Diane Mathis and Christophe Benoist, the pancreatic lymph node, and TCR transgenic mice
By Adam Burrack
In our “immunology of type 1 diabetes” series I’ve previously described critical roles for CD4-expressing T cells of the immune system in the process leading to the development of type 1 diabetes (T1D). CD4+ T cells are critical to the immune response due to their central role in providing “help” to both (1) activating “killer” CD8 T cells – which physically interact with beta cells promoting beta cell death – and (2) in providing “help” to activate B cells. As I described in our post about the career of George Eisenbarth, activated autoreactive B cells produce autoantobodies, which do not transfer disease on their own, but are a leading clinical indicator of developing T1D. In my previous post about Jeff Bluestone’s research program I described the strategy of transfer of regulatory T cells – which also express the molecule CD4 on their surface – to stop new-onset T1D. Through this strategy or others, regulatory T cells may one day represent one arm of a clinical treatment to stop new onset T1D. For one last piece of context, our post describing the contributions of Katie Haskins and her laboratory to our understanding of the biology of autoreactive CD4 T cells which precipitate T1D introduced us to the BDC2.5 T cell line.
This post will focus on the research of Diane Mathis and Christophe Benoist, who jointly run an immunology laboratory at Harvard University and focus on T cell activation and type 1 diabetes onset. Drs Mathis and Benoist have spent their careers studying CD4 T cells in the pathogenesis of T1D and other autoimmune conditions. The Mathis-Benoist laboratory has contributed several breakthroughs to our understanding of T1D over the past 30 years through the study of T cell activation, T cell trafficking to sites of immune responses, and effector functions of T cells at these sites of responses. These discoveries include the seminal finding that the pancreas-draining lymph node (PLN) is the anatomic location where beta cell-derived proteins are first presented to T cells in an inflammatory setting such that an autoreactive T cell response is promoted and diabetes occurs in the NOD mouse. Interestingly, in genetically related mice expressing diabetes-associated genes autoreactive T cells are present in the pancreas-draining lymph node but do not precipitate T1D. Inflammation is an important “bridge” between the innate immune system and the adaptive immune system which is generally interpreted by immune cells as a “danger signal” and is a helpful adjunct to promoting a strong T cell response in the setting of viral infection and unfortunately appears to play a role in promoting autoreactive T cell responses as well. Therefore, levels of inflammation within the PLN may be a key difference in the NOD mouse – and people who develop T1D – that leads to development of inappropriate T cell responses.
Closely related, an important finding from the Mathis-Benoist group is that priming of T cells in the pancreatic lymph nodes of 2 week old NOD mice precedes invasive insulitis and autoimmune attack of beta cells. This result suggests that some developmentally-regulated event, such as pancreas remodeling around the time of weaning (weaning is the cessation of breast feeding and transition to a more adult, mixed, diet), is associated with significant beta cell death, inflammation, and the resulting T cell activation leading to autoimmunity. In basic research studies, mice are weaned at 3-4 weeks of age, so these experiments suggest that the pancreas in NOD mice is particularly prone to high levels of inflammation before this transition in diet and consequent changes in pancreas architecture and that the changes precipitated by weaning are a type of ‘tipping point’ for the development of a pathogenic T cell response. Interestingly, others have recently corroborated that inflammation within the pancreas during this time frame is more pronounced in diabetes-prone NOD mice than in diabetes-resistant but genetically similar mouse strains. This heightened pro-inflammatory response within the pancreas (and PLN) of NOD mice creates an environment within local lymph nodes which is more able to promote a T cell response. Importantly, endocrine self (digestive hormones including secretory granules carrying insulin molecules) and (food-derived) non-self interact in pancreatic lymph nodes, setting the stage for gut inflammation to “tip the scales” toward generating autoimmunity against self in the presence of gut inflammation. Future posts in this series will address in more detail factors that promote gut inflammation and set the stage for gut-directed immune responses.
To return to the Mathis-Benoist laboratory, a key reagent these researchers have generated, and which is used extensively in the study of the cell biology of T1D, is the BDC2.5 TCR transgenic mice. TCR transgenic mice are genetically engineered to express a single type of T cell receptor (TCR), which in theory is specific for only one target peptide. In this way, immunologists in the 1990s and early 2000s could study a single T cell receptor in mice, which allowed the study of T cell biology on the whole-animal scale. The Mathis-Benoist laboratory applied genetic technology used to create mice expressing a single TCR to study the TCR discovered by Katie Haskins’ group expressed by the BDC2.5 T cell. This TCR transgenic mouse has been used to study the processes of T cell activation (in the pancreas-draining lymph node), T cell trafficking (from the lymph node into the pancreas), and effector functions (killing beta cells within the pancreas) by dozens of researchers around the world. As described in our previous post about the career of Katie Haskins and work in her laboratory, we now know that BDC2.5 CD4 T cells specifically target peptides derived from the chromagranin A molecule, which is a component of the secretory granule within which insulin is secreted from pancreatic beta cells.
The Mathis-Benoist laboratory continues to study the cell biology of autoreactive T cells using BDC2.5 mice. One early question regarding these T cells is whether they represented regulatory T cells (and could therefore potentially be used therapeutically to stop disease) or pathogenic T cells (and could therefore be studied as a model of a ‘typical’ autoreactive T cell to develop ways to inhibit the activation and/or function of pathogenic T cells). Early studies from both the Haskins and Mathis-Benoist laboratories established that BDC2.5 cells are pathogenic (diabetes-causing or “diabetogenic”) autoreactive T cells. A corollary is whether these effector T cells could be converted to regulatory T cells in the pancreas (a theme we will return to in future blogs), thus down-modulating an active disease process at the relevant anatomic site. Results along this line of reasoning were not encouraging. Furthermore, upon seeing their self-antigen in the islets, the Mathis-Benoist laboratory showed that BDC2.5 TCR transgenic T cells do not convert into regulatory T cells. This suggests that the phenotype – the function – of these autoreactive T cells may be set when they leave the thymus. The Mathis-Benoist laboratory continues to investigate differences in T cell selection (which is the process that determines which T cells leave the thymus and have the chance to participate in immune responses) and activation between disease-prone NOD mice and non-diabetes-prone mouse strains using cutting-edge technology.These studies investigating single cells in NOD mice compared to non-diabetic mice are critical to begin to determine why T1D occurs in some genetic situations and not others. It is likely that many individually small factors combine to promote inappropriate T cell responses leading to autoimmunity. The more of these small steps we can understand, the more opportunities we generate to inhibit these processes and prevent disease onset.