Food For Thought August 20, 2002

You are what you eat. This is a phrase that we do keep hearing. Eat healthy. Stay away from empty calories. Americans exercise more today than they did in 1950 but we still have too many people who are obese. There was a recent report that announced that a single slice of pizza can have over 600 calories with several days worth of fat. So stay away from pizza.

You are what you eat. Of course, if you were stranded on a desert island, and you could only bring one foodstuff to eat, what would it be? It has to be well balanced, with all the things you need. According to Pizza Marketing Quarterly, the US government calls pizza a complete food that covers the basic food groups. And a Parade magazine survey named pizza the number one food people wanted if stranded on a deserted island. How ironic. So, dueling press releases have shown that you can't ever, ever eat pizza - unless, that is, you are stuck on a deserted island.

You are what you eat. Well, it appears that there are some very real data demonstrating this exact fact, in contrast to one of the current ideas promulgated following all the recent DNA sequencing - You are your genes. People tout the belief that once your DNA sequence is known, your future is determined. Craig Venter, Mr. Celera himself, is going to write a book about his genome (since it was the one sequenced), with each chapter dealing with a specific genetic problem he has. Soon, we may be able to screen everyone's DNA in minutes and then know what diseases they will be susceptible to, presumably so we can cure them rather than just deny them insurance. At least that is the argument.

But genotype does not completely determine phenotype. How your body functions is not completely defined by your genes. It comes from the INTERACTION between your genes and your environment. Your genes define your adaptability to the environment; your ability to withstand what the environment throws at you. But your environment may also determine which genes get expressed and what the resultant proteins do.

Two recent papers serve notice that this interaction is critical; that neither genotype or phenotype is always paramount. Your environment may help shield a large number of mutations present in your genes from selection, allowing high levels of variation to be present in a population living in a stable environment. Variation is the engine behind evolution. Having this variation available under the right circumstances could mean the difference between survival and extinction.

There has always been a conundrum posed by natural selection. This process works to make a species fit exquisitely into its ecological niche. The genes a species possesses create variable individual animals, ones with a range of possible phenotypes, and natural selection then removes those that do not fit well in the environment. One can see that, if the environment is stable, then as time goes on, variation may be greatly reduced. If the species produces too much variation, then many animals will be less fit than the optimal and will be lost. That is a lot of energy to expend for only a few survivors.

It would then make sense that, in a stable environment, the best strategy would be to create offspring that most match the genome of the parents, since they obviously bear genes that fit the environment. Bacteria often do this, with daughter cells being virtual clones of the parent. This is a wonderful strategy, until the environment alters drastically and the selection pressures dramatically change. If there is no variation, then the species will have few members fit for the new environment. During a time of great change, a species wants to have as much variation as possible, so as to make sure that there is a good chance that at least SOME of the organisms will live.

So, how can an organism maintain a large amount of genetic variability, needed for variable environments, while still presenting a stable phenotype when its surroundings are stable? One strategy might be to use the environment itself to help determine variability. Stable environment, stable phenotypes. Variable environment, variable phenotypes. If the genes were totally paramount when determining phenotype, this would be impossible. Yet over 60 years ago a paper in Nature by Waddington postulated the existence of a mechanism called "genetic buffering." During times of stability, the genetic variability would be hidden from view. It would be buffered, presenting a stable phenotype. Changes in the environment could then overwhelm this genetic buffering system, allowing previously unseen genetic variation to be seen in phenotypic alterations. The mechanism for this has been ill defined until recently. These two reports demonstrate how genetic buffering might be accomplished and suggest that it is a widespread means to control variation.

The first report deals with a group of proteins called heat shock proteins. Heat shock proteins have been known for 50 years. Many of them turn out to stabilize the structures of other proteins, often helping these proteins fold properly and preventing them from aggregating. Heat shock proteins are found throughout the metazoa and are highly conserved. They have been around for a long time. They must perform a vital function, to be found in everything from insects to mammals. A reason for this is the role they may play in genetic buffering.

One of these proteins, hsp90, found in organisms as varied as Drosophila, Arapidopsis and humans, is especially important to many proteins involved in growth and development. Proteins that would normally fold improperly, or might have altered activity because of their altered structure, are stabilized and perform normally because of hsp90. Hsp90 makes sure there is enough of these proteins present at certain times of development. If these proteins are not there, or in the wrong concentrations or with different activities, substantial changes in developmental pathways may occur. By stabilizing protein structure, hsp90 also helps keep some proteins functioning properly, even if they possess mutations that affect structure.

In a recent paper in Nature, Quietsch et al. used a variety of means to alter the expression of hsp90 in Arapidopsis, a type of mustard plant. The lab had previously shown that changing its expression in Drosophila had huge effects on development, with malformed legs, antennae, etc. Not very useful phenotypes, most likely because measures used to decrease hsp90 expression were pretty brute force. But an interesting observation they did make was that, after selecting for certain phenotypes over multiple generations of hsp90-depleted flies, the phenotypes could be maintained, even after hsp90 function was restored. That is, the effects of hsp90-depletion had allowed certain genes to be expressed in ways different from normal. Selecting for these new phenotypes would allow these altered genes to be set into the genome, permitting them to provide the new functions, even when hsp90 function was returned.

The main problem with the work in Drosophila was that so many of the changes were very harmful. The newest paper in Nature used milder environmental effects, such as drugs and temperature, to modulate hsp90 expression in plants. What they found will give them plenty to investigate for years. Hsp90-depletion resulted in alterations in leaf shape and color, root structure and function, even in seedlings from identical strains. That's right - plants that are supposed to be exactly the same genetically, displayed different phenotypes. In addition, different strains displayed a variety of effects, indicating that hsp90 was hiding genetic variability that was not normally visible in the phenotype.

They describe 3 important functions for hsp90. The first is simply providing genetic buffering in both the animal and plant kingdoms. It allows the organism to present the same phenotype, even if there are underlying differences in the genotype. The second is the degree in variation that it can control; the plasticity of the species it acts on. That is, the amount of genetic variability hsp90 can buffer is different from species to species. Finally, it also stabilizes random or stochastic effects of the environment on phenotype in the same species. Seedlings from the same strain grown in similar conditions demonstrated large degrees of variation when hsp90 was inactivated.

They were able to demonstrate that hsp90 can hide large amounts of genetic variation in a species, so that the organisms present a constant phenotype in a stable environment. But, altering its activity, which can happen with a simple increase in temperature or other environmental challenges, can now expose new variability to natural selection, shifting the species to a new fitness peak. And these new traits can become stable again once the environment stabilizes. A neat mechanism.

Natural selection appears to drive a species up a fitness peak, trying to make it as perfect for its ecological niche as possible. It would then seem that natural selection pushes all members of the species to possess the same phenotype, and, if phenotype is purely determined by genotype, to possess the same genotype. But now we can see that, although the phenotype might be the same, the genotype can be quite different, just waiting for the right environmental conditions to be expressed. When the environment changes, the species can now display its hidden genetic variability, and possibly overlap a new fitness peak for the new environment. Once things stabilize, it can again progress up the fitness peak, keeping its genetic variation hidden until it is needed again.

My guess is that hsp90 may not be the only protein involved in these types of mechanisms, it may not even be the most important. The ability to access hidden genetic variation would, however, be a very useful trait. We know that the presence or absence of certain vitamins from the diet can have severe metabolic effects, however, the ability of vitamins to hide or to expose genetic mutations is something that is beginning to appear in other disciplines, such as Evo-Devo.

What a great name. One of the earliest statements of this field is Ernst Haeckel's "ontogeny recapitulates phylogeny" (right up there with "eschew obfuscation" for the greatest amount of information understandable by the fewest number of people). It describes the idea that the development of an embryo exactly retraces the evolutionary path an organism took. So, a human embryo progresses through amphibian, fish and reptile stages in the development of a fetus. We may know it is not that simple but examining the interface of evolution and development is a rapidly expanding field, due to the tools of molecular biology now being adapted for it. So, naturally, its researchers work on Evo-Devo.

At a recent meeting, Claudia Kappen at the University of Nebraska Medical Center discussed an interesting system demonstrating that hsp90 is not the only potential molecule involved in masking variation. Her story begins with an altered Hox gene.

As I have mentioned before, Hox is an important gene for the developmental pathways in almost every metazoan. Kappen inserted a mutated Hox gene into a strain of mice. This gene normally works to control the presence and maturation of precursor cells in cartilage. This particular mutant Hox gene is hyperactive, resulting in the inability of the mice to properly form cartilage. The mechanism for this is still unknown and is, I would expect, a major focus for Kappen's research. Mice with this mutation die soon after birth. Their rib cage is not strong enough to withstand breathing.

Now, she did all this work at the University of Arizona in Tucson, not in Nebraska. She moved there 2 years ago, and took her mice with her, where they DID NOT die. Instead of shattering, their skeletons stayed together. Now, I do not know if her work at Arizona was as a post-doc or as a tenured professor, but it must have been disconcerting to move to another location and see all the results of your work completely disappear. I can sympathize with her state of mind. Was it all an illusion? Was it the phases of the moon?

No!! She is one sharp scientist and was able to show why the phenotype of her transgenic mice was different, even thought the genotype was the same. Something was different in their environment; something that was now masking the mutation. It turned out to be the corn cobs. The mice in their new home in Nebraska got cages with corn cob bedding. Could something in these cobs provide the environmental difference that was overcoming the mutation? On a hunch, she examined whether folate (one of the B vitamins) could be the missing ingredient.

Normally, the cartilage cells from the mutant mice, to quote Kappen, "shriveled up and died" when grown outside the animal. If grown in a folate-enriched medium, they grew just as well as wild type cells. If she fed transgenic mice extra folate, they had almost normal skeletons. Her conclusion says it best: "The faulty Hox gene can be modified by an environmental substance." So, in this case, a faulty gene could be hidden in the environment. It is not hard to envision how the environment could mask other variations, some beneficial, permitting them to be unleashed when the diet or environment changes.

We are what we eat. Phagotype determines phenotype. Our genes are exquisitely tied to our environment. Even small changes in our surroundings may have huge effects on what our genes do and how successful we will be when nature does its selection. Increasing genetic variability is a major advantage for any organism and hiding this from selection means that variability can be unleashed when needed.

It seems to me that this may have large ramifications when it comes to the whole idea of designer drugs. Knowing a person's DNA sequence may only provide part of the solution. Their environment will also need to be examined or modified. This could make some things quite complex. I know it makes my head hurt. I just know that when I go to that desert island, I'm bring along Flintstone vitamins containing lots of folate.