Codons and HIV

TRKeske trkeske at
Sat Mar 14 23:39:05 EST 1998


This discussion of mutation rates and HIV may seem boring, but
it is worth paying attention.  It relates to some critical issues- 
not only the real age of the virus and the possibility that it may
have been man-made, but also the future of the epidemic and 
how far wide-spread it may become.

One of the issues that I have been debating is whether HIV is
likely to evolve backwards, closer again to SIV.  I have said
that is not likely, but have been arguing with a couple people
who claim otherwise, as in the following:

" > A constant mutation merely means that the pattern is
   > different from the last pattern that it held.

   > Suppose you have a single point that can be only one of
   > two things.  If it mutates twice, it will be back to its
   > original state.

   > Now, a genetic codon is three bases long and can contain
   > one of four possibilities.  This gives a total of 64 possible
   > combinations for a triplet codon.  Now, a constant mutation
   > rate not only will have the codon "moving back" toward
   > the original codon sequence, it is practically guaranteed
   > to happen.  There are only 64 possible sequences.  If the
   > sequence mutates 64 times, you are guaranteed to have a
   > duplicate.

   > HIV can be quite old.  Just because some mutates a lot
   > doesn't mean that it mutates very far.
   > Influenza and cold viruses have been around for centuries,
   > and yet they are some of the most rapidly mutating viruses
   > known.  I have heard that you can't get the same cold
    > twice."

I would hope that most people would know intuitively why the
above argument is nonsense.  This was not posted to an
audience of microbiologists, so I think that it would help to
define some terms, first:

   co-don (ko' dan) n. a small group of chemical units, believed
   to be a sequence of three nucleotides, that codes the 
   incorporation of a specific small group of amino acids into
   a protein molecule during the synthesis of a protein.
   Codons are present in DNA and RNA.

The above argument is silly because it is talking about having
64 possible combinations for just ONE codon. 
The only problem is that HIV-1 has (a-hem) 401,231 codons.
The total possible combinations, I estimate, are on the order
to 10 to the 308-th power.

Yes, if you have one codon change, it certainly could change
back, and move you "closer" to SIV.  However, the subtype
O virus was SIGNIFICANTLY closer to SIV, genetically,
than the other known strains, and there is also significant
genetic difference between HIV and SIV.  

The odds against  a large number of mutations all moving back
to older strain would be astronomical.  It is similar to the fact
of how human beings may surely continue to evolve into many
different forms, but never precisely into monkeys.

The statement about "never getting the same cold twice" hints
at the very contradiction.  You develop immunity to each
strain that infects you, yet you keep getting new infections,
regularly.  Why?  Because you constantly get new mutations
that are unlike what you saw before, virtually unlimited in

As I said, the mutation issue may be critical to the entire future
of the epidemic.

For example, an HIV subtype was found in Thailand in 1996 
that spreads much more easily through heterosexual intercourse. 
The subtype grows much more easily in the female reproductive
tract than does the HIV subtype that currently predominates in the

Another major issue is whether some mutations might result in
failure of current blood tests to screen the virus.  We have been
comforted by the assurance that our blood banks are virtually
100% safe.  If we start having to play catch-up, revising the
tests to cover new strains, we may have windows of vulnerability.

Inquiring minds would want to know- how many possible strains
of HIV might we expect to see, in what timeframe?    This would
seem like a simple enough question, but my attempts to search
literature on the subject have not yet yielded a satisfying answer.

I'm still researching, but I am fairly confident that the
possibilities for mutation will be astronomical, essentially
limitless.  Direct mutation is only one aspect of virus change-
there is also recombination that occurs among different
virus strains.

In general, it is possible for nearly anything to evolve into
anything else, given enough time. Human beings evolved from
a primordial soup of basic raw materials.  Counting the
possible combinations of codons is beside the point, in
that sense.

What makes for rapid evolution, however, is rapid mutation.

It appears that the concept of HIV lingering in the body, dormant
for long periods, may have been inaccurate.  It now appears
that there is a constant battle with the immune system, 
which is partially effective, but it is the HIV mutation rate
that statistically works to the long-term advantage of the

It is a concern that the same phenomenon that applies to
the HIV battle within a single individual, might also apply
to our collective, long-term battle with HIV as a species.

Similar to how the immune system tries to identify
invaders, and purge them, we try in the blood banks
to identify invasion by contaminated blood.  

An individual with HIV can go for a long time without
realizing the true nature of the long-term problem.  If
mutations were ever to start evading detection in blood
supply screening, as it evades detection by the human
immune system, we could be in more serious trouble
than we realize.

It would seem to be advisable that we continue studying
these issues, until we can be confident that we have a
very good handle on them.

Tom Keske
Boston, Mass.

More information about the Microbio mailing list