The human body is, ultimately, a machine - an astronomically complex machine, of whose workings we remain pitifully ignorant - but still a machine. Like any machine, it accumulates 'damage' as a side-effect of its normal operation: molecular and cellular changes that occur throughout life are initially harmless, but eventually (when too abundant) increasingly impede the normal operation of the machine and eventually cause it to fail altogether. Conceptually, there are three strategies to postpone a machine's demise beyond its 'warranty period'. First, we can treat it really well throughout its life, thereby slowing down the accumulation of damage: but that can never stop the accumulation altogether, because to do so would require not operating the machine at all, and anyway it cannot address damage that has already occurred. Alternatively, we can combat the late-life symptoms, the dysfunction that eventually emerges: but that too is only a short-term approach, because the underlying damage that causes the dysfunction is still accumulating and making the dysfunction harder and harder to address. This is why the way in which machines that people love are in fact kept in good shape is the third strategy: repair and maintenance, in which we let the damage be created, but repair it before it becomes so severe as to cause dysfunction. In the case of the human body, this means using regenerative medicine against aging.
So... can it work? Are all the types of damage that contributed to age-related ill-health amenable to repair?
If 'repair' is interpreted strictly, there are two classes of damage for which the answer to this question is probably 'No'. They are mutations in the nuclear and mitochondrial DNA. However, a better way to interpret the term 'repair' is 'repair for practical purposes' - and here the scene is rosier. Mitochondrial mutations may not be feasible to remove, but it may be much more feasible to eliminate their consequence, dysfunctional mitochondria, by complementing their mutations with nuclear transgenes - a goal that has long seemed quixotic but has now largely been achieved by Corral-Debrinski's group. In the case of nuclear mutations (and epimutations, i.e. disruptions of the decorations to DNA that determine its expression pattern), a different end-run seems available: to focus on the consequences of those mutations at the level of cell number, which come down to inadequate cell death or excessive cell division. Both these classes of problem are potentially amenable to reversal - by suicide gene therapy and by telomere maintenance abrogation, respectively. (Mutations that deplete cell number do not accumulate, by definition: the loss of cells is a part of aging, but it can potentially be addressed by stem cell therapies.)
What about nuclear mutations/epimutations that do not affect cell number? My interpretation of the available data is that such damage simply does not accumulate fast enough to matter in anything like a currently normal lifetime: the imperative to avoid cancer until procreation has driven evolution to develop DNA repair and maintenance machinery that is 'unnecessarily good' in respect of other mutations.
Other types of damage exist that are probably also too slow to matter. Some are for the same 'protagonistic pleiotropy' reason: glycation-induced adducts such as carboxymethyl-lysine, for example, are limited in abundance because they arise by the same process that also causes protein-protein crosslinks (particularly glucosepane), whose effects are much more potent. (Crosslinks themselves need to be addressed, and much work is in progress to develop ways to cleave them.) Aspartate racemisation of long-lived proteins also accumulates, but its effects seem to be minimal.
In conclusion, regenerative medicine - if appropriately broadly defined - seems to have a fighting chance of combating aging within the next decade in laboratory mammals and within perhaps 25-30 years in humans. Since its health benefits would so immensely exceed those of the best possible 'gerontology' and 'geriatrics' approaches, there is a clear case for rapidly prioritising efforts to develop such interventions.
Aubrey de Grey is organising the 4th conference on Strategies for Engineered Negligible Senescence in Cambridge on 3-7 September 2009.