Not every startup with an interesting research tool needs to be a tool company. And not every company that studies many proteins at a time has to use mass spectrometry, or even call itself a proteomics company. One example is Merrimack Pharmaceuticals, which plans to develop drugs in the areas of immunology and oncology using its research on protein networks, with protein and antibody arrays as its core technology. In fact, the company already has what appears to be a promising drug in the pipeline.
Merrimack, based in Cambridge, Mass., was founded in 2000, with angel seed funding, by Ulrik Nielsen, Gavin MacBeath, Anthony Sinskey, Peter Sorger, Michael Yaffee, and Mike Cardone, researchers at MIT and Harvard who had developed protein microarray technology. But the company lacked sufficient capital, and an application area for its technology.
Enter Atlantic Biopharmaceuticals, another early-stage company founded in 1994 that had licensed worldwide rights to alpha-fetoprotein, or AFP, which had shown promising immunomodulatory effects. Atlantic, initially a “virtual company” working with consultants and contractors, funded research into AFP involving animal models of immunological disorders like multiple sclerosis, rheumatoid arthritis, and myasthenia gravis. After raising $13 million in venture capital in a Series B round Wharton Biotechnology Partners I, an affiliate of Wharton Equity Partners, in 2001, Atlantic acquired Merrimack, and took over its name.
Suddenly the company had not only access to capital, but also a drug, whose “value could be enhanced, and its development accelerated, through the application of Merrimack’s specialized tools,” said Tad Stewart, Merrimack’s director of business development. “It also diversified our risk,” said Peter Lewis, a senior managing partner at Wharton Equity Partners and a member of Merrimack’s board of directors. “Instead of reliance upon only one potential compound, which was AFP, we were able to have another possibility of drugs…and that was through the exploitation of this technology.” This year, Wharton invested another $2 million in the combined company, and is planning to add more money before the end of the year, according to Lewis. Merrimack has grown, meanwhile, to 16 employees, occupying about 17,000 feet of space near Kendall Square in Cambridge.
The company applies arrays with limited numbers of antibodies or proteins to several stages of drug discovery. “Rather than get every protein in the proteome down on a chip, we are focusing on smaller subsets” belonging to protein networks and families, said Mark Moody, the company’s director of technology development. These arrays are printed on the bottom of 384-well plates, thus allowing for automation.
To understand better how AFP works, for example, the company studies cytokine profiles in animals, human blood cells, and tissue culture cells treated with AFP using antibody arrays. “This tells us some information about what the mechanism of action might be and gives us some biomarkers to look for during our clinical trials and preclinical work,” said Moody.
Secondly, in order to find new drug targets, Merrimack exploits the fact that protein-protein interactions are often mediated by conserved domains. It identifies these interactions by arraying families of proteins — typically dozens — containing these domains in the bottom of microplate wells, and looking for interactions with proteins that contain complementary binding domains.
This type of array can then be adapted to a drug screening assay, looking for compounds that disrupt interactions. “If you are developing a drug, you want to target specific interactions and not have a compound that wipes out a whole class of interactions, because that might lead to side effects,” said Moody.
Other proteomics companies have chosen to study protein-protein interactions by yeast two-hybrid technology — like Myriad Proteomics and Hybrigenics — or pull-down experiments followed by mass spectrometry, favored by Cellzome and MDS Proteomics. According to Moody, the advantage of using arrays is the high-throughput, low false positive and false negative rate, and easy adaptation to drug screening.
But Merrimack does not limit itself to arrays; it has also developed proprietary cell-based assays to validate drug leads.
AFP, however, is beyond the in vitro validation stage. Animal safety studies of the protein — which is produced in transgenic goats that secrete it into their milk — are about to begin, said Lewis, and the first Phase I clinical trial in humans is planned to start next year. Initially, the company will focus on treating myasthenia gravis and has obtained orphan-drug designation from the FDA for this application. However, with the help of a pharmaceutical partner, it is hoping to start clinical trials for multiple sclerosis and rheumatoid arthritis.
According to Lewis, in addition to AFP, Merrimack has also developed another potential lead compound, in the areas of B-cell lymphoma and colorectal cancer, using its array technology. “We are in discussions with a pharmaceutical company that might be interested in working with us on this com- pound,” he said.
Collaborations are high on the list for Merrimack, both to bring its own molecules to market and to enable pharmaceutical companies to use its technology to study drug action. However, the company is not planning to sell its technology as a tool: “If we can develop a real lead from this technology, and we can bring our protein well into the clinic along its way, then we would be positioned as a significant drug discovery company,” said Lewis.