Biology Reference
In-Depth Information
Chapter 27 as a framework for discussing synthetic morphology; in that chapter, morphogen-
esis was reduced to the action of about a dozen different basic modules of cell behaviour such
as proliferation, migration, aggregation, and so on.
Mechanisms that may be considered to be 'morphogenetic modules', such as the assembly
of motile leading edges, or convergent extension, or activation of elective cell death, will be
associated with specific patterns of gene expression.
18
e
21
When these patterns have been
characterized in systems known to use these morphogenetic mechanisms, they can be sought
bioinformatically in microarray data obtained from systems undergoing morphogenetic
events that have not yet been characterized. Changes in the gene expression of the unchar-
acterized event that are similar to those seen when a known morphogenetic module is acti-
vated would naturally suggest that the same module is being invoked during the
uncharacterized event. This type of approach offers the apparent promise of a new and rapid
way to study morphogenesis, at least at the level of gaining hints that will help in the design
of rigorous experiments. It is not, however, foolproof because the whole idea of modularity is
only a metaphor, and life is not really so simple.
True multi-level, modular systems, such as those used in engineering and particularly in
software engineering, have a rule: high-level systems may use low-level modules in a variety
of ways, but they do not alter processes that are local to the low-level module. Computing
hardware is a typical example of a multi-level, modular system. Low-level systems, such as
electromechanical relays or semiconductor logic gates, are modules that can performprimitive
Boolean operations such as AND or OR. When connected together the right way, these
modules can produce a higher level module that can, for example, add two binary numbers
d
modules such as this can be combined together to create higher level machines right up to
a complete computer. At each new level, properties such as addition emerge that were not
present in the basic modules. Nevertheless, creation of different types of high-level machines
is done solely by modifying the connections between basic modules and never by altering the
workings of the modules themselves. Creation of a different high-level computing circuit in
this way does not involve, for example, the addition of an extra contact to a certain relay, or
of an extra transistor to a certain logic gate, but only the use of the same basic level modules
in a different way. Modern computer languages force programmers to use similar rules.
Low-level processes tend to be written as 'functions', and higher-level processes call them
self-contained functions by passing a defined set or parameters (for example, numbers to
be factorized) to them and accepting the results back. The writing of the high-level process
determines the order in which basic functions are called, but there is no means for a high-level
process to interfere directly with the way that a basic function works
d
to swap a '
' for a '
รพ
'in
a line of code in the function, for example.
Biology breaks this rule. New cell and tissue shapes and behaviours
can
evolve simply by
connecting existing basic morphogenetic modules in new ways, and that is the predominant
model for evolution at the moment.
22,23
It is clear, though, that the evolution of new forms can
also happen by tinkering with the functions of basic modules themselves, with or without
a change in the way that they are connected. Although the details of their evolution are
not known, the actin-polymerization mechanisms described in Chapter 6 are probably an
example of this. Addition of specific proteins such as fimbrin or espin to a 'make actin fila-
ments' module changes the effect of that module, and these proteins work as structural
components at the lowest levels of that module, not as high-level 'commands'.
