|
|
 |
DR. DAVID MYLES - Kosan Biosciences
September, 2004
The Research: Using natural products as starting points for drug design and discovery
Kosan Biosciences was formed in 1995 with the goal of manipulating natural biosynthetic processes to
produce analogs of biologically active natural products. In other words, they use natural products as
starting points for a complex drug discovery program. The company produces huge arrays of analogs
with the goal of identifying and developing new products with potential benefits such as improved drug
efficacy or reduced side effects relative to the original natural product. Though the company does not
focus exclusively on marine natural products, some of its most promising ongoing projects are based on
marine discoveries. The company's main focus is the class of compounds known as polyketides. David
Myles is the executive director of Kosan's Chemistry Group.
 - VIDEO CLIP 1: "Kosan Company Background"
 What's a Polyketide?
Many commercial drug products are polyketides, including the widely used antibiotic erythromycin, some
cancer and upper respiratory ailment treatments, and certain cholesterol lowering agents. Polyketides form the
building blocks of a wide array of natural products. The biosynthesis processes responsible for producing them are
very similar to those involved in the synthesis of fatty acids. These compounds contain large rings of carbon and
oxygen atoms and are derived from acetate or propionate. Bacteria produce polyketides primarily as chemical defense
mechanisms, and as a consequence they are almost universally biologically active. Polyketides are found in both
terrestrial and marine organisms, but the marine environment is an especially attractive place to look for them
because it's simultaneously biologically diverse and highly competitive and the competitive pressure forces
organisms to develop unique and powerful ways to defend themselves. This is particularly true of organisms that are
non-mobile like sponges, which have rich sources of protein but no means of defense other than tasting bad or being
poisonous to their potential predators.
- VIDEO CLIP 2: "Focus on Polyketide Biosynthesis"
Methods of Manipulation
Kosan's approach is to use known polyketide natural products as starting points in the production of large numbers of
analogs of that product in search of new compounds that out perform the original natural products.
Researchers begin by looking at the structure of a compound of interest and then working to determine if they can obtain at
least portions of the biosynthetic machinery that produces that compound. Then, at the genetic level, the company
manipulates that biosynthetic machinery to produce analogs of the initial natural product that had already shown some
promise. A variety of techniques are used to accomplish this.
The basis of much of Kosan's research is the creation of what Myles calls "Frankenstein genes." In this process, certain
genes responsible for the production of a given compound are isolated and combined with some of the genes in the
biosynthetic pathway of another compound and spliced together. Expression of this Frankenstein gene in a microbial host
produces a sort of hybrid natural product. "This, I think, is one of the more powerful things that we're doing here at
Kosan," says Myles, "we call it morphing."
The morphing process can be used to produce whole new compounds, or parts of them. In the
latter case, a portion of the biosynthetic pathway responsible for a particular natural product might be disabled, and
these parts--different related precursor products--fed into the pathway to see what novel products might arise. Though
research groups have yet to recreate the biosynthetic pathway of a marine natural product as a means of sustainable
production, many are working toward that goal, which has been accomplished in some cases with terrestrial natural products.
Working with biosynthetic steps over which they already have control insures that Kosan can produce the resulting products
relatively simply, which is a key concern. Very few commercially available natural products are produced completely
synthetically; most depend at least partially on biological production methods.
Ultimately, Kosan is able to produce a wide variety of natural products to work with and modify in the medicinal chemistry
laboratory. Those analogs that show the greatest potential are tested in in vitro and in vivo models, and, ideally,
developed as drugs. To determine which analogs to pursue as drugs, Kosan follows standard drug discovery practices, first
looking for potency in cytotoxicity assays, then through pharmacokinetic analyses to predict how a compound will behave in an
animal or human. Finally, they test a compound's ability to cure animals of cancer, and, if all goes well, begin conducting human clinical
trials.
Overcoming the Supply Challenge
One great challenge in working with analogs, as with research on natural products, is obtaining enough of a given compound
to study. "It's very often the case that in the early stages almost vanishingly small quantities of the desired
product are produced in the fermentation process," says Myles. To deal with this challenge, Kosan has developed strong
analytical capabilities that allow researchers there to detect products in fermentation systems down to picogram levels
and, perhaps more importantly, to confirm that the material of interest has in fact been isolated. The analytical chemists
who do this work then provide feedback to the fermentation group, allowing them to optimize the process to produce more
material. In this way the group has been able to go from very small quantities of material to gram quantities.
Transforming Discodermolide
Kosan has conducted substantial research on the production of analogs based on discodermolide, which is licensed to Novartis. However, due to concerns raised during the first phase of Noavtis's
discodermolide human trials, the future of this analog research at Kosan is currently unclear. Discodermolide was discovered by Harbor Branch
Oceanographic Institution Harbor Branch Oceanographic Institution and showed early promise as a
potential treatment against various forms of solid tumor cancers. Kosan became interested in discodermolide because discodermolide is a
polyketide and shares the same mechanism as KOS-862, Kosan's lead anti-cancer compound. In collaboration with the University of Pennsylvania,
Kosan has synthesized and tested over 100 discodermolide analogs. Discodermolide is an extremely complex polyketide, which makes its production
difficult, so one goal of the project was to produce analogs with simplified structures that retain discodermolide's biological activity.
Several of Kosan's discodermolide analogs are quite potent and simplified, relative to discodermolide, says Myles. "One of the really fortunate
things about working with compounds like polyketides that are already potent compounds is that if you make conservative modification you
generally end up with compounds that are also active," he says. In fact, more than 80% of the analogs showed activity levels acceptable for
pre-clinical compounds.
Calystatin and Laulimalide
Another marine project focuses on the compound calystatin, which has a terrestrial analog, leptomycin, which is a potential
cancer treatment that made it to human clinical trials and was later cancelled due to toxicity issues. These compounds work by
binding to the critical p53 protein in cancerous cells and preventing its movement out of the nucleus, leading to programmed
cell death. Kosan is working to produce analogs of calystatin that might show reduced toxicity compared to the original
product, but the project is in early stages and the key focus at this point is better understanding exactly how calystatin
works. To that end, Kosan researchers are also producing analogs of the product to use as probes with fluorescent tags so that
the activity of the products inside cells can be studied. "We're trying to understand the biology of calystatin--how it
functions,and how best to harness that function for potential uses in chemotherapy," says Myles. One key concern was whether
calystatin was in fact binding to other proteins besides the p53 in cancer cells, causing toxicity in some way. Experiments
so far have suggested this is not the case, meaning that the toxicity is likely rooted in the compound's interaction with p53.
Laulimalide is a potential anti-cancer compound discovered by Phil Crews at the University of California, Santa Cruz, with
whom Kosan collaborates . Laulimalide, also known as Fiji analide, works in a similar fashion to discodermolide,
interrupting the process of mitosis and causing cell death. The current challenge with laulimalide is that it is unstable,
which leads to a side reaction that renders the compound inactive. Besides identifying the genes that produce the compound to
address standard supply issues, Kosan's goal is to produce and identify analogs that retain the anti-cancer properties but are
more chemically stable. Kosan's work on this project is in very early stages.
- VIDEO CLIP 3: "Compound Evaluation"
- VIDEO CLIP 4: "Laulimalides"
Natural Appreciation
"The impact of natural products and, as a subset of that, marine natural products, on drug discovery cannot be
underestimated," says Myles, "It's been huge." Nonetheless, he says commercial biotech companies' views on natural products
have oscillated
considerably in recent years. Years ago, natural products were the primary starting point in drug discovery, and they had a
great impact on the field. Erythromycin and other antibiotics to fight infections, as well as numerous other drugs aimed at a
variety of other conditions, were discovered and commercialized as a result. In time, though, total chemical synthesis of
compounds, or combinatorial chemistry, began to gain favor. Myles himself worked in this field before arriving at Kosan.
Combinatorial chemistry allows creation of matrices of analogs by filling cells in a grid and allowing precursors of potential
drug compounds to combine in a systematic-but not specifically directed-way. "It's a great way to make diverse structures,"
says Myles, but those structures fall into a fairly narrowly defined region of physical property space. Natural products lie
outside that space, says Myles. "These are compounds that are wholly unlike materials that can be obtained from combinatorial
chemistry," he says, "I think it's pretty clear that time has shown that combinatorial chemistry isn't the answer to all
questions; that natural products and the diversity available in natural products is a terrific way to start off a drug
discovery effort." As some of the marine natural products currently in clinical trials cross the commercialization threshold,
the importance of natural products will only become clearer.
Many of the projects Kosan pursues involve collaborations with academic researchers. Myles says the goal for such projects is
to let the science drive the process. "We clearly are coming to these collaborations with different agendas and different
objectives, but if we let the science drive the process we can generally reach our separate objectives together," says Myles.
Kosan supports its academic partners financially when appropriate, and shares as much information with them as possible. One
of the company's greatest sharable assets, according to Myles, is the ability to evaluate analogs very quickly. So, if a
collaborator is making analogs, Kosan can evaluate them and identify compounds of interest much more quickly than most
academic researchers would be able to on their own.
- VIDEO CLIP 5: "Kosan Collaborations: Letting Science Drive the Process"
Educational Background: An Interesting Odyssey
"I was one of those undergraduates who couldn't make up his mind," says Myles. He knew he wanted to do something in science,
but wasn't sure if it should be Earth science, physical science, or mathematics. He even entertained the notion of getting a
degree in history as an undergraduate. By his third undergrad year, at Occidental College in Los Angeles however, he settled
on chemistry, though somewhat by accident. He had taken an advanced language course in high school and exempted out of a
language requirement, creating a gap in his schedule. So, he went to his chemistry professor's office and said he'd like to do
some research. "This fellow took pity on me, I suppose, and said 'OK, you can join the lab,' and that was the beginning of a
really interesting odyssey. Myles was attracted to organic synthesis and ultimately decided to go to graduate school in
chemistry at Yale University in New Haven, Conn. While there he even worked on polyketides, as he does now at Kosan. After spending four-and-a-half years
at Yale, Myles was looking to expand his focus with some biology as a
postdoc, and ended up doing work in biocatalysis, or studying the use of enzymes as catalysts for organic compound
transformations.
"What we were really trying to do there was kind of coerce a biocatalyst to accept a substrate that wasn't
its native substrate, and process that substrate to produce a useful raw material for a chemical process," says Myles. It was
during his postdoc years that he also discovered the first paper on discodermolide, which he works on still. His first faculty
position was at Univ. of California, Los Angeles, where he studied discodermolide and a number of other natural products,
including amphidinolide B, which is a marine natural product from a dinoflagellate. While at UCLA Myles developed an interest
in medicinal chemistry and took a leave of absence to pursue research in that area with the Chiron Corporation near San
Francisco, where he ended up doing combinatorial chemistry. His experiences there and with polyketides led him to his current
position as executive director of the Kosan Chemistry Group.
 -
Print-Friendly Page Format
|
 |
