Dartmouth Engineer - The Magazine of Thayer School of EngineeringDartmouth Engineer - The Magazine of Thayer School of Engineering

The Innovator

Biotech entrepreneur Tillman Gerngross reveals how he turns the impossible into multi-million-dollar companies.

By Karen Endicott

Chapter 1: “You Are Crazy!”

As the human genome project entered its final stages of mapping and sequencing every human gene, Tillman Gerngross was among the thousands of scientists worldwide anticipating a new era of protein-based therapeutics — drugs that could treat anemia, cure inflammatory diseases like lupus, or stop cancer in its deadly tracks.

Gerngross, a bioengineer and professor of engineering at Thayer, was also anticipating the next question: How are we going to make all those drugs?

The year was 2000, and he was already thinking about an answer.

In the pharmaceutical industry, the culmination of the human genome project represented a beginning rather than an end. “With the human genome you had a blueprint of every human protein,” says Gerngross. But turning that knowledge into drugs would be a major challenge.

“Protein-based drugs have to be made in living cells,” Gerngross explains. Complicating matters further, about 70 percent of human proteins are coated with sugar structures that affect their function. Called glycoproteins, these sugar-covered proteins present a particular challenge for drug-makers. “If you want to make the proteins for therapeutic purposes, you have to make them in a system that puts the human sugar structures on them,” Gerngross says. That process is called glycosylation.

The conventional method for making protein-based drugs used animal cells — a slow and expensive method prone to uncontrollable variations and inconsistencies. Scientists were eager to come up with alternative methods that would yield more consistent results and could be scaled up in large manufacturing plants to handle the number of potential drugs that the human genome project was expected to unlock.

Gerngross saw flaws in most of those alternative approaches. “People tried to solve the problem in a number of ways,” he says. “But if a hammer is all you have, everything starts looking like a nail. If you are a plant geneticist, you know how to put genes into plants. So people were putting human genes into corn and tobacco plants. I thought that approach was orthogonal to how the drug industry actually makes drugs: in a highly controlled environment, not a field. In addition, corn and tobacco were not able to put human sugars on the proteins they made. People were also putting human genes into goats, chickens, and cows to express the proteins in the milk, then purifying them from the milk. You can keep a goat farm clean, but that is very different from a manufacturing suite for a pharmaceutical drug. I thought those approaches were unlikely to succeed.”

Gerngross took a different tack. “If you could choose an organism that is very good at making proteins to start with and teach it how to put the human sugars on, you would have something much better.”

He had an organism in mind. “As a bioengineer I said, why don’t we take yeast, which is a very good organism to make proteins — it’s cheap, grows fast, and has very powerful genetic tools to manipulate it — and genetically engineer it in such a way that we can teach the yeast how to make proteins that have these human sugars on them.

“And most people said: ‘You are crazy.’ At a minimum you’d have to put in about a dozen genes and knock out a whole bunch of additional genes in the yeast. It would be a cell engineering project of a magnitude beyond anything that had ever been done.”

But Gerngross thrives on matching wits with nature. Growing up in Austria, he loved to garden. “As a teenager I had a huge vegetable and herb garden. I liked working with plants, setting up the whole thing and having control over where you plant your cucumbers, your tomatoes, and all that,” he says, with a chuckle. “There’s a lot of chemistry involved: the nitrogen cycle, issues related to chemistry. And I read a lot about science — biology, physics, chemistry.” He graduated from a science high school, and then headed to Paris to study French at the Sorbonne. But he soon returned to science, studying chemical engineering in Vienna. “Chemical engineering was generally viewed as a very hard subject matter, and I was attracted to the challenge,” he says. He earned a master’s degree in biochemical engineering from the Technical University of Vienna and then joined Arnold Demain’s microbiology lab at MIT while continuing to work on his Ph.D. in molecular biology. At the start of his professional career, Gerngross focused on making plastics from corn rather than fossil fuels, but concluded that the process would require too much energy and produce too much greenhouse-gas emissions. He published his analysis in Nature Biotechnology and Scientific American in 2000. “Both attracted significant media attention and demonstrated Gerngross’ willingness to ask hard questions even when they are unpopular, and more importantly, to deal with the consequences when the data do not support his assumptions,” reports Demain. Gerngross, who had begun teaching at Thayer School in 1998, says he did some scientific soul-searching. “The more I thought about what science is all about, particularly engineering, I saw it as a way of connecting science to human needs. The work you do has to benefit humanity in some way. My creativity ended up being more and more focused on how you make drugs, discover them, and ultimately on how you cure diseases.”

At the beginning of his work on yeast glycosylation, Gerngross took a traditional academic approach. “I went to the typical funding agencies — NIH, the Whitaker Foundation for Bioengineering, NSF — and they said no, this is not something that is really feasible,” he recalls. He understood their reluctance to fund him. After all, people had been working on pieces of the glycosylation puzzle for a decade without much success. “It was pointed out to me that I was neither a yeast geneticist nor a glycobiologist,” he says, “There was more than a healthy dose of skepticism.”

Being turned down “forced me to consider other options,” he says.

The path led to former Thayer School Dean Charles Hutchinson, who had hired him in 1998. An experienced entrepreneur, Hutchinson had faith in the man and the idea. “Tillman is very smart, focused, and energetic,” says Hutchinson. “The concept of making biologics in yeast made sense. Of course we’d also have to have clear milestones and hit them and be on budget.”

Hutchinson and Gerngross co-founded GlycoFi in 2000. (That’s Glyco for glycosylation and Fi for the high-quality fidelity of the product.) As CEO Hutchinson was responsible for raising funds and overseeing legal and administrative matters. As chief scientific officer Gerngross led the scientific team and, as Hutchinson says, “set the tone and vision” for the company.

“I became an entrepreneur by default,” says Gerngross.

Chapter 2: Like a Fox

Yeast cells.  Image courtesy of Adimab.
Yeast cells. Image courtesy of Adimab.

Gerngross taught himself what he needed to know. “The beauty of science is that you can read all the important papers and develop an understanding of what is really going on. It was months of reading to understand what people had done and speculating about why they had failed, then coming up with alternatives to overcome the deficiencies of their approaches.”

For example, he says, “the first enzyme that you need to make a human glycoprotein in yeast is an enzyme that takes off a form of sugar called mannose. In essence you have to ‘teach’ the yeast to carry out this reaction by introducing the right enzyme. All the prior work that had been done in Japan and Europe used an enzyme from a particular fungus that was engineered to go to a particular location in the yeast. They could prove the enzyme reached the right spot, but it had no effect on the removal of mannose. They argued that using more of the enzyme would solve the problem, but it still didn’t work.”

He discovered why. “I found an old paper from the ’70s where someone had taken that very same enzyme and described its activity when it is exposed to different pH environments and found that the enzyme is only active in very acidic conditions.”

No amount of the enzyme would make a difference if it wasn’t suited to the environment, he concluded. “It’s like sending guys in bathing suits to the North Pole to perform a task. It doesn’t matter how many you send — they’re not equipped to do the job in that environment.”

He outlined his strategy: “We need to find enzymes that have different pH optima and match up each enzyme with the environment we’re sending it to.”

He was right. “Sure enough, we tried different combinations and when we tried ones that have a better pH optimum, all of a sudden the reaction worked,” he says. “We got something to work that other people had literally been working on for 10 years and couldn’t get to work.”

GlycoFi ran through a vast array of permutations at each step to discover how to eliminate the sugar structures the yeast normally makes and engineer it to produce the kind of sugar structures humans make. “We developed the tools to repeat this proces over and over again to finally come up with a humanized yeast that makes fully human proteins,” says Gerngross. It had taken six years, but Gerngross and his team had done the impossible.

Gerngross credits GlycoFi’s achievement to the efforts of his scientific team. “Much of the success we’ve had is based on having been able to attract very strong talent early in their careers,” he says. “I’ve picked people based on raw talent. They may or may not have experience in this particular area, but it was clear to me that they stand out.” Then he lets people use their talents. It’s good for the individuals and for the enterprise, he believes. “Companies are a microsociety,” he says. “You have to articulate what your values are and you have to rally people around those values. And those values are: People will be treated fairly. People will be rewarded based on their contribution to the organization.”

Those aren’t just empty words. “Tillman is a true leader with self-reflection and a selfless commitment to the task at hand that people are eager to follow,” says Dr. Huijuan Li, a former post-doc in Gerngross’ lab at Thayer who worked with him at GlycoFi. “He is always willing to look to others for their opinion, open to change, and ready to go the extra mile to get the job done,” she says. “He helped me professionally to become more than I ever thought possible. He unlocked the potential of the scientists at GlycoFi to accomplish this scientific achievement.”

GlycoFi ended up being more valuable than either Gerngross or Hutchinson anticipated at the onset — because the company produced a more consistent product than conventional methods. “It turned out that when you made things in the conventional method in mammalian cells you got fairly heterogeneous mixtures of sugars on your protein. Yes, they were all sort of humanlike, but they were always different, and they all had different pharmaceutical properties. From a drug discovery perspective, it’s terrible,” Gerngross explains.

“Our engineering yeast gave us a level of control that wasn’t possible in the industry before. While GlycoFi was originally set up to solve a manufacturing problem, it ended up being a company that could make a drug more potent and more effective — and that is really what became the value, in addition to the manufacturing piece. If you can make a drug that is 100 times more potent, then you can dose it lower. That’s why Merck bought the company.”

Merck paid $400 million for GlycoFi in 2006, the third highest price ever paid for a private biotechnology firm, according to the National Venture Capital Association.

Chapter 3: The Labyrinth

Yeast cells. Image courtesy of Adimab.
Yeast cells. Image courtesy of Adimab.

After selling GlycoFi, the man who had become an entrepreneur by default became an entrepreneur by choice. He zeroed in on another fast-growing pharmaceutical area: discovering and optimizing human antibodies that could be used to develop new treatments for tough conditions such as infectious, inflammatory, and auto-immune diseases, central nervous system disorders, and cancer.

The problem, however, was not purely technical. He also needed to maneuver around the maze of patents in the antibody area that would trap the technology in legal corners and costly third-party payments.

“A patent allows you to exclude others. It doesn’t necessarily mean you can practice your invention, because there may be elements of what you do that actually are patented by someone else,” he explains. “So while I have a patent, I still may need rights to this and to this piece so I can make my piece work. The end user ends up having to pay me and this guy and this guy. That makes it very cumbersome and costly, if you have a technology and you have to go to five other people and ask for rights so you can actually use it.”

By then Gerngross was joined by his former student, Errik Anderson ’00 Tu’07, who was finishing up his M.B.A. at the Tuck School of Business at Dartmouth. The two met regularly and started reviewing the patent literature, digging through hundreds of patents in their quest to come up with a better antibody discovery platform. A pattern emerged as they kept on stumbling over the same name: MIT Professor Dane Wittrup, whose work on displaying antibody fragments on the surface of yeast was legend. So Gerngross called him. “We had independently converged on the conviction that a comprehensive antibody discovery capability was sorely needed in the pharmaceutical industry, so when he called me we were basically completing each others’ sentences,” says Wittrup. “We went from first meeting to incorporating Adimab in just a couple of months.”

Gerngross and Wittrup founded Adimab (it stands for Antibody Discovery, Maturation and Biomanufacturing) in 2007.

“In the GlycoFi case, the big hurdle was technical. In the Adimab case, the big hurdle was legal — coming up with a path that was legally unencumbered and then on top of that building something that is better than competing technologies,” says Gerngross.

He and Wittrup wanted Adimab to be a new platform for identifying and delivering antibodies for drug companies. “We want them to be able to walk in and say, ‘I want you to give me human antibodies against this cancer target,’ and literally eight weeks later they will have a hundred antibodies against that target. Previous technologies would take half a year and they would give you five antibodies,” says Gerngross.

While developing Adimab’s technology, “we looked at close to a thousand different patents,” he says. “We had five people working on this, with two legal consultants, for nine months, just understanding what other people had protected, what they didn’t have protected, what you could do, what you couldn’t do, and then based on those constraining factors, we came up with something very elegant that bypassed all that and was still better in the end. We seem to have come up with the optimal solution to our design problem.”

“What they have accomplished in this short period of time is remarkable, and anybody in the business of developing antibodies should pay close attention — these guys are rapidly changing the technological landscape of antibody discovery,” says Mike Ross ’71, managing general partner at SV Life Sciences, a venture capital firm that invested in Adimab and GlycoFi. Ross, a Thayer Overseer, is on Adimab’s Board of Directors.

“It’s staggering how much more effective this technology is, how short a time it takes, and how many drug leads it generates,” says Terry McGuire Th’82, co-founder of Polaris Venture Partners, an investor in Adimab and GlycoFi. McGuire chairs Thayer’s Board of Overseers and is on Adimab’s Board of Directors.

The industry is responding. Adimab has already landed antibody discovery deals with pharmaceutical giants Merck, Roche, and Pfizer and recently sold about 3 percent of the company to Google for more than $8 million.

“It’s really a game-changing way of doing this kind of drug discovery,” says Gerngross.

One thing that hasn’t changed is his management style. “People love working with Tillman because they know that he is supremely ethical and fair,” says Errik Anderson, who is now Adimab’s chief operating officer. “Every decision he makes is grounded in the facts of the situation. He doesn’t let his personal ambitions get in the way of doing the right thing, which is probably why he’s been so successful.”

Gerngross is too busy running Adimab — now up to 45 employees — to stop and reflect for long on his successes. He rarely takes time any more to go sailing or skiing. His one obvious indulgence is a machine that matches his own sense of precision performance — a silver Porsche that he drives between Adimab and Thayer School, where he continues to teach courses on biotechnology and biochemical engineering. The real rewards, he says, are in the work. “Making money becomes a minor issue over time,” he says. “Having changed lives in a meaningful way, giving people opportunities to do something big — those are the things I cherish much, much more than the economic impact of the companies I’ve started.”

In the end, Gerngross ties his entrepreneurial accomplishments to the reason he became an engineer. As he puts it, “Entrepreneurship is an extension of my academic career that allows me to take basic discovery into a realm where it actually impacts real people.”

— Karen Endicott is editor of Dartmouth Engineer.

Owner’s Manual

Entrepreneur Tillman Gerngross on founding a successful company

COMMAND POST: Tillman Gerngross directs a 45-member team at his Adimab lab in Lebanon, N.H. Photograph by John Sherman.
COMMAND POST: Tillman Gerngross directs a 45-member team at his Adimab lab in Lebanon, N.H. Photograph by John Sherman.

Establish clear milestones. “When you want to put together a complex scientific program or project with many moving pieces, you have to articulate a strategy of how you’re going to do that. You can’t just say we want this and hope for the best.”

Hire smart people. “Much of our success is based on attracting strong talent early in their careers. I’ve picked people based on raw talent. They may or may not have experience in this particular area, but it was very clear to me that they stand out. They have to be scientifically strong, but beyond that they also have to be able to work with other people, and recognize their own strengths and weaknesses.”

Articulate company values: “People will be treated fairly. People will be rewarded based on their contribution to the organization.”

Be clear about what you do and don’t know. “We didn’t know exactly how some things worked, but we knew what some of the influencing factors were. And that was enough. We could say let’s change all the parameters and see how that impacts the system.”

Analyze everything. “In the process of entrepreneurship — creating something of value — you cannot limit your analysis just to technical superiority or improvements. It has to hit all levels. It has to be technically better, you have to have freedom to operate, you have to have a legal path forward, and on top of that you have to be able to protect what you have. It has to be sufficiently novel that there are elements other people can’t reproduce. All these things have to come together to make something of utility and value.”

For more photos, visit our Research and Innovations and sets on Flickr.

Categories: Features

Tags: engineering in medicine, entrepreneurship, faculty, innovation, research

comments powered by Disqus