Autoimmunity and tumor immunology: two facets of a probabilistic immune system

Abstract

Background: The immune system of vertebrates has evolved the ability to mount highly elaborate responses to a broad range of pathogen-driven threats. Accordingly, it is quite a challenge to understand how a primitive adaptive immune system that probably lacked much of its present complexity could provide its bearers with significant evolutionary advantage, and therefore, continue to be selected for.

Results: We have developed a very simple model of the immune system that captures the probabilistic communication between its innate and adaptive components. Probabilistic communication arises specifically from the fact that antigen presenting cells collect and present a range of antigens from which the adaptive immune system must (probabilistically) identify its target. Our results show that although some degree of self-reactivity in the immune repertoire is unavoidable, the system is generally able to correctly target pathogens rather than self-antigens. Particular circumstances that impair correct targeting and that may lead to infection-induced autoimmunity can be predicted within this framework. Notably, the probabilistic immune system exhibits the remarkable ability to detect sudden increases in the abundance of rare self-antigens, which represents a first step towards developing anti-tumoral responses.

Conclusion: A simple probabilistic model of the communication between the innate and adaptive immune system provides a robust immune response, including targeting tumors, but at the price of being at risk of developing autoimmunity.

Figure 1

Conceptual scheme of the immune response. Ags are collected from the environment by specific cells (APCs) that belong to the IIS. Following activation of the IIS by danger signals, probabilistic communication between the IIS and AIS occurs via Ag presentation, permitting the AIS to mount an Ag-specific immune response. The overall immune response is the result of innate and adaptive effector mechanisms.

Figure 2

Effect of central and peripheral tolerance on immune self-reactivity. The x-axis represents the effective abundance of a self-Ag, in arbitrary units. Top: probability that self-reactive lymphocytes towards a given self-Ag remain after the induction of central tolerance. The curve corresponds to 1 – f _i from eqn. 1 and it can be also interpreted as the fraction of self-Ags with a given effective abundance that remain untolerated. Middle: probability that an Ag (self or non-self) is efficiently presented by an APC. Bottom: combined effect of central and peripheral tolerance on the probability of targeting a self-Ag. In our simple model, peripheral tolerance is determined by the threshold a ^*, the minimum effective abundance required for efficient Ag presentation. In the absence of peripheral tolerance, the probability of targeting a self-Ag becomes greatest for those self-antigens with effective abundances close to a ^max.

Figure 3

Probability of targeting self-Ag in basal conditions (curve) and after pathological over-expression (isolated points). The arrows indicate multiplicative changes in the effective abundances of several self-Ag (increase factors are written on the arrows). Only those Ags with effective abundances greater than the threshold a ^* can induce a response (red symbols). Note that, due to peripheral tolerance, the targeting probability for Ag with abundances below a ^* is effectively zero; the values represented in black (dashed line and open symbols) correspond to the targeting probabilities in the absence of peripheral tolerance as given by eqn. (8b). The MDI in this example has been set to 5, which means that an Ag with basal abundance equal to a ^max has to be over-expressed by a factor 5 in order to reach the threshold abundance a ^*.

Figure 4

Discrimination ability ∆ as a function of the minimum detectable increment MDI (eqn. 3 ). The ability of the IS to detect tumors depends on the region where it works. In region 1, the IS responds to small changes in effective abundances with low discrimination, which implies a higher risk of autoimmunity. In region 2 (shaded area), the IS responds to moderate variations in self-Ags with high discrimination making it optimal to detect tumors correctly. In region 3, large increases in effective abundances are required to make self-Ag detectable. Actually, the IS of vertebrates seems to be working in this third region. Notice that the border between regions 2 and 3 is somehow arbitrary, since a better characterization of fluctuations in Ag effective abundances would be required in order to decide what a “large” change is (see Discussion).