By David N. Leff

Try this word-association game on family, friends and colleagues: Say "cigarette smoke" and count how many think of lung cancer.

And how few think of bladder cancer.

To be sure, pulmonary carcinoma is the No. 1 cause of cancer deaths in the U.S. Urinary bladder carcinoma ranks fifth among all cancers in this country, with more than 50,000 cases diagnosed each year and 10,000 deaths. Its first sign is blood in the urine.

Bladder cancer's cause-and-effect link to smoking has been known for a long time, but just how dangerous cigarettes are to bladders is a hot research topic.

One such researcher is Jack Taylor, who heads the Molecular and Genetic Epidemiology Section at NIH's National Institute of Environmental Health Sciences (NIEHS), in Research Triangle Park, N.C.

"[Among] the many cancer-causing constituents of tobacco smoke," Taylor told BioWorld Today, "are the arylamine bladder carcinogens." Here's how he explained the predilection of this particular benzene-ring molecule for the epithelial tissue of the urinary bladder:

"Like many other carcinogens, when an arylamine comes into the body, it doesn't by itself cause that much damage to the DNA. It's a procarcinogen, which has to be activated by the body's own enzyme system in order to turn it into a really bad actor," Taylor explained.

"So when an arylamine gets into the liver," he continued, "it can go one of two pathways. It can be deactivated if you've got an enzyme called NAT2 around in that organ. NAT stands for N-acetyl transferase. Normally, the liver highly expresses NAT2, and deactivates it.

"But another pathway exists," Taylor continued, "the blood. So that arylamine gets into the bloodstream, goes to the kidneys and gets filtered into the urine. Then it can be taken up from the urine by the bladder epithelium, where NAT1 is highly expressed."

Each of the NAT genes, Taylor pointed out, exists in 20 or so polymorphic variations, or alleles. One version of NAT1 occurs in an allelic form called NAT1*10. "You can take that arylamine and NAT1*10 and turn them into a highly reactive species that can bind to DNA," he said. "So, in that way, being slow for NAT2, the deactivating process in the liver puts you at increased risk for bladder cancer, and being fast for NAT1*10, the activating step puts you at increased risk as well."

For Taylor, "slow" and "fast" have special meanings.

"People have done phenotyping where they look at a person's genotype and then at the various detoxification things the enzyme does," he explained. "So, when we say someone is 'slow' for NAT2, that means he or she has a genotype that would confer a slow phenotype, implying slowness of the detoxification pathway.

"A variety of different enzyme systems have arisen through evolutionary time," he observed. "They probably reflect in part the evolution of strategies for detoxifying various dietary constituents. That is, the continuing struggle between plants, to keep animals from munching them, and animals munching plants, and figuring how to detoxify the nasty substances that the broccoli comes up with."

Bladder Cancer Risk Factor More Than Multiplicative

Taylor is senior author of a paper in the current issue of Cancer Research, dated Aug. 15, 1998. Its title is "The role of N-acetylation polymorphisms in smoking-associated bladder cancer: Evidence of a gene-gene-exposure three-way interaction."

Those three ways consist of the two NAT genes plus environmental exposure, i.e. cigarette smoke. In a generic example, he put it this way:

"Let's say you've got an exposure that increases your risk of developing disease twofold," he said. "And let's say that there's a gene that increases that risk threefold. If the two factors were independent of one another, than you'd think they would have a two-times-three, or sixfold, increase. If they interacted, the risk would be greater than multiplicative.

"So we're demonstrating the novel finding that there is interaction between smoking and NAT1*10. In fact, it went further than that. There was interaction between smoking, NAT1 and NAT2. So these things are doing more than multiplying the risk of one another.

"What we show is that smaller and smaller groups of people in the population have substantially increased risk of developing bladder cancer, associated with smoking."

Taylor and his NIEHS team began by obtaining blood samples from 230 bladder cancer patients and 203 non-cancerous controls. Using restriction-fragment-length polymorphisms, they identified the commonest NAT1 and NAT2 gene alleles in their DNA, and recorded the smoking habits of all participants in the survey.

"The take-home message," Taylor said, "is that roughly half of the population is walking around with the NAT2 slow genotype, and 35 percent or so with the NAT1*10 fast genotype. These two at-risk subgroups are relatively common in the population. They are independent of one another."

Practical Utility Of Findings Equivocal

Asked about potential clinical application of these findings, Taylor responded, "That's difficult to say. It's not that everyone carrying the NAT1*10 allele is walking around with a 'bad' allele. It's bad in terms of bladder cancer, but may well be good in terms of other sorts of disease risks, or various aspects of life.

"One of the things that we critically need is the ability to do rapid genotyping of large numbers of individuals," he said. "And, although the chip technologies of Affymetrix [Inc., of Santa Clara, Calif.] may be a great way of looking for a lot of different alleles all at the same time, it doesn't respond to the critical scientific need at the moment of being able to cheaply and effectively do a smaller number of different alleles on an on-demand basis." (See BioWorld Today, Aug. 26, 1998, p. 1.)

Taylor concluded: "We can't expect Affymetrix to make a new chip that scans the five new alleles we discovered yesterday." *