Genetic memory vs natural selection

Pierre sonigo at
Sat Oct 9 05:46:25 EST 1999

Jay and Nancy Mone <jaymone at PAONLINE.COM> wrote :
> Pierre,
> In all of my years as a biology teacher, I have never heard of the term
> genetic memory.  What in the world is it?  As long as we are on the
> subject, exactly what is a non-valid evolutionary history??
> >From a purely evolutionary standpoint, genetic adaptation is a shprt term
> phenomenon.  Genetic variability can not anticipate future selective
> pressures.  Rather, natural selection for any phenotype provides
> advantages to the cell or organism only in the present environment.  If
> that environment changes, any particular pre-existing adaptation may or
> may not remain advantageous.
> Jay Mone'

Of course I agree with you ! Sorry I need such a long post to explain.

Many disciplines of biology are based today on the existence of a "genetic
program" : for example this program is held responsible for building the
organism from an egg, and then for organizing and regulating the
physiological functions. Interestingly, in this description, the program
precedes the functions. The program is the cause and the function is the

Now, let's shift to the evolutionary point of view you mentioned. I know
that genotype is only selected through the corresponding phenotype. When the
phenotype is selected, the genotype is "kept with it" and will favour the
reproduction (reproducibility) of the selected event. Thus, from the point
of view of natural selection, the genotype might be considered a memory of a
previously selected phenotype. This is why I used the term genetic memory,
which has been used by many authors already in place of terms such as
genetic program, genome, etc. to insist on the evolutionary point of view.
But it is the same concept, except that here, the function precedes the
"memory" . The function (ie the phenotype) is the cause and the program (ie
the genotype) is the effect.

Most scientists will agree that the function was first in evolution. However
in their understanding of embryonic development and cell functions for
example they consider the program is causal/ie the first. The question of
whether the function or the program was first might be considered as futile
because, in present organisms, it is a cycle : function precedes program
which precedes function etc. But the problem is tough for the first cycle,
ie the origin of the genetic program. This is why there are so many unsolved
controversies about the origin of genetic information.

A possible solution is to establish a direct connection between genotype and
phenotype, which is IMHO the tendency of modern neodarwinians like Dawkins.
In this conception, genotype and phenotype fuse. To be coherent with this,
the origin of life has to be explained by a fused genotype/phenotype : for
example the ribozyme of the RNA world. But this is not so easy to agree
with. In case you think this is a solved point, see for example discussions
on or The main problem is that in
multicellular organisms, between genotype end phenotype, we have so many
events (ontogenesis).

What is a non valid evolutionary history ? Exactly what you suggested ! In
my example, it was a story where natural selection does not work in the
present, here and now, to favour something that will occur later.

Let's apply this to cancer by considering evolution at the *cellular* level
: the question is why don't we have much more cancers ? Cells which
reproduce fast (cancer) should be strongly selected over slower cells
(normal cells). Why don't cells choose to cancerize much more frequently ?
Why do neurons stop reproducing ? It might be answered that cells do not go
to cancer because they will kill the organism and die too. In other words
they will not do something favorable now (cancer) to obtain an advantage
later (the survival and efficacy of the organism). The other solution (as
proposed for example by John Edser on is to refuse cells
as a valid selected unit.

I found no satisfactory answer to this problem. I mean in the litterature or
by asking specialized colleagues. On I was answered that
it works in models when the individual is mortal. However if individuals are
supposed to die approximately at the same period when cancer is likely to
occur, it cancels the difference between the present and the future and
there is no more problem. The selection still works in the present.

In the organism, there are billion of cells growing for decades and a highly
selected (high fitness) event like cancer should occur at every minute due
to selection at the cell level. The cancer should be "controled" by
something. If not we should all die from cancer at  the embryonic stage. The
current theories of cancer suppose that cancer is controled by the genome,
and occur only when the genetic control is "destroyed". I can provide
references (recent and respectable) if you want.

Thanks for your patience

More information about the Immuno mailing list