SCIENCE, Page 62The Gene Hunt--COVER STORIESScientists launch a $3 billion project to map the chromosomesand decipher the complete instructions for making a human being
By LEON JAROFF
Know then thyself...the glory, jest, and riddle of the world.
--Alexander Pope
In an obscure corner of the National Institutes of
Health (NIH), molecular biologist Norton Zinder strode to a
30-ft.-long oval conference table, sat down and rapped his gavel
for order. A hush settled over the Human Genome Advisory Committee,
an unlikely assemblage of computer experts, biologists, ethicists,
industry scientists and engineers. "Today we begin," chairman
Zinder declared. "We are initiating an unending study of human
biology. Whatever it's going to be, it will be an adventure, a
priceless endeavor. And when it's done, someone else will sit down
and say, `It's time to begin.' "
With these words, spoken in January, Zinder formally launched
a monumental effort that could rival in scope both the Manhattan
Project, which created the A-bomb, and the Apollo moon-landing
program -- and may exceed them in importance. The goal: to map the
human genome and spell out for the world the entire message hidden
in its chemical code.
Genome? The word evokes a blank stare from most Americans,
whose taxes will largely support the project's estimated $3 billion
cost. Explains biochemist Robert Sinsheimer of the University of
California at Santa Barbara: "The human genome is the complete set
of instructions for making a human being." Those instructions are
tucked into the nucleus of each of the human body's 100 trillion
cells* and written in the language of deoxyribonucleic acid, the
fabled DNA molecule.
In the 35 years since James Watson and Francis Crick first
discerned the complex structure of DNA, scientists have managed to
decipher only a tiny fraction of the human genome. But they have
high hopes that with new, automated techniques and a huge
coordinated effort, the genome project can reach its goal in 15
years.
The achievement of that goal would launch a new era in
medicine. James Wyngaarden, director of the NIH, which will oversee
the project, predicts that it will make "major contributions to
understanding growth, development and human health, and open new
avenues for therapy." Full translation of the genetic message would
enable medical researchers to identify the causes of thousands of
still mysterious inherited disorders, both physical and behavioral.
With this insight, scientists could more accurately predict an
individual's vulnerability to such obviously genetic diseases as
cystic fibrosis and could eventually develop new drugs to treat or
even prevent them. The same would be true for more common disorders
like heart disease and cancer, which at the very least have large
genetic components. Better knowledge of the genome could speed
development of gene therapy -- the actual alteration of
instructions in the human genome to eliminate genetic defects.
The NIH and the Food and Drug Administration have already taken
a dramatic step toward gene therapy. In January they gave approval
to Dr. W. French Anderson and Dr. Steven Rosenberg, both at the
NIH, to transplant a bacterial gene into cancer patients. While
this gene is intended only to make it easier for doctors to monitor
an experimental cancer treatment and will not benefit the patients,
its successful implantation should help pave the way for actual
gene therapy.
The very thought of being able to read the entire genetic
message, and perhaps alter it, is alarming to those who fear the
knowledge could create many moral and ethical problems. Does
genetic testing constitute an invasion of privacy, for example, and
could it lead to more abortions and to discrimination against the
"genetically unfit"? Should someone destined to be stricken with
a deadly genetic disease be told about his fate, especially if no
cure is yet available? Does it demean humans to have the very
essence of their lives reduced to strings of letters in a computer
data bank? Should gene therapy be used only for treating disease,
or also for "improving" a person's genetic legacy?
Although scientists share many of these concerns, the concept
of deciphering the human genome sends most of them into paroxysms
of rapture. "It's the Holy Grail of biology," says Harvard
biologist and Nobel laureate Walter Gilbert. "This information will
usher in the Golden Age of molecular medicine," says Mark Pearson,
Du Pont's director of molecular biology. Predicts George Cahill,
a vice president at the Howard Hughes Medical Institute: "It's
going to tell us everything. Evolution, disease, everything will
be based on what's in that magnificent tape called DNA."
That kind of enthusiasm is infectious. In an era of budgetary
restraint, Washington has been unblinkingly generous toward the
genome project, especially since last April, when an array of
scientists testified on the subject at a congressional committee
hearing. There, Nobel laureate Watson of DNA fame, since picked by
the NIH to head the effort, mesmerized listeners with his plea for
support: "I see an extraordinary potential for human betterment
ahead of us. We can have at our disposal the ultimate tool for
understanding ourselves at the molecular level . . . The time to
act is now."
Congress rose to the challenge. It promptly allocated more than
$31 million for genome research to the NIH and to the Department
of Energy and the National Library of Medicine, which are also
involved in the quest. The combined appropriations rose to $53
million for fiscal 1989.
Even more will be needed when the effort is in full swing,
involving hundreds of scientists, dozens of Government, university
and private laboratories, and several computer and data centers.
With contributions from other Government agencies and private
organizations like the Hughes institute, the total annual cost of
the project will probably rise to $200 million, which over 15 years
will account for the $3 billion price tag.
The staggering expense and sheer size of the genome project
were what bothered scientists most when the idea was first broached
in 1985 by Sinsheimer, then chancellor of the University of
California at Santa Cruz. "I thought Bob Sinsheimer was crazy,"
recalls Leroy Hood, a biologist at the California Institute of
Technology. "It seemed to me to be a very big science project with
marginal value to the science community."
Nobel laureate David Baltimore, director of M.I.T.'s Whitehead
Institute, was one of the many who feared that such a megaproject
would have much the same impact on biology that the shuttle had on
the U.S. space program: soaking up so much money and talent that
smaller but vital projects would dry up. Others stressed that the
technology to do the job in a reasonable time was not available.
But by 1986 some opponents realized they were fighting a losing
battle. "The idea is gaining momentum. I shiver at the thought,"
said Baltimore then. Now, however, he approves of the way the
project has evolved and has thrown his weight behind it.
What really turned the tide was a February 1988 report by the
prestigious National Research Council enthusiastically endorsing
a project that would first map and interpret important regions of
the genome, then -- as better technology became available --
proceed to reading the entire genetic message. Most of the
remaining critics were silenced last fall when the NIH chose the
respected Watson as project director. Still, some scientists remain
wary of the project. Says David Botstein, a vice president at
Genentech and a member of the Human Genome Advisory Committee: "We
need to test its progress, regulate its growth and slap it down if
it becomes a monster. Jim Watson understands the dangers as well
as any of us."
The concern, as well as the cost, reflects the complexity of
the human genome and the magnitude of the effort required to
understand it. DNA is found in the human-cell nucleus in the form
of 46 separate threads, each coiled into a packet called a
chromosome. Unraveled and tied together, these threads would form
a fragile string more than 5 ft. long but only 50 trillionths of
an inch across.
And what a wondrous string it is. As Watson and Crick
discovered in 1953, DNA consists of a double helix, resembling a
twisted ladder with sidepieces made of sugar and phosphates and
closely spaced connecting rungs. Each rung is called a base pair
because it consists of a pair of complementary chemicals called
nitrogenous bases, attached end to end, either adenine (A) joined
to thymine (T) or cytosine (C) attached to guanine (G).
Fundamental to the genius of DNA is the fact that A and T are
mutually attractive, as are C and G. Consequently, when DNA
separates during cell division, coming apart at the middle of each
rung like a zipper opening, an exposed T half-rung on one side of
the ladder will always attract an A floating freely in the cell.
The corresponding A half-rung on the other section of the ladder
will attract a floating T, and so on, until two double helixes,
each identical to the original DNA molecule, are formed.
Even more remarkable, each of the four bases represents a
letter in the genetic code. The three-letter "words" they spell,
reading in sequence along either side of the ladder, are
instructions to the cell on how to assemble amino acids into the
proteins essential to the structure and life of its host. Each
complete DNA "sentence" is a gene, a discrete segment of the DNA
string responsible for ordering the production of a specific
protein.
Reading these genetic words and deciphering their meaning is
apparently a snap for the clever machinery of a cell. But for mere
scientists it is a formidable and time-consuming task. For
instance, a snippet of DNA might read ACGGTAGAT, a message that
researchers can decipher rather easily. It codes for a sequence of
three of the 20 varieties of amino acids that constitute the
building blocks of proteins. But the entire genome of even the
simplest organism dwarfs that snippet. The genetic blueprint of the
lowly E. coli bacterium, for one, is more than 4.5 million base
pairs long. For a microscopic yeast plant, the length is 15 million
units. And in a human being, the genetic message is some 3 billion
letters long.
Like cartographers mapping the ancient world, scientists over
the past three decades have been laboriously charting human DNA.
Of the estimated 100,000-odd genes that populate the genome, just
4,550 have been identified. And only 1,500 of those have been
roughly located on the various chromosomes. The message of the
genes has been equally difficult to come by. Most genes consist of
between 10,000 and 150,000 code letters, and only a few genes have
been completely deciphered. Long segments of the genome, like the
vast uncharted regions of early maps, remain terra incognita.
To complicate matters, between the segments of DNA that
represent genes are endless stretches of code letters that seem to
spell out only genetic gibberish. Geneticists once thought most of
the unintelligible stuff was "junk DNA" -- useless sequences of
code letters that accidentally developed during evolution and were
not discarded. That concept has changed. "My feeling is there's a
lot of very useful information buried in the sequence," says Nobel
laureate Paul Berg of Stanford University. "Some of it we will know
how to interpret; some we know is going to be gibberish."
In fact, some of the nongene regions on the genome have already
been identified as instructions necessary for DNA to replicate
itself during cell division. Their message is obviously detailed
and complex. Explains George Bell, head of genome studies at Los
Alamos National Laboratory: "It's as if you had a rope that was
maybe 2 in. in diameter and 32,000 miles long, all neatly arranged
inside a structure the size of a superdome. When the appropriate
signal comes, you have to unwind the rope, which consists of two
strands, and copy each strand so you end up with two new ropes that
again have to fold up. The machinery to do that cannot be trivial."
One of the most formidable tasks faced by geneticists is to
learn the nature of that machinery and other genetic instructions
buried in the lengthy, still undeciphered base sequences. To do so
fully requires achievement of the project's most challenging goal:
the "sequencing" of the entire human genome. In other words, the
identification and listing in order of all the genome's 3 billion
base pairs.
That effort, says Caltech research fellow Richard Wilson, "is
analogous to going around and shaking hands with everyone on
earth." The resulting string of code letters, according to the 1988
National Research Council report urging adoption of the genome
project, would fill a million-page book. Even then, much of the
message would be obscure. To decipher it, researchers would need
more powerful computer systems to roam the length of the genome,
seeking out meaningful patterns and relationships.
It was from the patterns and relationships of pea plants that
a concept of heredity first arose in the mind of Gregor Mendel, an
Austrian monk. In 1865, after studying the flower colors and other
characteristics of many generations of pea plants, Mendel
formulated the laws of heredity and suggested the existence of
packets of genetic information, which became known as genes. Soon
afterward, chromosomes were observed in the nuclei of dividing
cells, and scientists later discovered a chromosomal difference
between the sexes. One chromosome, which they named Y, was found
in human males' cells, together with another, called X. Females'
cells, on the other hand, had two copies of X.
But it was not until 1911 that a gene, only a theoretical
entity at the time, was correctly assigned to a particular
chromosome. After studying the pedigrees of several large families
with many color-blind members (males are primarily affected),
Columbia University scientist E.B. Wilson applied Mendelian logic
and proved that the trait was carried on the X chromosome. In the
same manner over the next few decades, several genes responsible
for such gender-linked diseases as hemophilia were assigned to the
X chromosome and a few others attributed to the Y.
Scientists remained uncertain about the exact number of human
chromosomes until 1956, when improved photomicrographs of dividing
cells clearly established that there were 46. This revelation led
directly to identification of the cause of Down syndrome (a single
extra copy of chromosome 21) and other disorders that result from
distinctly visible errors in the number or shape of certain
chromosomes.
But greater challenges lay ahead. How could a particular gene
be assigned to any of the nonsex chromosomes? Scientists cleverly
tackled that problem by fusing human cells with mouse cells, then
growing hybrid mouse-human cells in the laboratory. As the hybrid
cells divided again and again, they gradually shed their human
chromosomes until only one -- or simply a fragment of one -- was
left in the nucleus of each cell.
By identifying the kind of human protein each of these hybrid
cells produced, the researchers could deduce that the gene
responsible for that protein resided in the surviving chromosome.
Using this method, they assigned hundreds of genes to specific
chromosomes.
Finding the location of a gene on a chromosome is even more
complicated. But over the past several years, scientists have
managed to draw rough maps of all the chromosomes. They determine
the approximate site of the genes, including many associated with
hereditary diseases, by studying patterns of inheritance in
families and chopping up their DNA strands for analysis. With this
technique, they have tracked down the gene for cystic fibrosis in
the midsection of chromosome 7, the gene for a rare form of colon
cancer midway along the long arm of chromosome 5, and the one for
familial Alzheimer's disease on the long arm of chromosome 21.
One of the more dramatic hunts for a disease gene was led by
Nancy Wexler, a neuropsychologist at Columbia University and
president of the Hereditary Disease Foundation. Wexler was highly
motivated; her mother died of Huntington's disease, a debilitating
and painful disorder that usually strikes adults between the ages
of 35 and 45 and is invariably fatal. This meant that Wexler had
a 50% chance of inheriting the gene from her mother and contracting
the disease.
In a search coordinated by Wexler's foundation, geneticist
James Gusella of Massachusetts General Hospital discovered a
particular piece of DNA, called a genetic marker, that seemed to
be present in people suffering from Huntington's disease. His
evidence suggested that the marker must be near the Huntington's
disease gene on the same chromosome, but he needed a larger sample
to confirm his findings. This was provided by Wexler, who had
previously traveled to Venezuela to chart the family tree of a clan
of some 5,000 people, all of them descendants of a woman who died
of Huntington's disease a century ago. Working with DNA samples
from affected family members, Gusella and Wexler in 1983 concluded
that they had indeed found a Huntington's marker, which was located
near one end of chromosome 4.
That paved the way for a Huntington's gene test, which is now
available. The actual gene has not yet been isolated and since
there is no cure at present, many people at risk for Huntington's
are reluctant to take it. "Before the test," Wexler says, "you can
always say, `Well, it can't happen to me.' After the test, if it
is positive, you can't say that anymore." Has Wexler, 43, taken the
test? "People need to have some privacy," she answers.
Tracking down the location of a gene requires tedious analysis.
But it is sheer adventure when compared with the task of
determining the sequence of base pairs in a DNA chain. Small groups
of scientists, working literally by hand, have spent years simply
trying to sequence a single gene. This hands-on method of
sequencing costs as much as a dollar per base pair, and deciphering
the entire genome by this method might take centuries.
The solution is automation. "It will improve accuracy," says
Stanford's Paul Berg. "It will remove boredom; it will accomplish
what we want in the end." The drive for automation has already
begun; a machine designed by Caltech biologist Leroy Hood can now
sequence 16,000 base pairs a day. But Hood, a member of the Genome
Advisory Committee, is hardly satisfied. "Before we can seriously
take on the genome initiative," he says, "we will want to do
100,000 to a million a day." The cost, he hopes, will eventually
drop to a penny per base pair.
Hood is not alone in his quest for automation. That is also the
goal of Columbia University biochemist Charles Cantor, recently
appointed by the Energy Department to head one of its two genome
centers. "It's largely an engineering project," Cantor explains,
intended to produce tools for faster, less expensive sequencing and
to develop data bases and computer programs to scan the data. Not
to be outdone, Japan has set up a consortium of four high-tech
companies to establish an automated assembly line, complete with
robots, that researchers hope will be capable of sequencing 100,000
base pairs a day within three years.
Is there a better way? In San Francisco in January, Energy
Department scientists displayed a photograph of a DNA strand
magnified a million times by a scanning tunneling microscope. It
was the first direct image of the molecule. If sharper images can
be made, the scientists suggested, it may be possible to read the
genetic code directly. But that day seems very far off.
Even before the Human Genome Project was begun by the NIH,
others were deeply involved in probing the genome. Building on a
long-standing program of research on DNA damage caused by
radiation, biologist Charles DeLisi in 1987 persuaded the Energy
Department to launch its own genome program. In addition to the
sequencer and computer-hardware engineering projects, Energy
Department scientists are focusing their attention on mapping seven
complete chromosomes.
Victor McKusick, a geneticist at Johns Hopkins University, was
in the game much earlier. He has been cataloging genes since 1959,
compiling findings in his regularly updated publication, Mendelian
Inheritance in Man. In August 1987 he introduced an electronic
version that scientists around the world can tap into by computer.
At the end of December it contained information on all the 4,550
genes identified to date. Says McKusick: "That's an impressive
figure, but we still have a long way to go." Several other
libraries of genetic information are already functioning, among
them GenBank at the Los Alamos National Laboratory and the Howard
Hughes Medical Institute's Human Gene Mapping Library in New Haven,
Conn.
McKusick also directs the Human Genome Organization (known
informally as "Victor's HuGO"), a group formed last September in
Montreux, Switzerland, by 42 scientists representing 17 nations.
"The U.N. of gene mapping," as McKusick describes it, plans to open
three data-collection and -distribution sites, one each in Japan,
North America and Europe.
Geneticist Ray White, formerly at M.I.T., has established a
major center for genetic-linkage mapping at the University of Utah
in Salt Lake City. In 1980 he began a study of 50 large families,
collecting their blood samples, extracting white blood cells, which
he multiplies in cell cultures, then preserving them in freezers.
Working with family pedigrees and DNA extracted from the cell
bank, White and his group have identified more than 1,000 markers,
each about 10 million base pairs apart, on all the chromosomes.
They have also been major contributors to the Center for the Study
of Human Polymorphisms, set up in Paris by French Nobel laureate
Jean Dausset to coordinate an international effort to map the
genes. Of the 40 families whose cell lines reside in CEPH's major
data banks, 27 have been provided by White's group.
How and if these and other genetic research efforts will be
coordinated with the Human Genome Project is a question being
pondered by director Watson and his advisory committee. "Right
now," says Watson, "the program supports people through individual
research grants. We have to build up around ten research centers,
each with specific objectives, if we want to do this project in a
reasonable period of time."
The effort will also include studies of genes in other
organisms, such as mice and fruit flies. "We've got to build a few
places that are very strong in mouse genetics," Watson says,
"because in order to interpret the human, we need to have a
parallel in the mouse." Explains Genentech's Botstein:
"Experimentation with lower organisms will illuminate the meaning
of the sequence in humans." For example, genes that control growth
and development in the fruit fly are virtually identical to
oncogenes, which cause cancer in humans.
One of the early benefits of the genome project will be the
identification of more and more of the defective genes responsible
for the thousands of known inherited diseases and development of
tests to detect them. Like those already used to find Huntington's
and sickle-cell markers, for example, these tests will allow
doctors to predict with near certainty that some patients will fall
victim to specific genetic diseases and that others are vulnerable
and could be stricken.
University of Utah geneticist Mark Skolnick is convinced that
mapping the genome will radically change the way medicine is
practiced. "Right now," he says, "we wait for someone to get sick
so we can cut them and drug them. It's pretty old stuff. Once you
can make a profile of a person's genetic predisposition to disease,
medicine will finally become predictive and preventive."
Eventually, says Mark Guyer of the NIH's Human Genome Office,
people might have access to a computer readout of their own genome,
with an interpretation of their genetic strengths and weaknesses.
At the very least, this would enable them to adopt an appropriate
life-style, choosing the proper diet, environment and -- if
necessary -- drugs to minimize the effects of genetic disorders.
The ever improving ability to read base-pair sequences of genes
will enable researchers to speed the discovery of new proteins,
assess their role in the life processes, and use them -- as the
interferons and interleukins are already used -- for fighting
disease. It will also help them pinpoint missing proteins, such as
insulin, that can correct genetic diseases.
Mapping and sequencing the genes should accelerate progress in
another highly touted and controversial discipline: gene therapy.
Using this technique, scientists hope someday to cure genetic
diseases by actually inserting good genes into their patients'
cells. One proposed form of gene therapy would be used to fight
beta-thalassemia major, a blood disease characterized by severe
anemia and caused by the inability of hemoglobin to function
properly. That inability results from the lack of a protein in the
hemoglobin, a deficiency that in turn is caused by a defective gene
in bone-marrow cells.
To effect a cure, doctors would remove bone-marrow cells from
a patient and expose them to a retrovirus* engineered to carry
correctly functioning versions of the patient's faulty gene. When
the retrovirus invaded a marrow cell, it would insert itself into
the cellular DNA, as retroviruses are wont to do, carrying the good
gene with it. Reimplanted in the marrow, the altered marrow cells
would take hold and multiply, churning out the previously lacking
