Evolution of a Company and a UR PhD Graduate Student

Career Story Blog Post By Ernest Smith, PhD, Chief Scientific Officer at Vaccinex

I entered graduate school in the summer of 1994, a couple weeks after graduating from St. John Fisher College with a Biology degree. I joined UR’s Department of Microbiology and Immunology, which afforded me a wide range of labs to work in. Over the following nine months, I did rotations in labs that focused on Immunology, Virology and Biochemistry. Even as a graduate student, I have always been more interested in developing new technologies than I have been in discovering new biology. I see discovery research as a form of detective work; you uncover the facts, follow the leads and present your case. You don’t change the facts; you just try to understand them. In contrast, I found inventing new technologies exciting because you are changing the reality and making something exist that didn’t exist before. Although building a knowledge base through discovery research and understanding biology well enough to intervene and treat diseases are both critical, I was most excited at the prospect of inventing technologies that are used to solve intractable biology questions.  

I picked the Immunology lab for my thesis project because I liked the person I would be working for and I found the project fascinating. The goal was to invent a new gene discovery technology that would enable the rapid screening of gene libraries in mammalian cells. The technology was based on the ability to invent methods to create gene libraries in a mammalian virus vector (vaccinia virus). The initial use of the technology would be used to identify antigens recognized by tumor-specific T cells. The ultimate goal was to have a powerful technology that would lead to the discovery of shared cancer antigens that would be incorporated into broadly effective cancer vaccines. And you need to remember, we started the project at a time before the genome had been sequenced and before microarrays had been developed. Gene discovery technologies were at a very early stage and primarily dependent on vary laborious and inefficient screening methods, particularly if the screening needed to be done in mammalian cells. Our new technology was designed to be a mammalian cell-based selection method, not a screening method, and this would bring increased efficiency and sensitivity to gene discovery compared to current methods. 

Fortunately, in just a few years, we invented a very powerful and elegant library-based gene discovery technology that drew universal praise when we presented it. We felt very strongly that this new library technology would enable ourselves and others to ask and answer questions that other technologies could not address. This was a very exciting time! The financial world was a very different place as well back then, with dot-com and biotech stock reaching all-time highs in the mid-late 1990s. My advisor, Maurice Zauderer, appreciated the commercial potential for this new technology, and with investors eagerly investing in biotech start-ups, he was able to raise the seed money to start a new company, Vaccinex, in October 1997. 

I defended shortly after the company was formed and became a manager for one of Vaccinex’s two research groups. I was of course fortunate that I did not have to do a post-doc before getting a “real” job. Because a lot of my work was the foundation of the company, I could not see passing up the opportunity! And if I had left, since I was the expert in this technology, I really think it would have hurt the company’s success. 

Our initial business plan was to use our technology to identify tumor antigens that could become the basis for broadly effective anticancer vaccines. Over the initial few years, little changed as we morphed from academic lab into a start-up. Our technology was very well received, and we were awarded a number of grants from the NIH and a very prestigious technology grant from NIST (National Institute of Standards and Technology) that allowed us to stretch our investors’ dollars.  At the time we started Vaccinex, very few people had heard of therapeutic antibodies and even fewer thought an antibody specific for a single target antigen could outperform a vaccine. But during the late 1990s, it became apparent that therapeutic antibodies were more clinically effective than therapeutic cancer vaccines, with several antibodies such as Rituxan, Remicade and Herceptin getting approved, for indications ranging from oncology to transplant rejection to rheumatoid arthritis, and a number of cancer vaccines failing in the clinic. Our technology was working well, we had identified a very interesting vaccine candidate, and Vaccinex had formed a collaboration with a true thought leader in the cancer vaccine field at NCI (National Cancer Institute). We were gearing up to take our proprietary vaccine into IND (Investigational New Drug)-directed preclinical studies and had flown down to NCI to discuss the study with our collaborator. I remember the meeting very well, how he explained that the vaccines were not working, and how “we are missing something” that was keeping the vaccines from being effective. This meeting made us re-think our strategy, and we elected not to move forward with this vaccine trial.

Fortunately, as the antibody field took off, we appreciated that our technology was versatile enough that we could make some modifications to it, to allow us to become a service provider by screening our libraries and selecting fully human antibodies for paying clients.  Early technologies for therapeutic antibodies relied upon the chimerization or humanization of mouse-derived antibodies. As the field developed, new technologies came into play, including strains of transgenic mice that expressed human antibody genes, and phage or yeast cells expressing human antibody binding fragments. The resulting antibodies were “fully human”, and predicted to be of lower immunogenicity than antibodies selected with the earlier technologies that retained some mouse sequences.  Companies developing these technologies were generating millions of dollars in annual revenue, and several were public companies with hefty valuations. Around this time, I was promoted to run the entire research department, first as Research Director, and then later as Chief Scientific Officer. We had been working for some time on these modifications and technology development. Specifically, our initial technology involved the creation of tumor-derived gene libraries in our proprietary vector, and then screening the library to identify the interesting antigens. In order to modify our technology to select antibodies, we switched and began making antibody gene libraries in our vaccinia vector. We designed our vector so that the antibodies would be expressed on the surface of mammalian cells (one antibody specificity per cell), and we could sort out the interesting antibodies using fluorescently labeled antigen and cell sorting. By 2003, we had something that was functional, and we began to start actively marketing our technology. I got to fly all over the world making technology presentations to potential customers. Over the next year or two, we did successfully complete a few fee-for-service partnerships, and one of these projects has advanced to clinical stage in India. The reality is that we were late to the game, and the competition was well established. Although our technology had advantages, it wasn’t quite distinct or strong enough to get a sufficient number of companies to switch to us. We got good value for the relationships that we did form, but it wasn’t enough to sustain us. 

As it turned out, investor expectations were also changing during this time. More and more, investors wanted to see small biotechs developing clinical products that had substantial sales potential, rather than working on service technologies (they wanted billions not millions in revenue). So we once again shifted our strategy and began focusing our technology and know-how on our own targets that would be turned into proprietary therapeutic antibodies to be developed as new medicines. This shift really allowed the company to expand the breadth of capabilities that we still have to this day. We were running as many as four discovery programs in parallel, which was very challenging. Each program came with the need to develop novel functional assays and animal models. Over time, we settled on our VX15 program as our lead.  The target for VX15 is named SEMA4D (Semaphorin-4D). This molecule has very interesting biology in both central nervous system disorders and oncology. We are the first company to develop an antibody against this target, or even this class of targets. This required an enormous effort on our part to develop functional assays, select the antibody for animal proof-of-concept studies, and choose the lead antibody for clinical development. We then had to test the antibody in numerous functional assays and animal models. During this time, we uncovered some very interesting new biology that has led us in some exciting directions. 

Once we had generated a sufficient body of data, we elected to move forward with our lead human antibody into clinical development. This step engendered a number of new changes in the company, including hiring quality assurance and regulatory people. We also hired a number of people with expertise in CMC (chemistry, manufacturing and controls) and toxicology, and having some of our existing people switch over and develop product release assays and bioanalytical assays. Fortunately, the manufacturing and toxicology work went as planned, and we received FDA clearance for our first clinical trial in early 2011. As you can imagine, it was really a sense of accomplishment to know that we made this drug and did all of the work that was necessary to get to that point. This drug has now completed Phase I trials in both oncology and Multiple Sclerosis, and phase 2 trials are imminent for both Huntington’s disease and oncology. 

One of the things that becomes apparent as “your” drug moves into clinical trials is that your ability to influence the outcome rapidly disappears. A clinical trial is really the ultimate discovery biology experiment. We think we made the right molecule and did the right experiments to show that there is relevant biological activity, but only time will tell. It really is out of our hands at this point. It will work or it won’t. A few years from now we will either look like pioneering geniuses or… something else. 

Clinical development is shockingly expensive. I joke that in research we estimate costs in thousands of dollars, while the development team estimates in millions of dollars. It is funny because it is true! Companies and investors take on these risks because the payoff for a successful drug is enormous and is truly transformative for a small company. The downside to this is that a failed study is generally catastrophic for a small company. As VX15 moved into development, and as more of our internal research capabilities became freed up as VX15-related tasks were completed, I started devoting more and more of my time and efforts to trying to think of additional ways that we could create value for Vaccinex (outside of VX15). Over the intervening 5 or 6 years since we began to focus on VX15, a lot had changed in the antibody-selection service space. A number of the well-known players in that space have been acquired by big pharmas, and a number of studies have pointed out the technological shortcomings of some of the other existing technologies, in particular phage display. The main technology competitor was/is a company named Adimab that utilizes yeast display and that had quickly grown into a major success, with millions in current and future revenue, lofty valuations, and major partnerships. We felt that opportunities in this space had actually opened up, and that if we could build something that was competitive, we could actually create a lot of value for Vaccinex. So we went “back to the beginning” with our technology and did an almost complete redesign.

One big advantage with our first generation technology was that we expressed full length IgG in mammalian cells. Because mammalian cells have a dedicated quality control machinery to remove mis-folded or poorly active proteins, using mammalian cells for selection gives the best possible chance that the selected antibodies will have acceptable biophysical properties as a therapeutic (expresses well in mammalian cells, low aggregation, low off-target cross-reactivity, stability, etc). Antibodies selected by Phage Display (E. Coli expressed) or standard Yeast Display (eukaryotic but not mammalian) have unpredictable properties because bacteria and yeast do not have mammalian cell quality control mechanisms. The problem that we had with our first generation technology was that the throughput of sorting mammalian cells was 10 – 100 times lower than what others could do with phage and yeast. What we selected was of good quality, but the throughput was lacking.  In order to be competitive, we had to figure out a way to improve the throughput while maintaining mammalian cell quality control.  In our new technology, we have figured out a way to express full length IgG on the surface of the vaccinia virus itself. We can easily harvest the vaccinia-monoclonal antibody virions from the mammalian cells and rapidly select out specific binders using antigen coated magnetic beads. We can rapidly screen billions of combinations (equivalent to other technologies), and all of the antibodies that we select have passed mammalian cell quality control and have favorable downstream properties. I am really proud of what we have accomplished over the last couple years. We now have a technology that is highly competitive both in quality, throughput and timeline. Over the last year or so, we have begun to actively market this technology by giving conference presentations, webinars, and face-to-face meetings, and I am very happy to say that we have signed several important deals. Hopefully there are many more to come. 

This is a very exciting time for Vaccinex as we continue to aggressively develop our VX15 drug, while building a parallel service business that we hope will continue to bring in both short term and long term revenues. 

Tracey Baas | 7/16/2015