🚀 go-pugleaf

Gregory Turenchalk (GD 1999) (gturench@minerva.cis.yale.edu) wrote:
: Ari-Matti Sar�n (saren@operoni.helsinki.fi) wrote:

: : Count again. We have two cromosomes each (one from father, one from mother). 
: : Each chromosome has (during part of the cell cycle) two chromatids. That makes 
: : 4 alleles. From practical point of view we do have two alleles (one in each 
: : chromosome), but with point mutation and couple of billion truckloads of (bad) 
: : luck, we may talk about 4 different alleles. 

: : Yes, I'm nitpicking - no hard feelings, eh? :-)

: : Ari-Matti 

: Well, even if the universe aligned to create the incredibly unlikely 
: events that you describe, it would still not change the fact that a 
: normal diploid individual is never expressing more than two alleles of a 
: gene at a time. 

Okay, let's go for the killer example that shows just how delicate
the locus and allele definitions are in the face of biological reality
(i.e. abstract meets concrete and doesn't fare well).

The immunoglobulin (Ig) and T-cell receptor loci are both responsible for
generating enormous amounts of diversity in the immune system.  Each one
of us expresses thousands and thousands of different Ig and TCR proteins.
In addition, for reasons a little too much to go into here, it is
absolutely imperative that the relevant cells of the immune system 
express _one and only one_ Ig (or TCR, depending on the cell type).
How is this accomplished?  First, only one of the two copies of the
locus in question can be expressed; the other is shut down completely
(allelic exclusion).  Second, the variable portions of the antibody are
determined by an irreversible DNA splicing event.  Third, certain constant
regions are later altered by another irreversible DNA splicing event.
Finally, the antibody genes themselves are subject to a heightened mutation
rate and actually evolve during your lifetime.

So, how many Ig alleles and loci does a person have.  The germ line cells
don't undergo any of these processes, and so you could say we each have
a large, elaborate Ig locus.  But at the somatic level, where alleles really
make the difference, we have thousands of different phenotypes, each of 
which should have an allele -- and indeed there is a different DNA molecule
responsible for each different protein.


: A Downs syndrome individual would have three alleles for certain 
: genes since they are triploid for chromosome 21 


Another situation worth mentioning is tumor suppressor genes.  These
are genes whose normal functioning prevents cancer.  Mutant tumor suppressor
genes can lead to cancer, but only if both copies are knocked out.
If you inherit one nonfunctional allele and one functional one, you
are at a high risk for cancer because somatic mutation might change
the functional allele to a non-functional one, and while the chance
of this happening in any one cell is small, multiply that by oodles
of cells and you have a probability near 1.  But there's no strong
reason to expect the new mutant allele to be the same as the old,
and so the unlucky individual now has 3 alleles in all their body, though
any given cell has only 2.

: but the question was focusing on the status quo not the exceptions.

This is the key take-home message.  Locus and allele are abstract concepts
that work well for most situations most of the time if you generalize
a little bit.  But, they are simply not capable of encapsulating all
the beautiful complexity that exists in natural genetic systems.  So
use them in good health, but don't try too hard to squash them into
tight places.  They're just like all those wonderful simplifications
in physics -- after all, physical objects can't really be 0-dimensional
points, yet this approximation works pretty well for a lot of 
calculations.