Nature Biotechnology
Published online: 3 April 2007; | doi:10.1038/nbt0407-365
Pharma consolidates its grip on post-antibody landscape
The next wave of protein therapeutics, based on non-traditional scaffolds, may offer advantages in the range of conditions they can treat, delivery, and manufacturing. But they are just beginning to prove themselves in the clinic.
Cormac Sheridan
The contours of the post-antibody landscape are starting to emerge, as evidenced by a recent spate of deals involving large pharmaceutical firms and biotech companies developing the next wave of protein therapeutics based on nontraditional scaffolds. Although mostly preclinical or in the early stages of testing in humans, these next-generation compounds hold out the promise of being able to reach targets that are inaccessible using current antibody technologies. They also offer additional technical advantages in terms of both delivery routes and ease of manufacturing, as well as additional commercial advantages because of their freedom from third-party patent claims. That said, their developers will have to check off several technical and regulatory boxes before these technologies can realize their clinical and commercial potential.
Monoclonal antibodies (mAbs) and fragments thereof, mostly but not exclusively based on the IgG1 immunoglobulin structure, have been center stage in the biotech industry for more than a decade. The 21 such drugs that have so far gained US Food and Drug Administration approval are approaching annual sales of around $20 billion. Meanwhile, R&D in the field has shifted from murine to chimeric entities, to humanized, and to fully human antibodies. Antibody fragments, which are smaller and less complex than full-sized mAbs, are now an increasingly visible feature of late-stage drug development pipelines.S. San Francisco, California–based Genentech's successful introduction last year of the wet age-related macular degeneration treatment Lucentis (ranibizumab), a miniaturized version of the previously approved cancer drug Avastin (bevacizumab), exemplifies their promise. The product generated $380 million in sales in its first six months on the market.
Even pharma companies that have already acquired or licensed classic antibody platforms, or those that did not choose to participate in the first wave of antibody drug development, are now willing to look further back up the pipeline at these more exotic but less tested technologies. Last December, London-based GlaxoSmithKline paid $454 million cash for Cambridge, UK–based Domantis, a firm established in 2000 as Diversys by Ian Tomlinson and Greg Winter to commercialize their invention of domain antibodies, the smallest functional binding units of human antibodies (Nat. Biotechnol. 25, 153, 2007). Last October, Amgen, of Thousand Oaks, California, paid $290 million—and it could pay a further $90 million in milestones—to acquire Avidia, in Mountain View, California, which was established in 2003 by serial entrepreneur Pim Stemmer as a spinout from Redwood City, California's Maxygen. While at Maxygen, Stemmer invented avimers, a novel class of binding proteins based on A-domains, which are binding units comprising 35 or so amino acid residues that occur as repeating units in several cell surface receptors (Nat. Biotechnol. 23, 1556–1561, 2005). The first of these, an anti-inflammatory interleukin-6 inhibitor called C326, has already entered a phase 1 trial. New York-based Pfizer has also decided to acquire BioRexis Pharmaceuticals, of King of Prussia, Pennsylvania, for an undisclosed sum.
On the partnering front, Waltham, Massachusetts–based Adnexus Therapeutics recently inked a deal in oncology, potentially worth over $1.2 billion, with New York-based Bristol-Myers Squibb for the development of up to six of the former firm's adnectins, which are derived from the tenth domain of the human structural protein fibronectin. In January 2007 and November 2006, Ablynx signed target deals worth up to $265 million and $212.5 million, respectively, with Ingelheim, Germany-based Boehringer Ingelheim, and Madison, New Jersey–based Wyeth. The Ghent, Belgium–based firm is developing a new class of antibodies based on variable heavy chain antibody fragments derived from llamas. Last year, Wyeth and Seattle-based Trubion Pharmaceuticals signed a potential $800-million deal involving the latter's rheumatoid arthritis drug TRU-015, which has just completed dosing in a phase 2b clinical trial. The data, which will become available later this year, could provide an important fillip to other players that are at an earlier stage of development.
Venture investors are also noting these developments with interest. After Copenhagen-based Genmab, best known as a developer of monoclonal antibodies, lifted the lid on its UniBody technology at a company R&D day in London last October, CSO Jan van de Winkel found 12 different venture capitalists on his voice mail when he returned from the event to his base in Utrecht, in The Netherlands. "They wanted me to spin this out straight away," he says. A UniBody comprises a modified IgG4 immunoglobulin that lacks a hinge region and corresponds to half a wild-type antibody, as if the 'Y' shape traditionally used to depict an antibody had been bifurcated. The result is a stable, immunologically inert format that recognizes a single binding site.
The source material and the specific structural details of these new protein recognition molecules all differ (Table 1). However, they share (or aspire to) several key characteristics: high diversity; low molecular weight; high target selectivity and affinity; lack of immune effector functions, such as complement activation; and ease of expression in prokaryotic systems.
In a recent review of the field, Andreas Plückthun, of the University of Zurich, and coauthors identified over 40 nonantibody scaffolds in development in total, including molecules with potential applications as diagnostics or reagents (Nat. Biotechnol. 23, 1257–1268, 2005). Some companies, he suggests, are simply trying to avoid intellectual property issues associated with mAbs. But they'll need to show more. "I think that's too thin a proposition. I think you also have to have advantages in the molecule," he says.
Plückthun, a cofounder of the antibody development company MorphoSys of Martinsreid, Germany, has set up a new firm in Zurich, Molecular Partners, to take forward a series of scaffolds, and the company has signed research deals with Basel–based Roche and with Schering (now part of Berlin–based Bayer Schering Pharma). Its most advanced platform is based on designed ankryin repeat proteins (DARPins), which are derived from ankyrin repeat proteins, a family of membrane-associated proteins involved in attachment to the cytoskeleton. "This has really been an exercise in design from first principles," he says. DARPins offer many of the advantages Plückthun says these next-generation technologies will need to succeed. They are highly stable small molecules (with a molecular weight of 15–20 kDa) with picomolar binding affinities that have been produced in Escherichia coli expression systems with yields of over 10 grams per liter. "In molar terms this would be the equivalent of 30 grams per liter of an antibody fragment, which is just not achievable," says Plückthun.
"I think the table stakes in this area is creating high-affinity binders," adds Adnexus CEO John Mendlein. After that, "The technology battle is much more related to the mundane aspects of drug discovery," he explains, addressing issues such as solubility, aggregation, manufacturing yield, pharmacokinetics and immunogenicity—and ensuring that they apply across the entire class. "That's how you get a drug," he says.
On paper—and in in vitro experiments—many of these proteins appear to offer a superior proposition to antibodies, in certain settings at least. However, different molecules have different shortcomings that need to be overcome. Domain antibodies, for example, have stability issues, claims Genmab's van de Winkel. "All of them require modification to get molecules with a longer half-life," he says. Pegylation, for example, which is commonly used to increase stability, can cause additional immunogenicity, he says, and it also exposes developers to third-party royalties.
All of the protein classes in development will need to demonstrate stability and lack of immunogenicity in people, says Adnexus's Mendlein. Because its nanobodies are of nonhuman origin, Ablynx will be particularly focused on immunogenicity data from its first human trial, which was about to commence as Nature Biotechnology went to press. However, Ablynx's CEO Edwin Moses says that preclinical studies indicate that it should not be a problem.
So far, there's been minimal evidence of specialization among the competing nonantibody scaffolds. Most firms are focused on applications in cancer and immunology, and in some cases imaging applications in cancer. For many, the low-hanging fruit is represented by validated targets that are already the focus of existing mAbs therapies. "We can go into markets where there's antibody IP [intellectual property], and we can avoid it," Mendlein explains.
Business strategies will probably dictate the evolution of the field, rather than the specific attributes of the various protein classes under development, observes Plückthun, adding that any potential for specialization is not apparent from their formats. However, he says, the DARPins may have particular potential as targeted carriers of other therapeutic molecules. "It's there, in my opinion, where they will make the greatest impact because there is not stable technology [in the field of immuno conjugates] yet."
There appears to be agreement that no one company will dominate the post-antibody space. Right now, a negative result—in the form of a third-party clinical trial failure—is the biggest threat to the companies involved. "This would backfire on the whole field," cautions Patrick Amstutz, chief business officer at Molecular Partners.
Table 1 Selected companies developing protein therapeutics based on novel scaffolds
Company Molecule Scaffold, clinical status
Ablynx Nanobodies Llama heavy chain antibodies, phase 1
Adnexus Therapeutics Adnectins Fibronectin domains, phase 1
Affibody (Stockholm) Affibodies Protein binding domain of Staphylococcus aureus protein A, preclinical
Aptanomics (Lyon, France) Peptide aptamers Synthetic peptides, preclinical
Avidiaa Avimers A-domains derived from cell surface receptors, phase 1
BioRexis Pharmaceuticalsb Transbodies Transferrin, phase 1
Borean Pharma (Aarhus, Denmark) unnamed Trimerized tetranectin domains, preclinical
Domantisc Domain antibodies Heavy or light chain antibodies, preclinical
EvoGenix Therapeutics (Sydney, Australia) Evibodies Derived from V-like domains of T-cell receptors CTLA-4, CD28 and inducible T-cell costimulator, preclinical
ESBATech (Zurich) scFV fragments Stable single chain antibody fragments, preclinical
Genmab Unibodies Monovalent IgG4 mAbs fragments, preclinical
Micromet (Munich) BiTEs Bispecific, T-cell activating single-chain antibody fragments, phase 1
Molecular Partners DARPins Designed ankyrin repeat proteins, preclinical
Pieris (Freising-Weihenstephan, Germany) Anticalins Derived from lipocalins, preclinical
Scil Proteins (Halle, Germany) Affilins Derived from human lens protein gamma crystalline, preclinical
Trubion Pharmaceuticals SMIPs Custom-designed small modular immunopharmaceuticals, phase 2b
