Building Better Proteins

A virus (blue) surrounded by immunoglobulin (IgG) molecules. The Y-shaped antibody molecules have two arms that can bind to specific antigens, marking pathogens for destruction by immune cells.

© TIM VERNON / SCIENCE PHOTO LIBRARY

There are approximately 24 monoclonal antibody therapies on the US market at present, and approximately 240 more in various stages of clinical investigation. Nearly 30 were in Phase III trials as of last October. Between 2003 and 2004 the market for these compounds, which treat everything from cancer to multiple sclerosis, grew by more than 48 percent to more than $10 billion, and has since quadrupled to some $40 billion.

But the rapidly expanding market is getting crowded, making it an unfriendly place to introduce a whole new antibody therapy into the mix. At the same time, the patents for many of the first antibodies, developed in the early 1990s, are about to expire. This has some drug makers engineering existing products into new and improved antibodies, or whole series of similar antibodies, to extend the patent lives of those therapies.

“That concept is valid,” says Janice Reichert, senior research fellow at Tufts University’s Center for the Study of Drug Development and editor-in-chief of the mAbs journal, which publishes research articles on monoclonal antibodies. “There are definitely ways to improve the molecule and come up with something that’s better.”

Engineering antibodies may extend their lifespan in the body or make them more potent, say several companies seeking to use the approach to develop better antibodies to add to their own pipelines or license engineering technology to larger drug makers to apply to their products. Some smaller companies have already attracted multimillion-dollar deals with Big Pharma hoping to make novel antibody therapies just different enough to warrant a new patent—giving their therapies new life as optimized treatments.

Systematically altering the makeup of proteins is a relatively young science, but applied to the development of optimized antibody therapies, it may yield the next generation of treatments for an array of diseases and disorders. “The ability to do protein engineering is not big news,” Reichert says. “It’s just a matter of how creatively you use it.”

MedImmune, the Maryland-based AstraZeneca subsidiary, has a few antibody therapies in its pipeline that came into being thanks to the company’s protein engineering platform. Like most monoclonal antibody–based approaches, MedImmune’s therapies use immunoglobulin G (IgG)—the most common antibody type, which recognizes pathogenic viruses, bacteria, and fungi. MedImmune engineers the Y-shaped IgG molecule’s tail, called the Fc or constant region, which interacts with the immune system to clear pathogens or prevent their entry into cells. Changing the sequence in the Fc region, even if only by a few amino acid residues, can cause big differences in how long antibodies persist in the blood or how strongly they attract immune cells.

Herren Wu, vice president of antibody discovery and protein engineering at MedImmune, says that the company’s half-life extension platform, called YTE for the three amino acids they swap into the Fc region, has yielded promising compounds, one of which is currently undergoing clinical testing. “We engineer the Fc region through mutations,” he says. “Those mutations enable us to have a molecule that has a much longer half-life than the normal antibody.” Wu says that the triple mutation in the Fc region extends half-life by increasing the antibody’s binding affinity for the neonatal Fc receptor, which binds circulating IgG molecules and recycles them by rescuing the antibodies from lysosomal degradation. The more IgG antibodies that bind to this receptor, the longer they persist in the blood.

MedImmune’s IgG-based MEDI-557—which protects against respiratory syncytial virus (RSV), the most common cause of bronchiolitis and pneumonia in US children under the age of 1—is in Phase I trials. It contains the 3 YTE amino acid changes in its Fc region, and preliminary data in monkeys suggests that its half-life is 3 to 4 times longer than the half-life of palivizumab, MedImmune’s approved RSV preventive from which MEDI-557 was derived (J Biol Chem, 281:23514–24, 2006).

Wu says that MedImmune also engineers antibodies to manipulate how they interact with pathogens, making the molecules more disruptive to defined targets. Motavizumab, one such binding kinetics-enhanced compound in MedImmune’s pipeline, is 10 to 20 times more potent than palivizumab, its unengineered precursor. The compound, which has completed Phase III trials for the treatment of RSV, includes changes in its variable region (outside the Fc region) that increase the stickiness of the antibody to a protein that RSV requires for entering a host cell.

In addition to using its protein engineering techniques to improve its own pipeline, MedImmune’s antibody half-life extending technology is attracting interest from other drug makers, Wu says. “We are talking with industry,” he says, without naming names. “There are companies interested in using our technology.”

MedImmune’s YTE half-life extension technology uses three amino acids (threonine, glutamic acid, and tyrosine) in the Fc region of IgG.

Courtesy of MedImmune

Another biotech company, California-
based Xencor, has already had success convincing Big Pharma that protein engineering is a fruitful exercise. Last year, Xencor struck licensing deals with two pharmaceutical giants for its Xtend Antibody Half-life Prolongation Platform which, like MedImmune’s YTE technology, manipulates the Fc domain to increase the lifespan of antibody therapies in patient’s bodies, thereby lowering dosing requirements.

Pharmaceutical companies are “taking notice of methods to optimize the performance of the proteins,” Xencor’s president, CEO, and cofounder Bassil Dahiyat says. By optimizing the functionality of their antibody therapeutics, drug companies large and small can enhance the longevity and diversity of the products in their pipelines. “[That] is a huge aspect of why people want to do protein engineering.”

In March 2009, Xencor licensed the technology to Merck for a $3 million fee with additional payments promised upon selection of a successful variant, achievement of clinical development milestones, and royalties on sales of an approved product. That same month, Xencor entered into a licensing agreement with Pfizer, also aiming to use the company’s half-life prolongation technology. Neither the pharmaceutical companies nor Xencor have disclosed exactly which Merck or Pfizer compounds will be engineered using the Xtend technology, but Dahiyat says that the drug companies plan to “use it widely.”

This could change the way drug companies develop antibodies, Dahiyat says. “Our hope is that people start using our Fc domain for any antibody drug that they’re going to make.”

Dahiyat adds that the field of protein engineering has advanced rapidly in recent years. “Protein engineering, in the last 5 or 6 years, has gone from being poor to middling to becoming explosive.” This has allowed biotechs such as Xencor to rapidly get their own antibody pipelines up and running by using the emerging technology.

Xencor has demonstrated some significant improvements in antibody function by changing only 2 amino acids—out of the 225 in the Fc region. For example, using the company’s half-life extension technology, Dahiyat says that Xencor scientists have modified an IgG antibody to live three times as long as the wild-type antibody. Also, Dahiyat says that they’ve improved the binding of IgG antibodies to particular Fc receptors by 50- to 100-fold, which can increase the strength with which the antibodies bind to effector cells and the vigor with which they attack pathogens or tumor cells.

Macrogenics, a biotech based in Maryland and San Francisco, is yet another company tinkering with the makeup of existing antibody therapies to create new drugs. Scott Koenig, CEO of Macrogenics, says that the company is expecting to enter clinical trials soon with MGAH22, an Fc-optimized antibody that the company hopes will treat tumors that overexpress the human epidermal growth factor receptor 2 (HER2)—a hallmark of some breast, bladder, lung, and gastric cancers. MGAH22 is an engineered version of trastuzumab, an approved antibody treatment currently in use for breast cancers, which typically overexpress HER2 at high levels. Macrogenics optimized the trastuzumab’s Fc region to dramatically increase how strongly it binds to a variety of Fc receptors on effector cells in cancer patients. By making these adjustments to the protein, “you can dramatically improve the killing of tumor targets,” Koenig says.

Even academia appears to be getting in on the action. George Makhatadze and colleagues at the Rensselaer Polytechnic Institute in upstate New York recently detailed a targeted strategy to substantially increase the thermodynamic stability of nearly any protein, while preserving its unique function (PNAS, 106:2601–6, 2009). Their computational technique, which alters amino acid sequences by less than 5 percent, creates proteins that remain stable at temperatures 10°C higher than normal. Improving the stability of proteins has important ancillary effects, Makhatadze says. “By increasing stabilization, we are also offsetting aggregation in these proteins,” improving the antibodies’ effectiveness, he says. “We increase [the antibodies’] resistance to proteolysis at the same time.”

Pharma companies took notice of Makhatadze’s work, and he says he was close to a deal with an unnamed corporation when it fell through over “IP issues.”

Makhatadze adds that drug companies have paid too little attention to protein engineering in their rush to develop novel therapeutic proteins and get them through clinical trials. “Once you go through clinical trials, you’re stuck with whatever sequence you have,” he says. The mutation of even a single amino acid sequence requires a complete redo of clinical testing en route to a new FDA approval.

But with a new FDA approval comes a fresh patent, and the second time around “you already have the experience of that first molecule,” says Reichert. “That knowledge should help speed and reduce the costs of clinical trials. On the other hand,” she adds, “you already have a competitor on the market.”

Reichert adds that the burden of proof for antibodies that are derived from existing products through protein engineering can be higher. “This has to be a competitive calculation as to whether that would be sufficient to merit putting a molecule through this whole clinical studies process.”

But on the plus side, drug makers could conceivably engineer a suite of antibodies, all with different and useful therapeutic properties, that are all based on the same basic molecule. “You can now own a franchise,” Dahiyat says.