LE Magazine December 2001
Page 2 of 4
L.E.: How reliable are the magnitudes of the changes you reported? If you saw a 4-fold change in a given result with short-term calorie restriction and a 2.5-fold change with long-term calorie restriction, how likely is it that these changes are really different and don't just differ because of random statistical fluctuations?
|A recently published study indicates that if people will lose ten pounds, regardless of what their weight is before they start the diet, then many of their physiological parameters of health will improve. So, even losing weight for a short period of time has beneficial effects.|
S.S.: We found that the changes we detect with Affymetrix gene chips (the ones used in the PNAS study) are really pretty reliable. It is true that you can't do very good statistics with these Affymetrix chip studies. The reason for this is that the chips and chemistries for the chips are so costly. Also, old animals are expensive and difficult to get. So we've been in a situation where we're able to measure 1,000 to 10,000 times the number of genes we used to be able to measure, but we don't routinely do as many samples as we used to do. So traditional statistical tests are difficult. We've used a technique called Northern blotting in order to validate our changes. What we found was that at least 95% of the changes that we detected with the Affymetrix chips were reproduced using Northern analysis. So, my guess is that the results we found are reliable.
Any gene chip study is really an initial screening event. It can show you that things that you might never have expected to change do change, and it can point you in new directions that you might never have known about, because it's an unbiased screening of a large percentage of the total genes that the organism expresses. When you have those changes, then you can be pretty sure, if you're looking at changes that are 1.7-fold or greater, as we did, that the changes are real.
L.E.: Here's a difficult question. We know that long-term calorie restriction extends life span, but since short-term calorie restriction only reproduced some of the changes that long-term calorie restriction produced, how can you predict the effect of short-term restriction on life span?
S.S.: My assumption is that the effects of calorie restriction are linear. By this I mean that if you are able to improve the expression of 70% or 60% or 50% of the genes and return them to the level of expression that life-long calorie restriction produces, then my assumption is that you're going to get 70 or 60 or 50 percent of the effects of long-term calorie restriction. My reason for thinking this is that it seems that what matters is general gene patterns rather than specific genes. Between different tissues, you don't find that the exact same genes change, but you find that genes that are in the same kinds of pathways and that are involved in the same kinds of responses change. And so my guess is that calorie restriction creates beneficial patterns of gene expression. At this point, though, it's just an opinion. Ultimately, we'll have to do more studies and follow up to find out whether or not these short-term changes really do also delay the onset of age-related diseases and extend life span.
L.E.: This starts to get us into the area of how you can apply your findings clinically and what the implications are for clinical intervention. Short-term calorie restriction is obviously more attractive to most people than long-term calorie restriction, but we've also heard that yo-yo dieting, in which you repeatedly get rid of a lot of weight for a while and then gain it back again, might be harmful.
S.S.: My current understanding is that the studies say that it's not harmful to do that. My personal opinion is that you should do whatever you need to do to get the weight off, short of something that would hurt your health. I don't think that you should try bulimia or drugs that could harm you, but I think that people would be well advised to do whatever they can to get the weight off.
A recently published study indicates that if people will lose ten pounds, regardless of what their weight is before they start the diet, then many of their physiological parameters of health will improve. It improves your glucose sensitivity, lowers your blood glucose, lowers your blood insulin levels, improves your heart rate, improves your blood pressure. So, even losing weight for a short period of time has beneficial effects.
L.E.: Okay, so is it theoretically possible to use short-term calorie restriction to partially reverse aging in very old humans?
S.S.: I've had people ask me about using caloric restriction for cancer patients or for very elderly people, and my advice is always not to try it. Calorie restriction is something that's very well characterized in animals and rather poorly characterized in humans. We are not animals in a vivarium. We have to go out and cope with a very complex world, and we have to have energy and strength to do it. There's no question that under-eating improves our health, but I don't think that you should take sick people and try to improve their health by under-feeding them.
L.E.: This suggests the need for a more practical alternative. Since almost nobody wants to be on calorie restriction anyway, and since it does have its safety issues and inconveniences, there is a desire on the part of many people to develop what are called calorie restriction mimetics, in other words, drugs that imitate the effects of calorie restriction. You refer to that in the paper, and the fact that your results provide an opportunity for screening drugs and for finding the magic pill that would simulate calorie restriction.
S.S.: To the best of my knowledge, gene chip analysis would be the fastest way of doing a first screen for drugs and treatments that mimic the effects of caloric restriction.
L.E.: Are you in fact screening any potential candidates at this time?
S.S.: Yes, and we're preparing to screen others, too. We're also preparing to screen compounds for their effects on gene expression for other companies. Subsequent studies can be done to verify that compounds that pass this biomarker screening are in fact effective in preventing the onset of age-related diseases and extending life span. Excitingly, gene chip biomarker studies can be done in humans after the preliminary studies are done in animals, and then the screening of humans for delaying the onset of age related diseases with these compounds is a very real possibility.
L.E.: So if you find a compound that's effective in a mouse or in a monkey or whatever, you can find out right away if it has the same gene expression effects in a human.
S.S.: Yes, you can.
L.E.: So not only do you have a technique here that can give you a comprehensive look at virtually the whole genome, but what's even more revolutionary is that it can do so for aging intervention tests in a short time.
S.S.: That's really the contribution of our technology-it makes the initial screening rapid.
L.E.: This has really been the bottleneck that has held up the entire field of interventive gerontology: it's just not practical to test somebody for their whole life to see if they live longer or not. Now you have a solution to that problem.
S.S.: Even to test an animal for an effect directly on life span takes more than three years in a mouse. Others have produced screens that they say would take a year or ten months. But if you can screen in four weeks, the number of screenings that you can do increases enormously, and the cost goes down dramatically.
L.E.: Talk a little about the company you are working for that is developing the commercial applications of your work.
S.S.: We founded a company called LifeSpan Genetics, which is currently funded by the Life Extension Foundation. LifeSpan Genetics has licensed the commercial rights to two seminal patents that have been applied for in the field, one from the University of Wisconsin, one from the University of California. We are testing drugs for their calorie restriction mimetic effects and are continuing to look at the effects of different periods of caloric restriction in many tissues in both mice and monkeys, and we are also planning studies in humans.
L.E.: It used to be that the Life Extension Foundation would quietly support research in other labs that produced no fanfare. So it's very gratifying that now we're able to report on research that was funded by the Foundation and that has not only made it into the prestigious Proceedings of the National Academy of Sciences, but has also been singled out for special attention by that journal. I understand that stories on your work are appearing on CNN and Science Now as well as in the Wall Street Journal, The Washington Post, the journal Nature, the Reuters News service, and other media sources. Congratulations!
S.S.: For many years the Life Extension Foundation has funded studies of ours and of others that were published in scientific journals. But now through LifeSpan Genetics we have two papers, not only the PNAS paper but also one soon to be published in the Journal of Nutrition, which represent work that is patent-pending and licensed by the company for work in identifying anti-aging CR mimetic compounds and treatments.
L.E.: From a practical point of view, how would you approach clinical application of your observations? I assume it's hard to take liver biopsies from humans, so what would you do?
S.S.: There are tissues in humans that you can assay readily. And I think there's a tremendous amount of information that can be gleaned from gene chip assays of those tissues, especially since studies are being done in both monkeys and mice.
L.E.: What are the prospects for using gene expression profiling to develop a biological measurement of aging? In other words, a way of giving an individual a reasonable assessment of what his or her biological age is.
S.S.: Gerontologists have been interested in developing a biomarker system for measuring biological age for at least twenty years, that I know of. And they want to do that for precisely the reasons we've talked about in this interview. One would like to be able to know when a treatment or compound is affecting aging without having to look at life span. Also, just from every human's point of view, we'd like to know how close we are to death. I don't know how well gene expression profiling will serve as a yardstick for biological age. It's a little early to tell yet, but I think because we're looking at so many things, gene expression levels eventually of 30,000 genes and 3.1 splice variants each, some of which will change with age, I can't help but think it's likely we'll find highly reliable biomarkers of aging.
L.E.: Agreed. The magnitude of the problem of dealing with aging may depend on how many genes are involved. Can you speculate about how many genes you think may change with aging in a whole mouse or a human?
S.S.: It is an amazingly small number. Most gerontologists agree that there are on the order of tens to hundreds of genes that are involved in changing the life span of an organism from shorter to longer, based on selection experiments (directed breeding to produce long-lived animals). It has to be a relatively small number of genes because you're able to select for long life over a relatively short period of time.
L.E.: In the original Weindruch and Prolla paper on muscle, the number of genes whose expression went up with age by more than a factor of two was 58 out of the 6347 genes examined, or 0.9%. The number of genes whose expression went down by more than a factor of 2 was 55, or also about 0.9%. The total percentage of genes that went either up or down by over 2-fold was 1.8%.
S.S.: You have to remember that that was 1.8%. of the total probes that were on the chip. I don't know about those studies in particular, but normally in gene expression studies done with micro-arrays, the number of genes that actually report a signal (are active) is significantly less than the total number of genes on the chip. I don't know what their results were in terms of active genes. But I think the number of genes that change with age is probably a larger percent, several percent, of the genes that are active.
L.E.: OK. How many of the genes were active in your study?
S.S.: Oh, around 4,000, or about one third of the total looked at.
L.E.: OK, now let's try to compare the muscle results with your liver results. First, you examined over 11,000 genes and found only 46 known genes that went either up or down with age.
S.S.: Yes. I was surprised to find, looking at the liver, that in control mice there were only 20 that went up and 26 that went down. And there were far fewer changes when we imposed calorie restriction.
Back to the Magazine Forum