Clinical Micro
Sensors: Star Trek Meets the Human Genome
By
Michael Rogers
Clinical Micro Sensors founders are,
from left to right, Jon Faiz Kayyem, Tom Meade, and Scott Fraser.
Not too long ago,
in the once hyperbolic world of Internet start-ups, it was a common occurrence
for thirty-something entrepreneurs to make their fortunes almost overnight.
But the riches have always been more elusive in biotechnology, where the
development of a proven drug or healthcare product can take several years,
which made the market for biotech initial public offerings less explosive
than the dotcom IPO market.
But for Jon Faiz Kayyem,
PhD ’92, the money came surprisingly fast, as Clinical Micro Sensors,
the biotech company he cofounded in his garage six years ago, was bought
by Motorola last year for approximately $300 million. That deal made Kayyem
an instant multimillionaire at the age of 36.
While the buyout may
have been big, the primary goal of Clinical Micro Sensors is equally grandiose:
to change the way medicine is practiced. The company hopes to do this
with a hand-held device that will instantly analyze the DNA in a sample
of a patient’s blood, searching for a wide range of diseases. Kayyem
says that when people hear about his DNA detector, they have visions of
the palm-sized device that Star Trek’s Dr. McCoy would use to treat
patients with only a touch of a button. But Kayyem’s product is based
on nuts and bolts science and engineering and not fantasy. He hopes that
it will give medical labs a new way of doing business, since—by his
account—their expensive and time-consuming methods of analysis will
be obsolete once his product becomes commercially available.
Although Kayyem says
that he never would have guessed that the payoff for Clinical Micro Sensors
would have come so fast, getting there did not happen overnight. Kayyem,
who grew up in Los Angeles, came to Caltech in 1986 as a graduate student
in molecular biology, after getting his bachelor’s and master’s
degrees at Yale. He worked with Professor of Biology William Dreyer on
molecules that help the brain wire up during development and on their
relationship to molecules that identify foreign pathogens.
To conduct their research,
Kayyem and Dreyer had to develop new methods for detecting molecules.
Kayyem, who received his PhD in 1992, says that his work with Dreyer taught
him that the tools one develops to make scientific discoveries are just
as important as the discoveries themselves.
“When I finished working with Bill, I wanted to make new tools, as
opposed to being a leading scientific scholar in the field,” he says.
Kayyem adds that he was also influenced by Lee Hood, the former Caltech
biologist who invented a method for the automated sequencing of DNA, a
key to the success of the Human Genome Project. “It became really
clear that you could have a highly leveraged impact on a field if you
made a tool that everyone used versus making a discovery that may have
been an additional piece in a puzzle, but wasn’t really helping others
put the other pieces together,” says Kayyem. “I found that really
exciting. And Hood said to me, ‘If you want to do anything in this
field, just make sure that you’ve got the Human Genome Project in
mind.’
“A lot of people
in my class went to work on the discovery process of genes and their functions,
and I wanted to work on the testing side,” Kayyem says. “I thought,
‘If we discover all these important genes, then we can change the
way commerce works, we can change the way medicine works, and we can affect
environmental issues and food safety issues.’ But we couldn’t
do that without using really expensive tools and highly trained people.
So I thought I would try to bring DNA testing down to a level where technicians
could do it, perhaps even in the field.”
After getting his
PhD, Kayyem went to work as a postdoc with Scott Fraser, the Anna L. Rosen
Professor of Biology. Kayyem figured it would make sense to develop the
DNA testing tools in Fraser’s lab, since he thought that the tests
would likely involve fluorescent dyes and Fraser was an expert on fluorescence
systems.
Kayyem figured that
the easiest way to test for a specific DNA sequence would be for a doctor
to take a patient’s blood sample, stick it in a tube and watch the
tube change color if the patient had the sequence associated with a disease,
since fluorescent dyes change color when there is a specific molecule
in a solution. “I got it to work on proteins and then tried to get
it to work on DNA, and it didn’t work at all,” he said. “It
failed so utterly, that I didn’t know where to go.”
Fraser then sent him
to Tom Meade, a senior research associate in biology and a bioinorganic
chemist who, like Fraser, works in Caltech’s Beckman Institute. Meade
is an expert in electron transfer, an area in which Caltech has made numerous
discoveries. He had been investigating electron transfer through DNA since
late 1988. Meade told Kayyem that the transfer of electrons during DNA
binding events was likely preventing the action of the fluorescent dyes.
But rather than end the investigation, Meade saw this as an opportunity,
figuring that they could study the changing electrical signals as a way
to evaluate DNA. At about that time, researchers were putting DNA on silicon
chips, so Meade and Kayyem figured that they could take advantage of the
electrical properties of the chips to do electronic detection of DNA binding
events.
“I spent three
years working with Tom trying to make it work,” Kayyem recalls. “Tom
wanted the story bulletproof, because there were different theories on
DNA as to whether it would be a really bad conductor or an unbelievably
good conductor.” It turned out to be somewhere in between. Meade
and Kayyem discovered in 1993 that electrons can race from one end of
a DNA strand to the other as long as the two strands of the molecule are
bound together. When a single strand of DNA was used, the electrons didn’t
travel as fast, and this difference proved to be a key to the development
of a DNA sensor.
At the time that Meade
and Kayyem were working on electron transfer in DNA in the early to mid-1990’s,
other scientists were busy discovering the genes responsible for certain
diseases. Meade and Kayyem figured that one could identify a disease by
taking the section of the DNA unique to a particular disease, splitting
the DNA apart, and putting it on a chip. Then if you took a DNA sample
from a patient and it matched the DNA on the chip, the two strands would
bind together and an electron would speed down the double-stranded molecule.
If there was no match, the electrons would travel more slowly. A sensor
could distinguish between the two conditions and thus reveal whether a
patient was infected or not.
“Every time I
thought we had enough to write a paper, Tom would make us go back and
measure ita
different way,” Kayyem recalls. “We had to develop chemistry
and test the chemistry to validate that the molecules were stable. The
work was time consuming and meticulous and you couldn’t have impurities
in the system or the chemistry wouldn’t work.”
After getting it to
work in solution, Kayyem says that he decided that it was time “to
take a dive and see if we could swim across to a product: a chip.”
That would be more expensive and would require industrial production techniques,
so Kayyem told Meade that he wanted to start a company. Meade and Fraser
convinced Caltech to give Kayyem a one-year license to develop the technology
and raise enough money to turn it into a product. In exchange, Caltech
got royalties and equity in the company, called Clinical Micro Sensors.
Fraser and Meade joined Kayyem as co-founders. “This was 1995,”
Kayyem says. “I started working out of my garage. I wrote a business
plan and started looking for money.”
Meade played a significant
role in finding financing. After he gave a talk on the technology at a
conference, a reporter for a science journal wrote an article trumpeting
its commercial possibilities. An investor who read the article then contacted
Meade, offered to invest in the technology and helped line up other potential
investors. Within a year, Kayyem and Meade had raised $6 million.
Kayyem moved the company out of his garage and into a building in the
Old Town section of Pasadena. Rather than move to the Bay Area—the
location of choice for most biotech start-ups—Kayyem chose to remain
in Pasadena, in part, because he wanted to work outside the start-up spotlight.
“I wanted to
be a little bit stealthy about this,” Kayyem says. “I didn’t
want word of our incremental improvements going out into the world because
then expectations for us would be high. I wanted to show up one day with
a system that works. And that’s what we did.
“We asked investors
to give us 30 months, and told them that we’d build a handheld prototype
by then. We then thought we’d raise more money. People say that these
companies never turn out the way you lay them out, but by 30 months, we
were just about out of money, and we had a handheld prototype.”
The prototype includes
a sensor and biochips, on which probes of single strands of DNA are deposited.
When a sample of DNA is injected onto the chip, binding occurs with the
chip DNA if the DNA halves are complementary. The system also contains
DNA sequences, called signaling probes, with proprietary electronic labels
attached to them. When the DNA binds together, the electronic labels release
electrons, producing a signal that can be detected by the handheld sensor
when the chips are inserted in a slot in the sensor.
In 1998, Meade and
Kayyem showed off the prototype at a conference at MIT. They took prepared
samples of blood which were tainted with non-infectious fragments of either
HIV or hepatitis C, injected them into the DNA chip, and their handheld
sensor indicated the presence of the viruses. Also at the meeting were
executives from electronics companies. While viewing the other displays
and talking to officials of these firms, Kayyem says that he began thinking
that Clinical Micro Sensors was really part of the electronics industry
rather than the biotech industry.
“That meeting
made me think that life sciences might not be that far removed from electronics
if you can do life sciences electronically,” says Kayyem. “Our
product may not be a consumer electronics device, but it’s certainly
an electronic device for professionals.”
In 1999, when Kayyem
started thinking about moving the DNA sensor from the lab to the marketplace,
he turned to Motorola, an electronics’ industry giant, to forge a
partnership. Motorola first invested several million dollars in Clinical
Micro Sensors, but then decided to buy out the company. “The whole
information worlds and biotech worlds are on convergent paths,” says
George Turner, vice president and general manager of Motorola Life Sciences.
“Motorola thought that Clinical Micro Sensors had the most advanced
and well-thought-out plan to create a laboratory on a chip.” Turner
estimates that the size of the clinical evaluation market could be worth
$10 billion a year in sales when the sensors become common tools for medical
practitioners, which he says should happen within the next 10 years.
“Our plan is
to dominate the bio-chip sector,” says Kayyem. “Motorola is
making high-density -arrays of DNA chips that are used in the discovery
of new genes, and we’ve got the products that are fairly uniquely
positioned to compete on the diagnostic side. I don’t see any realistic
competition. Other biochip players use scanners and optical systems with
lasers that make the tests expensive. You can do an occasional cancer
test or an occasional HIV viral load test using fairly expensive equipment,
but you’re not going to alter widely the way medicine is practiced
with those, compared to our method, using very inexpensive scanners and
very inexpensive disposable chips. So I can comfortably say that we are
uniquely positioned to take advantage of those widespread applications
of genomics.”
Within five years,
Clinical Micro Sensors hopes to make its sensors available to doctors.
Kayyem figures that they will cost no more than $200 or $300, and the
company may even give them away, since it hopes to make most of its money
by selling the disposable chips that will be used to test for -diseases
and viruses. “If a doctor is testing for strep throat or influenza,
that’s a simple test, so maybe we’ll charge only $20 a test,”
he says. “For other diseases, the chips may cost more.”
Another application
for the sensor is helping doctors determine which drugs will work best
on specific patients. “If you have a pain in your stomach, your doctor—in
addition to doing an oral history and an external examination—is
going to test you to see if you have signs of any genetic disorders or
signs of any infection,” Kayyem says. “We can test to see if
you have a bacteria in you. The doctor can also test you to determine
what your likely response will be to one of the different medications
that he has. He’s got everything from antacids to Tagamet to Pepcid
to Zanax. People respond differently to those, so he’s going to do
a test with our chips that will tell him which drug you’re going
to respond best to, because it’s actually all in your genes.
“We’re starting
to be approached by research labs that are telling us that they have patients
with a particular genetic makeup, who do really well when they are put
on a particular Alzheimer’s medication. But others, with a different
genetic makeup, didn’t respond to it at all. So now, before they
deliver this medication, they will want to test people first. So that’s
exactly what we’re going to do. That’s a perfect test for us.”
Often, doctors prescribe
many medications based on trial and error, and that can be time consuming.
“It’s a huge cost to society and on health care for doctors
to get people on the right drugs,” Kayyem says. “For example,
in the case of depression, it turns out that certain doctors always give
Prozac first and then wait a few months. It works on some people but not
all. If not, the doctor may suspect that the patient is not taking the
medication regularly, since compliance is often bad. So the doctor has
to take forever before he can really figure out if the patient is responding
or not. Six months may go by. The patient may be the same or may have
an adverse response, so the doctor may put the patient on something else.
Some people do great on Prozac, but what if you could have saved the six
months? By the time of the last decision, the doctor may have sent the
patient to a psychiatrist. It can take a year before a patient is on the
right medications. There’s no reason for this.”
Kayyem says that the
sensors can also be used to test for pathogens in the environment, to
test for the presence of fungi and bacteria in crops, and for use in animal
health applications. And Meade says that the product is being considered
by NASA to test for water contamination aboard the space shuttle. Two
current projects are in animal husbandry, where companies want Clinical
Micro Sensors to test for a particular genetic code so that they can determine
which animals to breed.
“In livestock,
the code might signal a healthy cow or a fast rate of growth or a low
fat content of the meat,” Kayyem says. “Companies have identified
these codes and want ways to test for them. Current tests cost up to $1,000,
which may be more than what the cow is worth. These projects will give
us practice making chips for under $20 and will allow us to sell them
at the same time that we’re gaining experience learning how to adhere
to FDA guidelines for manufacturing. Then we will be well positioned to
address the -clinical markets.”
Over the next year,
Clinical Micro Sensors will be perfecting its chips and sensors so that
its detection system can eventually handle dozens of tests simultaneously.
“Once you’ve got a critical mass, it becomes more likely that
people will start wearing these things around on their belts,” Kayyem
says. “If a computer only runs one piece of software, you’re
not going to buy that computer. You want to buy a computer that has an
operating system that allows lots of different programs to run on it.
“We’re now changing from an R&D company to a manufacturing
company,” he adds. “We’re trying to establish our chip
technology as a platform in the industry by getting it in the hands of
users, so that by 2002 or so, we can have a number of tests and good manufacturing
practices for the FDA. And hopefully, all of these genes that every-one’s
been discovering will start to filter down into the healthcare community.
“We’re already seeing it in infectious disease detection where DNA-based testing has really replaced antibody-based testing. All these infectious disease organisms, like chlamydia and HIV and herpes and gonorrhea, have a DNA component. So I think that by 2002, doctors will call us saying, ‘I think I’m giving half my patients the wrong medication. I heard that there are genes that will tell me which drug I should be giving. Do you have a product that tests for that?’ Hopefully I’ll be able to say ‘Yes’ by then.”
Article Links:
RAINFINITY: From Outer Space to Interspace
 |
|||
 |
 |
 |
 |