Signal Processing in the Immune System

This web site applies signal-processing paradigms, well-known in engineering theory and practice, to understand possible behavior of vertebrate immune systems.

Do T Cells Correlate Danger-Signals to Antigen?   My first paper proposes that the immune system derives maximum information about the health of a class of host cells by comparing their danger-signals to those from putatively healthy cells. Three-party cytotoxicity assays show that T cells challenge only syngeneic "bystanders" after killing cognate targets. The conventional interpretation of bystander-killing is that it protects the host from viruses that have disrupted their host cells' ability to present viral peptides. However, a requirement for syngeny would be very counterproductive to that function, since viruses could (and do) silence syngeny signals (MHC molecules) as well as peptide display; a syngeneic requirement would allow viruses to escape Tc killing by completely silencing their host cells. Since the FasL challenge of bystanders therefore does not benefit the host primarily by killing virally silenced cells, one must suspect other primary functions.

I propose the FasL challenge provides information to the T cell, a function quite plausible in light of the proven reverse-signalling of FasL. My proposal that Fas display on non-cognate targets reduces activation and cytotoxicity-- the unproven component of the hypothesis that T cells correlate danger to Ag-- seems the plausible mechanism to prevent upregulation of Tc not specific to disease Ags. This proposal is the immunological equivalent of the common (and in the engineering community, universally accepted) design-pattern of using differential measurement against a reference signal, to improve the quality of a signal bearing needed information but contaminated by noise or false-signals.

Contrast this comparative danger model against earlier models. The self-non-self-discrimination theory posits that thymic encounters with cognate Ag prevent cytotoxicity against it. The original danger model posits that long-term encounters with cognate Ag in the absence of costimulation, prevent cytotoxicity against it. Dependence on either long-term or organ-specific history would severely limit the ability of the immune system to quickly quench autoimmune reactions. Furthermore, during real infection, when costimulation signals abound, the immune system would have difficulty preventing upregulation of cytotoxicity against self-peptides presented alongside pathogenic peptides, if it did not use comparative danger-sensing.

Neuroscience has long recognized that the nervous system is tuned to detect differences and changes, not absolute levels. Therefore besides having parallels in humanly designed signal-processing systems, the assertion that the immune system recognizes danger by detecting differences on many scales and in many contests, also has experimentally proven precedent in other fields of biology that investigate complex systems. Self vs. non-self discrimination could be construed as one kind of difference measurement. The hypothesis that individual cells compare danger signals between cognate vs. non-cognate targets, would be an application of the comparison principle to the danger model of immunology.

I would be happy to give you a free hardcopy reprint copy of the paper as published. Send your physical mailing address to the email address at the bottom of this page. I am also happy to send .doc (Microsoft Word), .txt, or .rtf versions of the paper via email. Medical Hypotheses owns the copyright and is the sole distributor of the .pdf version, which they will sell you.

Testing the Hypothesis that reference-cells facilitate danger-correlation by individual T cells.

Main Power Point slides for talk on comparitive danger sensing

Other recent research

Does the immune system identify necessarily conserved pathogenic epitopes by finding similarities to the host proteome?

Experiments to canonically discover conserved tumor epitopes

Send comments, questions, or reprint requests to: .

Useful immunology links:

  • Immunobiology by Janeway, Travers, Walport, and Shlomchik
  • Molecular Immunology by Focosi
  • On-Line Medical Dictionary