Analysis techniques for Treg residency using single-cell RNA-seq

Identification: Gomes, Tomás

Analysis techniques for Treg residency using single-cell RNA-seq

Ricardo J Miragaia1,2, Tomás Gomes1, Agnieszka Chomka3,4, Laura Jardine5, Angela Riedel6, Thomas Krausgruber3,4, Ahmed Hegazy3,4, Jacqueline Shields6, Muzlifah Haniffa5, Fiona Powrie3,4, Sarah Teichmann1

1Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SA, UK; 2Centre of Biological Engineering, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal; 3Kennedy Institute of Rheumatology, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Roosevelt Drive, Oxford OX3 7FY, UK;4Translational Gastroenterology Unit, Experimental Medicine Division Nuffield Department of Clinical Medicine, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK; 5Institute of Cellular Medicine, Newcastle University, UK; 6MRC Cancer Unit, University of Cambridge, Cambridge, UK

Cellular heterogeneity is a key aspect of the adaptive immune system for response to pathogens and homeostasis maintenance between tissues. T regulatory (Treg) cells have crucial roles in homeostasis, and the differences and relationships between resident Treg populations in different tissues are not fully understood. To explore and characterize different resident T cell populations, we generated scRNA-seq data for Tregs and T memory (Tmem) cells from various tissues of mouse and human individuals.

To define genes characteristic of cell clusters, we compared marker gene prediction methods based on either linear modelling, random forests or Shannon entropy. These allowed us to retrieve previously described residency markers and highlight novel ones. We used this methodology to compare cell populations within and between tissues, ranking genes by their relevance in defining tissue Treg and Tmem in both species.

From marker genes in each cell population we constructed gene co-expression and regulatory networks, which present an integrated characterization of these cells. Finally, we compared populations between human and mouse, to assess evolutionary conservation of genes and pathways in immune cell residency.

To characterize differences between tissues in immune response, we compared scRNA-seq from mouse Treg and Tmem cells collected from the skin and lymphoid tissues in the presence or absence of melanoma. We derived cell response from latent variables using Bayesian Gaussian process latent variable modelling (BGPLVM). Contrasting the melanoma challenge model with the changes in steady state cells exposed key differences between modes of T cell activation.

In summary, single cell RNA-seq combined with high precision methods can provide an accurate picture of T cell residency in different tissues and species.


Credits: None available.