Dear Violeta and Pentcho:
I think that the evidence is that immunoglobulin (Ig) hypermutation is a
"Darwinian" evolution.
Certainly the (poorly understood) mutational process is directed or
targeted, in
the sense that it only occurs in the variable region of active Ig genes.
There
appears to be an elaborate machinery for the targeting of this mutational
process, and there are no examples to date of it being targeted to any other
parts of the genome.
However, the apparent targeting of mutations in CDRs is for the most part
due to
selection after mutation for those cells that are more "fit", that is, that
now
encode antibodies with higher affinity. The different mutated offspring of a
initially clonal cell arise in a special structure called a "germinal center",
and compete with each other within the germinal center for binding to a
limiting
amount of antigen. Only mutations in CDRs are likely to make a difference to
binding of antigen, while mutations elsewhere in the gene are likely to
harm the
antibody by affecting folding/stability; cells bearing the latter mutations
will
tend to be eliminated from the population by selection.
In artificial cases where genes are subject to the mutational process without
subsequent selection, the mutations are widely distributed (though still
only in
the targeted region).
Certain locations, termed "hotspots", are more prone to mutation than others.
This seems to be due to some sequence specificity of the mutational process:
certain four-base sequences are more frequently mutated than others. Ig
germline
genes have evolved to take advantage of this by placing hotspot sequences
within
some CDRs, where mutations are more likely to be useful. This is the closest
thing I can think of to the "Lamarckian" idea, but I would argue that it is
still
the result of Darwinian processes.
I did not understand what Violeta meant by:
> hypermutations in the CDRs are neutral (in Kimura's sense) with respect to a
> posteriory established improved antigen binding properties of the antibodies?
As regards Pentcho's idea of environmental factors "finding" and
interacting with
specific parts of the genome: in one sense this is obviously true -
organisms go
to great lengths to arrange via protein sensors and signalling cascades that
external events lead to changes in gene expression (and less commonly,
alterations in primary DNA structure). But perhaps one can imagine situations
where the "finding and interacting" occurs in the absence of
pre-programming by
the organism?
Cheers,
- Peter
---------------------------------------------------------
Peter Wang, M.D., Ph.D.
MRC Centre for Protein Engineering,
Hills Road, Cambridge, CB2 2QH, England
Tel (01223) 402104 (international calls +44-1223-402104)
Fax (01223) 402140 ( " " +44-1223-402140)
---------------------------------------------------------
