Biology Reference
In-Depth Information
In this model, these characteristics of embryogenesis are therefore
determined by the structure of the DNA which acts as a random
protein generator and not like a deterministic genetic programme.
Natural selection optimises the way this model functions by sorting
the genetic recombinations, which are produced in each generation
and change the relative position of the genes so that the probabili-
ties of activation sequences corresponding to cell lineages of the
embryo are themselves optimised and the overall viability of the
organism increases. However, we must emphasise the fact that this is
a random phenomenon which only leads, as regards gene expression,
to statistical frequencies of expression in the cells, i.e. it allows varia-
tions in expression between the cells which give rise to differentiation
between the cell lines but it is not the sole agent of it. The probabil-
ities for gene expression are optimised, but differentiation also involves
cell selection which makes the process even more effective.
We have considered the case of a protein regulator. The model
functions identically whatever the chemical nature of the regulator
element. If it is a DNA sequence situated far away on the chromo-
some (Fig. 31), it is the random folding of the DNA molecule by
diffusion which determines the probabilities of gene activation.
Obviously this is an extremely simple example, but it nevertheless
reflects the case of the genes for globin. Their expression is regu-
lated by a DNA sequence situated at a distance on the chromosome
called the locus control region (LCR). In the course of development,
they are sequentially expressed in the same order as their positions
on the chromosome, relative to this LCR. This regulation is due to
a phenomenon in which the globin genes randomly compete for the
LCR (Townes and Behringer, 1990) depending on their position on
the chromosome (Hanscombe
et al
., 1991; Kupiec, 1996).
Of course, there are many more transcription regulators and
genes in a cell nucleus than in our example, but the same proba-
bilistic mode of functioning can be applied overall whatever
the number of molecules or their chemical nature. From the
moment there are fewer transcription regulator molecules than
their DNA binding sequences, these regulatory molecules must
compete for their DNA binding sequences, leading to numerous
