ddn Cancer Research News Exclusive: Mapping metastasis

BALTIMORE—Even in the face of an assortment of increasinglyeffective treatments, breast cancer continues to claim several thousand liveseach year, ranking as the third most deadly type of cancer. The primary causefor breast cancer's high death rate is metastasis, but despite its importance,few specifics are known regarding the process.

Thanks to a recent study from a team of researchers at JohnsHopkins, however, there are a few more answers. The team's work revealedhow a specific protein that ensures cell survival in low-oxygen situations cancause cancer cells to spread into the lymphatic system, with the resultspublished in the Sept. 10 edition of the Proceedings of the National Academyof Sciences Online Early Edition.

The study focused primarily on a protein known ashypoxia-inducible factor 1, or HIF-1. In the case of solid tumors, in whichcancer cells can grow so densely that they starve themselves of oxygen, cancercells will activate this protein to initiate the growth of new blood vessels.The HIF-1 protein was discovered nearly 20 years ago by Gregg Semenza, M.D.,Ph.D., director of the vascular program at Hopkins' Institute for CellEngineering, and his team, as noted in a Johns Hopkins press release.

Semenza says that HIF-1 has been studied by numerousinvestigators in a variety of cancers, with the finding that those with the highestlevels of HIF-1 alpha in primary tumor biopsies are most likely to die of theirparticularly cancer. There is an association between expression of HIF-1 alphaand increased patient mortality in several cancers, he notes, includinguterine/cervical, colon, lung, pancreatic, prostate and bladder cancers.

The research team, led by Semenza, the C. Michael ArmstrongProfessor of Medicine and a member of the McKusick-Nathans Institute of GeneticMedicine, has worked with HIF-1 before in examining how breast cancer tumorsspread from the breast to the lungs. In a previous study, the team found thatHIF-1 interference in the mice with the breast cancer tumors reduced growth ofprimary tumors and also prevented metastasis through blood vessels leading tothe lungs.

Building off of that discovery, Semenza and his colleaguesinjected mice with human breast cancer cells that had been geneticallyengineered to knock down HIF-1 protein levels. After 24 days, the mice wereexamined and it was discovered that compared to mice with unaffected HIF-1levels, the mice with knocked-down HIF-1 had lymph nodes with 76 percent fewerhuman breast cancer cells.

"We've known that increased levels of HIF-1 areassociated with increased tumor vessels and with patient mortality," Semenzasaid in a press release. "Now we've found that HIF-1 activity is directlyresponsible for the spread of breast cancer to the lymph vessels."

The team's next step was to starve human breast cancer cellsof oxygen to see which genes would respond to HIF-1, and they found that theplatelet-derived growth factor B gene (PDGF-B) was five times more active whenoxygen was scarce. PDGF-B increases the delivery of oxygen to cells that arenot currently getting enough oxygen by stimulating blood vessel growth, andalso enables individual cells to survive in low-oxygen conditions. Closerexamination showed that the DNA sequence surrounding the PDGF-B gene presentedregions of DNA known to be recognized and bound by the HIF-1 protein, andfurther tests demonstrated that the HIF-1 protein binds to the PDGF-B gene andactivates it. In addition, they discovered that PDGF-B produced by breastcancer cells is pumped out of the cells and stimulates the growth of lymphvessels.

Semenza and his team treated the cancer model mice witheither digoxin, which blocks HIF-1 activity, or imatinib, a cancer drug. Thesetreatments were found to reduce tumor size by 78 percent and lymph nodemetastasis by 94 percent.

The two ways that breast cancers spread are by getting intolymph vessels and then spreading to the lymph nodes, or by getting into bloodvessels and then spreading throughout the body, says Semenza. And women who hadmetastases all over their bodes, he notes, always present with lymph nodemetastases.

"So we were focused particularly on trying to understand howhypoxia within the cancer might promote the metastasis of breast cancer cellsto the lymph nodes," says Semenza, "and what we found was that the expressionof PDGF-B was increased by hypoxia, that it was dependent on HIF1, and whatPDGF-B did was that it stimulated the endothelial cells that make up thelymphatic vessels. It stimulated them to divide, so there's more lymphaticvessels around in the presence of PDGF-B. And the other thing is that thePDGF-B can actually attract the lymphatic endothelial cells, so in other words,the lymphatic endothelial cells will move towards the tumor, and the lymphvessels will come into the tumor, and that will again bring them in proximityof the cancer cells. The cancer cells can then invade into the vessels andspread to the lymph nodes."

Moving forward, Semenza says he and his team will bestudying HIF-1 expression in triple negative breast cancer, which ischaracterized by a lack of expression in three receptors: the estrogenreceptor, the progesterone receptor and the HER2 neu receptor. As a result,these cancers cannot be treated by anti-estrogen therapies or the popular drugHerceptin, and instead are treated with cytotoxic chemotherapy. Unfortunately,as Semenza notes, "they tend to relapse very early after treatment and to dovery poorly, so there really isn't good therapy for this group of patients atthe moment. We're interested in testing whether inhibitors of HIF-1 might beuseful as part of their therapeutic regimen."

Oncologists at Hopkins will be initiating a clinical studylater this year to treat women with breast cancer with digoxin to see if thedrug can inhibit HIF-1 activity in breast cancer patients the way it did in theanimal models.

"This first trial is just a very simple trial to ask whetherat the doses that can be used in patients, whether digoxin is an effectiveinhibitor the way it is in mice," says Semenza. "And then if we have evidencethat it's working to block HIF-1, then the next step would be to perform atrial probably in combination with other cancer drugs to see if there is anactual therapeutic effect."

Funding for the study came from grants from the AmericanCancer Society and the National Cancer Institute, as well as from funds fromthe Johns Hopkins Institute for Cell Engineering. The paper's authors include,in addition to Semenza, Luana Schito, Sergio Rey, Huafeng Zhang and CarmentChak-Lui Wong of Johns Hopkins; Marco Tafani and Matteo Russo of SapienzaUniversity of Rome; and Andrea Russo of Istituti Fisioterapici Ospitalieri.