Insoluble in more ways than one, a little-known breed ofnuclear proteins is broadening its tumor-marker potential frombladder to colon cancer, then on to breast and prostate cancerand beyond.

Today's issue of PNAS (Proceedings of the National Academy ofScience) carries a report by cell scientists at Matritech Inc. ofCambridge, Mass., titled "Nuclear matrix proteins in humancolon cancer." It describes an in vitro experiment that showed100 percent specificity in distinguishing malignant fromnormal colon tissue.

Last summer, Matritech launched clinical trials of its newlydeveloped monoclonal antibodies to the nuclear matrix proteins(NMPs), for diagnosing post-therapy recurrence of bladdercancer by a non-invasive urine test (see BioWorld, June 23,1993). That controlled study of 1,000 patients at 16 medicalcenters is nearing its end point, the company's president andchief operating officer, David Corbet, told BioWorld. "We'relooking for FDA submission during the first half of this year,"he said.

At about that time, added Ying-Jye Wu, Matritech's vice-president for research, "we hope to make public additional,newly acquired information elucidating the molecular geneticbasis of the NMPs. Thereafter, once we have the sequencing,we'll be significantly closer to finding out the proteins'function."

Distorted Nuclei PP Age-Old Clue

Cancer bends the nuclei of tumor cells all out of shape,explained Paul Odgren, a graduate student in the cell biologylaboratory of Edward Fey at the University of Massachusetts inWorcester. Fey is a pioneer in discovering the transformation ofNMPs from healthy to malignant cells, which PNAS published in1986. French microbiologists noted the abnormal morphologyof tumor-cell nuclei in the mid-19th century, Odgren added,and pathologists still use it as one of the criteria to diagnose,stage and grade cancer.

But Ogdren said the actual mechanism of how a canceroustransformation alters the tissue activity of nuclear matrixproteins is still unknown.

Now NMP's long-shrouded molecular basis is beginning to yieldto cell biologists studying nuclear matrices at a double-handfulof centers worldwide.

Like the cell's cytoskeleton, the matrix serves as a sort ofscaffolding to keep the nucleus in shape. But beyond thatstructural purpose, Wu observed, "the matrix plays a morecomplex and more important role in terms of gene regulation."Indeed, it's been compared to Mission Control at NASA's SpaceCenter.

In this way, Wu and other NMP scientists now speculate that"tumor development is the result of a multistep processinvolving mutation in at least five genes, including the p53gene, plus activation of a ras oncogene." Where, how and whyNMP genes are expressed, said Corbet, "is certainly somethingwe're focused on at this very moment."

In the just-reported colon cancer NMPs, he added, "our nextstep is to develop antibodies to these proteins to detectexpression for diagnostic purposes. From the antibodies we'veprepared already, we can find out the gene sequence byhybridization or other sequencing approaches, includingprotein microsequencing, and then drive toward a commercialproduct."

Besides the colon cancer diagnostic now in the works and thebladder cancer test in clinical trials, Matritech is developingNMP diagnostics for breast and prostate cancer.

When it reaches the point of commercialization, one ofMatritech's commercial partners will take over for marketingand distribution. Yamanouchi Pharmaceutical of Tokyo hasrights to the Far Eastern sale of a colon cancer product, Corbetsaid, with the U.S. and European rights unpartnered at thispoint.

Other product-development allies include Sangtec Medical ABof Sweden, to which Matritech has ceded selected breast-cancertechnology, and Hybritech Inc., its bladder-cancer partner.

As for patents, "we have literally a patent fortress coming fromthe basic work out at MIT, which led to Matritech's creation,"Corbet said. The company has worldwide exclusive licenses tothree MIT-held U.S. patents in the NMP area and a prostatecancer patent issued to Johns Hopkins university. "As eachdisease-specific NMP is discovered," he explained, "thecompany strategy is to exclusively license that work or patentit itself." Matritech filed late last year to protect its colon cancerdiscoveries to date.

A Billion-Dollar Market

Colon cancer is the second-leading cause of cancer-relateddeaths in the U.S. and other Western countries, Matritechpointed out, citing American Cancer Society estimates that 1.2million Americans have been diagnosed with the disease -- anumber that grew by 152,000 last year with 57,000 deaths.

Added to prostate and breast cancers (the leading malignancymortalities), these three cancers afflict half a million Americansannually and kill 138,000. Matritech technology, Corbet said,has identified tumor-specific markers in all three.

NMPs released into urine or blood by cancerous cells differfrom those released by normal ones of the same type. In theexperiment reported in PNAS, Wu and his colleagues comparedNMPs from 18 adenocarcinomas of the colon to 10 healthycolon samples. In the former, they found six tumor-specificnuclear matrix proteins ranging in molecular weight from20,838 to 58,600. Not one of them occurred in healthy cells.

The 10 normal specimens contained four tissue-specificproteins, MW 17,400 to 40,640. Not one of them appeared inthe malignant cells PP a specificity score of 100 percent.

At present, the only FDA-approved serum-based clinical assayfor monitoring colorectal cancer patients is carcinoembryonicantigen (CEA). When discovered some decades ago, CEA washailed as a "tumor-specific antigen," the holy grail ofimmunotherapists. Soon, however, CEA turned up in othertissues as well, so it was downgraded to "tumor-associatedantigen." But it's still the only game in town.

Matritech hopes that when its nuclear matrix proteins are fullyelucidated and commercialized, they will overtake and perhapsreplace CEA to monitor post-therapy cancer recurrence andperhaps to screen populations at risk. The company estimatedthe worldwide market for such assays at $1 billion.

Corbet cited a recent Mayo Clinic study to assert that "in spiteof its extensive use, CEA does not detect colorectal cancer inmore than 40 percent of late-stage patients and falselyidentifies about 16 percent of persons who do not have coloncancer." Moreover, "an estimated 25 percent of colorectalcancer patients cannot be effectively monitored with CEA."

Corbet welcomed the recent identification of a colon cancergene (see HGS story) as "certainly an important discovery." Buthe added that "it merely puts an individual in a particular riskgroup, which accounts for only 15 percent of patients. On theother hand, identification of a (nuclear matrix) colon cancermarker has immediate significance."

-- David N. Leff Science Editor

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