Tim writes, following me:

>> However, it would seem very unlikely that it could act as an error-recovery
>> mechanism. When you only have two copies of anything -- and they disagree
>--
>> there is no simple way to determine which, if either, copy is correct.
>
>That doesn't matter:
>
>If you have suffered a point mutation and then swap genes at random when 
>generating sperm, you will produce 1/4 double-mutated offspring, 1/2 
>single mutated offspring - and 1/4 offspring with no mutations.
>
>That is enough to undo deleterious point mutations - and reverse the 
>effect of Muller's ratchet.

There's been a great deal written on this subject here in the last week, but
not all of it was quite correct. In that regard, let me correct Tim's few
sentences.

Firstly, there is no recombination on the Y chromosome, thus there can be no
"gene swapping." Recombination occurs between sister chromatids of two
homologous chromosomes, of which the Y has none. 

The most parsimonious explanation for the general phenomenon of recombination
is that it allows the probabilistic breakage of linkage disequilbria, providing
at least the possibility of freeing a particularly beneficial allele of a
particular gene from eternally being condemned to residing near a particularly
deleterious allele of another gene. Recombination is not so much an error
recovery mechanism as it is simply a repackaging process. If every gene were
encoded on its own chromosome, recombination would be unnecessary.

Nevertheless, this mild form of error recovery isn't possible on a single
unpaired chromosome, absent of a homolog. Nor is such gene swapping possible on
a palindromically encoded gene sequence, at least to the best of anyone's
knowledge. Palindromic genes form in this fashion:

  http://e107.bioinfo.nctu.edu.tw/~biology/images_g/Fig%2010-21b.GIF

The lower figure of the two shown is called a "cruciform"
construct. Cruciforms
readily form in *in vitro* DNA, but they are generally thought not to form *in
vivo*, although no one knows that to be true with great certainty at the
moment.

Palindromic cruciforms readily form in vitro because (i) the two sides of the
palindromes are a perfect match for one another and are thus attractive to one
another, and (ii) they are given the structural freedom to do so.

If one side of the palindrome should suffer a point mutation, a perfect
crucifomic match would no longer be possible, as shown by the sequences of the
two bases at the center of the illustrated palindrome. Is error recovery
possible in such a case? No. The question following the mutation would be:
which is the correct base and which is the one in error? There's simply no way
of telling.

However, a perfect palindrome offers the remote chance of offering an error
checking pass/fail methodology. If a perfect palindrome were expected but was
detected to no longer be perfect by some error checking process, the entire
chromosome could be discarded without significant loss. There are certainly an
abundance of alternatives in every ejaculate wad, but discarding known
defectives would greatly slow the entropic decay of the few remaining,
presumably critical genes left on the Y.

There is absolutely no evidence at the moment that such an error checking
mechanism exists for the Y's palindromes and it may well be found that the
palindromes serve no purpose at all.

Wirt Atmar
---
þ RIMEGate(tm)/RGXPost V1.14 at BBSWORLD * Info{at}bbsworld.com

---
 * RIMEGate(tm)V10.2áÿ* RelayNet(tm) NNTP Gateway * MoonDog BBS
 * RgateImp.MoonDog.BBS at 2/24/04 8:52:03 PM