Biology Reference
In-Depth Information
6.5.1
Networks with noise
The stochastic expression of genes is a phenomenon that has been
undeniably demonstrated. However a great many researchers are
divided in their adherence to two different interpretations. The first
interpretation takes into account the properties of networks of bio-
chemical reactions while the second ensues from analysis of the
structure of chromatin. Neither of these conceptions accounts for
the whole phenomenon. According to the first interpretation, sto-
chastic expression is caused by disturbances which affect the way
gene networks function. We know that between two cells there are
always inevitable little differences in the concentration of molecules
and in the speeds of chemical reactions. These fluctuations may
have considerable consequences if they affect molecules present in
low concentrations, the activity threshold of which is itself fairly
low. The probability of exceeding the activity threshold in certain
cells and not in others is all the greater the lower the activity
threshold of the molecule relative to the fluctuations in concentra-
tion of that molecule. For a transcription factor which has to acti-
vate genes, the transcription noise which results from the
fluctuations can lead to the stochastic expression of these genes and
produce heterogeneity of types in a population of cells which were
identical at the start.
This differentiation will be all the easier if there are regulatory
loops in the gene network (Fig. 28; McAdams and Arkin, 1997,
1999; Kepler and Elston, 2001). This conception thus accepts
stochastic expression without challenging the theory of genetic pro-
gramming. It reduces randomness to noise that improves the way
the programme works by permitting cell lines to bifurcate. Yet it
still assumes that there is a microscopic network of intermolecular
relationships determined by the stereospecificity of proteins, and
that this network is an order underlying macroscopic cellular organ-
isation. It does not therefore challenge the molecular biology princi-
ple of order from order. Here, the way the genetic programme
functions is as a deterministic phenomenon with noise (chapter 2,
ยง2.2.4). Now, we have seen that it is difficult to hold such a view
given the generalised non-specificity which affects proteins, including
