USING STEM CELLS TO IDENTIFY BETTER AND CHEAPER DRUGS - Bioscience Americas

USING STEM CELLS TO IDENTIFY BETTER AND CHEAPER DRUGS

John Farrell of Forbes Magazine recently interviewed Lee Rubin of the Harvard Stem Cell Institute.

Here’s Mr. Farrell’s interview in its entirety.

For an academic scientist, Lee Rubin at the Harvard Stem Cell Institute (HSCI) has a great deal of biotech industry experience, and this has inspired some ingenious experiments with induced pluripotent stem cells (iPSCs) for disease modeling and drug discovery.

I visited his lab recently to talk about his approach in more detail.
Rubin has a PhD in Neuroscience from Rockefeller University. He was a chief science officer for a company called Ontogeny, founded by current HSCI co-director Douglas Melton, where he helped develop a drug in partnership with Genentech.

The Erivedge capsule (Vismodegib) was approved by the FDA last year for the treatment of basal cell carcinoma.

But the particular research that led to this treatment … also steered Rubin toward stem cells.

“I got interested in stem cell biology –not because I’m a stem cell biologist, but because I’m a neurobiologist interested in disease. And this seemed practically like a good way to study disease.

“I found, however, when I was in industry, that working on the orphan diseases I was interested in, ALS, Spinal Muscular Atrophy (SMA) first–it was very difficult to do it in industry then, because people had not yet bought into the notion that it might be commercially viable to work on a so-called rare disease, although it’s not that rare.”

About 30,000 kids in the U.S. have SMA, according to Rubin. “And I thought, obviously, if we kept them alive longer you would increase the market size. Seventy percent of the kids die before they’re one or two, so you could dramatically increase the market size by treating the disease.”

This has indeed turned out to be the case, as now several different pharmaceutical companies are working on SMA, although they were not at the time Rubin came to HSCI.

“So I came here to pursue this interest in disease, which may sound funny in a way, but I thought it was more appropriate to do it in an academic setting.”

He was also determined to use a stem-cell based approach: By marrying what he did in industry with what he wanted to do in pure research, to pursue drug discovery and small molecules.

The result is what Rubin calls a sort of biotech-style screening lab. “I do a lot of collaborations with Doug Melton and many, many others here to work on, basically, how to make differentiated cells, reprogrammed cells, expanded cells, surviving cells, those kinds of things.”

Rubin works mostly with iPSCs, derived from real patients. If his lab can develop drugs that they can identify in a screen as candidates that might be useful, they can test them across ranges of patients –many individual patients.

“And that can pretty much only be done, practically speaking, using an iPSC approach,” he said.

“So, as an example, we discovered a compound, something we were excited about for ALS. Kevin Eggan (also at HSCI) and I collaborated to test that compound on 60 different motor neurons in parallel, using a very interesting technology we developed in my lab from live cell imaging which was required.”

The technology is very similar to the Nikon system I wrote about (sidebar) being installed for the Toronto based Center for the Commercialization of Regenerative Medicine.

Rubin’s is an older version of the BioStationg, customized for imaging cell studies in the very same way. “It’s a kind of standard issue Biostation that Nikon and DRVision have been modifying for our needs,” he said. And it allows him to screen multiple patient cell studies at once.

“More and more it turns out practically, as we try to develop drugs,” he said, “that having access to a wide variety of patient samples is very important for us.”

For example: if you had a drug for a disease like ALS or Parkinson’s or Alzheimer’s –where there are many causes for those diseases– then it’s likely that not every drug will work on a hundred percent of every case of the disease.

And further, for those diseases, there are many different mutations in the genes that can give rise to them, said Rubin.

“So, what if you had a drug that worked on ten percent of ALS patients,” he said. “How would you know which ten percent?”

You wouldn’t, at the moment. You would try to treat as many people as possible and hope that you could identify a subpopulation of people that would respond to your drug. That would be the standard approach in a clinical trial.

But it would also be an expensive one. And for rare diseases, like ALS, there aren’t that many patients available to rely on for testing.

“Now, what if you could take a hundred ALS patients, make their motor neurons in-vitro, test a potential drug on all one hundred, and see which patients cells respond the best?” said Rubin.

“And then treat those patients –only those patients.”

In this scenario, he explained, you have a drug that works on 10% of the cases. But then you would downsize the scale of the clinical trial ten-fold, if you directed the drug mainly at patients most likely to respond to it based on the iPSC patient studies.

“So this has, to me, the potential of having an effect, not only in modeling the disease better by doing all of your studies in diseased motor neurons, which was never possible before.”

In this scenario, he explained, you have a drug that works on 10% of the cases. But then you would downsize the scale of the clinical trial ten-fold, if you directed the drug mainly at patients most likely to respond to it based on the iPSC patient studies.

“So this has, to me, the potential of having an effect, not only in modeling the disease better by doing all of your studies in diseased motor neurons, which was never possible before.”

In addition, he said, you’ll have better drug candidates, not only because scientists can potentially test the safety of these drugs on human cells before they put them into people.

“But also, you can select subsets of patients that will respond the best in terms of efficacy to safety ratio to your particular drug.”

For Rubin, not only is this challenge scientifically interesting, if it’s right, it will enable drug companies to identify better drugs cheaper.

“So, a lot of us are really motivated to see if this idea is right.”

Share this post