disease (RRMS/NEDA).

    fig. S2. Percentages of intestinal TH cell subsets in RRMS patients stratified according to immunomodulatory treatment.

    fig. S3. Correlative statistical analysis between percentages of intestinal TH17 cells and age, EDSS score, and disease duration.

    fig. S4. Characterization of Streptococcus species by amplification and sequencing.

    fig. S5. Correlative analysis of relative abundance of Prevotella and Streptococcus strains in mucosa-associated microbiota and age, EDSS score, and disease duration.

    fig. S6. Relative abundance of Prevotella and Streptococcus strains in HCs and RRMS patients stratified according to immunomodulatory treatment.

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. Percentages of double-positive IL-17+IFN-γ+ and IL-17+IL-22+ T cells in the intestinal mucosa of HCs and RRMS patients with active disease (RRMS/EDA) or inactive disease (RRMS/NEDA).
    • fig. S2. Percentages of intestinal TH cell subsets in RRMS patients stratified according to immunomodulatory treatment.
    • fig. S3. Correlative statistical analysis between percentages of intestinal TH17 cells and age, EDSS score, and disease duration.
    • fig. S4. Characterization of Streptococcus species by amplification and sequencing.
    • fig. S5. Correlative analysis of relative abundance of Prevotella and Streptococcus strains in mucosa-associated microbiota and age, EDSS score, and disease duration.
    • fig. S6. Relative abundance of Prevotella and Streptococcus strains in HCs and RRMS patients stratified according to immunomodulatory treatment.

    Download PDF

    Files in this Data Supplement:

Navigate This Article