Biology Reference
In-Depth Information
majority of cases, the wild type eventually out-grew the mutants, though it took
many generations before this happened, implying that the selective disadvantage
suffered by the mutants was very small, such that their effects on the phenotype,
though unmeasurable in any short-term comparison, are sufficient to account for
the maintenance of the genes in the wild type. Even if this is only true for a
subset of the genes with no overt phenotype, its profound implications do not
seem to have been absorbed. How can a hypothesis for the link between the
activity of a gene product and the selective advantage it offers the organism be
framed and tested when there is such a small difference in the properties and
behaviour of an organism whether it is present or absent?
There is no reason to believe that such results are unique for yeast. Mice with
the myoglobin gene knocked out (Garry et al., 1998) had an unchanged exercise
capacity and a normal ventilatory response to hypoxia; their muscles appeared
to function normally. It turned out that there were multiple differences in the
anatomy and biochemistry of the mutant mice that compensated for the lack of
myoglobin (Grange et al., 2001), but that does not change the fact that it is now
much more difficult to make a simple statement about the function of myoglobin
in mammalian muscle.
It is possible to take this argument further and ask why there should be
any expectation that it ought to be possible to link the activities of any and
every component of a biological system with a discrete function. The fact that
biologists think that this is a reasonable approach to understanding function
was satirised by Lazebnik (2002) in his article 'Can a biologist fix a radio?'.
His point was that a radio can only be understood in terms of its characteristic
functional modules, but within the modules, the functions were implemented by
the interactions between the components, and most of the characteristics of the
components had no relevance to their contribution to the module. His criticism
of biologists for their obsession with minute description of the characteristics
(colour, morphology, behaviour of the system when the component is removed)
was apt, but does it follow that living cells ought to be analysed by delineating
the functional modules within? The function served by a radio, as for any other
artefact, is known from the start; its modular structure results from both the
designer and the constructor consciously choosing to break the problem of
making the radio into a set of discrete sub-problems. This is not the route taken
by biological evolution. In fact, when biologically inspired genetic algorithms are
used to computationally evolve electronic circuits to fulfil a specified function
(Bennett et al., 2000), the performance of the sub-tasks becomes a single unified
process, and although known circuit designs can be 'rediscovered' in this way,
entirely novel designs of a dense, non modular nature can be produced (Koza
et al., 1999). If this is a fair model of the evolution of biological networks to
implement functions, it illustrates there is no reason why we should expect to
recognise any similarity between the arrangement of components in the network
