Latest Immune Models and Hybrid Approaches - Immunological Computation: Theory and Applications - page 155

triggered. Another work (Laff erty and Cunningham, 1975) proposed that the T h
cell itself needs to be activated by signal 1 and signal 2 coming from APCs (see
Figure 6.2c). It is important to note that T h cells receive signal 1 from two sources:
B cells and APCs.
h e next model introduced by Janeway (1992) used the notion of INS, such as
bacteria, which “primes” APCs causing the production of signal 2; in Figure 6.2d,
priming signal is labeled as signal 0.
However, Matzinger (1994) proposes that the priming of APCs is due to
a danger signal from stressed tissues or cells (Figure 6.2e). She also proposed
extending the e cacy of T h cells by routing signal 2 through APCs. h is signal
is marked as “signal 3” in Figure 6.2f. According to her, the antigen seen by the
T k cell does not need to be the same as the T h ; the only requirement is that both
must be presented by the same APC. h is allows T h cells to prime many more T k
cells than they would otherwise have been able to. h e DT can be seen as a natu-
ral extension of immune signal models proposed earlier. Figure 6.3 exhibits the
partitioning of the antigen universe based on three models: self/nonself (SNS),
INS, and DT (Matzinger, 2002).
According to Matzinger (2007), there is a category of damage-associated molec-
ular patterns (DAMPs) that encompasses both pathogen-associated molecular pat-
terns (PAMPs) and alarm signals. h e ultimate control lies with the tissues in which
the response occurs, rather than with the pathogen against which it is directed.
Particularly, tissues use all sorts of mechanisms to keep the cells and molecules trig-
gering immune responses in order to control the invaders. Many of these cells also
seem to recognize stress-induced “self ” molecules rather than foreign pathogens.
According to this view, at least some immune responses are initiated by tissue-
derived signals that activate and educate APCs to control the eff ector class of an
immune response. Accordingly, some danger signals such as tissue damage trigger
a myriad of immune reactions and responses.
One of the problems of DT, however, is that the exact nature of danger signals
is unclear. Also, there are some danger signals that should not trigger an immune
response such as cuts or transplants. In addition, DT is not able to explain autoim-
mune diseases.

6.1.1 Danger Theory-Based Algorithms

Aickelin and Cayzer (2002) include the following aspects of the DT in their artifi -
cial immune system (AIS) design principles:

Appropriate APCs to present danger signals need to be modeled.
A danger signal can be either positive or negative, which means the presence
or absence of the signal.
Although in biology the danger zone is spatial, in computation model other
notions of proximity, such as temporal proximity, may be used.