Linda Sherman, Scripps Institute, Thematic connections between autoimmunity and cancer research
by Adam Burrack, PhD
Thus far in our tour of world-class researchers who study autoimmunity and type 1 diabetes (T1D), I have described a number of researchers studying the cell biology of the immune response and the genetics leading to the propensity to develop autoimmunity. In particular on the genetics side of the equation, in our blog about the career of George Eisenbarth, I described genetic risk determined by the HLA molecule (and the analogous MHC molecule in NOD mice) which shows peptides to T cells of the immune system. Also described in that post are experiments demonstrating that HLA/MHC itself does not drive disease development, but rather facilitates the development of T cells with the potential to be autoreactive. In both mice and people presumptive autoreactive T cells which develop in disease-prone individuals require further signals to mediate disease. Today’s featured researcher studies a genetic perturbation leading to “over-active” T cell signaling, a separate risk factor from HLA associated with development T1D and other autoimmune conditions.
Linda Sherman at the Scripps Institute in San Diego is the 2014-2015 President of the American Association of Immunologists. In her lab Dr. Sherman studies both autoimmunity and cancer, an interesting approach which facilitates insights from one field to be directly applied to research in its mirror image biological setting. Much of cancer immunotherapy is focused on stopping development of immune tolerance to proteins expressed by tumor cells or blocking this tolerance once it has been established by a solid tumor, whereas immune-based therapy for autoimmune disease focuses on re-establishing immune tolerance after these mechanisms have broken down, leading to disease. In both settings, T cells are key effectors and must be either coaxed into acting when they are inhibited locally (cancer immunotherapy) or prevented from acting or their mode of activity changed to stop disease processes (autoimmunity).
In her work investigating the immune-mediated mechanisms of T1D, Dr. Sherman has focused her efforts on T cell activation and how individuals prone to develop autoimmune disease may have perturbations in this general pathway. In this scheme for thinking about T1D, T cells which are activated “too easily” can lead to beta cell damage due to disease-associated HLA molecules which present peptides from beta cells in a manner which looks “new” to the T cells. The combination of “too easy to activate” T cells with the so-called molecular mimicry hypothesis could partially explain how autoimmunity develops. In addition, this sequence of events could explain how multiple types of autoimmunity develop, not just T1D. In particular, Dr. Sherman studies the signaling molecule PTPN22 which has a key role in the T cell activation process during germinal center reactions in lymph nodes. As described previously, the relevant lymph node in T1D is the pancreas-draining lymph node, in the weeks/months/years prior to T1D onset.
In a 2012 manuscript in the journal Diabetes, Dr. Sherman’s group took a genetic approach to investigate CD8 T cell tolerance. Dr. Sherman utilized genetically engineered mouse strains generated in part through efforts at Jackson Labs to interrogate the break-down in T cell tolerance to beta cell-derived proteins. In short, researchers at Jackson labs have created a series of NOD-derived mouse lines with various regions of the genome from C57/Bl6 mice (which do not develop T1D or any other autoimmune conditions) knocked “into” the genome of NOD mice. These “knock-in” mice do not develop T1D due to these very subtle genetic differences causing changes in T cell development and activation. Through studying the differences in disease outcome – and T cell activation – between these strains of mice, researchers can develop insight into how the NOD mouse loses its “self-tolerance” to insulin and other beta cell derived proteins.
To summarize, Dr. Sherman found that the Idd3 genetic loci from diabetes-resistant C57/Bl6 mice (and more importantly proteins encoded by these genes including T cell growth factors) was required on lymphocytes (T cells) to prevent diabetes development, whereas Idd5 from the diabetes-resistant mouse strain C57/Bl6 was required within the antigen-presenting cells was required to prevent disease. For the most effective disease prevention in this study, both genetic changes were required: T cells had to express Idd3 (so they could be more effectively controlled) and dendritic cells had to express Idd5 (to promote less inappropriate T cell activation). This type of elegant genetic approach allows us to conclude that to develop diabetes (or autoimmunity in general) there are most likely “problems” on both the T cell (Idd3) and antigen-presenting cell (Idd5) side of the equation. The additional knowledge of which disease-related proteins are encoded within particular Idd loci can yield targets for therapeutic approaches to either inhibit T cell activation (proteins encoded within Idd3) or change the context of antigen presentation (proteins encoded within Idd5) in NOD mice in further experiments.
Lastly, a 2014 paper from Dr. Sherman’s group in the Journal of Immunology reports that mice deficient in PTPN22 are unable to effectively down-regulate the germinal center reaction, leading to increased production of antibodies and increased expression of the T cell activating cytokine IL-21 within germinal centers. This combination of events may increase the propensity for inappropriate T cell and B cell responses. These findings provide mechanistic evidence demonstrating how autoimmune disease-associated mutations in PTPN22 (which impairs ability to down-regulate T cell responses within lymph nodes) have the potential to synergize with HLA/MHC-driven generation of autoreactive T cells to promote aberrant T cell activation. Disease-associated HLA/MHC makes it more likely that these individuals will have potentially beta cell-reactive T cells in their lymph nodes than individuals without disease-associated HLA/MHC. Interestingly, this work dove-tails nicely with the topic of a previous post in our series: clinical observations recently reported from Lucy Walker’s group in England demonstrating a chronic tendency toward T follicular helper cell polarization of CD4 T cells, in general even in peripheral blood samples, from subjects who had developed T1D.
In summary, animal models of human disease have their uses. As discussed in previous posts, it may be unrealistic to expect treatments which prevent disease in (inbred) NOD mice to prevent disease in genetically diverse humans. However, genetic models in the NOD mouse combined with cell biology experiments using mouse cells (and perhaps human-derived cell lines) can yield knowledge of the cell biology involved in the development of autoimmunity which is broadly applicable. Through this approach therapeutic targets to be explored in the mouse model before clinical application can be cultivated. Over time, we hope this rational scheme for target development and validation in autoimmune disease yields consistent clinical benefit.