BjornsMethods/DevelopmentPrinciples

=3D Spatial Programming=

Problem:  How to program 3D spatial structures:
 * Is it possible to discover and utilize the proper hierarchy of organization and dynamics to compile global 3D spatial structures through the use of smaller cellular objects?  -->  Biology does it.  Can we learn from developmental biology?

Sub-Problems:
 * What are the levels of abstraction?
 * What are the interfaces between the levels of abstraction?

Definitions:


 * Agent = individual cell

Hierarchy Level II: Agent Behavioral Primitives

 * 1) replicate (cell division)
 * 2) die
 * 3) change shape/structure
 * 4) crawl
 * 5) adhesion to substrates/other cells
 * 6) emit chemical signals (diffusible or membrane bound)
 * 7) receive chemical signals via receptors; also electrical and mechanical sensors?
 * 8) store information (memory)
 * 9) process signal and current state information


 * NOTE:  All of the above primitives are accomplished via a lower level of abstraction: biochemical primitives

Hierarchy Level I: Biochemical Primitives

 * 1) Gene Regulation
 * 2) *Transcription Factors
 * 3) **activators
 * 4) **inhibitors
 * 5) *Chromatin modification ==> global gene accessibility
 * 6) **Methylation
 * 7) **Acytlation
 * 8) Protein modifications producing allosteric structural changes influencing protein function (Ex: on/off switch, change functionality)
 * 9) *Kinases/Phosphatase (Phosphorylation state)
 * 10) *Nucleotide Exchange Factors (GEF's, ATP additions, etc...)
 * 11) *Methylation & Acytlation
 * 12) Polymerization of identical subunits (Ex: tubulin --> Microtubules)
 * 13) Dimerization (homo/hetero-)
 * 14) Protein complexes to perform higher level function not capable via single proteins

Differentiation
One salient difference between amorphous computing and bio-cellular computing is that biology has devised an elegant hierarchical differentiation scheme to allow specialization and sub-specialization of cell types to distribute functions to a non-homogenous group of agents. Ultimately we want to meld differentiation into an amorphous-like computing system to utilize the power of hierarchical organization an heterogeneous populations.

A key component in the process of differentiation in biology is the idea I call a decision network. The purpose of a decision network is to perform a computation between communicating cells to determine the fates of the cells involved. A simple example of a decision network is the Delta-Notch ligand/receptor system found in nearly all organisms that performs a lateral inhibition role to prevent all cells within a local environment from differentiating into the same cell type. It is essentially a competition network where one cell wins and then subsequently inhibits the losers. Another decision network example is found in Dictyostelium prespore/prestalk (Psp/Pst) cell differentiation which is composed of a small negative feedback network acting as a homeostat that robustly produces a particular ratio of Psp/Pst cells (~75/25) within a given colony.

The utilization and understanding of decision networks are going to be a key interface in ultimately programming cellular differentiation within colonies of similar cells.

I argue that once a cell has made a decision what fate to take on, it wastes no time to act on that decision. For example, it is generally excepted that once a cell has decided it will commit suicide it will enter a fixed-action pattern (term stolen from behavioral neuroscience) to follow through to completion. What good is a half dead cell? Furthermore in Dictyostelium, once the cells have decided they will enter the social phase of their life, they immediately switch over to a new gene expression state (~25% of gene expression is modulated). This gene expression modulation occurs many times as the cells proceed with different morphogenetic transformations through development, but maintain a relatively stable gene expression pattern while no morphogenetic transformations are occurring. If it is true that significant state transitions occur in a crisp, decisive manner then it gives hope that genetic programming may be possible through the interaction of small decision networks and more global gene expression modules.

If you haven't noticed yet, I have absolutely no idea what I'm talking about and encourage any input/comments via email to: millard@mit.edu

Some Principles of Developmental Biology
Original List:


 * 1) Life vs. Death
 * 2) *Cell proliferation (replication)
 * 3) **Stem Cells - pleuri-potent cell replacers
 * 4) *Cell death (apotosis)
 * 5) *NOTE: Need both proliferation and death
 * 6) Differentiation
 * 7) *Cells change function and structure in a hierarchical way to build up a complex organism originating from a single cell
 * 8) Cell morphogenesis
 * 9) *Cell shape/structure change to facilitate global morphogenesis
 * 10) Hierarchical Organization
 * 11) *Cell differentiation results in reduction of cell potential
 * 12) *Many possible cell fates all stemming from initial zygote (root of tree)
 * 13) Induction
 * 14) *Cell communication that alters/induces cell fates
 * 15) *Often times there are inducing centers --> a particular cell or group of cells emitting the induction factor(s)
 * 16) *Community effect - if cell placed in certain environment, the cell may be induced to take on same fate as its neighbors
 * 17) Semi-modularity
 * 18) *Sub-division of tasks (tissues and organs), help organize functional units
 * 19) *Boundary formation - creates compartments to further order cells
 * 20) Cell sorting
 * 21) *Cells can sort into groups of like cells through selective adherence to each other and to the extracellular matrix
 * 22) Strong Attractor Systems
 * 23) *Allow fate determination to be relatively stable once arrived
 * 24) Competency
 * 25) *Differential cell abilities to respond to signals due to its current state
 * 26) Epigenetics
 * 27) *Chromatin structure and regulation of gene expression --> mechanism for changing competency and differentiation
 * 28) Combinatorial code of gene expression
 * 29) * With limited number of genes/signals cells reuse same signals to mean different things given different competencies of cells listening
 * 30) Maternal factors
 * 31) *Inherited directly from mother directly from mother cell
 * 32) *Inherited regulatory molecules can effect cell fate autonomously without induction (determinants)
 * 33) Lateral inhibition
 * 34) *Neighbors prevented from taking on same fate
 * 35) Analog to digital conversion through nonlinear thresholding
 * 36) *Cells respond differently to the same signal when it is presented at different concentrations (morphogen).
 * 37) *Continuous information contained within a concentration gradient is converted to a discrete outcome through thresholds of activation.
 * 38) Cell migration (motility)
 * 39) *Often regulated by signaling molecules
 * 40) Cell adhesion
 * 41) *Cells capable of adhering/pulling on other cells, surfaces, and extracellular matrices through surface proteins
 * 42) Cell sensing of self and environment
 * 43) *Senses and interprets chemical, electrical and mechanical signals from other cells and environment
 * 44) Cellular information processing
 * 45) Cells process information based self state and environmental state
 * 46) Morphogenesis
 * 47) *Taking on 3D form
 * 48) *Typically embryos form 3D structures by first forming 2D sheets and subsequently folding
 * 49) *3D structure is often required for organ functionality
 * 50) Genetic pattern formation
 * 51) *Spatially and temporally regulated differential gene expression
 * 52) *Stimulatory vs. Inhibitory signaling (positive & neg feedback)
 * 53) Evolution
 * 54) *Conservation of certain gene functions through evolution (homologs, orthologs)
 * 55) Cell memory
 * 56) *Cell has ways of remembering past events by changing its current state based on experience