Adimab, a Lebanon, N.H., biotech startup developing a “platform” for the discovery and commercialization of yeast-derived antibodies, raised $6 million in a first funding round (hat tips to the In Vivo Blog and VentureWire). The company was founded by Darthmouth’s Tillman Gerngross — who co-founded GlycoFi, a biotech acquired by Merck last year for $400 million — and MIT’s Dane Wittrup. Both researchers are chemical engineers with a longstanding interest in protein expression and engineering.
The concept behind Adimab is kind of intriguing, although it’s also complex and limited to solving a particular set of business-process issues — which, when you think about it, is just about what you’d expect a pair of engineers to come up with. The problem, at heart, is that monoclonal antibodies are a pain for many pharmaceutical companies to work with, due to the fact that discovering them and preparing them for use as drugs involves a variety of disparate technologies, many of them owned by a hodgepodge of other companies and institutions. Working out licensing agreements to acquire rights to all these technologies is possible, but still something of a headache.
As I understand it from an interview Gerngross gave to In Vivo, Adimab plans to address this problem by developing its own bottom-up system for discovering new antibodies and moving them along the development process. (The company’s name, which it prefers to capitalize as “ADiMaB,” is a mash-up of several of these development steps: Antibody Discovery, Maturation and Biomanufacturing.) Ideally, this yeast-based “platform” would yield antibodies that aren’t tied down by the web of intellectual property that covers many of today’s antibodies, making a potentially attractive fit for the first Big Pharma company that comes along. In fact, Gerngross seems explicitly mercenary about his intentions, telling In Vivo that “we’re building a business that will service pharma better than anyone else and [one] that could very quickly trigger an acquisition.”
Which is great so far as it goes, I suppose, although it’s hard to get too worked up about a new company when one of the co-founders seems to want it to disappear into some big-company bureaucracy as quickly as possible. The technology behind Adimab, however, is pretty interesting, involving as it seems to fairly recently developed techniques for forcing yeast cells to produce human antibodies (a 2006 Wittrup paper describing this technique of “yeast surface display” is here). Although since this also sounds a lot like what GlycoFi was doing, it will be interesting to see how Adimab’s approach differs.
In addition to the advantages described above, yeast-based production of human antibodies would presumably also be considerably faster and more efficient than current techniques, which generally involve mouse antibodies that require additional “humanizing” so they aren’t eliminated by the immune system when used as drugs. Of course, the technology is still at an early stage, and to the best of my knowledge, no yeast-derived antibody has even been tested in humans as an experimental therapeutic, much less turned into a functioning drug. (As always, if you know otherwise, please let us know in comments.)
Adimab’s first-round investors are SV Life Sciences and Polaris Venture Partners.
Tags: biotechnology, co:Adimab, deal, inv:Polaris-Venture-Partners, inv:SV-Life-Sciences, monoclonal-antibodies8 Comments
-
Roy said:
I think most antibodies are not glycosylated. Prof. Wittrup’s display technology is quite different than the secretion of correctly glycosylated erythropoietin which was the basis of GlycoFi. The two companies (GlycoFi and Adimab) should be pretty different beasts.
-
David P. Hamilton said:
Thanks for your comment. If glycosylation isn’t the issue for yeast-produced antibodies, do you have any idea what the Wittrup display technology is doing that couldn’t be done before?
-
Roy said:
I probably can’t say it better than the company web page: http://www.adimab.com/technology.html
Yeast aren’t “designed” to make human antibodies or glycoproteins and must be engineered. The beauty of yeast is that they can be genetically engineered quickly (unlike mice) and grow quickly. The GlycoFi strains were engineered to produce human glycosylation, the Adimab strains will be engineered and screened to produce certain antibodies. Probably without glycosylation at all (it’s pretty easy to get rid of glycosylation in yeast). If they copy the GlycoFi results, it would probably involve licensing the IP from Merck at this point. Sorry to be so long winded. Hope I’ve answered your question.
-
Roy said:
Here’s another company that apparently makes human antibodies in yeast. I don’t know how their method differs from Adimab’s though.
http://www.alderbio.com/default.asp?action=Technologies&pageName=AntibodyProduction
-
David P. Hamilton said:
OK, this has all been very enlightening on the subject of monoclonal-antibody production. My initial confusion resulted from not understanding which parts of traditional (i.e., hybridoma-based) Mab production are being superseded by yeast display, which I previously understood to be useful mainly in protein engineering/evolution, but not so much for protein discovery (which is what you need if you want to find novel Mabs, right?).
If I understand correctly — and this is a big if — the idea here is that yeast display can in fact also generate novel Mabs, presumably because the gene libraries that code for variable regions are understood well enough that it’s now possible to generate a broad range of epitopes through some form of mix-and-match procedure. That seems to be what Alder is describing here (thanks for the initial link to Alder, btw). I’d also go out on a limb here and guess that it’s possible to do this with purely human genes, so as to sidestep immunogenicity problems from the get-go.
The other question that came up as I looked more deeply into this is whether phage display is still a viable competitor to yeast display, or whether it’s been left in the dust for some reason.
Am I getting close yet?
-
Roy said:
My take is that hybridomas would be the best for making the right antibody since the cells are designed to do it but it’s a tedious process and covered by patents that become expensive to use.
I think you’re right about the yeast, using a broad library and probably some mutagenesis and selection, it would be possible to create the desired antibody. They’re probably using a human library and as long as there isn’t some yeast glycosylation, the immunogenicity is probably not an issue.
Sorry, I can’t help you about the phage display. I know a little bit about yeast, not so much about phage, or antibodies for that matter.
It’s all pretty exciting though, especially with the possibility of a nice buy-out.
-
Roy said:
Just got back from a meeting on advances in fermentation and there were quite a few talks on antibodies. Thought I would update with some information, rather than my previous speculation.
From talks by people at Genentech, GlycoFi and Alder (no talks from Adimab though); it seems that antibody glycosylation is not required for antigen binding or in vivo stability but for activation of downstream effects such as antibody-dependent cell-mediated cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC).
As far as phage display, Genentech uses it for antibody optimization.
Seems like GlycoFi was doing some antibody work, not sure how that plays into the Adimab work, since they didn’t present anything.
One of the speakers from GlycoFi mentioned that there was IP around in vitro glycosylation. Unfortunately he didn’t mention any specifics.
-
David P. Hamilton said:
Hey, thanks for the update. This has all been an interesting education in antibody production. Good to know that glycosylation is primarily important in complement and cell-mediated effects, since to my understanding, these are the primary routes by which many antibody drugs work in the first place. Rituxan, for instance, binds CD20, but the antibody’s constant region induces complement-related death of B cells. Even Avastin may have a complement or cell-related effect as well, in addtion to its VEGF blockage, since it seems to induce tumor shrinkage as well in some cases. This is one of the things that fascinates me about antibody drugs, and one reason I’m generally skeptical about attempts to mimic antibody function in other ways such as aptamers and other oddball molecular structures.
Thanks again for your input.
2 Trackbacks
3:14 pm
VentureBeat » Alder Biotherapeutics and its yeast-produced antibodies take in $40M said:
[...] in yeast. For an additional information on the merits of yeast-based antibody manufacture, see our previous coverage of Adimab, a startup developing its own yeast-production system for similar reasons. If [...]
3:17 pm
VentureBeat » Alder Bio and its yeast-produced antibodies take in $40M said:
[...] in yeast. For an additional information on the merits of yeast-based antibody manufacture, see our previous coverage of Adimab, a startup developing its own yeast-production system for similar reasons. If [...]