5 replies to this topic

[maestro949]

If aging is the failure of a genetic regulatory as well as a physiological network then curing the network as a whole will be necessary to fully halt and reverse the symptoms of aging. The fight against cancer has turned from looking for single cure-all silver bullets to studying the complex interplay between gene regulation. Aging research will likely require a similar approach but to an even more extensive level of understanding of the various interrelated components. These articles give some insight into the future battlefield where the fight against aging will be fought.

Gene Perturbation and Interventions in Probabilistic Networks

Identification of Genetic Networks

Binarization of Microarray Data on the Basis of a Mixture Model

[John Schloendorn]

curing the network as a whole will be necessary to fully halt and reverse the symptoms of aging

Halt aging - yes
Reversing its symptoms - probably no.

[maestro949]

Both are required. If the network continues to degrade at an exponential rate, even if the exponential factor is decreased, then the game of whack-a-mole on downstream symptoms just starts to go faster and faster over time. Tackling the symptoms and pathologies will likely push up average life expectancy significantly (good) and maximum life expectancy some (great) but it only buys us "some" time and that "some" is too difficult to quantify with such a complex number of systems and variables.

Slowing and eventually stopping the degredation and returning the network to an efficient steady state is just as imperative IMO. SENS doesn't contradict this as most, if not all of it's proposals would certainly alleviate stress on the network but there are likely numerous additional unforseen challenges or roadblocks that arise when attempting to implement not only SENS but every treatment we throw at the various problems (e.g. cancer is a prime example of this) that will require some fairly clever and/or sophisticated solutions to tackle. I suspect new and novel approaches will likely need to displace some of many, if not most of the proposed and specific solutions as we learn more about the genetic network as a whole.

[John Schloendorn]

What you call "degradation of the network at an exponential rate" might be caused by damage accumulating in the network's components. So it is hoped that fixing damage completely would also reset the network to its earliest, lowest rate of degeneration. If this were done periodically, pathology might be postponed indefinitely, without the need to make any change to the networks' topology. This idea is the only thing that gives me hope this can be done in anyone's lifetime.

[maestro949]

But what if much of the damage is less physiological damage but rather just an aggregate of gene regulation that are out of tune and increasingly and negatively affecting each other? e.g. hormone level A drops thus affecting gene expression B. B's adjusted gene expression affects C which in turn affects A. Add a hundred thousand or so variables between A and B and between B and C to the above example along with criss crossing nodes and you have one complex state machine hell bent on surviving whatever external factors can throw at it. It's resiliency is amazing in the short run but it's complexity becomes it's own downfall if some of the regulation switches get stuck at the wrong expression level. Yes damage may cause some of these "stuck" expression levels but undoing the damage doesn't necessarily trigger the expression levels to reregulate itself. Perhaps it does in some cases, perhaps not in others and worse may even cause unforseen problems up or downstream. [:o]

I'm more optimistic as far as what we can accomplish in our lifetime and perhaps even over the next 10-20 years even if we have to go the full reduction route. I do believe that the complexity is beyond the human mind to fully comprehend in regards to understanding all the possible reactions that are taking place simultaneously but it's not beyond what a machine can learn, process and boil down for us.

A few things need to happen.

1.) The research (especially the proteomic search) needs to be automated. Advancements in technology that will allow machines to do the searching will yield enormous advancements towards our goals. It's not hard to imagine proteins being monitored and tracked in real time from transcription to the point it is dissolved where evolutionary-based pattern recognition software is tracking each step of the way and collecting and analyzing the data. Will someone come up with a a clever way to automate gene knockout experiments across a wide range of genese? What, maybe 5-10 years off?

2.) The systems data from #1 needs to be better modelled and then finally aggregated for an entire range of processing whether it be statistical anlysis, drug discovery or next gen therapies from stem cells to gene therapies. Any repetive task a human can do, a machine can and will do faster, e.g. Assay analysis. The amount of data will be staggering though. Stochaistic algorithms will be needed to expedite prediction of the missing puzzle pieces for snapping them into place.

3.) The cats need some hearding. ie, the attack needs to be organized on a larger scale for efficiency. Microbiologists & Biochemists leading the way on the reduction. They are backed up by creative automation engineering and information management (DBAs, data analysts, software engineers, etc) on the back end. Systems Biologists analyse the data looking for patterns and make recommendations on where the next set of investments should be made across the board. The research, the automation, drug discovery, etc.

It might sound like pie in the skie but things are evolving in this direcion and already happening to some degree within private and academic institutions and through various networks of labs and universities but the levels of coordination to tackle the proteome will need to be much more significant than that of the genome. It will happen and it will probably follow the same pattern of accelerating returns that we saw with the genome.

[John Schloendorn]

damage may cause some of these "stuck" expression levels but undoing the damage doesn't necessarily trigger the expression levels to reregulate itself

This may be true in some, but probably not other cases, and is one of the reasons I was speaking of "hope" rather than certaincy. My "hope" rests on the intrinsic ability of an undamaged organism to rebalance itself to homeostasis, which has proven itself in scenarios such as ApoAI-mediated sterol efflux from the atheroma in humans [1] or beta-amyloid removal in mice and possibly also humans [2], but there are obviously no guarantees that this will always be so.

It is however pretty much expectable that re-regulating gene expression (if that were somehow possible) will in many cases not undo the molecular damage already done, because the respective types of damage build up slowly, but irreversibly, even in young organisms with well functioning "networks" (examples are atherosclerosis, amyloid, senescent cells, nuclear DNA mutations and others). Therefore, damage-removal should be the central part of aging therapies, and tweaking metabolism an extra if needed, which is perhaps only quite late on the escape velocity vector. Especially when considereing the dysbalance of current research towards tweaking metabolism (like you say: "It will happen" (anyway)), damage-removal therapies are certainly what immortalists should try to advance.