Because a big chunk of the genetic information that goes into building a mouse needed to put together a human, mice are often employed in the laboratory as our genetic stunt doubles. In fact, the similarities between sections of human and mouse DNA enable researchers working with mouse genes to make surprisingly accurate predictions about the location and function of their human counterparts.

View a video clip on human/mouse genetics research (QuickTime, 1.2 minutes, 1.9 MB)

For over 40 years, ORNL researchers have been studying the nuances of the genetic relationship between mice and humans, sometimes with considerable fanfare, other times in relative obscurity, but always with the goals of increasing our understanding of the subtle language of genetics and improving human lives.

Some of the Laboratory's latest research in this area, and its plans for a new 250,000-square-foot Center for Biological Sciences are described in this article.

Building a Better Mouse Model for Sickle-Cell Disease

"Often, things don't turn out as beautifully as you expect," says Ray Popp of ORNL's Biology Division describing the trials of research in the emerging field of genetic engineering. It also explains his excitement over the success his group has had in "building" a breed of mouse to serve as a proving ground for the treatment of sickle-cell disease in humans.

Affecting about 100,000 Americans with roots in Africa, India, and the Middle East, the sickle-cell gene is thought to have prevailed over the normal form of the gene because it provides a defense against another scourge--malaria, a mosquito-borne disease common to equatorial regions. Inheriting the gene from one parent provides an individual with some resistance to malaria, while inheriting the gene from both parents results in sickle-cell disease.

The sickle-cell disease affects the body's hemoglobin--the substance red blood cells use to absorb oxygen from the air breathed into the lungs and distribute it to the rest of the body. When hemoglobin gives up its oxygen, it becomes deoxyhemoglobin. In the blood of sickle-cell patients, deoxyhemoglobin is less soluble than its normal counterpart and causes the red blood cells to twist into stiff, sickle-like shapes.

"This condition has the effect of pouring irregular rocks, instead of marbles, through a small pipe," says Popp. "The distorted cells tend to block up the blood vessels." When this happens, muscles and other tissues around the blockage don't get enough oxygen, and outward signs of a condition called sickle-cell crisis occur. These signs include ulcers on the skin in the area of the blocked vessels and deterioration of the kidneys, lungs, liver, and the retina of the eye. In fact, older sickle-cell patients often go blind.

Over time these blockages cause small blood vessels of the kidneys to rupture, decreasing kidney function by 25 to 50%, and slowly poisoning the body with accumulated waste products. Similarly, sickle cells accumulate in the small capillaries of the lungs' air sacs, causing them to repeatedly break and heal. This process can build up scar tissue through as much as two-thirds of the lung, interfering with the efficient exchange of oxygen between the lungs and the blood. The leakage of blood into the lungs and back into the blood vessels can also lead to infections like pneumonia from contaminants being passed from air in the lung into the blood.

Hoping to help alleviate some of the hardships brought about by this debilitating disease, Popp and his group set out to build a mouse model of human sickle-cell disease to use for testing drugs that interfere with the sickling process. These compounds can't be thoroughly evaluated on blood samples because samples don't allow researchers to look at side effects on other parts of the body. "That's why we need to have a whole animal to test," notes Popp, "not just blood in a test tube."

The first step toward the goal of producing a mouse model for this disease was taken about 10 years ago when researchers found they could transfer a human sickle-cell gene known as beta S into a line of mice and have the mice produce human hemoglobin. Unfortunately, these mice exhibit the sickling trait only when their oxygen intake is reduced, as it occasionally is when the mice are tested in a special chamber with a controlled, low-oxygen atmosphere. The mice can be kept in this environment for a couple of weeks, but they can't be maintained or bred under these conditions.

The next step was taken when Popp's group was doing research for DOE on the mutagenic effects of chemicals and radiation on genes that control the production of mouse proteins, including mouse hemoglobin. One of the changes they created caused the resulting mouse hemoglobin to have a high affinity for oxygen, a characteristic of normal human hemoglobin. "We suspected," says Popp, "that the changes that had occurred as a result of the mutation had created a mouse hemoglobin that could be used to develop a mouse model for human sickle cell disease."

As these animals were developed, Popp and his colleagues found that if they crossed the beta S mice with normal mice, the red blood cells of their offspring were normal. However, as Popp suspected, if they were crossed with mice having altered hemoglobin, the next generation would have elongated cells or cells with spurs, a less-pronounced form of sickling. This sickling isn't as extensive as it is in human red cells because the mouse produces a mixture of mouse and human sickle-cell hemoglobin, and usually, only the human portion of the hemoglobin sickles. Also, like humans with sickle-cell disease, these mice suffer from anemia--an abnormally low number of red blood cells--although the mouse anemia is not as severe as that noted in humans.

"It's not perfect," says Popp, "but it's a model in which there is expression of the sickling phenomenon under normal conditions. We don't have to place these mice in a reduced-oxygen environment to induce the effect."

Despite these successes, Popp is clearly not laboring under the illusion that this model of sickle-cell disease in humans cannot be improved. "It would be ideal if one could produce 100% human hemoglobin inside the red blood cell of the mouse," he says. "In this mouse, we have combined mutations that favor sickling and production of a relatively large amount of sickle-cell hemoglobin. Unless someone can replace essentially all of the mouse hemoglobin with human sickle-cell hemoglobin, all animal models of sickle-cell disease still have the problem of the mixture of hemoglobins, which interferes with the sickling phenomenon."

Popp's current research is funded by a grant from the National Institutes of Health, and he is hoping to get further funding to test the effectiveness of compounds designed to combat sickle-cell disease in humans on these genetically engineered mice.

Meanwhile, work on improving the model continues. It's been three years since researchers began crossbreeding these mice, but only in this past year has the sickling begun to be expressed with regularity. Some mice express the sickling problem more severely than others because the sickle-cell trait shows up better on some genetic backgrounds than others.

"It takes about 20 generations, or about 6 or 7 years, for the sickle-cell trait to be bred into a new strain of mice and the population to become genetically inbred," says Popp. A new strain of sickle-cell mice could be used to test preventive drug therapies, which may eventually prove useful for treating sickle-cell disease in humans.

"It's true genetic engineering when you design an experiment based on what you know about human genes and put them in mice having the right genetic characteristics to maximize their expression," says Popp. "Our contribution to the field was to put human beta S hemoglobin on a background which favored expression of the sickle-cell trait.

"Exactly what we predicted occurred--it's rewarding that it worked out that way."

Closing in on Mouse Diabetes and Obesity

Can a bunch of fat, yellow mice help uncover the genetic roots of diabetes, tumor development, and obesity? Rick Woychik of ORNL's Biology Division thinks it's an intriguing possibility.

The straw-colored rodents in Woychik's laboratory possess a mutant form of the "agouti" gene, whose location was recently pinpointed by a team of researchers from ORNL and the University of Tennessee. The agouti gene, named after a large, South American rodent, is normally responsible for the grizzled coat color of wild mice, which results from the combination of black (or brown) and yellow areas on each hair. But mutant forms of the gene can cause animals to have not only yellow coats, but also diabetes, obesity, and tumors. The genetic variations that cause these effects are referred to as "obese yellow" mutations.

Researchers are most interested in the inability of obese yellow mice to effectively use insulin--a hormone produced by the pancreas to process sugar and other carbohydrates. This biochemical shortcoming is also characteristic of Type II, or non-insulin–dependent, diabetes in humans. This form of diabetes affects one of every 20 Americans, producing symptoms ranging from blurred vision to chronic infections and fatigue. Unlike Type I diabetes, which is potentially deadly and far less common, the non-insulin–dependent form of the disease does not result from a lack of insulin. Instead, it stems from the body's inability to use insulin efficiently.

"There are many instances where people have produced diabetic mice by genetically knocking out the insulin-producing cells of the pancreas," says ORNL researcher Rick Woychik. "The reason for this type of diabetes is quite straightforward--the mice are unable to produce insulin--period. Those animals are valuable for certain things, but they don't tell you much about the mechanisms of the Type II form of the disease."

The obese, yellow mice, on the other hand, start developing a diabetes-like condition and obesity when they are just out of adolescence--at about 4 to 5 weeks old. These animals still produce insulin. In fact, to maintain normal blood sugar, they produce a lot more than normal mice--but they can't use it as efficiently."

The ORNL-UT team was successful in both isolating the agouti gene from the rest of the mouse's DNA and in cloning it. Cloning the gene involves inserting it into a bacterium where it is copied over and over as the bacterium reproduces. This process provides a relatively large quantity of the gene, enabling researchers to study them in minute detail.

Although this latest work breaks new ground, the Laboratory is no stranger to the agouti gene. For years, ORNL genetics pioneers Bill and Liane Russell have studied the effects of radiation and chemicals on the gene. "Based on their work," says Woychik, "we know quite a bit about how damaging or deleting the agouti gene affects the animal."

It appeared that the agouti gene was unusually complex, accounting, in part, for the 16 different classes of mutations it produces. For example, inactivating the gene or deleting it entirely results in a totally black mouse because the protein that is supposed to be produced by the gene in the mouse's skin is missing. This deficiency causes a breakdown of the normal molecular communication of the gene with the melanocytes, or pigment-producing cells, and, in the absence of other instructions, they produce only black pigment. Another mutation produces two-tone mice, black on their backs because the agouti gene is not activated in this area and tan on their bellies because the gene is overactivated in this region. Obese, yellow mice are entirely yellow because, in this mutation, the agouti gene has been changed in a way that causes it to be overabundant when attached to part of another gene, causing the protein produced by the gene to be drastically overproduced in their skin and also in other tissues and organs where it normally doesn't occur.

Because there are so many different mutations of the agouti gene, it was thought for years that these differences were the result of interactions among a complex family of genes, or one gene that behaved in a variety of different ways under different circumstances. "So, for example," says Woychik, "it was thought that black-and-tan mice resulted, perhaps, from one agouti gene that regulated coat color on the back and another that controlled coat color on the belly. We have now discovered that agouti is a relatively simple gene, although it does have different regulatory regions that normally function in different areas of the skin."

Of the 16 varieties of the agouti gene, the mutation linked most closely to these health effects is known as "Lethal Yellow." This mutation, which was described by the French geneticist Cuenot in 1905, is a recessive embryonic lethal mutation--meaning that, if a mouse inherits the Lethal Yellow gene from both parents, it will die before it is born. On the other hand, if a mouse inherits the mutant gene from only one parent and a normal gene from the other, it will develop the yellow hair and health problems described previously.

"The other reason this is so exciting," Woychik continues, "is that, although the Lethal Yellow mutation was the first lethal mutation ever characterized in the mouse, no one had any idea of the biochemical nature of the mutation or which gene was responsible." Now that Woychik and his team have uncovered the structure of the gene, they are beginning to understand, at the molecular level, why it can have such a devastating effect."

Woychik believes that the real story behind the lethal yellow mutation unfolds something like this: In normal mice, there are two genes sitting next to each other--one that appears to regulate how genetic information is copied during the early stages of embryonic development, and the agouti gene, which controls hair color. These genes normally don't interact, but, in Lethal Yellow mice, a large deletion causes the regulatory portion of the development gene to be moved closer to the agouti gene. This arrangement puts the agouti gene under the control of the development gene's regulating signals, which are normally activated throughout the body. As a result, the protein produced by the agouti gene also appears throughout the body.

Just as the absence of the gene's protein is thought to interrupt communication with pigment-producing cells in the skin of black mice, its presence in places where it should not be may also interfere with normal intercellular molecular communication in the liver, pancreas, skin, and breast of these animals, triggering the development of tumors. "The same thing may be true in adipose, or fatty, tissue," says Woychik. "Normally, after adolescence, the production of fat tapers off, but not in mice having one of the Obese Yellow mutations. We think the presence of the agouti protein interferes with the normal signaling mechanisms that tell adipose tissues to stop producing fat molecules."

Similarly, Woychik speculates that the mutant expression of agouti in muscle tissue in these mice is responsible for their resistance to absorbing insulin. "In fact," he says, "it is known that obese people with a lot of body fat somehow become insulin-resistant in their muscle--and, if they lose weight, their insulin resistance improves."

Agouti isn't your standard coat-color gene. This characteristic is primarily regulated by several other genes acting in the color-producing cells called melanocytes. Agouti, on the other hand, is not expressed in the pigment-producing cells, but in the neighborhood of the hair follicles, and it is somehow communicating with the melanocytes to tell them to produce black or yellow pigments. "So, at the level of cell biology," Woychik says, "agouti is not just a pigmentation gene. It also provides an insight into how cells communicate with each other."

So, given all of these intriguing findings, is the research team ready to pack up and move on to other projects? Hardly. Woychik still has plenty of questions he wants answered: "Can we express the agouti gene specifically within muscle and induce the type II diabetes without the obesity?" he wonders. "And can we induce the obesity without the diabetes by only expressing agouti in adipose tissue? Moreover, can we express the gene in liver specifically and induce the liver tumors without the obesity and without the diabetes?" Affirmative answers to these questions would enable researchers to begin to understand cellular signaling mechanisms in these tissues and how they cause some cells to form tumors.

To accomplish this goal, Woychik and his team will build DNA constructs--variations on the agouti gene with specific instructions for where the gene should be expressed. These constructs will then be introduced into fertilized mouse eggs, producing genetically engineered, or "transgenic," offspring.

"If we make a transgenic mouse with the normal agouti gene," says Woychik, "the gene would be expressed only in the skin because there are regulatory elements associated with the gene that tell it to behave that way. What we are going to do is trick the gene into being expressed in the liver, for example, by building a construct that connects the DNA sequences that control expression of genes in the liver to the agouti gene. Then we'll use that construct to make transgenic mice in which the agouti gene product is expressed only in their livers."

The problems created when the agouti gene shows up in too great a quantity or in places where it shouldn't be at all continue to drive this research effort. "This is an important research area," says Woychik. "Understanding the agouti gene and the consequences of its overexpression may provide keys to understanding human health problems like diabetes, obesity, and the growth of tumors, as well as giving us insights into how communication among cells occurs."

Tracking Down the Cleft Palate Gene

One of the areas of DNA that mice and humans have in common is located on chromosome 7 in the mouse and chromosome 15 in humans. This region first came to the attention of researchers because it contains one of the genes responsible for determining the color of a mouse's coat, making genetic alterations in the area easy to track.

Over the years, researchers have employed chemicals and radiation to remove dozens of different pieces of DNA from this region in an effort to isolate and study the coat-color gene, as well as other genes located nearby. An unexpected effect of these deletions was that some of the genetically altered offspring developed cleft palates in addition to their coat-color changes, suggesting that a gene involved with forming palates was also in the neighborhood.

By reviewing the results of previous research on this area of mouse chromosome 7 and selectively causing and studying other mutations in the area, members of the Biology Division's Mammalian Genetics and Development Section, including group leader Dabney Johnson, have isolated a gene that is at least partly responsible for the development of normal palates. They have demonstrated that the protein it produces prevents cleft palates--a birth defect in which the roof of the mouth fails to develop completely.

This gene, known as cp1, produces a protein previously thought to function only in the brain, where it acts as part of a neurological inhibitor--moderating brain activity by keeping neurons from constantly firing. However, during palate devlopment, the protein steps outside this role to contribute to the process of making normal palates. "We checked the pharmacology literature," says Johnson, "and we found that people had known for years that exposing pregnant rats, and their developing embryos, to valium and related neurodepressors could result in offspring having cleft palates. To prove that they had indeed isolated a gene that played a crucial role in the production of normal palates, Johnson and her staff used a technique known as "phenotype rescue" to inject normal copies of the gene into fertilized eggs of mice that carry the defective palate gene. Some of these normal genes succeed in taking over the function of the defective genes, resulting in offspring that have normal palates. As a result, the newborn mice, which would otherwise die within a day presumably because they suck milk up into their lungs and drown, are "rescued" from the lethal effects of the cleft palate.

Phenotype rescue experiments are by no means foolproof. Sometimes the inserted gene doesn't make any of the product it's supposed to--or it may make too much, too little, at the wrong time, or in the wrong place. In the case of this cleft palate gene, however, results of these experiments support the contention that the cp1 gene is critical to palate formation.

"This gene is just a small part of the complex developmental process involved in forming a palate," says Johnson. One in 100 of the mice that have the faulty palate gene still manages to form a normal palate, but these mice exhibit neuromuscular nervousness, they are small, and they usually survive for a only few weeks. Even mice that develop from fertilized eggs that have been fixed by adding a normal gene are still small, although they don't show the same nervousness.

Johnson believes that these mice have more problems than just a damaged cp1 gene. "We think there may be another missing or damaged gene that only shows up when the mice survive the palate problem, she says."

Johnson and her group are looking forward to using the knowledge they've gained from isolating the cp1 gene to find other genes involved in the palate formation process by studying groups of mice that are genetically very similar except for the cp1 gene.

"It is almost never just one gene that does a complex job, like palate formation," Johnson says. "It always seems to be a cascade of genes, so when one gene is missing, the cascade is interrupted and the end product is abnormal."

Johnson and her group will have ample opportunity to explore the nuances of the cp1 gene thanks to a three-year grant from the National Institutes of Health supporting their research. This bodes well for determining the cause of cleft palates in humans. "Mouse mutations that resemble classic human birth defects get people's attention," Johnson says. "People have looked at cleft palates for years and never found evidence that a single gene was responsible for the condition. The situation may be more complex in humans, but the genetic similarity between the regions of mouse and human chromosomes is very strong.

"This is a recessive gene that works in classic Mendelian fashion," she continues. "It must be missing in both parents to cause a cleft palate, and when you replace the gene, the cleft palate goes away. This finding suggests that the cp1 gene alone can correct the defect--there is no other gene missing from the deleted region of mouse chromosome 7 that is involved in palate formation. If we can prove this result experimentally, it will be very exciting."


At the genetic level, mice and men share much more than a passing resemblance, as these efforts to unearth the genetic roots of sickle-cell disease, diabetes, and cleft palate demonstrate. For almost half a century, ORNL scientists have cultivated this similarity to provide a living model for the structure, the rhythm, and the functioning and malfunctioning of the genetic engine that powers the human species.

More and more frequently, these efforts are yielding insights into the genetic causes of human disease and holding out the promise of treatments or even cures for these ailments in the not-to-distant future.

Where to?

[ Next article | Search | Mail | Contents | Review Home Page | ORNL Home Page ]

Other human genome work covered in the Review

Search Human Genome News