Re: Mitochondrial Eve
Fri, 21 May 1999 13:24:47 EDT

In a message dated 5/20/99 7:26:24 PM Mountain Daylight Time,

> On Wed, 19 May 1999 wrote:
> > I didn't say that a gene can change and still be exactly the same gene.
> > In fact I said the exact opposite: "Of course if genes change then so
> > can their mutation rates, but **if they change they are technically no
> > longer the same genes they were before they changed**." I'll say
> > that again: a gene that changes is no longer the same as it was
> > before. So a gene that changes is NOT exactly the same gene. If
> > a gene changes then it becomes a new gene, and so may have a
> > new mutation rate. But if a gene does not change then it stays
> > the same old gene and its mutation rate does not change.
> I agree with Steve Clark. This argument is circular. I look
> forward to see your reply to his comments.

I already replied to him the day before yesterday; perhaps you saw it after
reading my reply to you. So far he has not responded, probably because he
realizes he jumped to alot of silly conclusions and prefers not to call
attention to that fact. However, he said nothing about what we are
discussing, nor did he say that any of my arguments are circular. He
objected to my statement that alleles of microsatellites are different genes;
as I admitted he was right to criticize that because I had been ignorant
about what microsatellites were and had misinterpreted it because of the term
allele. I am not familiar with the use of the term allele for anything other
than true genes, so when I saw allele I assumed the authors were discussing
true genes. As it turned out I was wrong.

However, that has no bearing on what we are debating. Let me try one more
time to explain it. Let's use a hypothetical example. You have three genes.
Gene A has a mutation rate of 2, gene B has a mutation rate of 3 and gene C
has a mutation rate of 4. We let each gene sit for a million years then look
at them again. We find that gene A is unchanged and that its mutation rate
is still 2. We also find that gene B has changed into a new gene, D, but
that its mutation rate is still 3. And we find that gene C has also changed
into a new gene, M, and that its mutation rate has increased to 4.5.

Does it make sense now? Because the mutation rate for any gene is largely
constant, the only way the mutation rate can change is if the gene changes in
some fashion. If a gene does not change then its mutation rate will not
change; hence in our example the mutation rate of gene A does not change even
after a million years because gene A has itself remained unchanged. If a
gene does change, it may or may not undergo a change in mutation rate; hence
in our example gene B mutates into gene D but its mutation rate remains
unchanged, whereas gene C mutates into gene M and as a result its mutation
rate changes from 4 to 4.5. As Steve Clark pointed out different genes do
not necessarily have different mutation rates (see more about this later),
but otherwise the only way a gene could change its mutation rate is either by
mutating into a different gene or by moving to a different place in the
genome (see later). If the gene in fact does not change, its mutation rate
will not change on its own. Nothing about this is circular.

> > > The definition that I've learned and that is in Futuyma's textbook
> > >
> > > "Allele: One of the several forms of the same gene, presumably
> > > differing
> > > by mutation of the DNA sequence, and capable of segregating as a unit
> > > Mendelian factor. ... ...DNA sequence variants, that may differ at
> > > several or many sites, are usually called haplotypes."
> > >
> >
> > Ah, now I understand the source of your confusion. In the above
> > definition
> > the phrase "one of several forms of the same gene" refers to the fact
> > they all produce the same basic product, though the products may differ
> > slightly depending upon the mutation involved. It does not literally
> > however, that the alleles are all the same gene. The phrase "capable of
> > segregating as a unit Mendelian factor" tells us that each allele can be
> > treated as a separate inheritable unit, subject to Mendelian laws just
> > like
> > any other gene. If in fact alleles were literally all the same gene
> > they could not act as separable Mendelian units, because they would be
> > indistinguishable. An allele is fundamentally a separate gene, so it
> > have its own mutation rate.
> I'm sorry, but this is not what the text is saying.

That is exactly what the text is saying; see below.

> In fact, I am not an ignorant in this matter....

But you do seem to be unaware of certain basic concepts that should be known
to any trained geneticist; see below.

> ...and I've NEVER seen any paper or book in which
> your definition is used.

That's what I mean about being unaware of certain basic concepts. I am not
proposing a new definition; rather I am taking the definition you have
offerred and explaining what it means in light of these basic concepts.
Perhaps I should get real elimentary here. Let me try one more time to
explain all this.

One of Mendel's breakthroughs was the recognition that traits are determined
by a pair of inherited factors. Take for example the trait of seed color in
peas. This trait takes one of two forms: yellow (determined by factor Y) or
green (determined by factor y). As such, an individual plant could have two
yellow factors (YY), two green factors (yy) or a yellow and green factor
(Yy). In other words, each trait was determined by two or more factors, each
of which could be thought of as a different form of the same trait. Yet
clearly Mendel considered each factor to be a separate inherited particle
which when combined in pairs determined what form the trait would take based
on which factor was dominant over the other. Later these factors were called
genes, and the term allele was coined to refer to any gene that determined
one particular form of a trait. In other words, each trait was now said to
be determined by two or more genes, each of which determined a different form
of the trait, and these genes were referred to as alleles. Seed color is
determined by two alleles, Y and y, both of which nonetheless were separate
genes and acted as independent inherited factors. So from the very beginning
of modern genetics, alleles were considered to be separate genes, inherited
as independent particles of heredity.

Soon after this it was recognized that each gene resided in a specific
location (called a locus) on a specific chromosome, and that each cell
contained two copies of each chromosome. Though each chromosome could only
carry a single copy of any one gene, the two chromosomes together explained
Mendel's observations. They also explained how each allele could act as an
independent inherited factor, since the individual organism would acquire one
chromosome with one allele from one parent and the second chromosome with the
other allele (either the same of different) from the other parent. It was at
this time that the phrasing "different form of the same gene" for allele came
into common use, because the term gene in this case was being loosely used to
refered to the specific locus the alleles occupied. Hence if a specific
locus always contained a specific gene, then an allele was a specific form of
that gene that would reside at that specific locus. However, it was still
recognized that each allele was itself a separate gene, capable of being
inherited as an independent factor.

Later it was recognized that each gene coded for a specific protein (nowadays
we say that each gene codes for a specific functional domain within the
protein). That reinforced the idea of alleles being different forms of the
same gene, since people tended to started using gene to refer to the normal
protein and allele to refer to an altered or abnormal protein. For example,
if hemoglobin beta chain is the normal gene, then the mutated gene that
produces the altered beta chain that produces sickle-cell hemoglobin would be
the abnormal allele. Even so, despite the phrasing, it was still recognized
that the hemoglobin S chain allele was a separate gene from the hemoglobin
beta chain allele and could be inherited as an independent factor.

Finally of course it was recognized that a gene was a specific sequence of
DNA nucleotides. This lead to the use of allele to refer to a sequence that
differed from another by one or more substitutions. As with gene-as-locus,
alleles were referred to as different forms of the same gene; as with
gene-as-protein, the gene was seen as the standard (in genetics the term
"wild type" is often used to refer to this standard sequence) and so any
differing sequence was referred to as an allele to distinguish it from the
standard. Even so, molecular geneticists still accept the basic Mendelian
concept that all alleles are separate genes (obviously excluding the allelic
variation of microsatellites and other non-gene sequences) and that they are
inherited as independent factors according to the Mendelian mechanisms of
genetics (though they are also influenced by non-Mendelian mechanisms as

So the definition of allele used by Futuyma (one of several forms of the same
gene, presumably differing by mutation of the DNA sequence, and capable of
segregating as a unit Mendelian factor) means that it is one of several DNA
sequences, presumably differing by mutation, that occupy a specific locus and
determine the form of a specific trait by coding for a specific functional
domain (that differs from the functional domains produced by the other
allelic sequences) of a polypeptide, which can be inherited as an independent
Mendelian factor from the other allelic sequences (ie. as a separate gene).
The people who write the books and papers understand all that, even if they
do not spell it out; the molecular geneticists who read those books and
papers understand all that as well. Apparently, the reason why you do not
understand all that is because you are unfamiliar with the basic concepts.

> My limited mind can't understand how is it
> possible for a gene to have a constant mutation rate if every time it
> mutates (even a single IN-DEL in your definition) it is a different
> gene!

Exactly the point I am trying to make. As long as a gene does not mutate, as
long as it remains unchanged, its mutation rate should remain constant, but
once it does change then its mutation rate could change if the conditions are
right. However, Steve Clark has pointed out that different genes do not
necessarily have different mutation rates, so in fact a gene could mutate,
even regularly, and yet its mutation rate could remain unchanged. According
to Wen-Hsiung Li and Dan Graur in their book _Fundamentals of Molecular
Evolution_ (Sinauer Associates, Massachusetts, 1991, pg. 76-77), "[t]he
assumption of equal rate of mutation for different genes may not hold in this
case [the large variation in the rates of nonsynonymous substitution among
genes], since different regions of the genome may have different propensities
to mutate. Wolfe et al. [Wolfe KH, Sharp PM, Li W-H. "Mutation rates differ
among regions of the mammalian genome." _Nature_ 337:283-285] suggested that
different regions of the mammalian nuclear genome may differ from each other
by a factor of two in their rates of mutation." So if gene A sits at locus
in the genome where the mutation rate is 5 and then mutates into a new gene,
its mutation rate would probably change only slightly, if it changed at all.
However, if gene A were transposed to a new locus where the mutation rate can
be as high as 10, then mutates, its mutation rate could conceivably change by
a large amount, again assuming it changes at all.

So in fact it is very possible for a gene to mutate continuously over a long
period of time without significantly changing its rate of mutation, as long
as it does not change location in the genome. What is more likely to happen
is that by mutating the gene changes its selection intensity, thereby
changing its rate of substitution. This is most likely what you are
referring to, though you may not be aware of it. See Li and Graur for more
details (pg. 67-98).

> > >
> > > I've never read a paper where the authors referred to different
> > > haplotypes
> > > of the same gene as different "genes." Since this is not my area of
> > > expertise (I'm a behavioral ecologist), I recognize that I may be
> > This is because they usually are not thinking in terms of strict
> > inheritence, but in terms of the product being produced. This is one of
> > those cases in science where the scientists themselves use the wrong
> > phrasing
> > and nomenclature for convenience, but are nonetheless well aware of what
> > they
> > really mean. It can, however, be confusing to someone who is not
> > with the subject.
> Could you cite any paper in which your "definition" was applied? I don't
> think so.

Of course not. For one thing, it is not my definition. For another, few
books or papers feel the need to take the time to explain the basic concepts
to readers who are supposed to know them already. However, in _The Cell: A
Molecular Approach_ by Geoffrey M. Cooper (ASM Press/Sinauer Associates,
Washington DC/Massachusetts, 1997) on page 88 you will find a discussion of
gene and allele which uses the terms exactly as I have used them; see earlier

> I don't consider myself unfamiliar with the subject. Actually
> someone doesn't have to be familiar with the field to notice that your
> definition doesn't make sense.

That is probably the most specious argument I have ever seen. It is in fact
because you are unfamiliar with the basic concepts (see earlier) that my
explanation of the definition makes no sense to you. I would suggest that
you read some basic genetics texts to familiarize yourself with the subject
under discussion.

Kevin L. O'Brien