WASHINGTON _ In the ongoing effort to selectively kill cancercells rather than indiscriminately slaughtering all of a patient's cellswith chemotherapy, researchers are developing a new experimentalagent they are calling a "gene bomb." Though only at the test-tubestage of testing, one such gene bomb managed to target and kill lungcells containing a cancer protein at a rate 1,000-times higher than itkilled cells that lacked the protein.

Luis da Costa, a research fellow in Bert Vogelstein's and KennethKinzler's laboratory at the Howard Hughes Medical Institute, inBaltimore, announced the results at the annual meeting of theAmerican Association for Cancer Research (AACR) here onTuesday. Further details of the gene bomb experiment will bepublished in the Proceedings of the National Academy of Sciences onApril 30, 1996.

According to da Costa, the gene bomb consists of an explosive, thepurine nucleoside phosphorylase (pnp) gene taken from Escherichiacoli bacteria, and a trigger, an artificial gene engineered by da Costaand colleagues that is capable of binding to the p53 cancer proteinand signaling the pnp gene to kill tumor cells. The trigger gene isactivated only in the presence of a specific protein, in the case of lungcancer cells, the p53 cancer protein. Once activated by the triggergene, the pnp gene is converted from a non-toxic to a toxic force _but in normal cells, the gene bomb would remain quiescent andharmless.

"When the bomb goes off, it also kills surrounding cells which wecall the bystander effect," da Costa told BioWorld Today. "Thatcould mean that you need to put these gene bombs into a lowerpercentage of tumor cells in order to kill the tumor than waspreviously needed." One of the toughest challenges of cancer genetherapy to date has been inserting genes into a large enoughpercentage of cells in order to "kill" an entire tumor.

Another potential advantage of the gene bomb approach is that itcould be adapted, via appropriate engineering, to a number of othercancers and viral diseases that are associated with specific proteins.For example, liver cancer is linked to the hepatitis B virus andcervical cancer is linked to the human papilloma virus. Viral proteinscould thus be used to trigger cell death in tumors in these and otherdiseases.

The gene bomb developed by da Costa has yet to be tested in vivo,even in animals, and thus its promise in the test tube may not pan outin living organisms. He concedes that major problems remain to besolved before human testing could begin. Although the gene bombinnovatively tackles the problem of specificity _ since the bomb istriggered only by cancer cells _ it doesn't really solve the perennialproblem of delivery. Although its impact may be magnified by thebystander effect, the gene bomb still has to be delivered to asignificant percentage of tumor cells in vivo in order to complete itsdestructive mission.

To date, da Costa said his lab has no corporate collaborators on thegene bomb. He told BioWorld Today that the approach is based uponthe accumulated knowledge produced by years of cancer research onmutant proteins, including the finding that the p53 cancer protein isnot present in normal cells.

BRCA1 Protein Identified

Scientists have identified the protein produced by the breast cancergene BRCA1, marking the first time that a tumor suppressor gene hasbeen shown to produce a secreted protein. The protein inhibitsgrowth in breast and ovarian cancer cells but not in other types ofcancer cells, according to Roy Jensen, assistant professor ofpathology and cell biology at Vanderbilt University Medical Centerin Nashville, Tenn. Jensen announced the finding at the AACRconference on Tuesday.

Jensen said that because the protein appears to be secreted, it maywork on surrounding cancer cells. Theoretically, inserting a workingcopy of a normal BRCA1 gene into only a few cancer cells couldproduce enough of the protein to control an entire tumor. A Phase Itrial testing that hypothesis in patients with advanced ovarian canceris slated to begin in the next few months.

Jensen and his colleagues conducted preclinical tests that showedtumor growth was drastically diminished and survival was prolongedin mice with a normal working copy of the BRCA1 gene vs. thosecarrying a defective BRCA1 gene. In some mice with a "good" copyof BRCA1, tumors were completely eradicated. n

-- Lisa Piercey Special To BioWorld Today

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