The Cognate interaction Genomic arrays A new era for modeling the immune response Benoit Morel.

The Cognate interaction

Genomic arraysA new era for modeling the immune response

Benoit Morel

Today’s experimental data:

A NEW ERA

More information than we can digest, yet. (blurred…)

Clearly a revolution in the making, for the experimentalists, the modelers,

AND (more importantly) their relations…

Importance of Large Numbers and variability

• Immune system involves ~1012 cells• Each cell is a complex object (~109 molecules)• Functional repertoire > 108

• Cells act in populations; • Cells of the same population are significantly

different.• There are many factors and co-factors: many

kinds of cytokines, chemokines, receptors, CD’s.• Microarrays many genes are activated together

Importance of the Information flow

• Innate response: – Toll Receptors, PAMPs (=Pathogen Associated Motif Patterns), – Danger signal (LPS?, Viruses??)

• Cognate interaction (D-cells communicate a lot of information during the cognate interaction)

• CD4 cells information processors

• Immunological memory: the ultimate legacy of the response

The cognate interaction:

-Couples the innate to the adaptive response

-It is a protracted event, using the “immunological synapse” which leads to the differentiation of T-cells .

-TCR have their share of promiscuity.

-TCR engagement has a very low activation energy.

- Still the cognate interaction has a very high sensitivity and selectivity.

-A very information intensive biological process

The cognate interaction generates a complicated set of events inside the cells:

Signaling cascades.

The signaling cascades translate a pattern of activation at the cell membrane into the triggering of some genetic activity inside the nucleus

More information processing is taking place…than we seem able to “model”

Signaling cascades

• “Evolutionary stables” (animal human models).

• Parallel processing of information ? – Maps the footprint of a stimulus at the membrane into a

pattern of genetic activity– The information is conveyed and processed through

physical interactions between molecules– the same stimulus can have different effects on different

cells.

• What can be known about the dynamics of the cascades and the Interactions between cascades?

NF-BNF-

CascadesCascadesShared kinases

Input Layer

Neural Net architecture?

Parsimony?

Cascades do not seem to involve so many factors considering the variety of footprints of excitation they convey to the genes

Where is the difference?

Information processing in cascades?

Filtering?

amplification?

Integration?

analysis of signal?

Biological form of Computation through direct interactions of molecules and complex formation?

Are there biological gates??

Boris N. Kholodenko :Eur. J. Biochem. 267, 1583±1588 (2000) Negative feedback and ultra-sensitivity can bring about oscillations in themitogen-activated protein kinase (MAPK) cascades

Modeling MAPK cascades (Huang-Ferrell: PNAS 93 (1996), pp. 10078-10083)

Goldbeter-Koshland 81

Why MAPKKK??

One fundamental mechanism:Michaelis-Menten

Aadk

AEka

dt

dB

A + E {E.A} E + Ba

dk

“Switch” or “ultra-sensitivity”

hh

a

a“Hill equation”:

Why MAPKKK??

Effects of cascades quite variable and complicated

- Dynamical properties sensitive to parameters (relative concentrations of enzymes) that are difficult to measure and one can assume quite variable

- May lead to the activation of many genes

Nuclear factors like NF-B or NFAT are involved in the activation of many different genes.

Gene transcription is a complicated process whose regulation involves a lot of factors

4 possible levels of control

Transcription: quite a regulated and complex process

…which involves quite a lot of factors…

“DNA sequences act as nucleation sites for the assembly of protein complexes”

“DNA-protein interactions are among the tightest and most specific molecular interactions known in biology”

The genes are the central compiler of the cells

• The product of the signaling cascades is conveyed to a system even more complex and regulated: the chromatine.

• Several percent of the genes regulate the chromatine.

• A large chunk of the energy consumed by the cell is for that regulation.

Structure of chromatine decides which genes can be activated.

Chromatine made of Nucleosomes

Chromatine remodeling

• Chromatine remodeling leads to the expression of different genes.

• This process involves dedicated nuclear proteins.

• It tends to be slower than mere transcription.• It takes place during differentiation• It leads to different patterns of gene activations• It is known to take place during thymic

maturation of the cells.• A safe assumption: It takes place during the

cognate interaction

Cell, Vol. 114, 277–280, August 8, 2003, Copyright 2003 by Cell PressMinireview Nuclear Receptors: A Rendezvousfor Chromatin Remodeling Factors

A new perspective on the cognate interaction and T-cell differentiation?

• Now that genomic arrays exist, it is possible to monitor at the genetic level what happens during the cognate interaction

• … and after• We can try to learn to analyze T cell specificity

on the basis of the new genes that are activated or not accessible anymore as a result of chromatine remodeling.

• D-cells also interesting to analyze and the effect of their activation and the information they convey to the T cells

Conjecture

• T-Cells and D-cells adaptive agents with two levels of adaptiveness:– Chromatin remodeling signals a change of

nature of activation for T-cells– Otherwise, T-cells act as “repetitive”

processors of the information

• What factors influence D-cells?

The immune system a Multi-intelligent-agent system?

• Interfaces with (i.e. could learn from and benefit to) the study of – Distributed intelligence systems, and/or– Distributed control with learning, i.e.– Systems with intelligent agent accessing,

filtering, evaluating, integrating information.

• Architecture of the distributed processing of information

• Mobile agents (codes)?

And compare with random graphs