Dario Vignali, Retrogenic T cells, studying how T cells target beta cells
by Adam Burrack, PhD
As we have written about in our “immunology of type 1 diabetes” series, T cells of the immune system are required for disease onset in both humans and the mouse model, non-obese diabetic (NOD) mice. The advantage of using NOD mice in basic research is studying a spontaneous-onset disease and a “polyclonal” population of T cells comprising many different T cell receptors targeting many different peptides expressed by beta cells. While the polyclonal situation is more realistic, to study “one response” at a time allows researchers to construct experiments that yield results which are more straight-forward to interpret. If there were one peptide-specific T cell response to study in the NOD mouse it would be the response against peptides 9-23 of the B chain of insulin. We have previously described the central role that insulin-specific T cells play in the immune response. To study individual T cell populations specific for particular peptides derived from insulin-producing beta cells, researchers must use mice engineered to express only one T cell receptor, such as the TCR expressed by BDC2.5 mice developed by Diane Mathis. Unfortunately, TCR transgenic mice like the BDC2.5 take 6-12 months to generate (when everything works as anticipated). Researchers wanted to develop a better, more time efficient, model to study the disease-causing behavior of autoreactive T cell receptors in vivo. Dario Vignali used his training in biochemistry and molecular genetics to create a more efficient way to generate mice with only one T cell receptor. He then used these mice to study the diabetes-causing (diabetogenic) potential of 17 different T cell lines.
Dr. Vignali – then at St Jude’s Children’s Research Hospital in Memphis, Tennessee – revolutionized the study of T cell-mediated autoimmunity in 2006. Dr. Vignali described a method for rapidly generating mice with T cells bearing only one T cell receptor. In manuscripts in Nature Protocols and Nature Methods, Dr. Vignali’s group introduced and characterized a technique which comprised (1) sequencing the amino acid sequence of an isolated T cell receptor, (2) cloning that amino acid sequence into a retrovirus which was used to (3) transfect bone marrow cells in culture dishes, and then (4) transplanting that bone marrow into irradiated mice to (5) generate mice bearing only 1 T cell receptor whose actions in vivo (in the mouse) could then be studied . “Retrogenic” mice (a play on the terms retrovirus and transgenic mice), as this technique was named, could be generated in a 6-8 weeks, much more quickly than 6-12 months to generate TCR transgenic mouse lines through manipulation of embryonic stem cells and selective mouse breeding. Retrogenic mice containing T cells with exactly 1 T cell receptor facilitate directly comparing the disease-causing potential of different autoreactive T cells in mice. One can generate retrogenic mice bearing any of several different suspected disease-causing T cell receptors (and only that single T cell receptor) and then compare (1) timing of disease (standard NOD mice develop diabetes around 12-20 weeks of life), (2) penetrance of disease (% of mice afflicted, standard NOD colonies 60-80% of female mice develop disease), and (3) degree of beta cell destruction in diseased mice (by histology following the mouse’s death).
In a 2008 paper in the journal Diabetes Dr. Vignali’s group studied 17 T cell lines using retrogenic mice that were thought to mediate diabetes in spontaneous onset NOD mice. Dr. Vignali’s group showed that pathogenicity (or ability to cause diabetes) varies dramatically among T cell lines that were derived from diabetic NOD mice. For example, in the retrogenic system several T cell lines did not cause disease, including GAD65-specific T cells. Of the T cell lines tested using this in report, T cell lines generated by Katie Haskins – the BDC2.5 and the BDC10.1 – mediated the most penetrant and most rapid disease onset in retrogenic mice. As we’ve previously described in our post about Dr. Haskins’ research program, both of these T cell lines are specific for peptides derived from the chromogranin A peptide found within secretory granules released from beta cells. In short, this report demonstrated that not all disease-associated T cell receptors can mediate disease. This result highlights the primacy of the anti-insulin response in diabetes pathogenesis.
In a cutting-edge 2009 publication in the journal Immunity, Dr. Vignali’s group demonstrated that autoreactive T cells, and only autoreactive T cells, infiltrate the pancreas during T1D development in the NOD mouse. Prior to this work, it was controversial whether the inflammation produced by beta cell damage and destruction led to the recruitment of non-specific T cells, or whether T cell recruitment was restricted to autoreactive T cells. Results described in this paper were critical because they demonstrated that T cell recruitment was restricted to islet-specific T cells at all stages of the disease process, essentially disproving the “bystander recruitment” hypothesis. The conclusions drawn from the 2008 and 2009 papers may seem contradictory at first, but can be reconciled. The 2008 paper showed that not all T cells associated with diabetes onset can mediate beta cell destruction on their own, whereas the 2009 paper demonstrated that (1) islet-specific T cells are recruited to inflamed/diseased pancreas throughout the disease process, and (2) non-beta-cell-specific T cells are not recruited to the pancreas during beta cell destruction. All three conclusions were important in improving the field’s understanding of the process of beta cell destruction.
Lastly, in a 2014 Journal of Immunology paper, TCR affinity for self-peptides and impaired negative selection in the thymus promote T1D development in NOD mice. The primary conclusion of this report was that both low-affinity and high-affinity T cells specific for the B9:23 peptide of the insulin molecule can mediate both islet inflammation and beta cell destruction in mice. This result was important because the prevailing hypothesis had been that low-affinity T cell receptors escaped negative selection in the thymus and caused beta cell damage in the periphery, whereas high-affinity T cell receptors that escaped the thymus were thought to develop into regulatory T cells (which act in opposition to autoreactive T cells, as described in our post about Jeff Bluestone). This report disproved the second half of the conventional wisdom of the field, demonstrating that high-affinity T cell receptors can mediate autoimmune attack of beta cells as effectively as low-affinity T cell receptors. Whether this is a consequence of disease-associated MHC molecules in NOD mice (HLA molecules in humans) or is a general phenomenon of T cell receptors in autoimmune disease in general remains an open question.
In 2013 Dr. Vignali’s group improved but did not substantively alter the technique in a publication in the journal Nature Protocols. Retrogenic T cells are used to study T cells in laboratories ranging from cancer immunotherapy to autoimmunity. In July 2014, Dr. Vignali moved to the University of Pittsburgh to take on new roles as vice-chair of the immunology department and as co-director of the Cancer Immunology Program. At the University of Pittsburgh Dr. Vignali continues his work characterizing the immune suppressive functions of regulatory T cells in the tumor micro-environment and how this knowledge can be leveraged in settings of autoimmune disease. In particular, the role of the suppressive cytokine IL-35, which is released by regulatory T cells, has been a recent focus in the Vignali group in both autoimmunity and cancer. We look forward to Dr. Vignali’s continued contributions to the field of T cell biology.