Skip to main content

Biological News

Nanobodies Get Their Due

Hot new technologies! We have waves of them in this business, and everyone talks about them, spends money looking at them, and does deals with small companies who are formed around them. But then reality sets in: only a few of these things march forward into the clinic, and even fewer emerge on the other side of the trials. Case in point: nanobodies. It seems like only yesterday that these were a big topic for people all across drug discovery, but the big startup in that field (Ablynx) was founded in 2001. And the first nanobody-based therapeutic (caplacizumab) was only approved earlier this year.

Those of you outside the field may be asking yourselves, so what’s a nanobody? Well, members of the camelid family (camels themselves, llamas, alpacas) went off in their own direction when it comes to immune response. Instead of In addition to (edited to fix my own misunderstanding – DBL) generating full-sized antibody proteins with several domains, they make trimmed-down versions with only one (specifically, only the “heavy chain” part). A full-sized monoclonal antibody will weigh up around 150 kiloDaltons, but these can be about one-tenth that size (if you just narrow down to the variable region), and that allows them the chance to do some things that regular antibodies have a hard time with. Epitopes that just aren’t normally accessible can be targeted, there are better possibilities for blood-brain barrier and cell-membrane crossing, the cost of manufacturing could be lower, you could administer them by different routes, and so on. (I should note that sharks also have their own weirdo antibody alternatives, which are being looked at as well).

But as this article at Nature Reviews Drug Discovery details, all these advantages were still only possibilities, and had to be proven in the real world. What really kick-started the nanobody craze was intellectual property – many people saw these new proteins as a way to get around the heavy-duty licensing fees that the big players in monoclonal antibody generation were charging, with a specific example being the Roche-Genentech “Cabilly” patents that only expired in 2017. It’s safe to say that Shmuel Cabilly had no idea of what he was really kicking off when he filed the original application (from City of Hope hospital) in 1983 – the legal saga that resulted took so many bizarre turns and huge reversals (all the way up to the Supreme Court) that it’s the exact intellectual property equivalent of a crazed long-running soap opera, the kind with evil twins, fake deaths, and it-was-all-a-dream sequences. The “Cabilly II” patent (US 6331415) is surely high on the list of the most fee-generating biotech patents – from both directions – in industry history; a lot of people took cracks at it in court (at ferocious expense), but it continued to bring in vast amounts of licensing revenue ($840 million in its last year).

Avoiding all that was a very appealing prospect, but realizing the potential of nanobodies was, as they say, nontrivial. They tend to have shorter half-lives than their full-sized cousins (some of which are spectacularly long-lived after dosing), and their smaller size has an inevitable trade-off in potency. In a head-to-head competition against a monoclonal, they’re probably going to lose, unless you’ve got some specialized edge working for you. Which is what happened for a while – companies went after well-validated targets (in order not to have that risk at the same time as a new technology risk), but that meant running right into the traditional antibodies. Which didn’t go well.

And that’s why we have caplacizumab, actually – it targets von Willebrand factor, which is involved in thrombotic thrombocytopenic purpurea (TPP). Going after that with a monoclonal, it was thought, would run the risk of aggregating the protein in yet another way and making the disease even worse, and in this case the short half-life of the nanobody isn’t a problem (TPP patients are getting frequent plasma exchanges anyway). So after some detours and dead ends, nanobodies might finally be arriving, twenty years after they were a big buzzing deal. The NRDD article says that there are about a dozen ongoing projects in the area (as opposed to hundreds and hundreds of antibody projects in development), but if they can hit in areas where antibodies can’t, then that’ll be worthwhile.

14 comments on “Nanobodies Get Their Due”

  1. Barry says:

    This isn’t Genentech’s first effort in this direction. A huge amount of protein engineering went into slashing 2/3 off the molecular weight of Avastin (a conventiona human antibody) to make Lucentis (a single-chain “antibody”) to the same epitope/target. Any pay-off in transport properties was too subtle to document. Antibodies–even slimmed-down antibodies–are exactly the sort of molecules that the phospholipid bilayer was invented to compartmentalize.

    1. antiaromatic says:

      Just a few points of clarification. Caplacizumab is not a Genentech molecule–it is out of Ablynx. Second, Lucentis is NOT a single-chain antibody, it’s a Fab. Third, Lucentis is NOT just avastin trimmed down to a Fab. For the particular application in the ocular space, being able to cram in more material that is functionally active (the Fc brings nothing to the party) gives you the ability to have longer-lasting efficacy. Also, several mutations had to be made to get Lucentis to be stable in the eye over the time period between doses, so there was more that had to be done to get to Lucentis than just trim down Avastin.

      1. steve says:

        The problem is that all the advantages you list are theoretical. A direct head to head by NIH comparing Avastin to Lucentis showed that clinically there is no difference.

  2. Dr. Manhattan says:

    ” Well, members of the camelid family (camels themselves, llamas, alpacas) went off in their own direction when it comes to immune response. Instead of generating full-sized antibody proteins with several domains, they make trimmed-down versions with only one (specifically, only the “heavy chain” part)”

    Actually, camelids produce both full sized antibodies and the much smaller ~15 Kd nano body version.

    1. Derek Lowe says:

      Fixing that right now – thanks!

      1. Some Dude says:

        To comment even more on that: the camelid single chain antibodies in vivo are still attached to an Fc, so they are much larger than 15kD in the animals. However there is no light chain, so one can just produce the variable domain (VHH) on its own and it is a well behaved recombinant protein – no light chain needed. The competing antibody-based molecule would be an scFv, a fusion of light and heavy variable domain with a linker – but that is double the size of the nanobody , and the linker can cause all sorts of problem.

        1. Isidore says:

          And even more: The Ablynx nanobody comprises not the entire heavy chain but only its variable region.

  3. BaddeleyRun says:

    Way back in December 2006 GSK, with great fanfare, bought a nanobody company called Domantis for about £230 million. The only problem was that the antibodies didn’t work. A PK/PD person quipped that they combine the worst features of small molecules with biologicals. Of course Tomlinson and Winter cried all the way to the bank and left GSK with yet another turkey! Or perhaps I’ve got this all wrong and Agilist would like to comment and defend the GSK “strategic vision”!

  4. Don Monroe says:

    I was expecting your post to be about the interesting work described by Carl Zimmer in the NYT last week, involving E. coli engineered to release nanobodies against CD47 upon quorum-induced lysis (ideally in tumors) The publication in Nature showed positive results IN MICE.

  5. Scientists are arrogant and should not have jobs says:

    I think people shouldn’t be assistant professors until they learned their lessons after an 8 year postdoc — 80% of current professors

    1. Da Vinci says:

      Nobody should be called a professor until they hold a chair. You’re either a professor or you’re not. There are not qualifiers. This is what Americans always get wrong. There is no such thing as an “assistant” professor. NOt everybody who teaches gets to call themselves a (something) professor

      1. Scott says:

        So, only the department head gets to be a professor?

        Because that’s the only ‘chair’ in an American university to get, and “Department Chair” looks better on your resume.

  6. john adams says:

    Aren’t nanobodies rapidly excreted via the kidney due to their low molecular weights?

    1. david borhani says:

      From the prescribing info:
      “The half-life of caplacizumab-yhdp is concentration and target-level dependent,” i.e. it binds tightly to von Willebrand Factor, and thus half-life depends on how much vWF a patient has.

      Nonetheless, drug appears to be cleared rapidly, because it is dosed daily. (Typical for a monoclonal antibody is biweekly).

      “CABLIVI should be administered upon initiation of plasma exchange therapy. The recommended dose of CABLIVI is as follows:
       First day of treatment: 11 mg bolus intravenous injection at least 15 minutes prior to plasma exchange followed by an 11 mg subcutaneous injection after completion of plasma exchange on day 1.
       Subsequent days of treatment during daily plasma exchange: 11 mg subcutaneous
      injection once daily following plasma exchange.
       Treatment after plasma exchange period: 11 mg subcutaneous injection once daily continuing for 30 days following the last daily plasma exchange. If after initial treatment course, sign(s) of persistent underlying disease such as suppressed ADAMTS13 activity levels remain present, treatment may be extended for a maximum of 28 days.”

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.