In health, the pancreas secretes powerful digestive enzymes, while itsIslets of Langerhans make insulin. In sickness, carcinoma of thepancreas is among the most vicious cancers known, and fifth cause ofmalignant death in the U.S. In this country, it kills more than 25,000men and women annually.

By the time a patient's carcinoma of the pancreas is diagnosed _usually in the sixth to seventh decade of life _ his or her tumor hasbeen growing for as long as five years, and life expectancy isestimated in weeks, or months, not years.

"Smokers have a two- to three-fold increased rate of contractingpancreatic cancer," said genetic oncologist Scott Kern, of JohnsHopkins University School of Medicine. "Even though only one-thirdof the American population smokes," he observed, "they account forabout three-fourths of the pancreas cancers. This means apreponderance of men over women."

Recent research in tumor suppressor genes has turned up unexpectedconnections between pancreatic tumors and breast and colorectalcancer and melanoma. Most recently, a report by Kern and his co-authors in today's Science announces their discovery of a deletedgene on human chromosome 18's long arm. It apparently accountsfor more than half of all pancreatic cancers.

Their article bears the title: "DPC4, A candidate tumor suppressorgene at human chromosome 18q21.1"

In 84 tumor samples, Kern told BioWorld Today, "The site of thep53 tumor suppressor gene on chromosome 17 had deletions in 90percent of the tissue samples, and mutations in 70 percent." Also, hecontinued, "losses or mutations of the ras gene amounted to 90percent, and of the melanoma-linked p16 gene of over 80 percent."(See BioWorld Today, Jan. 2, 1996, p. 1.)

Besides these three known tumorigenic genes, they found extensiveunexplained missing sequences in the DNA of chromosome 18. "Thismeant something big was going on there," Kern observed, "so wedecided to first describe that chromosome, to answer the question: Ischromosome 18 intact, or is it broken?"

In one small area of its genome, they found deletions affecting boththe maternal and paternal chromosomes, implying a homozygousgene defect. Another 17 to 20 percent had mutations of a single geneat that site. "That gene was the fourth locus we had looked at that wasdeleted in pancreas cancer, so we called it DPC4," Kern said.

Lucking Out At 11th Marker

Then he and his team turned to the international Human GenomeProject for genomic markers on chromosome 18. "They enabled us,"Kern said, "to describe that chromosome in great detail. That doesn'tmean that we didn't get lucky. We were prepared to do it usinghundreds of markers. What happened was that after the 11th marker,two of them turned out to be very close to each other, and in an areaof homozygous deletion."

They found that 30 percent of their 84 tumor samples hadhomozygous deletions at the DPC4 site, while another 22 percentcarried mutations that could inactivate the gene.

That gene, Kern said, "encodes a protein that does not have a domainresembling any other known domain. It doesn't seem to have anytransmembrane portion, or any leader sequences that would put it intothe cell's endoplasmic reticulum."

But this oddball protein's gene did match the DNA sequence of agene called Mad in the genome of Drosophila melanogaster, the fruitfly. In that insect, it apparently controlled embryonic development.And Mad encoded proteins in a pathway similar to that of the humantransforming growth factor-beta (TGF-b).

"These always seem involved in controlling a cell's behavior, and wethink that TGF-b pathways probably suppress cell growth in general,"Kern said, adding, " We don't know yet whether in humans itcontrols development. But the idea of a developmental protein beinga key tumor suppressor pathway for pancreatic ductal epithelium isan attractive way of looking at growth control."

This hints at diagnosis and therapy of pancreas cancer. Kern said that"There's nothing of clinical utility right around the corner." Butseveral corners down the road of time he foresees such applications.

"For people involved in the biotech companies," Kern suggested,"someone ought at some time to bring up an interesting question: Isthere a circulating factor, or factors, that could restore the brokencellular pathway involved in pancreatic tumorigenesis?"

He foresees two ways to circumvent such pathway blockage: "One isto find perhaps a parallel, TGF-b alternative pathway that the celldoesn't happen to use, but a physician could. Or, try to mimic theDPC4 gene's action."

Diagnostics With Strings Attached

Kern continued, "I don't like talking about gene therapy, because Ithink that's way too far off in the future. But I think the idea ofstimulating a pathway is something that's well within the reach ofcurrent pharmacology.

"You could imagine," he said, "a patient swallowing a capsule, with astring attached. The clinician would pull on the string and grab someduodenal juice, which would show whether there are mutations."

Such a diagnostic device is definitely not around Kern's corner. "Thetechnology is going to hold us back," he observed, "because there'sno way of determining the sequence of, say, 1,000 cells at one time_ and keep `em all separate. In the future, you can imagine certaincloning techniques which would allow you to do that. That's not at allunreasonable." n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.