www.sciencemag.org/feature/data/genomes/237-4813-358.pd

Who Owns the Human Genome?
Questions are mounting about whether anyone can "own" the human genome--whether it
can be copyrighted or patented and what e2 ct that might have on a j deral collaboration
This is the first of two articles exploring
current developments in the initiative to map
and sequence the human genome. The second
will fieus on organization and funding.
H ARVARD biologist Walter Gilbert
did not attend a recent workshop
on mapping and sequencing the
human genome, but he was clearly on everyone's
mind. As the Department of Energy
(DOE) and the National Institutes of
Health (NIH) continue jockeying over
which agency should lead the federal effort
and how it should be structured, a new set
of questions has emerged. What will be the
effect of this proposed project--the biggest
yet undertaken in biology--on open scientific
communication? Will researchers hold
close their results because the stakes--both
financial and professional--are so high,
thereby slowing the search for medically
important genes?
Can anyone "own" the human genome? If
a company sequences a gene or chromosome,
does it have proprietary control? And
specifically, can Gilbert really copyright the
human DNA sequence, as he says he plans
to do with his new company, Genome Corporation.
"There is scientific apprehension that materials
won't be available, that researchers
will have to repeat work, and that the government
will have to keep fi.mding duplicative
work," said Robert Cook-Deegan, an
analyst at the Office of Technology Assessment
(OTA) who organized the workshop
as part of the project he is directing on the
human genome initiative.*
Some say the effects of the proposed project
are already being felt throughout the
genetics community. "Until now U.S. researchers
have been generous in exchanging
clones," said C. Thomas Caskey of Baylor
College. "Every molecular biologist in the
United States knows the term 'cloning by
phone.' But now I'm definitely detecting a
tightening of this attitude as I call my
friends."
* The workshop, "Issues of Collaboration for Human
Genome Projects," was held 26 June in Washington,
D.C.
The workshop, which was cosponsored
by OTA and the Howard Hughes Medical
Institute, focused on barriers to collaboration
in a large federal project. The Hughes
Institute, a nonprofit organization, is now
spending about $1.5 million on research
related to mapping and sequencing. In the
free-ranging discussion, which included key
researchers, ethicists, lawyers, and representatives
from industry and public and nonprofit
agencies, there were clearly more
questions than answers. But the general
consensus seemed to be that problems will
worsen unless mechanisms are set up in
advance to ensure the open exchange of
information and materials.
The questions raised at the workshop are
not particularly new, now that the majority
of the nation's leading molecular biologists
have corporate ties of some kind. Yet they
seem particularly worrisome in regard to
efforts to map and sequence the human
genome. As Leroy Hood, a Caltech biologist
who is one of the leaders in developing
automated technologies for mapping and
sequencing, told Science, "Some people are
willing to share information and some are
not. That hasn't changed over the past 10
years. What could affect it is if you can
copyright or patent sequence data."
Unlike the rest of biotechnology, in
which patenting engineered microbes and
animals is becoming commonplace, there
seems to be something inherendy different--
and emotionally charged--about anyone
laying claim to the human genome, or
even a chunk of it. "Being able to copyright
the sequence would make me very uncomfortable,"
said Frank Ruddle of Yale. And
Caskey asked if there is a precedent for
saying, "This information is so important
that it cannot be proprietary. This is the first
time we'll ever get this information on
man---can we make a special case?"
In addition to these issues surrounding
ownership of the genome, other things set
this project apart from the rest of biology
and fuel concerns about scientific exchange.
One is the vast clinical applications and
huge profits expected to emerge from such a
project, which may exacerbate tendencies to
withhold results. Mapping the genome will
enable investigators to pinpoint the exact
location of genes associated with the 3500
or so known inherited disorders and may
also provide insight into numerous diseases,
including cancer, diabetes, and heart disease,
in which genetic predisposition plays a role.
Working out the exact nucleotide sequence
of these genes and the regions that control
their expression may reveal techniques for
early diagnosis or perhaps treatment.
"Working out the sequence and mapping
genes to certain areas has a ~r~eal monetary
value," said Ruddle. "It can be sold. That's
all to the good, it gives people an incentive
to pursue it. But at the same time, not all
will be given access, some may be locked
OUt."
With a project of such immense potential,
George Cahill of the Hughes Institute
summed up succinctly, "we have to look at
the bucks to ethics ratio."
This undertaking also differs in intent
from most biological projects. "The goal of
this project is to create a national resource of
information on the human genome--avail-
358 SCIENCE, VOL. 2.,37
able to all," as Cook-Deegan describes.
"That implies a type of data sharing that
might be different from normal science. If
we have a concerted effort, then we need
people to talk to each other."
As yet there is no centralized, interagency
human genome project--DOE and NIH are
pursuing their own initiatives--but there is
general agreement on how to proceed. The
first stage would be to develop a physical
map of the genome--a set of overlapping
DNA fragments that span the entire genome--
and then to locate genes and markers
on it. (The latter process is often referred
to as developing a genetic map.) This would
be accompanied by a simultaneous effort to
develop technologies for rapid mapping,
cloning, and sequencing. The second stage,
which might follow in 5 years, would be to
work out the nucleotide sequence of regions
of interest, if not the entire genome. As
sequence data and materials accumulate,
they would be put into a repository where
they would be available to other researchers.
Charles DeLisi of DOE, who instigated
the entire effort a couple of years ago, uses
the analogy of an accelerator. The goal, he
says, is not to answer fundamental questions
but to develop a tool to make that work
possible. And if this massive and expensive
project is to be completed in a reasonable
time, new information and methods must be
rapidly disseminated among the numerous
collaborators.
"In the normal scientific mode the researcher
is under no obligation to send out
materials or information before he has published,"
Cook-Deegan says. "But in this
case, the agencies might want something
different, for investigators to be more
open."
DOE and NIH have begun talking about
how to set up a database and repository, but
numerous nitty-gritty questions concerning
access, intellectual property protection, and
how to ensure that collaborators enter their
data promptly remain to be addressed, he
says.
These questions will be central not just to
the human genome project but to the rest of
biology as well, said David T. Kingsbury of
the National Science Foundation. "More
large centers of data generation are beginning
to emerge, and they will generate more
data than they can interpret. In response, the
role of scientists will change: they will become
more interpreters of data. If those
centers we are starting don't put their data
on-line immediately, we are in trouble."
Throughout the OTA meeting the conversation
kept coming back to Gilbert and
his plans to copyright sequence data. There
was palpable unease, as well as considerable
uncertainty, about what he actually intends
2.4. JULY I987
to do. Much of the discussion centered on
whether he legally can claim copyright protection
for the sequence. Opinion varied,
even among the lawyers. Does deciphering
and then writing down the sequence meet
the test of originality necessary for a copyright?
"One view, which is not widely shared, is
that you can copyright sequences," said Susan
Rosenfeld of the science and law committee
of the Association of the Bar of the
City of New York. "This view holds that
DNA is like a computer program, so it can
be copyrighted." Rosenfeld challenges this
view, however, and says that most other
attorneys do as well.
Gilbert told Science that he can copyright
the sequence because "someone worked it
out and wrote it down--so the order of the
letters is copyrightable, like a string of letters
in a book." A number of other participants
ceded him the point. As Rachel Levinson of
the NIH director's office noted after the
meeting, "If people didn't take Gilbert seriously,
they wouldn't be worrying about it."
Cahill, for one, believes that Gilbert can
copyright the format in which the sequence
appears, if not the anatomy--the sequence
itself. It's not the copyright per se that
disturbs him, he told Science, but what it
means for rapid exchange of information. "If
Gilbert's data would be of value only before
the sequence is out as public information, I
don't see how he can make any money
unless he sits on it, in which case he will
engender hostility. It goes against all tradition
in scientific philantrophy."
Gilbert does not see what all the fuss is
about. "The idea of the company is to be a
service to the biotech and pharmaceutical
industries and to the research community
• . . to answer questions that biologists have
in doing research," he told Science.
His company, Genome Corporation,
"will create a catalog of all human genes,"
probably starting with DNA from a placenta.
The map and sequence data would be put
into a database, along with other useful
analysis, "where it would be made available
to everyone--for a price." He declined to
speculate on what the fee might be.
As he envisions it, researchers will log
onto the database and ask any question, such
as, where does this piece of DNA belong?
As Gilbert explains, "the company will say,
for a price, that the gene is on chromosome
21, 1,300,000 bases from the left . . . . A
user can call up any part of it and read it. Or
a pharmaceutical company might like a copy
of the whole sequence; we could license it."
He emphasizes that people would be free to
use this information however they choose--
except, of course, to reproduce it and sell it.
"You can buy a book but you can't sell it. It
is exactly that distinction."
What he is selling, Gilbert says, is ease of
access. "Ease of access creates value. It does
not have to be free to be of great use. It is
like making restriction enzymes. Everyone is
free to make their own, but they choose to
buy them because it is cheaper. Here, it will
be cheaper to ask the question than to work
out the entire sequence yourself. "
He concedes that once someone else sequences
the genome, the value of his database
might decline. "That's a business risk. A
competitor could move into the field. But
where is the weight of information? Whoever
starts first will end up by owning, by
having in his possession, the whole database.
Once someone has done it, it is in no
one else's interest to do it again. It would be
cheaper to pay for it."
All this depends, of course, on Gilbert
getting there first. He is still shy of the $10
million in venture capital he says he needs,
but he expects to be in business by midsummer.
And, with a "reasonably sized
company, about 200 people," he expects to
complete the sequence in about 10 years, after
spending the first couple of years on mapping
and developing new technologies in several
areas of activity, including sequencing.
He openly admits to being a "technologic
optimist." Most other researchers believe
that sequencing cannot be done quickly or
economically until various cloning and sequencing
technologies are automated,
which is often estimated as at least a 5-year
effort. To Gilbert, however, "it's not a question
of new technology development, it's
RESEARCH NEWS 359
technology application."
If Gilbert's plans were seen as the only
challenge to open exchange of information,
the issue might not be the subject of such
intense debate. But other questions are arising
because work that is key to mapping and
sequencing the genome is also fundamental
to developing commercial products. As a
result, academic researchers are competing
directly with corporate scientists, who by
necessity operate under different rules of
disclosure. This situation may be common
to scientists in other fields. However, to
many geneticists, who are now finding that
they are denied access to scientific data, the
situation is new and often extremely frustrating.
A relatively new type of marker--restriction
fragment length polymorphisms
(RFLPs)--is a case in point. RFLPs (pronounced
rif-lips), which detect natural genetic
heterogeneity among people, are an
invaluable tool in searching for disease-cansing
genes, and they are indispensable in
mapping the genome. RFLPs can also be
fashioned into prenatal screening tests (linkage
tests) to detect genetic disorders, which
• explains the commercial interest.
These markers indicate the approximate
chromosome location--a region of perhaps
5 million to 15 million base pairs--of an
unknown gene. From there, finding the
gene itself is no small task, but it is far easier
than searching for it throughout the entire 3
billion base pairs of the human genome.
Indeed, these markers made possible the
recent localization of genes associated with
Huntington's disease, cystic fibrosis, Alzheimer's
disease, Duchenne muscular dystrophy,
manic-depressive illness, and chronic
granulomatosis disease.
As might be expected, competition is stiff
to develop these markers. Much of this work
is being done by Collaborative Research
Inc. of Bedford, Massachusetts, and by Raymond
L. White and his colleagues at the
Hughes Institute at the University of Utah,
although a few other companies are also
gearing up.
Collaborative, whose declared goal is to
be the leader in the field diagnostic tests for
genetic disease and cancer, has spent $10
million to date on this work and has developed
500 to 600 markers, according to
Thomas O. Oesterling, the president. David
Baltimore of MIT is the chairman of Collaborative's
science advisory board and is on the
board of directors.
White's group also has some 600 markers:
about 400 RFLPs and just shy of 200
variable number tandem repeat (VNTR)
probes, a newer and, he believes, more
informative type of probe. Both markers
detect genetic heterogeneity: RFLPs by
360
finding a single point mutation--a substitution
in one base pair; VNTRs by finding
repeated DNA sequences.
White's probes, along with those of other
Hughes Institute researchers, are deposited
in the American Type Culture Collection,
where they are available to other researchers;
Collaborative's are not. Collaborative
does lend its probes to some 40 to 50
academic collaborators, says Oesterling, under
confidentiality agreements.
Linking a gene to a chromosome is a
valuable first step, but it is not suificient for
developing a patentable commercial product.
The problem is that the marker may be
relatively far from the gene and thus may
become separated from it during recombination.
To develop an accurate screening test,
closer markers are needed, preferably flanking
ones.
That's where tension arises. If an investigator
announces that he has linked a gene to
a specific chromosomal region, others can
pull out their markers for that chromosome
and use them to find more informative
markers. The competitive advantage to
keeping quiet is undeniable.
In the grand scheme of things, a delay of 6
months or so would have little effect on the
pace of scientific advance. But many find it
disturbing nonetheless because it butts up
against deeply held scientific norms. "What
if someone finds the locus of a disease gene,"
Cahill asked at the workshop, "and then
holds onto it so others can't do the work.
How long can something be kept secret if it
is for the general good? I don't know who is
sitting on what now." Caskey pointed to the
recent cluster of papers announcing the locus
of the Alzheimer's gene. "It is unlikely
that everyone discovered it at the same
time."
Geoffrey Karny, a lawyer with the Washington,
D.C., law firm Dickstein, Shapiro,
and Morin, challenged the notion that industty
will sit on something solely for proprietary
reasons. "Researchers sit on stuff all
the time until they are sure."
Similarly, George Gould, a patent attorney
with Hoffmann-La Roche, does not
anticipate problems with withholding. "The
history of the past 6 years is that biotech
companies have made information available
to the public very quickly--far more quickly
than patent attorneys would like." Gould
says that he and most patent attorneys
would like their scientists to delay publication
for 18 months so as to improve the
chances of getting a patent overseas. But
typically, he says, "people will publish early
to get scientific priority as well as patent
priority."
"A key question for industry," said Bernadette
Alford, a lawyer with Collaborative,
"is how to protect our investment and collaborate
with scientists. We need new mechanisms
to protect our rights or it will not
work." She says there are few obstacles to
sharing clearly patentable products. "But
other pieces of information, like the 300 to
400 probes we have developed, are hard to
share. We won't file patents on 300 to 400
probes. For 90% of the work we do, we
don't see how we can share it."
Collaborative was the first group to link
the cystic fibrosis gene to chromosome 7. At
the meeting Alford held out their handling
of this discovery as ideal. "We found it on
Monday and announced it on Wednesday."
They have subsequently developed a panel
of 12 markers, flanking the gene, that are
being used in diagnostic tests, available in
their laboratory for $1100.
However, a number of researchers told
Science that Collaborative held back their
results for several months until they could
find more informative markerS.
Alford denies the charge. "We had been
working on cystic fibrosis for months. But
once we could get positive scientific confirmation,
we immediately announced it." Immediate
announcement was possible, according
to Alford and Oesterling, because
they could envision a diagnostic test and
thus filed for a patent. As Oesterling later
explained, they filed a patent application for
"the probe and any other probe between
that one and the gene."
Their broad-ranging patent application,
like Gilbert's copyright claims, does little to
SCIENCE, VOL. 237
relieve the simmering tensions.
"Their patent application seems unreasonable
to me," says White, who points out that
Collaborative essentially patented his work
before he even did it. Since Collaborative's
discover),, he and others have found closer,
more informative markers. And the whole
question may be moot anyway, White says.
Robert Williamson of St. Mary's Hospital
Medical School in London recently reported
the identification of a candidate cystic fibrosis
gene. IfWilliamson is right, says White,
"it will completely blow Collaborative's test
out of the water. It will become a useless
panel of probes."
There is no knowing how the patent
office will view Collaborative's application.
The situation regarding patents for these
products and processes, as for the rest of
biotechnology, is hazy at best. Indeed, a
patent application of the process for making
RFLPs has been pending for 7 years. The
application, filed by Stanford University,
has been licensed to Collaborative.
Alford, however, is fairly confident of
their claim. "We found the linkage to chromosome
7, and others began to look there,"
she told Science. "But one would argue that
it is our teaching that led them there. We'll
wait to see what the patent office says. But
we believe that by identifying the locus with
the RFLP, We are telling the world where to
look for the cystic fibrosis gene."
Although these new arrangements pose
challenges to scientific collaboration, several
workshop participants were sympathetic to
the predicament of Collaborative Research
and other companies. "It's a Catch-22," says
Cook-Deegan. "If a company behaves in
what scientists believe is a socially responsible
manner, they can't make a p?ofit."
"collaborative Research envisions themselves
as world leaders in markers in the
human genome," Kingsbury told Science.
"They are worried about sending out probes
that could be commercialized by their competitors.
Frankly, I'm somewhat concerned
about this. But I understand their concerns;
probes are easy to duplicate. I'm not sure
what kind of safeguards we need."
Clearly, new mechanisms will be needed
as DOE, and perhaps NIH, step up their
mapping efforts. DOE is already supporting
mapping work on chromosomes 16, 19, and
21 at Lawrence Livermore National Laboratory
in California, Los Alamos National
Laboratory in New Mexico, and at Columbia
University. And for fiscal year 1988,
DOE has requested $12 million for the
genome project, up from about $5 million
for the previous year. David Smith, who
runs DOE's human genome project, declines
to comment on the budget request for
fiscal year 1989. But a recent report by an
24 JULY I987
advisory committee to the Office of Health
and Environmental Research recommended
that DOE's genome project receive $40
million for 1989, with stead), increases up to
$200 million a year by 1993.
Within a few months DOE will be issuing
requests for proposals for mapping and
technology development. And DOE would
like to use Collaborative's markers in developing
their map, Smith told Science. "There
is a tension between maintaining the rights
of people to realize that monetary value and
between the goal of getting information
into the public sector where it can do some
good. We are talking with Collaborative.
We are optimistic that we can maintain their
proprietary interests and have markers available
for mapping. It is not an insurmountable
problem."
"We have a lot of probes, and we would
like to participate," says Afford. However,
without safeguards of some kind, Collaborative
will not place its probes in a repository
where they can be used by investigators
working on the physical map, she says. In
her view the best initial approach would be a
repository that contained a complete list of
information on the known probes, what
they do, and who has them--but not the
probes themselves.
One mechanism discussed at the workshop
is a requirement, imposed on all researchers
participating in the federal project,
to make materials and information available
at the time of publication or perhaps with a
1-year grace period. A possible model is the
Centre d'Emde du Polymorphisme Humain
(CEPH), the French database. CEPH sends
out its materials with the stipulation that
users in turn send their data back to CEPH.
It also offers a degree of proprietary protection;
a researcher may request that his materials
not be made public for 1 year, though
they are available to other CEPH collaborators.
Other large federal collaborations may
also present use~l models for meeting the
dual objectives of ensuring information exchange
while guaranteeing some form of
proprietary protection. Two mentioned at
the meeting were the federal programs to
develop a malaria vaccine and AIDS drugs.
The key, the agency officials, lawyers, and
scientists at the workshop agreed, is for
agencies to set up in advance the requirements
for information exchange and then
condition grants on them. "Unless arrangements
are created in advance," Cook-Deegan
says, "some people will not collaborate."
• LESLIE ROBERTS
A Younger Universe Is
Seen in the Stars
A tricky observation of radioactive thorium in nearby stars
leads to a surprisingly young age fbr the galaxy and fbr the
universe; many astronomers are skeptical--but some are
enthusiastic
A T a time when most astronomers
would agree that the oldest stars in
our galaxy have been shining for
some ] 5 billion years, anyone claiming that
the universe itself is only 11 or 12 billion
years old is bound to raise a few eyebrows.
Yet Harvey R. Butcher, director of the
Kapteyn Astronomical Institute in Groningen
*, the Netherlands, is claiming just that--
and some researchers think he is right.
Butcher bases his conclusion on recent
spectroscopic observations of the radioactive
nuclide thorium-232, which he has
measured in the sun and in 20 nearby stars
that are like the sun. The 14-bi!lion-year
half-life of this isotope has long made it a
favorite of researchers trying to do radioactive
dating on cosmic time scales. Indeed,
the standard figure for the age of our own
solar system, 4.6 billion years, is in part
derived from the relative abundance of thorium-
232 and various uranium isotopes in
meteorites and in moon rocks.
In Butcher's case, however, he compared
thorium-232 with a different nuclide, neodymiurn-
142, which happens to have a spectral
line that originates in the same part of a
star's atmosphere as the thorium line. This
location makes it a particularly usefial calibration
standard, since there are no correc-
RESEARCH NEWS 36I