Biology Reference
In-Depth Information
the component itself)
1
are not necessarily certain; attributed functions are
themselves hypotheses that need to be tested before they are used as axioms
for the development of further theories. In the case of feedback inhibition,
the initial assumption was made too readily that the control of metabolic
rate (an observable property of the system formed by an enzyme network)
was the function that the inhibitory activity served. Designation of enzymes
as the rate-limiting steps of pathways was another such error, attributing a
function to such enzymes that in fact they could not perform. This issue of
identification of function will be considered further below. In attempting
to show how the generic properties of metabolic systems could have arisen
through evolution by natural selection (as described in the previous point),
it is important not to commit the same mistake: any claim for a potential
contribution of some aspect of the activities of metabolic networks to the
fitness of an organism (the system of which metabolism is a component
part), such that it has therefore been subject to favourable selection, is also
a hypothesis that needs to be tested.
I have stated that I believe systems analysis of metabolism has improved our
understanding, but could this be over-confidence? What are the limitations to
our analyses? One limitation is that the models we analyse, whether theoretically
or experimentally, are very much more simple than the metabolism of a cell,
in many respects. Even the structure of the pathways we consider is simplified;
much of the time we consider a few steps, most often in a linear sequence, and
occasionally with a single branch-point or in a cycle. Further, most of the steps
are represented as conversions of a single substrate to a single product, whereas
enzymes that work in this way are a minority; three quarters or more use more
than one substrate or yield more than one product, or both. The cosubstrates
and coproducts in such reactions create links between different metabolic path-
ways and couple the fluxes of carbon to fluxes of nitrogen, sulfur, phosphate
and electrons. We often assume that we can isolate purely metabolic levels of
control from the control and regulation of metabolism by signal transduction,
gene transcription and translation, post-translational modification, and enzyme
degradation. It is true that it is possible in principle to extend our analyses to
include such phenomena (e.g. Kahn & Westerhoff, 1991; Small & Fell, 1990;
ter Kuile & Westerhoff, 2001), but experiments to test the conclusions become
more difficult. On top of this, it is impossible to adequately represent the full
1
'Function' can be used in different senses in biology; at one level it can be used to refer to activities
performed by components of a biological system, but it also extends to properties the system gains from
having the component. I am not including the former at all in my use of the term function, only the latter.
Wouters (2003) divides these higher level aspects of function into three further categories, but I do not wish
to be that specific here given that we are not being prescriptive about what level of organisation constitutes a
system in systems biology.
