October 02, 2007
Micro-Incubator For Cells Automates Experiments
Silicon technology applied to microfluidics is going to revolutionize biological science.
Integrating silicon microchip technology with a network
of tiny fluid channels, some thinner than a human hair,
researchers at The Johns Hopkins University have developed a
thumb-size micro-incubator to culture living cells for lab
In a recent edition of the journal IEEE Transactions
on Biomedical Circuits and Systems, the Johns Hopkins
researchers reported that they had successfully used the
micro-incubator to culture baby hamster kidney cells over a
three-day period. They said their system represents a
significant advance over traditional incubation equipment
that has been used in biology labs for the past 100
"We don't believe anyone has made a system like this
that can culture cells over a period of days autonomously,"
said Jennifer Blain Christen, lead author of the journal
article. "Once it's set up, you can just walk away."
Note the lack of need for daily labor-intensive care. The system is automated. Automation speeds progress, cuts costs, increases consistency and quality.
I expect that the rate of advance in biological sciences and biotechnology is going to greatly accelerate in the next few decades because of microfluidics and computer simulations. Experiments will get done more rapidly and with larger numbers of experiments done in parallel as cheap devices lower the material and labor costs of each experiment.
This ability to accelerate advances makes me very optimistic about the prospects for the development of rejuvenation therapies. Automation will enable the development of much more powerful manipulations of cells and tissues. The automation and miniaturization will enable cheap ways to introduce experimental conditions and measure the results automatically.
This is good news, of course. But recently I have given some thoughts to the parallels between recent advances in the biological testing equipment and the advances in microelectronics generally during the last several decades.
Simply put, predictions such as Moore's Law (which estimates the density of circuits, and thus the computing power of the chip) have been fairly accurate. However, predictions as to what those more powerful chips would allow us to do have been wildly and repeatedly wrong. The example that most easily comes to mind is General AI. When we said we'd have chips 1000x more powerful is Y years, we were right. When we thought that would allow for GAI, we were wrong. It turns out it's a far harder problem than we realized. Computer vision has seen similar setbacks.
We were also 'wrong' by not predicting what would be possible. No one foresaw Google, or the Internet generally, turning out the way it has. People are constantly amazed when the 'next generation' computer games come out - "Look at those graphics! Wow!"
The simple point I am getting at is that we may be confident in predicting advances in processing capability, but we ought to be more cautious in predicting real-world effects capability. Longevity/SENS may prove far, far harder than any of us thought; and many things we never dream of today may be commonplace tomorrow.
Perhaps your two observations regarding processing power and the Internet are synergistic... in accomplishing things like GAI (as opposed to ever stronger Chess players). Presumably the non trivial parts are the enabling a great number of complementary algorithms (based on something like pandemonium theory or 'genetic' optimization) to accomplish it... the creation of the semantic web may well hold a key to all of this. The other question mark is whether there is a complexity issue... that is whether or not these sort of systems are heuristic and not design centric.
You make a good point.
I'm most worried about brain rejuvenation. We should be able to, for example, develop cell therapies to inject into joints to repair joints. Ditto cell therapies for replacing lost muscle mass. Ditto for cell therapies for the vascular system. Plus, some organs (e.g. liver) will be relatively easy. Others will be harder.
But what about the nervous system? We might end up with young bodies and old minds.
Don't worry Randall, the nanobots will come and repair our brains and hopefully replace them with carbon nanotube circuitry.
also there's the time limit of needed human testing(FDA tests) , which take a large part of a drug development time.
so i think a lot is dependent in improvment in tech to predict the safety and toxicity of a drug before clinical trials.
I'm highly critical of FDA policies and would like it to have much less power. But I do not see the FDA as something that would slow things down that much if we could only figure out biological systems more rapidly.
Our biggest problem with drug development now appears to be that so many drugs fail in stage II or stage III clinical trials. Yes, the FDA is an obstacle. But look at cancer drugs. Tons have been tried and failed. A drug that wiped out half of all cases of cancer in phase I would get thru phases II and III pretty quickly if they continued to wipe out half of all cancer cases. The FDA tolerates orders of magnitude more side effects from cancer drugs.
Highly automated and highly sensitive small scale experimentation would allow much more rapid searches of solution space and also much more rapid figuring out of how biological systems work.
Randall, any comment on this article:
It is not the same as using human subjects, but I wonder if it would act as a cost-effective screen to prune out drugs that wouldn't make it through Phase II and III.
I hope it would prevent a torcetrapib fiasco. 800 million dollars down the drain to get to phase III. :( Catholics, of course, would become choleric. However, the pro-life scientist, Aubrey de Grey advocates embryonic stem cell research in Ending Aging.
Regarding your Reuters link: I am puzzled as to why those drug companies want to start with embryonic stem cells, differentiate them into liver cells, and then use those liver cells to study drug toxicity. Why not just extract some liver cells from human donors or from humans who volunteer to donate their livers in case of death? What advantage derives from differentiating stem cells to create the liver cells?
Do they have a hard time getting liver cells from donors? Maybe. Or do those liver cells not live as long and divide as much as they'd expect liver cells freshly made from embryonic stem cells to do? I don't get it.
So my comment is I'm puzzled. Curious to know the answer if anyone knows.