Of Mice, Canaries and Oncogenes
Imagine you're an eighteenth century Welsh coalminer. Every day you descend into dark tunnels, seeing by only the light you bring with you. You know explosions sometimes flash through tunnels just like these, perhaps sparked by a lamp, incinerating those just like you -- with no warning. They say these explosions happen because there is something in the atmosphere in these tunnels, something that can't be seen, or smelled, or tasted, something that burns like tinder to the spark. And you've heard stories of bodies found, people who suffocated because the tunnel's atmosphere had too little oxygen to sustain life. The unseen gases in the mine are like evil spirits whose presence isn't realized until too late. And aren't those dangers undetectable by human senses, the most fear-inducing of all?No one now remembers how those miners discovered that canaries are extraordinarily sensitive to the presence of the dangerous unseen gases that accumulate in the tunnels of coal mines, or thought of using their sensitivity as a signal of dangerous levels of those gases. Canaries are especially sensitive to the colorless, odorless gas methane, dying within minutes when exposed to sufficient levels. Methane levels high enough to kill a canary also pose a high risk of explosion. When a canary died, the miners knew it was time to take action -- time to get out of the mine.Since the day of the miner's canary, testing for hazardous substances has been enormously refined. As human society has grown both in the number of chemicals and other agents poured into our environment, it has also grown in awareness of the toxic and disease-causing effects of many of these substances. And in developed countries, one of the most feared diseases is cancer. Sixty years after these words were spoken, they remain true:"For everybody that is killed by the fact of cancer, multiplied thousands of minds are unnerved by the fear of cancer. This is an incidence of 'surgeon's knife, or any organized advice against panic. --(Glenn Frank, introductory remarks at the 1936 A Symposium on Cancer, University of Wisconsin School of Medicine, pub. by Univ. of Wisconsin Press, Madison, 1938)The genetically-engineered Harvard "oncomouse" and its relatives are among the modern successors to the miner's canary -- these specialized mouse strains can be used to assess the risk of cancer posed by exposure to particular chemicals. But in addition, they can help us understand carcinogenesis - the process leading to cancer.The story of these engineered mice begins with the earliest demonstrations that cancer could result from exposure to environmental agents. Cancers of various kinds were recognized and categorized by physicians as early as 400 B.C., but the causes of cancer, like those of most diseases, were unknown. One of the earliest recorded clues to the cause of cancer is the 1775 observation by the English physician Percival Pott that patients with a rare type of testicular cancer, had in nearly every case been heavily exposed to soot in childhood, while working as chimney sweeps.Ironically, another important clue that cancer might be caused by an environmental agent came out of one of the earliest non-surgical treatments for cancer -- radiation. Roentgen, Madame Marie Curie and other physicists involved in the first studies of the radioactive properties of elements like uranium often carried chunks of the minerals around in their pockets. They noticed that prolonged exposure to highly radioactive minerals caused burning of the skin and eventually, the development of skin tumors. And in the first half of the century, radiologists developed leukemia at twice the rate of the general population.But the fact that cancer can be induced by exposure to certain chemicals was not conclusively demonstrated until 1915. The Japanese pathologists Yamagiwa and Ichikawa painted crude coal tar extracts on the ears of rabbits; repeated applications over a period of months resulted in skin tumors. Voila -- the first test for carcinogenicity was born.In later studies, animals fed a suspect chemical were also found to develop cancer. Rats and mice became preferred cancer testing animals by virtue of their small size, short natural life span, and because they bear some similarity to humans in their diet and nutritional metabolism. As of the late 1940s, however, the mechanism by which cancer was induced was not understood. The subsequent refinement of tests for carcinogenic potential followed closely on the discoveries made concerning that mechanism.By the late sixties and early seventies, scientists had learned that the development of cancer was closely linked to changes in DNA (deoxyribonucleic acid, the polymer that encodes the all the information needed to produce living cells). Tests were developed using bacteria, yeast, and mammalian cells cultured in the laboratory, to identify agents which could cause damage to DNA, mutations, etc.However, comparing the results of microbial and cell culture tests to whole-animal tests showed that not all agents which could damage DNA also caused tumors. The converse was also true -- some chemicals that had been shown to cause cancer in animals did not appear to cause DNA damage or mutation. In the latter case, it is often because the chemical is biochemically processed to a new, "activated" form which does attack DNA, by enzymes in the host's own liver. Thus, these newer tests did not render mouse studies obsolete.Scientists also suspected that it was damage or mutation to specific genes - not just damage to DNA as a whole -- that was the likely culprit. Comparing the DNA of tumors to those of non tumor cells from the same patient, they identified specific mutations (changes in the coding sequence, which often produce a non-functional gene) and translocations (rearrangement of the order of DNA genes and gene segments) that were strongly associated with certain forms of cancer. Still, it wasn't clear at first whether these genetic alterations were a cause, or a consequence, of cancer.Attempting to answer this question, some researchers focused on the study of certain human syndromes featuring an increased incidence of cancer. Studies of retinoblastoma (Rb) tumor patients by Dr. A.W. Knudson were particularly fruitful, generating a major paradigm for the genetic history of individual tumors. Retinoblastoma is a rather rare tumor of the retina (part of the eye). Doctors observed that in about 40% of cases, the susceptibility to retinoblastoma had apparently been inherited. Those individuals tended to develop tumors in both eyes, not just one, and/or multiple tumors in one eye.Close analysis of the chromosomes in these multiple tumors revealed that they were of independent origin -- they weren't produced by metastasis (the spread of one original tumor to new locations in the body). This analysis also showed that in nearly all retinoblastomas, whether of the heritable form or not, the same chromosomal segment was missing. In retinoblastoma-prone families, children who developed the tumors were found to have inherited one chromosome in which the key region was already deleted.Dr. Knudson proposed that, at least for retinoblastoma, a key event in the production of cancer could be the complete deletion of a gene. Continued research verified Dr. Knudson's hypothesis and showed that the gene whose deletion or inactivation caused a retinoblastoma was what is known as a tumor suppressor gene. That is, correct activity of the gene regulates cell reproduction so as to prevent tumor development. Another tumor repressor gene was soon identified, termed P53. Non-functional p53 genes are found in many types of human tumors including colon cancer.Scientists continue to expand their study of families with high incidences of other types of cancer, and have now isolated a number of different genes whose deletion or altered regulation can lead to cancer. Not all of these are tumor repressors: in cases like that of the myc gene, it is an increase in the number of copies of the gene or the amount of product made from the gene that correlates with the cancer. In other cases, it is a rearrangement of the gene that separates it from genetic elements that normally regulate the gene. Collectively, all these cancer-related genes are known as oncogenes (onco -- is a prefix meaning tumor).And now, we come to the "Harvard mouse" -- the first mouse genetically engineered to produce an animal test system mimicking the oncogenic process deduced from human cancer-prone syndromes. This animal also has the distinction of being the first mammalian organism on which a U.S. patent was granted; in fact the nickname "Harvard mouse" was coined by patent attorneys. They referred to it as the "Harvard mouse" because the inventors (headed by Philip Leder) did the work at Harvard University, and thus Harvard owns the patent rights.So what exactly is the Harvard mouse? It's actually a strain of mice, in which each mouse carries a form of the myc oncogene. The C-myc oncogene is one of those for which too much of the gene's product (the protein the gene encodes) is the culprit, rather than too little as for Rb-1. Human and most or all other mammalian cells carry a useful, non-tumorigenic version of the myc gene. The Harvard Mouse strain additionally carries an altered, de-regulated copy of the gene. This "activated" myc gene was introduced using techniques of genetic engineering.Normally, the production of protein from the myc gene is regulated to occur as a cell moves out of the quiescent (non-reproducing) phase into the reproductive cycle. When the myc oncogene is present in multiple copies, or is rearranged so that it is separated from its normal regulating segments, the protein is over-produced in cells that ordinarily would remain quiescent. The 'excess of myc proteins appears to stimulate cells to continue reproducing. De-regulation of the myc gene can be visualized as rather like the Disney cartoon of The Sorcerer's Apprentice -- all those broomstick fragments jumping up to become whole brooms, mopping and mopping out of control.De-regulation of my alone isn't sufficient to result in malignant cancer, but it is one step along the path. We now know that in most cancers, multiple genetic alterations involving several 'different oncogenes have occurred by the time the cancer is full-fledged. Altered mycgenes are found in many human and animal tumors, usually along with a mutated or deleted tumor suppressor gene like Rb-1 or P53. Other activated oncogenes may be present as well. But the roles of the various oncogenes in promoting the development of specific types of cancer (such as the BRCA-1 gene which is found in breast cancer-prone families), isn't yet understood.However, the mice engineered to carry a de-regulated mycgene are more prone to develop tumors, even without experimental exposure to carcinogens. Research using these mice should prove very useful in sorting out the many factors, such as diet and hormones, known to play roles in the carcinogenic process. The mice can also be used to evaluate the effectiveness of agents thought to be anti-carcinogenic, such as the anti-oxidants Vitamin E and beta carotene. Since the Harvard Mouse in 1984, other "oncomouse" strains have been developed. One example is of the type called a "knockout" mouse, meaning that one or both copies of a particular gene has been deleted by the use of genetic engineering. For example, a mouse that mimics the inherited susceptibility to retinoblastoma has been produced. In this mouse, one copy of the Rb-1 gene has been deliberately sliced out, leaving only one functional copy, just as in the human syndrome. Interestingly, if both copies of the Rb-1 gene are knocked out, the mouse embryos die, indicating that function of the Rb-1 gene is necessary for normal fetal development, not only as a suppressor of tumors.Studies using these specially-engineered mice may help us learn to what extent modifications in diet and living habits affect the carcinogenic potential of chemicals. The mice may also help researchers dissect the steps in the process leading to fullblown malignant cancer. But, unlike the miners with their canaries, our situation is different - we can't evacuate this mine. And, our ability to clean up the many chemicals we discharge into our environment is limited at best. Many of them, including PCBs, chlordane, and other now-banned pesticides persist for years in the environment, finding their way into drinking water, accumulating in tissues of fish, making them unsafe for us to eat.xWe, as a society, do still have a lot more choices than eighteenth century miners did, and a lot more knowledge of the hazards we face. But won't we feel stupid if down the line, we come to understand every detail of the carcinogenic process, but are unable to stem the flood of new cancers arising from exposure to chemicals that we ourselves dumped into our only home?