杏吧原创

IF YOU had a dollar for every time the word 鈥済enome鈥� has been used in news
items in the past year, you鈥檇 be sitting on a tidy pile. But although the age of
genomics appears to be with us, its benefits haven鈥檛 yet made it to the
pharmacists鈥� shelves.

So when researchers at Human Genome Sciences in Rockville, Maryland,
announced in February that the pharmaceuticals giant GlaxoSmithKline had begun
clinical testing a few months earlier of the first medicine to be developed
using genomic information, it made a big splash. 鈥淚t鈥檚 a very important event
for the industry,鈥� says William Haseltine, the chairman of HGS, whose enthusiasm
for the role of genomics in drug development makes him sound like a combination
of prophet and huckster.

Could this be the first drop in a torrent of genome-derived drugs that will
revolutionise medicine? Well, possibly. Genomics is a welcome tool in the drugs
companies鈥� kit, especially when it comes to tracking down prospective drug
targets. But there鈥檚 a lot more to developing a drug than finding a target for
it to hit. Genomics can do little to relieve the grindingly hard work of taking
a drug from the lab to the marketplace.

GlaxoSmithKline鈥檚 experience with the drug SB-435495, developed using genomic
information supplied by HGS, is a case in point. Since the 1980s scientists at
SmithKline (which merged with Glaxo Wellcome in 2000 to form GlaxoSmithKline)
had been interested in an enzyme called lipoprotein-associated phospholipase
A₂ (Lp-PLA₂), which they believed was involved in heart
disease.

Researchers suspected that as the enzyme travels through the blood in tandem
with low-density lipoprotein (LDL)鈥攖he so-called 鈥渂ad
cholesterol鈥濃�攊t triggers inflammation and the formation of the plaques
which clog coronary arteries. But to find out more, they needed relatively large
amounts of the enzyme, and the best way to get that is to find the gene that
makes it. Before genomics, that job could drag on for years with no promise of
success. By 1993, researchers at SmithKline had worked out a few fragments of
its amino acid sequence, but they still didn鈥檛 have the gene in hand. 鈥淎t that
point we would鈥檝e begun attempting to identify the full sequence using standard
methods,鈥� says Colin Macphee, the leader of SmithKline鈥檚 Lp-PLA2 team.

But they didn鈥檛 have to. That same year, SmithKline signed a collaborative
agreement with HGS, which already had an extensive genetic library that included
fragments of the vast majority of human genes. The SmithKline scientists simply
took their partial amino-acid sequence and compared it with all the fragments in
the library to find one with the corresponding DNA sequence. With the fragment
to hand, it took only a few weeks to identify and clone the entire gene. 鈥淚t was
very timely,鈥� admits Macphee. 鈥淚t certainly facilitated the drug discovery
辫谤辞肠别蝉蝉.鈥�

So far, so good. But the real work was just beginning. Though Macphee鈥檚 team
could now produce the enzyme at will, they still needed to verify that the
enzyme really did promote inflammation and plaque formation. Only then could
they begin searching for the best drug to block its effects. And once they had
found a candidate drug鈥� SB-435495鈥攖hey still had to test it in
rabbits to see if it would prevent plaque formation without producing
unacceptable side effects. These tests ate up the next seven years. It wasn鈥檛
until late last year that GlaxoSmithKline began testing SB-435495 on people.

The initial results look promising, but a rocky road still lies ahead. 鈥淯nder
the best of circumstances, it will be three or four years before they can market
their new drug,鈥� says David Flockhart, a pharmacologist at Georgetown University
in Washington DC. And the final years are the most perilous in drug development.
鈥淭he majority of new chemical entities fall apart in clinical trials because of
the problem of side effects,鈥� he says.

This lengthy birth is typical, says industry analyst Debra Yu, managing
director of Bay City Capital, a biotechnology venture-capital firm in San
Francisco. Though Yu鈥檚 job is to convince investors to buy into genomics and
biotech, she is pragmatic about the impact of genomics on drug development.
鈥淩emember that the gene is not the product,鈥� she says. A gene, or its protein
product, can only provide a potential target for a new drug. But there are many
hurdles after that, and genome databases cannot eliminate the hard work involved
in clearing them. 鈥淕enomics is a tool鈥攁 great tool鈥攖hat we use at
the beginning of the drug discovery process,鈥� agrees Holland. 鈥淏ut it doesn鈥檛
change what we do as a business. The end product is the same as it has always
产别别苍.鈥�

One exception may be the use of antibodies as drugs. These can be produced
quickly once the target is known, and require less testing since they are less
likely to cause side effects. These advantages have led HGS, for one, to pour
most of its resources into antibody and other protein drugs. Critics, however,
point to a major drawback: protein and antibody drugs must be injected, since
stomach enzymes would digest them if they were swallowed as a pill. Since most
people hate injections, that has kept most companies focused on the so-called
鈥渟mall-molecule鈥� drugs such as SB-435495, which don鈥檛 have to be injected.

The real value of genomics for drugs companies is probably indirect. Whole
genomes and the technologies that use them鈥攕uch as gene microarrays that
can show which genes are active at a particular time鈥攚ill give scientists
a deeper understanding of how the body works. That in turn seems likely to
suggest new ways for drugs to correct problems. 鈥淩ight now, we鈥檝e identified
maybe 400 therapeutically relevant targets,鈥� says Terry Allen, a pharmacologist
at the University of Alberta in Edmonton, 鈥渂ut with the Human Genome Project
we鈥檙e looking at 3000 to 10,000 new targets.鈥�

Those new targets will be explored not only by pharmaceuticals giants like
GlaxoSmithKline, but also by hundreds of smaller companies, says Yu. These
smaller, nimbler companies, with their more focused approach and less grandiose
expectations, may accomplish things the big players can鈥檛. 鈥淏ig pharma is in a
straitjacket because of their commitment to blockbuster drugs,鈥� says Flockhart,
鈥渂ut there are lots of good small companies out there that don鈥檛 need a
产濒辞肠办产耻蝉迟别谤.鈥�

Don鈥檛 look for the results any time soon, though. 鈥淣ew targets will lead to
new drugs, but these new targets will just be recognised over the next two or
three years, so it will take a while to produce a real flow of new drugs,鈥� says
David Burt, a pharmacologist at the University of Maryland in Baltimore.