What does honey bee colony collapse disorder have to do with a potential new cancer treatment?
They both relate – in a convoluted manner – to an old antibacterial drug called nitroxoline.
True to my devotion as a natural product pharmacologist, I’m proud to say that new life would not have come to nitroxoline had not a fungal natural product called fumagillin been studied as an antiangiogenic anticancer drug – one that inhibits the formation of new blood vessels.
Let’s step back for a moment. Fumagillin, was first isolated as an antibiotic in 1949 from a culture of Aspergillus fumigatus. Most recently, fumagillin has been used in honey bee hives as an antiparasitic drug to control the microsporidium parasite, Nosema cerenae. You might recognize the name of this parasite from the controversial PLoS One paper last year by Jerry Bromenshenk’s group at Montana who hypothesized that the parasite synergizes with the iridovirus IIV to cause colony collapse disorder. (Disclosure: I write a blog at PLoS Blogs network, operated by the publisher of PLoS One).
Bee that as it may (pun intended), fumagillin was serendiptously identified again in 1990 by a group led by the late antiangiogenesis pioneer, Dr. Judah Folkman. (Note to the reader: I’m treading here on the expert ground of SBM editor Dr. David Gorski, so I’m sure he’ll weigh in to correct me somewhere herein).
In their Nature paper, Folkman’s group found Aspergillus fumigatus as a contaminant of a culture of capillary endothelial cells and noted that it caused cell rounding similar to other angiogenesis inhibitors. Indeed, the fungus was secreting fumagillin. However, fumagillin itself had too much systemic toxicity when used in mouse models of human cancer. So, with colleagues from the Japanese drug company Takeda, the group synthesized analogs of fumagillin, the most active of which was a compound called AGM-1470 or, later, TNP-470. However, the drug suffered from a poor bioavailability when taken orally and a short half-life in clinical trials. Folkman’s group continued to work on a new formulation of TNP-470 described in a 2008 Nature Biotechnology paper that appeared after he passed away at age 74.
Most noteworthy about fumagillin and its analogs was that they revealed an entirely new way of halting the growth of endothelial cells in tumors. In 1997, the laboratory of Craig Crews at Yale University took advantage of a chemically reactive group on the fumagillin molecule – an epoxide – to purify from human umbilical vein endothelial cells (HUVEC) a protein that was bound by the drug. As detailed in this PNAS paper, the drug’s target turned out to be a metalloproteinase called methionine aminopeptidase-2, or MetAP2. A group led by Jun Liu, then at MIT, demonstrated that TNP-470 also targeted this enzyme which acts by cleaving the initial methonine residue from a subset of proteins.
Most recently, Liu’s group, now at Johns Hopkins School of Medicine, screened a library of 175,000 compounds for agents capable of inhibiting MetAP-2 in parallel with another group of approved drugs from all therapeutic classes. Emerging from both of these screens was nitroxoline, a drug used to treat urinary tract infections in South Africa (PDF) as early as 1978, sold under the trade name, Nicene-N.
In the Liu group’s most recent paper appearing in November 2010 in the Journal of the National Cancer Institute, nitroxoline was investigated for its antiangiogenic activity against human xenografts of breast and bladder cancer in mice. At doses similar to those given to humans, nitroxoline had good efficacy – but not hugely impressive – in slowing tumor growth and inhibiting microvessel formation both models. The activity against the bladder cancer was interesting since the tumor cells were implanted orthotopically (in the bladder) rather than in the subcutaneous tissue of the animal. Why bladder cancer? Because over 70% of the administered dose of nitroxoline in humans accumulates in the bladder. These findings are, of course, preliminary – the investigators only used one dose of the drug in this work and did not try to combine it with any other commonly used chemotherapeutic drug.
Usually, I don’t get too excited about a drug with activity against human tumor xenografts until it shows good pharmacokinetic properties in people. However, nitroxoline has been used clinically and already has a history of safe human use. Repurposing old drugs with activity against novel targets presents a lower barrier of risk for pharmaceutical development when compared with an exciting novel compound of unknown human safety and bioavailability. Such compounds are more likely to proceed to clinical trials – although nitroxoline is not currently approved in the US, its approval elsewhere is likely to speed human studies.
No, no, nitroxoline will not be a cure for cancer. But it does provide a lesson of the value of natural products in the identification of new cellular targets. Nitroxoline’s antiangiogenic activity would not have been discovered had fumagillin not been identified as a MetAP2 inhibitor. And further study of nitroxoline has revealed that it has an additional activity not shared by TNP-470: it also inhibits the SIRT1 class of histone deacetylases. This finding opens another avenue for the study of SIRT1 inhibitors with agents that target MetAP2.