This is only partly correct. It is doubtful that any heritable mutation
would be adequate to provide a mammalian defense system capable of adapting
to newly arising pathogens. With their faster generation times, viruses et
al. often stay one convenient step ahead of even the most adaptable
component the immune system, which is the antigen receptor genes that
undergo a high rate of somatic mutation when they encounter foreign antigen.
However, this somatic mutation is not heritable.
A more proper view of the high rate of mutation in MHC genes is that this
provides a SPECIES with a highly polymorphic set of genes. With this the
POPULATION is able to respond to a greater repertoire of pathogenic insult
than any individual within that population could. Thus, the fact that the
MHC is highly polymorphic and mutable does not really mean that individuals
adapt to new pathogenic situations by mutating MHC genes.
>Steve Clark said something that struck me as interesting:
Well, I try.
>>MHC loci display what Jan Klein once claimed to be the highest
>>rate of GERMLINE mutation in the mammalian genome. This mutation occurs
>>via a process called gene conversion in which portions of other MHC-like
>>pseudogenes are exchanged with homologous regions of expressed MHC genes.
>>This is a special type of mutational mechanism that does not seem to be used
>>routinely by mammalians. However, it is more common in lower species like
>>fungus.
>
>And,
>
>>The lesson from the MHC system is that different genetic regions may have
>>different intrinsic mutation rates. This brings into play the different
>>mechanisms that can account for mutation, such as gene conversion which play
>>an important role in creating great heterogeneity in the MHC genes of many
>>(not all) species.
>Thus, MHC genes do not have just the normal "mutation" -- they do gene
>conversions. (Gene conversion is no ordinary mutation, rather it morphs
>information together from two separate sequences.)
What of the following are normal germline mutations. repair coupled
mutation, copy error, recombination, transposition, or conversion? The
point is that they are all functional and normal, and account for mutational
change in mammalian genomes.
>Recently I overheard that some pseudogenes are the result of, in some sense
>inserted by, viruses.
Transposable elements would be a preferred term. This includes retroviruses
as well as other things.
And that caused me to speculate a mechanism that
>'closes the loop' on the rapid acquisition of immunity.
The only immunity that is rapidly acquired via genetic processes is through
SOMATIC mutation of antigen receptor variable regions. These mutations are
NOT heritable. The rapid rate of MHC germline mutation, which IS heritable,
is only rapid relative to the mutation rate in other sequences of the
genome. MHC mutation does not occur anywhere near the rate needed to keep
up with antigenic drift of many pathogens. So there is no loop on the rapid
acquisition of immunity--as far as heritable changes go.
Imagine it like
>this. A viral infection might eventually spread to the germ cells, where a
>virus inserts some genetic material. Eventually (perhaps through a skip and
>a jump) some of this same genetic material is merged into the MHC genes via
>gene conversion. The MHC genes would then contain "information" about the
>virus that it did not have before -- information that would be of particular
>help to the IMMUNE system, of which the MHC genes are a part. The immune
>system would then have new information for identifying, and fighting future
>infections from that family of viruses.
>
>The information flow would be like this:
>
>virus ----> pseudogene ---> ... ---> gene conversion into an MHC gene --->
>immune system
>
>This cycle might have to be traversed many times for the immune system to
>fully acquire the necessary information on the virus.
>
>I suspect Steve may be shaking his head in dismay at my taking such
>liberties with the language and concepts of his field.
Yup. You get an A for originality, but originality is not much use if it is
not plausible. Sorry Walter, but you get a F for plausibility here.
First of all, while the origin of MHC genes is speculative, I don't think
anyone has described the formation of new genes, as proposed in Walter's
scenario. What the high rate of mutation does here is create a large number
of alleles. Humans have 6 class I MHC genes, each of which has many alleles
that likely arose via mutation.
Thus, it is unlikely that MHC genes originate from viral genomes as Walter
proposes above. From an evolutionary standpoint, the origins of MHC genes
would be considered to be an ancient event since mice, rabbits, etc have
gene complexes very similar to that of humans (this is perhaps one of most
remarkable examples of genetic homology between divergent species because
the similarities occur at multiple levels of complexity that are not present
in most genes). Thus, the role for gene conversion in this system, is to
modify the sequence of existing MHC genes and not to change a viral genome
into an MHC gene. By and large, the genes in the complex are fixed in
number, and new ones are not created spontaneously. This may be similar to
the distinction between micro and macro evolution.
Steve
____________________________________________________________________________
Steven S. Clark, Ph.D. Phone: (608) 263-9137
Associate Professor FAX: (608) 263-4226
Dept. of Human Oncology and email: ssclark@facstaff.wisc.edu
UW Comprehensive Cancer Ctr
University of Wisconsin "To disdain philosophy is really to
Madison, WI 53792 be a philosopher." Blaise Pascal
____________________________________________________________________________