Short Subject: The (Protein) Future Of Therapeutic Agents With Dr. John Amatruda
Based on a conversation between KSS and John Amatruda, MD, PhD, in February 2018. Published today as a digression in pub thread. Amatruda is a director of $CWBR.
Let me begin with a digression, one derived from a conversation I had with Merck ($MRK) luminary John Amatruda, MD, PhD, in February at a CohBar ($CWBR) investor function in NYC. A drug is a substance you want to give the body to rejigger its behavior to fix a disease. And Amatruda was expostulating on the future of drug development. It’s worth reading up on Amatruda at this link; he’s a scientific advisor to BioPub fave DiaMedica ($DMCAF) and board member at $CWBR, potentially the most disruptive and transformative biotech company on earth. Amatruda is among the most sought-after people in biotech.
He proposed that you make in your mind a simple 2 x 2 table. Across the top, the X-axis, would be Size of Molecule (Large or Small). Now, on the vertical axis, would be a question: Has the body seen this molecule before? Yes No.
According to Amatruda, the future is in protein therapeutics, all large molecules in our definition. And this is because large molecules rarely incur any of the toxicity small molecules do because they are catabolized by different, non-hepatic pathways. Indeed, if it’s a large molecule of a type your body has seen before, Bob’s your uncle. You just are very unlikely to have issues with toxicity from it. And this is in part because large molecular size has a soothing, molling effect on immune function. The system sees a leviathan, declares it familiar, says move along people, nothing to see here (Officer Barbrady, “South Park”). Amazingly much the same thing applies with a large molecule the body has NEVER seen. It has a way of getting a diplomatic passport. The industry loves this because things just don’t get so bogged down in phase 1.
Meanwhile, small molecules are RISKY. If it’s a small molecule the body has seen before, it could lead to allergy especially via a hapten-mediated pathway and cause horror. It goes to liver for degradation and could injure liver or release toxic metabolites. It may clear renally and zap the kidneys in so doing….something that will never happen with a protein therapeutic. A small molecule the body has NEVER seen before is a potentially very destructive situation because one has no idea how the system will behave. Organ toxicity? Myelosuppresion? Liver damage? Chromosomal damage? Protein therapeutics just don’t DO these things. I’m in no way implying the small molecule development work of a company like Aptose ($APTO) is somehow “dated,” as it’s not, but it is higher stakes, and expect the industry to shine for protein drug makers…think lovely houses like Pieris ($PIRS, originator of anti-calins), even like protein therapeutics master Novo-Nordisk ($NVO). Think of Molecular Partners and the DARPins. Think of CohBar and the MDPs as therapeutics….though small, in our paradigm they would be regarded as “large molecules.”
If the future of therapeutics lies in proteins, there’s a positive upside to that for investors. Small-molecule drugs have to be synthesized by a series of organic chemistry reactions (aldol condensations, nucleophilic substitutions, and the like). Aptose CEO Rice has told me that the synthesis pathway for CG’806 is really surprisingly complex, that not so much so that he worries the company will cure CMC issues as it did with APTO-253. But small-molecule drug manufacturing offers more opportunities for goofs, more junctures at which to introduce errors. What few readers may realize is that when a manufacturing issue arises and corporate has to deal with manufacturing, the atmosphere is seldom one of openness and cooperation. The manufacturing facility and its workers circle the wagons, get stiff backs and become highly defensive. They can’t imagine, no, that they’ve done anything wrong. The interaction is so acrimonious and so often degenerates into name-calling that a CMC-driven clinical hold commonly results in the present manufacturer being fired and a new one contracted….after interviewing and site visits. It can be easier to start from scratch than to troubleshoot with chemical workers covering their tracks.
Protein therapeutics, meanwhile, are generated these days generally by cultured adherent cells, usually hardy Chinese hamster ovary cells, transfected to secrete the protein of interest. This is old tech, in fact, and increasingly we will see conversion of protein drug manufacturing into liquid-suspension yeast-based systems, where yields are higher, protein elaborates more rapidly, and purification steps are fewer. This won’t make protein drugs any cheaper; the cost of making a protein drug is at most one percent of the sales price. But it will cut costs and improve margins. With either yeast or adherent cell facilities, generally a big pharma company rents the facility for a block of time, often 2-4 months a year, and produces in mere weeks that entire year’s “run” of protein therapeutic (which is the generally lyophilized for long shelf life). According to DiaMedica CEO Rick Pauls, his company can make a dose of KLK1 (DM199) for 10 cents.
We are likely to see, in the near term, a sweeping change from adherent cells to yeast suspension methodology for making protein drugs, and one interesting company standing poised to enrich shareholders from that changes is Dyadics International ($DYAI). We heard a most impressive presentation by the company at NobleCon14. We can’t say the drug manufacturing space interests us personally for investing in, as it’s rather applied science, but this isn’t to say money can’t be made there. As Dyadics ramps up and companies see the merits of its C1 methodology, we could easily envision this company becoming lucrative during a 3-5 year window. While yeast don’t necessarily have much in common with mammalian cells like Chinese hamster ovary cells, both are eminently capable of cranking out proteins not native to them following stable transfection. The underlying proteins, having identical amino acid sequences, are not always totally identical because of the important events of post-translational modification, specifically glycosylation, the attachment of sugar-related moities on selected side chains of amino acids prior to secretion from the manufacturing cell. It’s hard to generalize about the differences, but they can confer subtle alterations in protein structure, activity and antigenicity. What’s important about the yeast lines used by Dyadics is that their genetics have been tweaked so that glycosylation is less ebullient than when the protein is made in adherent mammalian cells, and more predictable. Dyadics is not without competition, but I anticipate a gradual sea change of protein drug manufacturing over to yeast-based systems because costs are less and throughput considerably higher. Here’s a recent corporate presentation by Dyadics management.
When you contemplate a native protein, the primary sequence order of amino acids is the most important definer of that protein…conferring shape, size, Stokes radius, behavior. It’s all built into the amino acid sequence. You cannot patent the natural sequence, ergo you toy with it, and you call this “optimization.” Among the most interesting protein optimization efforts ever undertaken by pharma pertained to a now nearly forgotten molecule called consensus interferon-alfa. The researchers recognized that while yes there is a defined archetypal primary sequence that variations do occur in nature. Sometimes a mutation plonks an alanine where there was once a valine. So they took interferon-alfa molecules from a large large number of patients, and sequenced all of them. They statistically analyzed. At amino acid position one, what amino acid was most commonly seen? And at position two. And so on. They arrived at sequence that had net difference of 16 amino acids from the commercial interferons of Roche and Schering-Plough. What was REALLY interesting was then testing it preclinically for measures of interferon function, such as receptor binding and post binding events. With these assays they found that their “consensus” molecule had 10 times the potency of the two competing products, yet was formulated based on a sequence that never occurs in nature (it was a consensus sequence, a conflated sequence…but the final sequence used arrived at by polling does not match any naturally occurring version).
Ken Cundy, Ph.D., of CohBar did something vaguely similar with MOTS-C.He was aware of some natural sequence variants and how those variants affect activity. He wanted something not occurring in nature so he could patent it. He wanted something he could modify to extend its half life and chose a method similar in effect to PEGylation but NOT PEGylation (their therapeutic analog can be injected twice weekly). This work was prior to discovery of the receptor, but Ken did have a cell assay model for analog efficacy (I know what it is and it’s not non-public information but the company has asked me not to discuss any of this til their big upcoming announcement). So, driven by intuition, suspecting there IS a receptor, recognizing that certain amino acids and not others are likelier to govern ligand-receptor interaction, he DID play with the sequence in an insight-driven way, swap up a few amino acids, and did arrive a something more active than native MOTS-c. I believe the drug being advanced has the same fundamental 16 amino acid sequence but has had two of those amino acids swapped out for biological performance optimization. The modifications will be disclosed at a later date. The implications of MOTS-c’s effects on humans being receptor-mediated are quite staggering really: it implies a high degree of interactivity and cooperativity between the two life forms (bacteria and eukaryotic cells) that comprise us. It implies courtship, consent, commitment…and a vast level of interdependence. Cundy and the company find themselves in biological terra incognita, venturing further into the unknown every day of the week, and mapping and defining it.
Will Cundy become a Nobel laureate?
The analog first faces a typical SADMAD phase 1a…and I have to be blunt: its chances of “failing” that or incurring toxicity are essentially nil. SADMAD is “single ascending dose/multiple ascending doses.”
Disclosures: I have long positions in $APTO, $CWBR, $DMCAF, $MRK and $PIRS. I will not trade in those equities for 7 days, reckoned in business days, after this prose appears. I have received nothing of pecuniary value from these companies in exchange for coverage. I have no short positions or options. I am grateful to reader zimmyzee for posing the question that led to this answer.