Genome transforms, challenges drug discovery

The transformation of drug discovery as a consequence of mapping the human genome started before that project was finished so that genomic medicines are already on the market, but scientists and pharmaceutical companies face daunting challenges on the way to “individualized” health care, panelists said during a session Wednesday at the Technology Review’s Emerging Technologies Conference at the Massachusetts Institute of Technology (MIT).

Individualized or personalized medicine involves creating drugs that target specific populations of people who are shown through their genes to be susceptible to particular diseases. In the lab, it involves determining which genes cause the onset of diseases and figuring out what chemical compounds can be used to stop that process from starting or from spreading. That will mean that individuals will increasingly be tested to determine which genes they harbor that might eventually cause, for instance, breast cancer or Alzheimer’s disease, and then they can take preventative drugs to stop the diseases from taking hold in the case of cancer or that lessen symptoms in the case of Alzheimer’s.

“We’re moving from what I’d call serendipitous drug discovery to predictive, deliberate drug discovery,” said Joanna Batstone, senior manager of IBM Corp. Life Sciences and a participant in the panel discussion “Turning Genomics into Medicines,” which started with presentations from the panelists.

In the past, drugs were at times discovered in a haphazard, or even accidental, way, with pharmaceutical companies encouraging employees to take vacations in exotic locations and bring back dirt, fungus and other organic material that could be taken into the lab to see what compounds could be extracted, with tests then done on animals to see the effects of the potential drug, said, Gavin MacBeath, an assistant professor in the department of chemistry and chemical biology at Harvard University. Then along came the early 1960’s and the advent of biochemistry’s effect on pharmaceutical companies and the eventual emergence of protein targets that led to understanding how proteins work at the molecular level.

The next step is to systems biology, which applies system-level understanding to cells, the panelists said. One example of that approach is to determine the effect compounds have on thousands of spots of protein at a time, rather than focusing so much on individual proteins. Researchers can immobilize tens of thousands of proteins on a glass plate divided into tiny individual wells and then apply small molecules of compounds to the proteins and watch which proteins the molecules bind to. If a molecule binds to all or a lot of the proteins, then the scientists know that the molecule isn’t being specific enough in its target. Translated into drugs, that would mean the medicine might kill cells it isn’t meant to kill. Such research early in the drug-discovery process has the potential to limit failure further along in the research chain and save enormous amounts of money, as well as to make trials safer for participants.

Vast stores of data about compounds and their effects already exist, as does a huge lot of information about individuals and their medical histories, so another challenge the industry faces is how to mine and analyze that information.

Scientists and pharmaceutical companies have had “a long history of the intersection of biology and science, but what we need to incorporate now is computer science,” said Todd Golub, director of cancer genomics at the Whitehead Institute/MIT Center for Genomic Research. “We are wholly unprepared for what is about to happen with the onslaught of information.”

There are 30,000 “give or take” genes in the human genome and scientists are uncovering which components of the genome make a difference when it comes to particular diseases, he said. In one experiment, his center took bone marrow samples from children with the two types of childhood leukemia. Seventy-two samples yielded about 10,000 genes and researchers found they could correctly diagnose which type of leukemia the patients had just by looking at the genome.

But does accurate diagnosis help with therapeutics? Golub answered his own question by saying that of the thousands of genes examined in a microarray, one was found to activate a particular type of childhood leukemia and that gene, known as a kinase, is of the kind that MacBeath is experimenting on. The researchers injected mice with the gene and within four weeks they developed leukemia. Then the researchers injected mice and at the same time also gave them a daily oral dose of a compound already developed by Novartis AG, and discovered that the development of leukemia in the mice was “radically” reduced.

That experiment indicates the promise of using the genome to predict which particular treatments will work and also shows how safer, more efficient and less costly clinical trials could be developed by using genomics to create computational models to select trial participants more likely to benefit from the drugs being tested, he said.

That example also demonstrates how existing compounds can be used for other treatments, which is an emerging trend in the pharmaceutical industry.

One worrisome trend has been tighter venture capital, particularly for startups and companies in the early stages of drug discovery, but that is about to reverse, said Daphne Zohar, founder and chief executive officer of PureTech Ventures LLC, a Boston-based company that identifies scientific breakthroughs at research institutions and provides management, business support and funding to form companies around those breakthroughs.

Venture capitalists get a better return on early-stage investment compared to later-stage funding, she said. One issue, though, has been the relative youth of genomics coupled with a woeful economy, so venture capitalists have held back funding early-stage companies. In the near term, they will partner more with universities and others to fund new ventures, she said, predicting “early stage investment will come back.”