Outrunning genetic determinism?

Tara Parker-Pope writes in the NYTimes that exercise keeps you young.  Or rather, she reports on the results of an experiment in mice that had been engineered to have malfunctioning mitochondrial repair enzymes, recently published in PNAS.  Mitochondria are cellular power generators, and as we age, if they become damaged, and are left unrepaired, the cells they power can falter or die, leading to all the effects of aging that most of us will experience sooner or later; muscles become less powerful, our cortex shrinks, hair turns grey or falls out, skin becomes wrinkled, and so forth. As Parker-Pope reports:

The mice that Dr. Tarnopolsky and his colleagues used lacked the primary mitochondrial repair mechanism, so they developed malfunctioning mitochondria early in their lives, as early as 3 months of age, the human equivalent of age 20. By the time they reached 8 months, or their early 60s in human terms, the animals were extremely frail and decrepit, with spindly muscles, shrunken brains, enlarged hearts, shriveled gonads and patchy, graying fur. Listless, they barely moved around their cages. All were dead before reaching a year of age.

Except the mice that exercised.

Some of the mice were made to run on a treadmill the mouse equivalent of a 10K race in 50-55 minutes, 3 times a week.

At 8 months, when their sedentary lab mates were bald, frail and dying, the running rats remained youthful. They had full pelts of dark fur, no salt-and-pepper shadings. They also had maintained almost all of their muscle mass and brain volume. Their gonads were normal, as were their hearts. They could balance on narrow rods, the showoffs.

But perhaps most remarkable, although they still harbored the mutation that should have affected mitochondrial repair, they had more mitochondria over all and far fewer with mutations than the sedentary mice had. At 1 year, none of the exercising mice had died of natural causes. (Some were sacrificed to compare their cellular health to that of the unexercised mice, all of whom were, by that age, dead.)

That is, genetic determinism isn't necessarily so deterministic after all!  This of course is one of our mantras here in MT.  And this is an interesting example because the kinds of single gene mutations with the strong, clear effects seen in these mice can be the mutations with the most deleterious effects of all genetic mutations.  Tay Sachs disease, cystic fibrosis, PKU, and so on are all single gene diseases with very deleterious effects.  Nobody--not even us!--denies this kind of genetic causation. 
And yet.....

Most cases of PKU are due to mutations in a gene called PAH.  These can cause severe mental retardation if left unchecked, but reducing a single amino acid from the person's diet will ameliorate or in many cases prevent the effect.  So here too, genetic determinism becomes environmental interaction, as with these mice. And, to complicate the notion of genetic determinism even more, there are many different alleles in the PAH gene and the actual phenotype varies very much and only for some alleles is it highly predictable from the genotype (and a couple of other genes are known to affect severity).  Most other  genetic disorders, even the 'Mendelian' disorders, have similar levels of variation, even when the major causative gene is known.

Indeed, it has been estimated that about 10% of very harmful alleles in humans are the normal allele in other animal species.  That means both environmental and genomic context are involved. And sometimes, as in PKU, one animal model (rat, monkey cell line, mouse, guinea pig) is 'better'--more apparently human-like--than others.  That implies differences elsewhere in the genome, but does not automatically imply that the model is thus more relevant to humans!  What and how we learn about even single-gene traits is not so simple.  So much for simple genetic determinism. 

If an obvious genetically determined trait like mitochrondrial disrepair can be altered by environmental factors, a serious variant in a ubiquitously vital gene, does this tell us anything about genetic non-determinism of non-deleterious traits?  Probably, though it's dangerous to generalize.  We know that risk of heart disease, type 2 diabetes, stroke, dementia, obesity and so forth -- all traits for which probably billions of dollars have been spent on searching for genetic causation -- can be altered by simple things like exercise.

So, what about normal variation in a trait like intelligence?  Or musical or athletic ability?  Or criminal tendencies?  The media and professionals alike persist in salivating over stories that do seem to indicate important genetic factors, and those are advances in knowledge. But stories like this one, that remind us that environmental factors can have a significant effect even on traits that have an identified single genetic component with a large effect, should be a sobering reminder that genes aren't always destiny.