Autoimmune responses in T1DM: quantitative methods to understand onset, progression, and prevention of disease

Abstract

Understanding the physiological processes that underlie autoimmune disorders and identifying biomarkers to predict their onset are two pressing issues that need to be thoroughly sorted out by careful thought when analyzing these diseases. Type 1 diabetes (T1D) is a typical example of such diseases. It is mediated by autoreactive cytotoxic CD4⁺ and CD8⁺ T-cells that infiltrate the pancreatic islets of Langerhans and destroy insulin-secreting β-cells, leading to abnormal levels of glucose in affected individuals. The disease is also associated with a series of islet-specific autoantibodies that appear in high-risk subjects (HRS) several years prior to the onset of diabetes-related symptoms. It has been suggested that T1D is relapsing-remitting in nature and that islet-specific autoantibodies released by lymphocytic B-cells are detectable at different stages of the disease, depending on their binding affinity (the higher, the earlier they appear). The multifaceted nature of this disease and its intrinsic complexity make this disease very difficult to analyze experimentally as a whole. The use of quantitative methods, in the form of mathematical models and computational tools, to examine the disease has been a very powerful tool in providing predictions and insights about the underlying mechanism(s) regulating its onset and development. Furthermore, the models developed may have prognostic implications by aiding in the enrollment of HRS into trials for T1D prevention. In this review, we summarize recent advances made in determining T- and B-cell involvement in T1D using these quantitative approaches and delineate areas where mathematical modeling can make further contributions in unraveling certain aspect of this disease.

Keywords: B-cells; Markov models; T-cells; T1D; autoantibodies; autoimmunity; avidity; mathematical models; predictive algorithms; β-cells.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd.

Figures

Figure 1

(A) A scheme showing the effect of high/intermediate/low avidity/affinity TCR-pMHC interaction. High affinity/avidity interaction leads to deletion of most autoreactive T-cells, creating a T-cell repertoire that is low in pathogenic and high in regulatory T-cells, resulting in healthy state without islet destruction. Intermediate affinity/avidity interaction results in a T-cell repertoire containing a higher portion of pathogenic T-cells, but also a high number of regulatory T-cells. This results in limited islet destruction, because the Tregs limit the destructive effects of the pathogenic T-cells. Low avidity/affinity interaction results in T-cell pool containing many autoreactive T-cells with few regulatory T-cells. This results in autoimmune state where most islets are destroyed. (B) A scheme showing the various components of the autoimmune response in T1D including the Copenhagen model. The uptake of β-cell specific proteins by APCs triggers APC recruitment and activation. This in turn leads to the activation of various classes of islet-specific CD4+ and CD8+ T-cells (i.e., Th-lymphocytes, regulatory T-cells (Tregs) and autoregulatory T-cells (Taut)), as well as B-cells. High avidity cytotoxic T-lymphocytes destroy β-cells by either secreting harmful cytokines, or by inducing apoptosis via cell-to-cell contact. Mature B-cells release islet-specific autoantibodies that may appear prior to disease onset.

Figure 2

Two possible Markov models describing the dynamics of phagocytosis in macrophages. Macrophages may have up to N undigested apoptotic cells inside them. Both models require an activation step of naïve macrophages with an activation rate constant ka. The forward transitions (excluding the first step), representing the engulfment of new apoptotic cells, occur with a rate constant ke, whereas the backward transitions, representing digestion, occur with a rate constant kd. (A) The activation step is assumed reversible. (B) The activation step is assumed irreversible.

Figure 3

Cyclic fluctuations in the pool size of effector T-cells with a given autoantigenic specificity, as predicted by the model in reference (80). (A) Periodic damped oscillations with high frequency. (B) Low frequency cyclic waves with few peaks that correlate with the number of clones considered in the model.

Figure 4

The steady state level of the fraction of surviving β-cells (βss) predicted by the model in (82) based on two independent hypotheses: (A) memory autoregulatory T-cells crowd the islets and block effector T-cells from reaching β-cells; and (B) memory autoregulatory T-cells kill APCs. Solid lines represent the physiologically attainable steady states, whereas dashed lines represent the physiologically unattainable steady states. The horizontal solid lines in both panels represent the healthy and unaltered βss=1, in the absence of autoimmunity, whereas the other solid lines represent βss<1 in the presence of autoimmunity. Notice that increasing the proliferation rate of memory autoregulatory T-cells in (A) causes a rapid rise in βss in the autoimmune case, whereas decreasing the scaled pool size of APCs induces a slow increase, suggesting that the latter is more physiological.

Figure 5

Increasing the proliferation rate of memory autoregulatory T-cells (denoted by m) may lead to the switch phenomenon. (A) At a low Taut proliferation rate, the level of IGRP206-214-reactive effector T-cells is elevated. (B) At a high Taut proliferation rate, the total level of other subdominant effector T-cells is elevated.

Figure 6

Increasing T-cell avidity on average leads to a faster β-cell destruction, a faster rise in cognate autoantibodies and a biphasic response in the peak amplitude of the autoantibody titer. Time evolution of (A) the average pool sizes of high (black), intermediate (dark gray) and low (light gray) avidity T-cells; (B) fraction of surviving β-cells in the presence of high (black), intermediate (dark gray) and low (light gray) avidity T-cells; and (C) scaled titer level of cognate high (black), intermediate (dark gray) and low (light gray) affinity autoantibodies.