Additional file 1 of The intestinal microbiome and metabolome discern disease severity in cytotoxic T-lymphocyte-associated protein 4 deficiency

Supplementary Material 1: Figure S1. Correlations between clinical parameters associated with gastrointestinal (GI) manifestations in patients with CTLA4 deficiency. Correlogram for clinical parameters in patients with CTLA4 deficiency. Circle values=coefficient of correlation (r value); circle size=strength of significance (red=positive correlation, blue=negative correlation, blank=no significant correlation). All presented r values have p<0.05. Figure S2. Alterations in phylum and genus abundances in patients with CTLA4 deficiency from NIH and CCI cohorts. (A) Heatmap of components of the core microbiome at the genus level that are detected in high fractions in CTLA4 deficiency groups (20% of the sample prevalence cut-off) (yellow = low prevalence, purple = high prevalence) (A1 = NIH cohort; A2 = CCI cohort). The generalized linear models (GLM) to find associations between microbial features and CTLA4 deficiency identified the phyla (B) and genera (C) that are significantly different in CTLA4 deficiency groups compared to healthy individuals. All comparisons for the genera are significant with p<0.05, unless a p-value is shown. Figure S3. Phylum- and genus-level differences in CTLA4 deficiency in the NIH cohort. Comparisons are provided for groups of patients with CTLA4 deficiency from the NIH cohort with different degrees of disease severity (Healthy n=16, Mild n=7, Severe No GI n= 6, Severe GI n= 19). (A) Box and violin plots indicating phylum abundances in each group. The name of the phylum is indicated in the top of each panel with the p-values for each comparison shown in the graph. Wherever the p-value is <0.05, the significance is marked with an asterisk (* = p<0.05, **=p<0.01, ***p<0.001). (B) Heat trees depicting the significant differential abundances (p<0.05) of bacterial genera between patients with CTLA4 deficiency with Severe versus Mild disease, and (C) Mild disease versus Healthy (red = higher abundance; blue = lower abundance). Figure S4. Distinct functional profiles in patients with CTLA4 deficiency. Heatmap of significantly different functional profiles inferred by PICRUSt2 performed to identify the pathways associated with changes in the microbiome in CTLA4 deficiency (blue represents higher abundance and yellow represents lower abundance). The relative abundance normalized to a Z-score was used to generate the heatmaps. Pathway comparisons of the CTLA4 deficiency group with Healthy are shown for the NIH cohort (A) (Healthy n=16, Mild n=7, Severe No GI n=6, Severe GI n=19) and the CCI cohort (B) (Healthy n=23, Mild n=9, Severe No GI n=4, Severe GI n=10). Figure S5. Phylum- and genus-level differences in patients with CTLA4 deficiency and a history of gastrointestinal (GI) manifestations from the NIH and CCI cohorts. Comparisons are provided for groups of patients with CTLA4 deficiency (CTLA4-D) from the NIH (A1, B1, C1, D1) and CCI cohorts (A2, B2, C2, D2) with a history of GI disease (NIH Cohort: No GI history n=9, YES GI history n=23; CCI Cohort: No GI history n=11, YES GI history n=14). (A) Phylum distribution and (B) principal coordinates analysis (PCoA) plot of beta diversity based on the Bray Curtis metric with p-values determined by analysis of similarities (ANOSIM). (C) Differentially abundant genera and (D) linear discriminant analysis (LDA) scores determined by the LDA effect size (LEfSe) analysis showing biomarkers at the genus level. Box plots of log-transformed counts for select genera are shown on the right. Figure S6. Differences in alpha and beta diversity measures in NIH and CCI cohorts based on clinical characteristics in the CTLA4 deficiency groups. Heat table with p-values listed for comparisons of alpha (Chao1, Shannon, Simpson, Fisher) and beta (ANOSIM, Permanova, Permdisp) diversity indices based on characteristics of patients with CTLA4 deficiency in the NIH (A) and CCI (B) cohorts. The darker the pink color, the higher the significance. Orange to yellow shades represent p-values between 0.05 and 0.08 (the lighter the color, the lesser the significance). ANOSIM tests whether distances between are greater than within groups. Permanova tests whether distances differ between groups. Permdisp calculates an F-statistic to assess whether the dispersions between groups is significant. Figure S7. Mechanism of inhibition of T-cell inflammation by abatacept (CTLA4 fusion protein), and sirolimus (mTOR inhibitor). Abatacept, a fusion protein of the Fc fragment of IgG1 and extracellular domain of CTLA4, binds to CD80/86 (B7.1. / B.7.2) in antigen presenting cells (APC) or B-cells, and prevents interaction with the CD28 receptor. Thus, it blocks the secondary signal required for immune cell activation following T-cell receptor (TCR) and Major Histocompatibility Complex (MHC)-II binding, thereby reducing T-cell activation and infiltration (left). The mammalian target of Rapamycin complexes (mTORC1 and mTORC2) are activated upon T-cell activation, growth factor or nutrient signaling, and trigger the 4EPB1 (Eukaryotic translation initiation factor 4E [eIF4E]-binding protein 1) and S6 kinase 1 (S6K1) pathways, and protein kinases Akt and PKCa involved in T-cell transcription, protein synthesis and cell cycle regulation. Sirolimus forms a complex with FKBP12 (FK506-binding protein), targets mTORC1 and mTORC2, and inhibits downstream pathways and associated functions (right).