YOUglycemia immunology of diabetes blog series
Rachel Friedman, 2-photon microscopy, early T cell infiltration of pancreatic islets
by Adam Burrack, PhD
I was in elementary school when the book and movie Jurassic Park were released, and I was a fan of Michael Crichton’s science fiction since. To be honest much of my curiosity about the world derived from a science fiction-like “what if” wonderment derived from the amazing tools that are constantly being developed to study biology. Sometimes I feel as if I’m living in a world where anything is possible: I’ve seen diabetes treatment progress from pig insulin to recombinant insulin with many options, and now to continuous glucose sensing and potential closed-loop and/or two-hormone systems. I’ve witnessed – and participated in – the advent of the field of directed differentiation of stem cells to pancreatic beta cells. One day these cells may become part of a beta cell replacement strategy to cure diabetes. My specific research focuses on tracking autoreactive T cells – which cause diabetes – in the mouse model of the disease. Who would have imagined any of this technology >30 years ago?
Today’s topic is another scientific technique that inspires that type of awe, at least for me, as a scientist: 2-photon microscopy. I don’t want to give a seminar on the physics and optics of how this technique works (see above Wikipedia link), but rather I will describe the type of information we can learn about how beta cells are destroyed in type 1 diabetes. Long story short, 2-photon microscopy allows us to watch cells interact, in real-time, in living animals. Current methods allow researchers to visualize up to 4 different “colors” of cells, as labeled by various fluorescent dyes and detected by their emission of specific wavelengths of light when excited by the 2-photon microscope. Applied to studying how type 1 diabetes happens, this means that researchers have the ability to watch T cells trafficking throughout the pancreas and ‘attack’ beta cells, which can be visualized at the same time.
This technology is not cheap: it requires significant infrastructure and funding support to maintain and use. Only a handful of laboratories in the country are capable of using this technology in the ways I have described above. One of these principal investigators is Rachel Freidman at National Jewish Health in Denver, Colorado. Dr Freidman is an early-career investigator who came to her faculty research position from the UC-San Francisco, which I have previously described.
In a 2014 manuscript in Proceedings of the National Academy of Sciences, Dr Freidman published initial findings investigating islet infiltration by T cells. This manuscript established that T cells within islets interact with antigen-presenting cells (dendritic cells) weeks before disease onset, establish stable contacts, and go on to produce the effector cytokine IFN-gamma. This process of T cell “priming” is critical to generating productive immune responses, and appears to facilitate the maturation and activation of autoreactive T cells within the islets. Other researchers (including Thomas Kay and Helen Thomas), have established that IFN-gamma achieves two key goals in ramping up immune responses: promoting antige-presenting cell maturation and up-regulating expression of MHC class I on non-immune cells. In other words IFN-gamma may set up a sort of positive feedback loop promoting beta cell death by (1) facilitating T cell priming, and (2) up-regulating MHC class I on beta cells which will promote physical interaction with activated CD8+ T cells. By the time diabetes occurs (blood sugar levels become elevated in the animal), T cells are no longer physically interacting with the antigen-presenting cells. By the time disease occurs it appears T cells do not require the same degree of “priming” by antigen-presenting cells. This may indicate that the T cell response has shifted from an “effector” response to a “memory” response by the time sufficient beta cells are destroyed to observe excursions in blood sugar level. This scenario suggests intervention will be required very early in pathogenesis to prevent T cell activation leading to disease. By the time disease occurs it may be the case that the relevant T cells are “memory” T cells. This difference in maturity level of T cells may require different therapeutic approaches than those that would inhibit “less mature” T cells.
In the 2014 manuscript Dr. Freidman primarily used a transgenic mouse, not the NOD mouse, to visualize generalized T cell interactions with antigen-presenting cells and beta cells. In a follow-up 2015 manuscript in the Journal of Immunology Dr Friedman and her graduate Robin Lindsay further interrogated this loss of T cell contact with antigen-presenting cells as disease progressed. In this paper the authors used a modified microscope set-up to visualize T cells interacting with antigen-presenting cells within the pancreas of NOD mice: true intra-vital imaging. In this paper they make the same core observation: later in the disease process (closer to the times in which intervention strategies are currently proposed to occur) T cells move more quickly and make less stable contacts with antigen-presenting cells.
Concordance between the transgenic mouse model and the NOD mouse is key: the biology of the T cell response appears to be consistent across these two models and suggests that T cell priming occurs early in disease pathogenesis – and then largely ceases. The implication of this finding is that interventions to prevent T cell priming must be initiated early in at-risk animals (or people), whereas later in pathogenesis – including in an islet replacement transplant scenario – strategies to inhibit activated/memory T cells must be employed.
To develop effective therapies to delay autoreactive T cells, this difference in “maturation status” is important. Preventing antigen presentation to “naïve” T cells would have represented a less stringent requirement than does preventing the actions of memory T cells. No one said it would be easy J