Cut it out, burn it out, poison it. These are the three traditionalmainstays of cancer treatment _ surgery to excise the visible tumor,ionizing radiation to consume residual malignant cells, and cytotoxicchemotherapy to kill metastatic tumor cells.

These external anti-cancer expedients were a little like an army thathires outside mercenaries, instead of mobilizing its own citizensoldiery to fight the enemy.

Then immunology came of age.

If the body's own immune defenses _ antibodies and T cells _ cansearch out and destroy invading armies of alien bacteria and viruses,why not train them to find and suppress uprisings of malignant cells?The answer: Tumors are not alien and they don't invade. Becausethey bore from within, the immune system sees cancers as friend, notfoe.

For decades, cancer immunotherapists have sought the holy grail oftumor-specific antigens _ molecular flags or badges to identifymalignant cells as fair game for antibodies and T cells. But tumorantigens turn out to be specific only to the tissue of their particularcancer _ of which there are over 100.

A just-reported experimental strategy to circumvent this seemingCatch 22 seeks to recruit a recently discovered antigenic proteinnamed p53, which occurs in a wide variety of cancers. "Targetingp53 as a general tumor antigen" is the title of a paper in theProceedings of the National Academy of Sciences (PNAS), datedDec. 19, 1995.

Its senior author is cellular immunologist Linda Sherman, of theScripps Research Institute, of La Jolla, Calif.

"Our goal was to try to find an antigen that might be useful fortargeting many different types of tumors, not just one tissue type,"she told BioWorld Today.

Many such antigens hang out along the 393-amino-acid chain of the53-kilodalton p53 protein. In its normal manifestation, p53 acts as atumor suppressor, keeping cells from multiplying when or where theyshouldn't. But mutations in the p53 gene, and they are numerous,removes the normal protein's function, which unleashes uncontrolledcell division. (See BioWorld Today, Sept. 9, 1995, p. 1.)

Mutant p53 Carries Telltale Tumor Antigens

Of 6.5 million cancers diagnosed in the world each year, half at leastharbor a mutated or missing p53 sequence in their tumorous cells. Itturns up in cancers of the bladder, brain, breast, cervix, colon,esophagus, liver, lung and ovary, among others. Typical of mutantp53's nefarious role, it transforms a benign polyp into colorectalcancer.

Sherman and her co-authors decided that because so many tumortypes over-express mutant p53, it "was a good candidate" for a wide-spectrum tumor antigen target. Their objective then became "to see ifwe could identify peptides that could be recognized by cytotoxiclymphocytes, so that these killer T cells could attack those tumorcells."

But rather than zero in on mutant oncogenic p53 sequences whichvary by tumor type, they took as their point of departure the normalwild-type human p53 tumor-suppressor protein. "Its commonantigens," Sherman said, "would be shared by many tumors, so wouldstill be available for recognition by killer T cells."

For starters, she and her co-authors synthesized 31 short peptides,nine to 11 nucleotides long, from the human wild-type p53 genome.The object of this exercise was to single out sequences carryingimmunogenic epitopes that would bind to an antigen-presentingmolecule called HLA A-2.1.

Class I human leukocyte antigen (HLA) molecules are a part of thehuman immune system's major histocompatibility complex (MHC),which bring selected antigens to the attention of killer T cells. "Weused some transgenic mice developed in our lab," Sherman said,"which express the human HLA molecule. They used the A2.1 allele[gene variant] of this molecule, she explained, "because it's so verycommon. Greater than 50 percent of Caucasians express HLA-A2.1,so it's highly shared."

Her team then injected those HLA-A2.1-expressing mice with theirpalette of peptides, "and determined whether or not they wereresponding with killer T cells that could recognize a particularpeptide, that is, tumor targets which were expressing the human p53."The T cells did indeed recognize and lyse human breast andcolorectal tumor cells.

Killer T Cells Tolerate Non-Mutant p53 As `Self'

Since their report in PNAS, Sherman added, "we've done a hepatoma[liver-cancer] cell line and a papillomavirus-infected cervical cancercarcinoma cell line. So we found that we could kill a broad variety ofdifferent types of tumor cells with these peptide-specific killer Tcells."

There's only one problem _ T-cell tolerance.

"The problem," Sherman explained, "is that humans would not beable to mount an immune response against non-mutated p53, becausetheir immune system sees it as a `self' protein. Because of sequencedifferences between mouse and human, we could get a response inour immunized mice against those peptide sequences that are foreignto them.

"Whether humans would be able to mount an immune responseagainst a large amount of these antigens," she pointed out, "issomething that would have to be tested experimentally.

"If using these peptides as antigens in humans," Sherman concluded,"doesn't work because of tolerance difficulties, then we would pursuethe possibility of using the mice as the source of the killer T-cellreceptors, presumably for use in human immunotherapeutics andcancer vaccines." n

-- David N. Leff Science Editor

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