202-OR: Automated Cluster Analysis by Machine Learning Reveals Important Dissimilarities in the Immune Phenotype of T1D, MS, and SLE
Background: Dysfunction of the immune cell control and break-down of self-tolerance are important in chronic conditions such as type 1 diabetes (T1D), multiple sclerosis (MS), rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE). Investigating common patterns and dissimilarities between t...
Saved in:
Published in | Diabetes (New York, N.Y.) Vol. 72; no. Supplement_1; p. 1 |
---|---|
Main Authors | , , , , , , |
Format | Journal Article |
Language | English |
Published |
New York
American Diabetes Association
20.06.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | Background: Dysfunction of the immune cell control and break-down of self-tolerance are important in chronic conditions such as type 1 diabetes (T1D), multiple sclerosis (MS), rheumatoid arthritis (RA) or systemic lupus erythematosus (SLE). Investigating common patterns and dissimilarities between these diseases can help identifying new therapeutic intervention strategies for T1D.
Methods: Unsupervised, automated gating and machine learning (ML) was used to identify clusters of immune cell phenotypes in fluorochrome stained PBMCs from patients with T1D (n=70), RA (n=70), SLE (n=50), MS (n=60) and healthy controls (n=70). The pipeline includes unsupervised pre-gating, FlowSOM clustering, and a statistical analysis (MANOVA) to check for significant differential abundances of cell populations among the conditions.
Results: The unsupervised analysis revealed important differences in the immune cell phenotype of peripheral CD4posT cells between the different auto-immune conditions. IL7alpha receptor (CD127) positive T cell clusters, such as CD3hiCD4hiCD15sposCD127hiCD161med cells are significantly decreased in T1D (p<0.001) compared to MS. Clusters of IL7Rpos memory T cells, expressing the co-stimulatory CD28 molecule are significantly increased in T1D (p<0.001) and SLE (p<0.001) compared to MS. Clusters including Treg-like phenotypes are significantly decreased in MS (p<0.001) compared to T1D, SLE and RA.
Discussion: IL-7R (CD127) and its gene polymorphism plays a major role in MS and mediates homeostatic proliferation in T1D. In SLE serum sIL-7R is associated with disease activity. Using unsupervised machine learning we were able to identify interesting clusters of immune cells expressing CD127 that show significantly different abundancies. Learning from immune modifying therapies used in MS and SLE treatment could help to develop new strategies for future T1D (combinational)-therapies targeting IL7/IL7R complexes.
Disclosure
B.Prietl: None. J.Vera-ramos: None. V.Pfeifer: None. M.Khalil: Advisory Panel; Novartis, Biogen, Gilead Sciences, Inc., Merck & Co., Inc., Research Support; Novartis, Biogen, Roche Pharmaceuticals. L.Herbsthofer: None. H.Sourij: Advisory Panel; Amgen Inc., Amryt Pharma Plc, Boehringer Ingelheim Inc., Eli Lilly and Company, Novo Nordisk A/S, Sanofi, Speaker's Bureau; Boehringer Ingelheim Inc., Novo Nordisk A/S, Bayer Inc., Sanofi. T.Pieber: Advisory Panel; Arecor, Novo Nordisk A/S, Consultant; Lilly, Research Support; AstraZeneca, Novo Nordisk A/S, Sanofi, Speaker's Bureau; Roche Diagnostics.
Funding
JDRF/Lupus Research Alliance/National MS Society Joint RFA (2-SRA-2021-1043-S-B) |
---|---|
ISSN: | 0012-1797 1939-327X |
DOI: | 10.2337/db23-202-OR |