Session 3: Lessons From Other Diseases

Jenna Pappalardo, David van Dijk, Le Zhang, Maggie Pecsok, Smita Krishnaswamy and David A. Hafler
Departments of Neurology, Immunobiology, and Genetics, Yale School of Medicine, New Haven, CT
The peripheral immune system is becoming increasingly appreciated as an integral component of tissue homeostasis. Tissues develop unique immune states through the regulated recruitment and retention of specific subsets of immune cells which then become further specialized based on surrounding tissue cues. Here, we explore the continuum of T cell states present in the peripheral blood and cerebrospinal fluid (CSF) of healthy individuals to define how the central nervous system (CNS) shapes T cell profiles and function using Seq-Well as a platform for massively-parallel single-cell RNA sequencing. We leveraged the natural progression of the single-cell data represented by Potential of Heat-diffusion for Affinity-based Trajectory Embedding (PHATE) to generate a tissue score that provided the basis for determining the acquisition of tissue-specific characteristics. We validated that this approach mirrored current knowledge about the state of T cells in CSF, including the increased expression of the brain-homing trafficking molecules ITGA4 and CXCR3. Moreover, we have identified metabolic pathways and other pathways necessary for maintaining brain homeostasis are enriched in CSF T cells. Further in vitro modeling of factors in the CNS that may shape T cell function are elucidating how these alterations converge on altering T cell function. Finally, this atlas of healthy CSF allows us to examine perturbation of the system by CNS autoimmune disease and how those perturbations can be corrected by immunotherapy. Thus, these findings will inform our understanding of immune-tissue interactions in the healthy brain as a critical basis for contextualizing perturbations in neuroinflammatory and neurodegenerative disease.

Multiple sclerosis (MS) is the commonest non-traumatic cause of neurological disability in young people in developed countries. There is currently no apparent cure for MS. Although the precise cause of MS remains speculative infection with EBV appears to be necessary and sufficient. In addition to EBV infection, MS is thought to result from other environmental effects acting in a genetically susceptible population. These environmental risk factors appear to act at the population level and interact with genetic risk factors to modify biological processes relevant to MS risk. Targeting EBV and other environmental risks associated with MS development offer a potential opportunity to reduce the risk of MS at a population level. In this talk, Professor Giovannoni will review the epidemiology of MS, including new clinical insights, that underpin the science of MS prevention trials. He will touch on the parallels MS has to type 1 diabetes mellitus and the overlap between these two autoimmune diseases.

Coeliac disease is a prevalent polygenic disorder caused by a harmful immune response to cereal gluten proteins. Previously the condition was considered an inflammatory, food hypersensitivity disorder, but presence of diagnostic autoantibodies, specific killing of enterocytes and autoimmune type of genetics advocate that coeliac disease rightfully can be considered an autoimmune condition with gluten as the driver of the autoimmune pathologies. As with most autoimmune diseases, the single most genetic factor in coeliac disease is MHC. The primary association is with HLA-DQ2.5, HLA-DQ2.2 and HLA-DQ8. These HLA molecules present gluten epitopes to CD4+ T cells – cells that conceptually serve as the master regulators of the immune reactions leading to disease. Gluten epitopes recognised by T cells are typically deamidated, and this deamidation is mediated by the enzyme transglutaminase 2 (TG2). Strikingly, coeliac disease patients have highly disease specific autoantibodies to TG2, and these antibodies are only produced in subjects who carry the coeliac disease associated HLA molecules when they consume gluten. It is hardly coincidental that TG2 is implicated in T-cell epitope formation and being the target for autoantibodies. Evidence suggests that B cells with their surface immunoglobulin antigen receptor, both cells specific for gluten and for TG2, serve as antigen presenting cells for gluten-reactive T cells thereby amplifying the anti-gluten T-cell response. Antigen receptors of disease-relevant T and B cells can be studied in detail by use of HLA-DQ:gluten tetramers and by use of labelled antigen (TG2 and gluten) that bind to surface IgA and IgM of gut plasma cells. The tracking of gluten specific T-cell clonotypes by HLA-DQ:gluten tetramer staining and antigen-receptor sequencing has revealed that such T cell clonotypes can persist for decades. The immunological scar imposed by gluten sensitisation has implication for design of immune therapies of coeliac disease, and this observation is relevant for treatment of other autoimmune diseases. 

The risk of developing type 1 diabetes is decreased in children born to mothers with type 1 diabetes as compared to children with a father or sibling with type 1 diabetes. We hypothesized that protection is provided by the increased proinsulin and insulin production by the fetus during a maternal type 1 diabetes pregnancy via an increased efficiency of (pro)insulin-specific T cell removal and regulation. To test this, CD4+ T cell responses to proinsulin and insulin were examined at birth in children who had a father or sibling with type 1 diabetes (n=14), a mother with type 1 diabetes (n=15) or without a first degree family history of type 1 diabetes (n=16). Multi-parametric flow cytometry analysis identified increased hematopoietic stem cell numbers in cord blood from children with diabetic mothers (p=0.002). The frequencies of responding CD4+ T cells to proinsulin (p=0.0046) and to insulin (p=0.0009) were increased in neonates with fathers or siblings with type 1 diabetes as compared to neonates of non-diabetic parents and to neonates with mothers with type 1 diabetes (p=0.0027 and 0.0046). Single cell gene expression profiling of proinsulin-responsive CD4+ T cells at birth showed reduced Th1 and Th17 profiles (CTLA4, p p<0.0001; CCR3, p<0.0001; CCR5, p=0.0001; CCR6, p=0.0008; AHR, p=0.0008 and RORA, p=0.0009) in neonates of mothers with type 1 diabetes compared to children with a father or sibling with T1D. In samples taken after birth, the ability of Treg cells to suppress proinsulin-specific CD4+ T cells was increased in children of mothers with type 1 diabetes compared to children with another first degree T1D patient (p=0.0016). These findings suggest that increased (pro)insulin production by the fetus during maternal type 1 diabetes gestation favors central and peripheral tolerance to these key targets of autoimmunity in type 1 diabetes.

High resolution mapping of genetic risk by GWAS, eQTL and imputation has identified 52 regions of genetic risk for type 1 diabetes, and these loci are highly enriched for immune function. When combined with functional annotation data it is also clear that the genetic risk of T1D is highly enriched in non coding regulatory elements. This suggests that the genetic risk alters gene regulation in a significant number of cases, and the target genes controlled by theses enhancers now must be fine mapped in order to identify the actual targets of genetic variation in T1D. Given that functional regulatory elements in immune cells are enriched for genetic risk,, and regulatory T cells are implicated in the loss of tolerance in T1D,, we have intersected our FOXP3 CHIP data onto the genomics data and find that >70% of the T1D risk loci contain at least 1 FOXP3 binding site.. In order to determine whether these enhancers are driving altered function in Treg or Tconv, a high resolution map of enhancer targets is needed. This requires a new technology called Chromosome Conformation Capture, which can biochemically trap DNA when looped over short or long distances. The genome wide approach is called HiC, and we have performed a HiC experiment to map all of the enhancer contacts in Treg and Tconv cells. Functional annotation of the targets of enhancers linked to genetic risk will be presented, and putative interventions to restore immune function are to be extracted for testing.