Printing Version
The House of IMMSIM

Humoral/cellular autopiloting during the response to viral infection

Franco Celada

Franco Celada

Darwin's original selection of the fittest acts on a population whose diversity is the product of germ line mutation. The extension of the mutation/selection concept to somatic differentiation (Fig. 1) is a milestone of Immunology, epitomized by the Clonal Selection Theory (1959). Since then it has become clear that Darwinian selective advantage is not limited to the specificity of repertoires, but governs the maze of stimulatory and regulatory interactions, and is ultimately shaping the entire immune system and the quantity/quality of individual responses.

It follows that we expect the immune system, when engaged by infection, to shape itself in the best possible way as to ensure its own survival (and, as a by-product, the patient's). I believe the recent findings about lymphokine-mediated regulation of Th cell responses to be evidence of the optimisation process.

Figure 1
Figure 1: Germ line (red arrows) vs Somatic differentiation (green arrows)

Of course, we cannot expect to find the best possible immunity always in place. There are many and well-known inadequate immune responses, such as allergies, Aids, granulomata, but there are many external reasons for failure, such as disparity of growth potential and active traps set to the immune recognition by the invaders. All in all, I, as an immunologist, can adopt the idea that optimisation is always sought, even if not always attained. For me, as a modeller, the basic question of the following: does the IMMSIM3 model that we developed over the years exhibit a similar tendency toward producing the most efficient response? To answer, or better, to put the possible answer in front of you, I must rapidly lead you through a rather complex simulation, and I will do it by considering the following points:

  1. IMMSIM3 incorporates both humoral and cellular immune responses, but humoral (H) and cellular (C) can also be run alone, against the same set of different infections, and their results compared with each other and to the combination of H+C (which is, predictably, the most efficient) (1).

  2. The set of infection agents used includes viruses identical in their immunologic specificity but varying along three "behavioural" characteristics. Since each parameter has 4 "levels", the total viral repertoire is 64. To order and to keep track of them we have numbered the individual viruses as nodes in a cube (Fig. 2), whose three dimensions represent speed, burst size and infectivity. Since the same parameter value can appear in different places of the numerical sequence, we did expect and did observe certain harmonic recurrences in the results.

    Figure 2
    Figure 2
  3. Fig. 3 contains several graphs carrying various information. The reason for combining this information is to allow the reader to compare the positions of the bars in the different graphs, relative to the virus index in the abscissa.

    Figure 3
    Figure 3:
    a) The first graph represents cures reached after infection, when only humoral response is allowed. The cures are limited to four segments of the index (1-4, 17-21, 34-37, 49-53). It can be seen on the cube that these segments correspond to viruses with low infectivity.
    b) The second graph shows the cures obtained by the cell-mediated response alone. In no case is there a 100% recovery. However, partial cures are grouped in 3 large and shallow peaks, which are positioned roughly in the valleys between the peaks of cures by humoral response. This means that the success of cellular response relies on viral characteristics clearly different from those that makes infections susceptible to humoral.
    c) The third graph is obtained by subtracting from the numbers of T effectors produced in both mode the numbers produced in Cellular only mode. What results is the amount of enhancement (in green) or the amount of suppression that T effectors experience in each infection, when both responses are active.

A very interesting finding born out of these experiments pertains to what happens when both responses are active at the same time. Cures exceed the sum of cures of H only + C only (to see the graphs, read (1). However, when examining the single responses, e.g. the antibody titer or the number of T effectors produced, we observed cases where in both the response was enhanced and cases where it was suppressed.

In order to find the underlying logic of these phenomena, in Fig. 3 we have printed section c) under a) and b). This graph is obtained by subtracting from the numbers of T effectors produced in both mode the numbers produced in Cellular only mode. What results is the amount of enhancement (in green) or the amount of suppression that T effectors experience in each infection, when both responses are active.

The position of the peaks of enhancement/suppression are revealing. By comparing them to cures by H only and cures by C only, one can clearly see that T effectors are suppressed in those infections where H response is effective but C response is inefficient; instead, they are enhanced whenever the cellular response is making progress towards a cure while humoral response is inept.

We have no clear idea yet of all mechanisms involved in these regulations, but I believe that this is a first stride in the direction of showing a capacity by IMMSIM3 to steer the type, or quality of the response towards the fittest.

References:

  1. Kohler B, Puzone R, Seiden P, Celada F, "A systematic approach to vaccine complexity using an automaton model of the cellular and humoral immune system". Vaccine (1999) - 19: 862-876

 

Up: The Approaches


Printing Version