The methuselah gene in fruit flies, which when expressed gives rise to the methuselah cell receptor (or mth), has been known for a decade now. It is one of a small number of single gene manipulations which can meaningfully extend longevity in this species - by 35% in this case.
When it comes to figuring out what's going on under the hood, even single gene manipulations in small creatures are terribly, enormously complex. This is one of the many truths that leads those researchers who focus on genetic and metabolic engineering to say that any significant extension of human life span through these methods is far in the future. It's a huge, huge undertaking, from our present position, to consider engineering any form of significant change to the way in which our metabolism works.
(Which is why I support far less complex and more certain approaches aimed at repairing the metabolism we have. The mountain of metabolic complexity is there, climb it if you must, but to get to the valley of enhanced longevity on the other side in good time, then use the level road of damage repair strategies that winds around the mountain's base).
You should take a look at a recent PLoS ONE paper on the methuselah gene for an impression of the complexity involved, as well as for a longevity gene in action in wild populations, where it varies from group to group and between individuals:
Williams' theory of antagonistic pleiotropy describes how pleiotropic alleles that increase fitness early in life may experience positive selection even though they incur a fitness cost later in life. Identified aging genes have consistently shown costs to lifespan extension, particularly in reproduction. Although there is some evidence that lifespan and reproductive success can be decoupled, multiple analyses have revealed previously undetected tradeoffs under specific conditions.
In addition to demonstrating negative effects on reproduction, longevity mutations are positively correlated with stress resistance. Such correlations may explain aspects of lifespan evolution, and why loss-of-function mutants can result in lifespan extension. However, it remains unclear whether identified aging genes are major contributing factors to the genetic variance for longevity that is routinely observed in populations.
The researchers then provide a good slab of evidence for variations in the methuselah gene to contribute to observed variations in longevity in wild fly populations. It makes sense for any successful species to have central controlling points in their biochemistry for such things as energy devoted to reproduction, stress resistance, longevity and so forth. Varied abilities across generations and individuals means a species more able to populate new habitats and survive environmental changes that favor one set of abilities over another - so we'll see more of that versus any other alterative biochemical arrangement.
The 30-50% range keeps coming up in experiments to extend life span in small mammals and flies; a number of different ways of achieving this degree of success have been demonstrated in the past 20 years. It seems that this is the natural plasticity of longevity for the biochemistry of many complex species. In some cases, it can be achieved by signficant changes in diet alone, while in others a change to one or a couple of genes is needed - a small change from the point of view of evolution. For the reasons given above, I suspect that this plasticity is a feature of most successful species, a buffer against change inherited from the ancestral past.