Download Alz genetics piece

Sue Halpern
802-388-3673
One morning last May, two white SUVs left the shade of the parking garage at the
Grand Almirante Hotel and Casino in Santiago, Dominican Republic as the gamblers and
prostitutes were calling it a night and headed north to rural Navarette. The lead vehicle
was driven by Angel Piriz, a thirty-six-year-old Cuban doctor who now lives in Harlem.
Beside him was Roserina Estevez, a recent graduate of medical school in Santiago who,
like Piriz, is working as research physician at Columbia University under the supervision
of Dr. Richard Mayeux. For nearly twenty years Mayeux, a neurologist, epidemiologist,
and director of the Taub Institute for Research on Alzheimer’s Disease and the Aging
Brain, has been compiling what has become the most comprehensive genetic library of
families with Alzheimer’s disease in the world. The family members are either residents
of the largely Dominican neighborhood of Washington Heights where Columbia
Presbyterian Hospital is located, or, like the people Piriz, Estevez and their colleague
Vinny Santana in the car behind theirs were hoping to see in Navarette, from the
Dominican Republic itself.
Alzheimer’s, which ultimately may turn out to be too broad a term to be of much
use clinically, is typically divided into two types. There is the early-onset variety, which
tends to strike between the ages of thirty and fifty and is very rare, occurring in less than
five percent of cases, and there is the much more common late onset disease, which tends
to afflict people who are sixty-five and older. Early onset Alzheimer’s is a genetic
disease that follows the simple law of Mendelian inheritance: if you are born with the
gene, you get the disease. Late onset Alzheimer’s is thought to be genetic, too, but in a
much more messy way: rather than being caused by a single gene, it appears to be the
result of a passel of genes that, individually or together, are believed to increase one’s
risk of dementia. So far only one of those risk factor genes has been conclusively
identified and confirmed. In May, though, as the Columbia researchers trolled the island,
drawing blood that was sent by courier each morning to New York, it looked like
Mayeux’s library, which contains DNA and neuropsychological data collected from over
13,466 participants, most of whom either have Alzheimer’s or are related to someone
with the disease, might soon yield a second.
“Our belief is that late onset Alzheimer’s disease is genetically determined
because it tends to aggregate in families,” Mayeux had told me five months earlier, when
I first visited him in his office on the 19th floor of Presbyterian Hospital. A deceptively
easy-going Louisianan who talks with a lingering nasal twang, Mayeux, who is 59,
oversees a staff of neuroscientists, geneticists, psychologists, epidemiologists,
neurologists, cell biologists, biochemists, genetic counselors, and animal modelers
spread over two floors of the hospital, as well as in outposts at the medical school and
Columbia’s Neurological Institute, where he also runs the Gertrude Sergievsky Center.
Many, like Joe Lee, who sifts through the 30,000 genes that comprise the human genome
looking for a genetic quirk that could explain the colonization of the brain by sticky
plaques and neurofibulary tangles—the signature pathology of Alzheimer’s disease--and
Scott Small, a neurologist who has developed a method of using magnetic resonance
imaging to look deep into the brain and see trouble before trouble has a name, took the
conventional academic route of medical degrees and doctorates. But a surprising number
never intended to chase what Mayeux calls “the great white whale of neuroscience.”
Angel Piriz, for instance, trained as a surgeon in Havana but, unable to practice clinical
medicine in the United States, had spent three years working as a bookkeeper for a
Manhattan construction company until he noticed the Columbia job on the internet.
Vinny Santana, a security guard in the Presbyterian Hospital emergency room, was
offered extra hours to escort Mayeux’s field researchers to interviews in Washington
Heights, eventually became one of those researchers himself, and then the research
coordinator and co-author of four scientific papers. Joe Flynn, sitting in a darkened room
in front of a computer screen making mathematical sense of MRI data, had spent seven
years as an accountant on Wall Street. Christian Habeck, who created diagnostic
algorithms, was an astrophysicist. Even Mayeux did not come to the hunt directly.
“I started off thinking this was not a genetic disease. I thought it was
environmental, associated with aging. But the accumulating data convinced me, seeing
that this disease tracks in families. It doesn’t always follow a pattern, but it does track in
families, so that if you have family members with the disease you have a much higher
risk, and siblings with the disease give you an even higher risk. The evidence was very
hard to counter.”
Mayeux took a bound volume the size of Webster’s Third from his bookshelf and
began flipping through the pages as he spoke. It was a encyclopedia of the family trees
of people in his study that showed who was related to whom and which ones had the
disease, who was disease-free, and who was living in the border town between lucidity
and dementia.
“What we wanted to do was find a population where we thought the rates were
higher, because the thing about genetics is that if you try to identify people who carry the
gene, you are looking for unusual people. It’s not like epidemiology, where you try to
get random samples of random people. Genetics is just the opposite. You want a biased
population. You want families where there is more of the disease because you have a
better chance of figuring out what the gene is.
“That’s how we stumbled into this study of people in the Dominican Republic.
We noticed when we were doing a general population study of elderly people who live
around the hospital that Dominicans had about three times the rate of Alzheimer’s disease
compared to the whites and blacks in the community. So you have to ask yourself why
that would be. Then it starts to explain itself that at least in the DR, Dominicans tend to
marry other Dominicans, and you don’t have different populations moving in there. You
have a smaller genetic pool, and first cousins marrying first cousins, so the gene pool
tends to stay enriched. Here’s one,” Mayeux said, pointing to a page in the book. “These
two people are twins. Look at how many people who are related to them are affected and
how many are beginning to experience symptoms.”
The family that Piriz, Estevez and Santana were hoping to track down in
Navarette had similar qualities. For one thing it was huge—one member alone had had
twenty-three children. For another, a number of relatives had already been diagnosed
with probable Alzheimer’s. (Probable because, so far, the only way to conclusively
diagnose the disease is at autopsy. But likely, because the doctors at Columbia, like those
at Alzheimer Disease Research Centers all over the country, are accurate about ninety
percent of the time.) “This is a branch of the original family we saw here last year,”
explained Santana, who is in charge of drawing the pedigrees, as well as directing all the
field work. “The proban—which means the first person in the family we saw—who was
in her late 60s, died in February. She’s confirmed with the disease. In her generation a
couple of cousins and siblings have AD. One of her children has it, and a couple are
borderline. There’s some first cousin intermarriage. These people we’re seeing today are
her cousins. If we can find them.”
Navarette isn’t much of a town—a strip of concrete shops on either side of the
road, and street vendors selling pineapples and mangoes and goat meat so fresh that the
goats are still tethered to the stall—and the family didn’t have much of an address. “It’s
called Ginger Alley,” Santana said, turning sharply into a narrow dirt track patrolled by
chickens. Piriz, who had briefly gotten lost, pulled in behind Santana, and without much
conversation, he and Estevez assembled their medical kits while Santana gathered the
clipboards, notebooks, and questionnaires they would need to administer the
neuropsychological tests that, along with the medical exam, were crucial for determining
who, at case conferences in New York a few weeks later, would be diagnosed with the
disease and who would be said, so far, to be exempt. Then, consulting Santana’s
provisional pedigree as if it were a map, they spread out along Ginger Alley, Estevez
knocking on the door of the house closest to the cars, Santana and Piriz walking farther
back, past houses made of concrete and tin, through someone’s dim, fly-infested kitchen
that led to a narrower alley and a warren of corrugated houses, asking for a man named
Matias de Jesus Vargas, the proban’s second cousin. As they went they attracted little
boy after little boy, a whole parade of them, who eventually led the researchers to a spare,
three-room dwelling: Matias’ house. Matias, a gaunt eighty-three year old man, tanned
from a life growing bananas and tending fields of rice, was in bed lying bare-chested in
blue shorts on top of yellow smiley face sheets, surrounded by two of his five wives, four
of his fourteen children, two of his daughters’ daughters, a great-granddaughter, a
brother, two cousins, the sister of one of his wives, and an indeterminate number of
children who might have been related to the assortment of adults. He wasn’t saying
much. A few months before he had been diagnosed with pancreatic cancer and wasn’t
expected to outlive the year. Even so, he had consented to Santana’s request to
participate in the study.
The neuropsychological tests that Santana prepared to administer to the man in
the bed were developed by Yaakov Stern, a psychologist who has worked with Richard
Mayeux for 21 years, and they were adapted for the Dominican population by Stern and a
young neuropsychologist at the Taub Institute named Jennifer Manley, to make sure they
were, in Stern’s term, “culture-fair.” The neuropsychological battery assesses a variety
of cognitive functions, many having to do with memory, since in Alzheimer’s disease,
memory—for the names of things, for words, for recent events—tends to deteriorate
profoundly early-on. Indeed, some researchers, notably the psychologists Mony de Leon
and Susan DeSanti at NYU, and the neurologist Rod Shankle at the University of
California at Irvine, were finding certain neuropsychological tests to be predictive of
disease. And in Yaakov Stern’s lab it had been observed that results on a verbal fluency
test appeared to foretell who, among the afflicted, would die the fastest.
“My feeling is that the best tool we have for diagnosing AD is neuropsych,”
Stern, a rangy man with a loping gait, said last winter as we walked back to the hospital
from the Neurological Institute, where he had just given grand rounds on the concept of
“cognitive reserve,” the reservoir of synapses and neurons that lets some people, even
those with plaques and tangles, resist dementia. “Neuropsych is sensitive to early
changes. But the point of transition where someone starts to develop problems, well, it’s
hard to make a fast rule about any one person that they are definitely going to develop
Alzheimer’s disease.”
To determine who, in the Dominican population, had the disease—as opposed to
who will develop the disease—Stern and his colleagues have developed an algorithm that
gives them a score that they have found to be clinically reliable.
“When someone comes to our Memory Disorders Clinic in New York, we try to
use the best norms for them based on their education, their age, even their gender,” Stern
explained. “That’s the typical neuropsychological approach. It asks: what can we
expect from this person. And if they deviate from the norm then we can assume there is
some deficit.
“In the genetics study in the Dominican Republic, we decided to use an algorithm
because, first of all, we didn’t know what the norms there were when we started, and
secondly we wanted to be consistent and unbiased. So for every test we have a cut score,
below which we’d say that’s bad, and if a person meets our criteria for cognitive
deficiency they have cognitive deficiency, and if they don’t, they don’t—for the purposes
of diagnosing dementia. The idea behind the algorithm is that we wanted memory to be
the gatekeeper. If someone didn’t have a bad memory, I didn’t want to say they had
dementia. Most of the time the algorithm works pretty well.”
“What day of the week is it?” Vinny Santana asked Matias de Jesus Vargas. “
What is the date? What year? Where are we? Can you count backward from twenty?
Can you name the months of the year in reverse order? I am going to tell you a name and
address and I want you to repeat what I’ve said: Juan Perez, Avenida Duarte cuatro y
dos, Samana.” This was the warm-up, and de Jesus Vargas was doing okay. He knew he
was in the bedroom, not the kitchen; he knew the year; he knew it was summer, though
technically it was spring.
Santana leaned in close. “I’m going to read you a list of twelve words, and when
I’m done I want you to repeat them back to me. Juevo,” he began. “Lava.” De Jesus
Vargas fingered the religious medal he wore around his neck and looked lost. “I can’t
remember,” he said, pointing to his head. In the algorithm developed by Yaakov Stern,
the cut score on this test is 25: if someone, given the opportunity to repeat any of these
twelve words six times, for a top score of seventy-two, can’t get to twenty-five, he can be
considered for “case status.” When the test was over, de Jesus Vargas’ score was fifteen.
In the next room, Angel Piriz was going through the same routine with one of de
Jesus Vargas’ wives, seventy-year-old Luz Maria Lopez. Lopez, a short woman in a
faded housedress and heel-worn flip-flops, looked at him suspiciously. “Do you ever
find yourself getting lost?” Piriz asked. “Hell no,” she said. He took her blood pressure,
looked into her eyes, tested her reflexes. Then he put on green latex gloves, took out a
syringe, and prepared to draw her blood.
Across the street from Columbia Presbyterian Hospital, at the Russ Berry
Biomedical Center, there is a room with a bank of industrial freezers containing 28,544
vials of blood. In 1986, when Richard Mayeux inaugurated what is known as the
WHICAP study, a sweeping epidemiological investigation of the health and habits of
2500 elderly residents of Washington Heights, he instructed the researchers to collect
blood in addition to recording demographic and medical data. The field of Alzheimer’s
genetics was in its infancy and putting blood on ice was either capricious or prescient,
depending on who was paying the bill. It had only been three years since George
Glenner and Cai’ne Wong, at the University of California, San Diego, succeeded in
isolating amyloid, the key constituent of the plaques that accumulate in an Alzheimer’s
brain, and two years since they discovered that amyloid was a peptide—a protein
fragment that came in different lengths. Glenner and Wong called the peptide beta
amyloid and proposed that a genetic mutation causing the overproduction of amyloid,
would be found somewhere along chromosome 21, the chromosome that goes awry in
Down Syndrome. The amyloid they had sequenced had come from Down’s patients, and
by middle-age, most people with Down Syndrome experience Alzheimer’s-like
symptoms, so it seemed logical that chromosome 21 would be implicated.
And that is where, in 1990, four years after Mayeux’s WHICAP researchers had
begun to fan out across Washington Heights, it was found by Alison Goate and John
Hardy, geneticists at the University of London. The first Alzheimer’s gene was called
APP, an acronym that stood for amyloid precursor protein. An APP mutation caused the
overproduction of beta amyloid in the brain; without exception, a person who carried the
mutation developed Alzheimer’s disease. APP gave scientists a way to begin to
understand what was happening in an Alzheimer brain, and a rudimentary hypothesis:
Alzheimer’s disease was caused by clumps of beta-amyloid that strangled neurons and
synapses. If, before, scientists were looking into a black hole of disease, they were now
peering into a tunnel, and they believed they knew, with some confidence, how a small
number people tumbled through its entrance.
“When it was first discovered, we thought that APP was it,” recalled Mayeux.
“That it was the Alzheimer’s gene.” But the math didn’t work: of the millions of cases
of Alzheimer’s disease, the APP mutation occurred in less than two hundred of them.
The search for additional Alzheimer’s genes took off during the nineteen nineties,
and by mid-decade three more had been found and confirmed. Two, called presinillins,
were deterministic genes that, like APP, caused early-onset Alzheimer’s. Like APP, too,
they interfered with amyloid production—though differently, by cleaving the amyloid
protein--which gave further credence to the idea that too much amyloid precipitated
dementia. Still, the presinillin genes weren’t it, either; PS1, which was found by Peter St.
George Hyslop, at the University of Toronto, occurred in about 200 families worldwide,
and PS2, which was discovered by Hyslop as well as Gerald Shellenberger at the
University of Washington, in only 30.
The fourth Alzheimer gene, though, was different. It wasn’t rare, it didn’t cause
early-onset Alzheimer’s, it influenced late-onset, common Alzheimer’s, and it wasn’t
even a mutation. Rather, APOE4, discovered by researchers at Duke, is a naturally
occurring variant of a gene that we all carry; the variant shows up in about 33% percent
of all people. APOE4, which interferes with the body’s ability to clear beta amyloid
from the brain, increases the risk of getting AD, but it doesn’t cause it. Many who have
the APOE4 gene never get Alzheimer’s disease, while many people who don’t, do, and
still, twelve years after its discovery, no one can say why that is.
When the Columbia team had been at work for more than nine hours along Ginger
Alley, Angel Piriz sat in Mathias de Jesus Vargas’ house interviewing the last volunteer
of the day, Maria Lopez Fernandez, the eighty year-old sister of de Jesus Vargas’ wife
Luz Maria. Fernandez, who was educated through the first grade, had borne twenty-six
children. When Piriz asked her how she was doing, if she forgot things or got
disoriented, Fernandez answered by telling him that she still threaded her own needles.
Then she failed every test he gave her—the months, backward, the word list, Juan Perez’s
address. “This is why we do a medical exam and a neuropsych exam,” Piriz said, packing
up. “Even though the tests indicate that she has a problem, she functions just fine.”
Meanwhile, outside, under a canopy of banana trees, Santana was chatting with about
fifteen members of the extended family, puzzling out their pedigree as if it were a
crossword. Establishing accurate family histories is essential to finding genes, and can be
just as challenging. People’s memories are faulty, public records are inaccurate, and
there are often inconsistencies in both. One time, after Santana had been following a
family for three months, drawing and redrawing its genealogy and never getting it right,
he discovered that one member of the family had had two husbands, both of whom were
named Rafael Vargas. And in one village where the Columbia team was currently
working, not only was every person related, they all had the same last name. Some of
them shared the same first name, too.
It was nearly six o’clock. The researchers had collected fifty-one vials of blood,
repeated the words on the verbal recall test 1,224 times, completed neuropsychological
and medical tests on seventeen people, reconnected Matias de Jesus Vargas’ intravenous
drip of nutrition and chemotherapy which had fallen out two days earlier, and suggested a
salve for one of his nephews who was suffering from a rash that had thickened the skin
on his arms and legs to a carapace. There was another cousin to see, a man who lived on
the opposite side of the highway that runs through Navarette, and possibly a relative in
Washington Heights. Santana made a note to track her down when he was back in New
York and called the cousin across town to tell him the team would see him the next day,
when they were returning from interviewing subjects in Puerta Plata. Then he backed his
SUV into a tree, which caused the bumper to crinkle and the crew of little boys to
materialize again. He checked to see that the blood in the cooler was all right, shooed the
boys away, then exited Ginger Alley, which he wouldn’t see again for a year or two, at
which point, he guessed, “some of the borderlines will have converted.”
In the lectures Richard Mayeux gives to people who may be unfamiliar with the
simplest facts of genetics—that genes are proteins, and proteins are made of a sequence
of chemicals called bases, and if any one base is out of sequence the protein may be
dysfunctional—he often ends up talking about cops and robbers.
“Our colleagues at the Genome Center in the United States tell us that there may
be three billion base pairs in the entire genome and that there are one hundred twenty
million base pairs in each of the chromosomes, and about two thousand to two hundred
thousand base pairs in a gene, and what we are doing is looking for a single base pair that
is different or out of sequence,” he told a group of sixty Dominican professors and
students at the Universidad Technologica De Santiago, the day after his field researchers
completed nine more interviews in the northern coastal city of Puerta Plata and four more
that morning in Jicome.
UTESA is the academic home of Dr. Martin Madrano, Mayeux’s chief
collaborator in the DR, one of only thirteen gerontologists on an island with a population
of nine million, and the person responsible for steering hundreds of Dominican
Alzheimer patients and their families in Mayeux’s direction. Back in New York that role
was played Dr. Rafael Lantigua, the deputy director of the Taub Institute and the medical
director of Associates in Internal Medicine, a group practice run by Presbyterian Hospital
for its neighbors in Washington Heights. Lantigua—who is short, dark, balding, and
doesn’t appear to spend a lot of time at the gym--had traveled to the DR that morning
with Mayeux—a tall, pale, curly-haired, zealously fit, former wide-receiver for
Oklahoma State who likes to tell people that he and Lantigua are brothers separated at
birth. The two men had sat shoulder to shoulder for hours by the hotel pool as Mayeux
translated the lecture into Spanish and Lantigua coached him through sticky points like
“if you use the word ‘relaciones’ you’re saying ‘making love.’”
“So we’re looking for that one base pair, which is like having the police know
that someone has committed a crime somewhere, but they don’t know where, and they
have to start looking for him all over the universe,” Mayeux went on. “That’s basically
where we are. The goal of genetic family studies is to try to get down to the earth, and
then into the neighborhood, and eventually to find the culprit.”
(Another time, trying to explain how the geneticist Peter Hyslop, who is not
typically credited with finding the APP mutation, was crucial to its discovery, he said, “If
I tell you the criminal is holed up in Hastings on Hudson, at least you have a good place
to start looking. Peter Hyslop said APP was in Hastings, and Alison Goate and John
Hardy said “Aha, it’s at 45 Main Street, and therefore it must be Mayeux.” Mayeux does
live in Hastings, a Westchester suburb, with his wife, Nancy Green, a pediatric oncologist
who is the medical director of the March of Dimes, and their two teenage daughters.)
If mutations are the bad guys, and scientists are the good guys, then since the
discovery of the presinillins and APOE4, the good guys had made something like a
hundred false arrests. In peer-reviewed paper after paper, research teams all over the
world claimed to have identified about one hundred unique genes that they claimed
influenced or in some way triggered late-onset Alzheimer’s disease, but not one of those
genes had been able to be replicated consistently by other researchers, if at all.
“There are no more APOEs, so people have to lighten up,” said Harvard professor
Rudy Tanzi, possibly the world’s most dogged Alzheimer gene researcher, and the author
of a few of those papers himself. But as John Hardy saw it, “There’s a tendency for any
individual lab chief to look so hard that he convinces himself that he’s found something
and other people check and they can’t replicate it. People have not replicated some of
our work, and we’ve not replicated anybody else’s work in these other genes. It’s been
twelve years and nobody has found anything. It’s been a big deal.”
There is one simple reason why no one has found a new Alzheimer’s gene in
more than a decade, and another reason that is less simple, and they both come down to
the same thing: statistics. Risk factor genes, the genes that will explain common, lateonset AD, are inherently elusive because carrying them does not automatically presage
disease. More challenging, there may be a lot of risk genes—five or six, in Tanzi’s
estimation, up to eight in Hyslop’s—each with a potentially minute effect. Risk genes, in
other words, are quiet, and it will take a very sensitive microphone to hear them. In the
case of Alzheimer’s, that microphone is a large family study like Mayeux’s, with reams
of information over time about each individual’s health issues, eating habits, work and
leisure pursuits, as well as genealogies that show the genetic path of AD. The big
numbers pick up on small effects, and the detailed histories parse the sample. Mayeux’s
thick books of pedigrees and database of dna allow researchers to define a person’s
genotype—what genes she carries—as well as a phenotype—what traits she embodies-and then to subdivide the phenotype according to which traits very specifically correlate
with the kind of dementia that characterizes Alzheimer’s disease. On the 19th floor of
Columbia Presbyterian Hospital, the crucial distinction was coming down to a person’s
performance on certain memory tests, and the researchers were seeing a pattern in
performance and disease that they hoped would show up in the genes.
“Age of onset is a wimpy phenotype,” Mayeux said to no one in particular at one
of the team’s weekly genetics meetings last January. “Memory is better.”
“Delayed recognition is the most sensitive test we have for AD,” Joe Lee, the
geneticist, told him. “That and another test that I can’t recall. I may be a subject for this
study soon myself.”
“We all may be,” said Mayeux, laughing. (In fact, both Mayeux’s mother and
aunt have Alzheimer’s Disease—he diagnosed them himself--and it is a measure of the
prevalence of AD that so does Yaakov Stern’s father, who is a patient at the Memory
Disorders Clinic, as did Vinny Santana’s grandmother. Santana realized she had the
disease on one of his research trips to the Dominican Republic, when he knocked on her
door to deliver groceries, put the first bag on her kitchen table, went out to the car for a
second bag, found the door locked on his return, knocked, and was greeted by his
grandmother—who had no recollection how the groceries had arrived in her house--as if
he had been gone for months, not minutes. In Alzheimer’s Disease, short-term memory
loss is the most obvious phenotype.)
A few months after the genetics meeting I attended in January, a number of the
same researchers who were at that meeting, plus Stern, Piriz and Estevez, were gathered
in Mayeux’s office when Santana rolled in a cart piled three feet high with the two
hundred interviews that the field researchers had completed in the weeks since their last
diagnostic case conference. In the stack were the files of first-time participants from the
Dominican Republic and follow-up examinations of subjects who had been seen on
previous visits, as well of as their relatives in Washington Heights. The day before the
meeting, which was about three weeks after their long day in Ginger Alley, Santana,
Estevez and Piriz had finally interviewed Matias de Jesus Vargas’ sixty-year-old niece in
her apartment a few blocks down Broadway from the hospital that offered the same view
of the Hudson River and the bend of Manhattan that they would have been seeing from
Mayeux’s office if their heads hadn’t been bowed over the case files like students
cramming for an exam. Despite being diabetic and overweight, the niece had no memory
complaints and no signs of impairment, even though diabetes itself is a risk factor for
Alzheimer’s. Her case was one of the ones on the day’s docket.
Santana, who was a college drop-out when he started working for Mayeux and is
now a few credits shy of completing an MBA, dealt out score sheets—officially called
“clinical core diagnosis”—that looked surprisingly like an IRS 1040 short form, with
various sections and schedules and subtotals, all leading to a bottom line: did the
participant have Alzheimer’s disease or not. To get there they had to rule out Parkinson’s
disease, prion disease, alcohol dementia, dementia with Lewy Bodies, frontotemporal
dementia and anything else that might mimic the symptoms of AD. They had exactly an
hour for the meeting because this was Mayeux’s last day at Columbia for three months—
though that last day turned out to be fungible—before he started a “mini-sabbatical” at
Rockefeller University where he was going to study genetics.
“We’ll just get through as many as we get through, and then get the data to Joe
Lee so he can put it into the computer,” Mayeux said, pulling a dozen folders off the cart.
“Here’s someone you saw,” he said waving a folder at Estevez, who was sitting next to
him. “What was his blood pressure?” He looked at her intently, not letting on that he
was pulling her leg—in fourteen days in the Dominican Republic the researchers had
examined ninety-eight people, and there was no way that Estevez could possibly
remember, nearly a month later, any individual’s vital signs. She looked at little stricken.
As the newest member of the team, Estevez had not yet grown accustomed to Mayeux’s
ability to tease, compliment and assert his authority all in the same sentence. “175 over
90,” she shot back. Mayeux looked stunned. “That’s amazing. How did you do that?
Did you know or did you guess?” But there was no time for an answer, “Okay, what else
do we know about this guy?” he asked.
Accurately diagnosing a subject is crucial to looking for genes, and the design of
Mayeux’s field studies, with their repeat visits every eighteen months, increased the odds
that they would get it right. According to Peter Hyslop, who used Mayeux’s Dominican
cell lines in his discovery of PS1, “the way most studies are done is that a person is seen
once and diagnosed as either having Alzheimer’s Disease or not having Alzheimer’s
Disease, and then not seen again. In the Washington Heights and Dominican studies,
people are followed up again and again. There may be some individuals on the
borderline when they were first seen, but when they were followed up later they had
developed a clear case of Alzheimer’s, so you can be quite certain about the diagnosis.
Conversely, there are people who are seen the first time who are normal, and if you
follow them for years and they are still normal there is a much greater chance that they
are really normal controls and not just pre-symptomatic carriers.”
To refine their diagnostic powers, the Columbia group convened an autopsy
meeting once a month to see how close their assessment of a person while alive was to
the incontrovertible pathological truth. The meetings were run by Dr. Lawrence Honig,
Dirctor of Clinical Care at the Alzheimer Disease Research Center, often to standing
room only crowds. Honig would present a patient’s history and tentative diagnosis, then
project slides of the brain, first whole, then in slices, stained in red and blue to show its
dominant features. Then the doctors would decide, based on the one crucial piece of
evidence they had been missing, if they had been right anyhow.
I had skipped the August autopsy meeting, and a week later, when I stopped by
Larry Honig’s office at the opposite end of PH 19 from Richard Mayeux’s, he showed me
the slides of a woman whose case had been discussed that day, a case that illustrated how
difficult it could be, even with years of data, to get a diagnosis absolutely right. Honig
walked me through the women’s history: a clerical worker with a year of college, she
had been first seen as a WHICAP participant in 1992 at the age of sixty-eight. At the
time she was besieged by numerous ailments—cirrhosis, gallstones, pulmonary disease,
carpel tunnel and facial palsy, among them—and, in addition, was a former smoker and
recovered alcoholic. In later years she was found to carry the APOE4 gene, and an MRI
showed some brain atrophy. She had done well on all her tests, though, the medical ones
and the neuropsychological ones, and not just that first year, but at every interval until
2000, when there was a decline in some of her memory scores. Two years later there was
a further decline in memory, and a spirited discussion among the clinicians whether or
not to move the woman from the non-affected category to a diagnosis of early
Alzheimer’s. The neurologists, led by Honig, were pretty sure, based on the fact that the
woman’s test scores had been stable for over a decade, that her recent memory problems
were the result of her multiple physical ailments. The neuropsychologists were sticking
by their algorithm. Unable to agree, they left the diagnosis unchanged, waiting to see
what would show up the next time around. But there was no next time. By late 2004,
when the woman was scheduled to be seen again by the WHICAP researchers she was
dying of congestive heart failure.
Honig called up a couple of slides of the woman’s brain on his computer monitor.
To my untrained eye the brain seemed pretty normal—there was nothing in the slices that
looked like measles, which is how plaques show up when they’ve been stained, nor did
the woman’s brain appear especially shrunken, as Alzheimer brains tend to be.
“We couldn’t even find a single plaque,” Honig beamed. “There were no signs of
AD. So I can crow that I was right. But we’re not always right, so we have to be
modest.”
Neurologists have spent the past hundred years waiting for pathologists to prove
them right, since 1906 when Alois Alzheimer autopsied the brain of a fifty-one-year-old
woman who was exhibiting the kinds of behavior that most of us now would reflexively
call Alzheimer’s disease, and found it riddled with plaques and tangles. Still,
pathologists had waited almost as long to find out if the plaques and tangles were actually
meaningful—if they caused disease or were just an artifact of some other biological
process. What were they to make of the people who died with all the pathological
evidence of Alzheimer’s but were not demented? And how to account for the presence of
both plaques and tangles? Were the plaques, which are made of amyloid beta, more
important agents of disease than the tangles, which are composed of a protein called tau,
or were the tangles the prime suspect, or were the two accomplices in fleecing memory?
These questions consumed researchers for decades, often contentiously, and still do to an
extent. But what Rudy Tanzi at Harvard called the debate between the “Baptists and the
Tauists”—those who believed in the supremacy of beta amyloid and those who favored
tau tangles as the primum mobile of forgetting—was becoming more civil all the time;
the tauists were getting more research money, which in science is a show of respect, and
no one was disputing the central role of beta amyloid in making an Alzheimer’s brain,
especially a form of beta amyloid called a-beta 42 , though what the plaques were doing
in that brain could still rouse a heated discussion. Oddly, the geneticists, who, as the
scientific literature shows, have never shied away from a fight, were the inadvertent
arbiter of these rows, for the answers were coming, not from the examination of slices of
gross tissue, but from investigations at the molecular level, from the interrogation of
genes.
“I think the controversy is over, in a way,” said John Hardy, the geneticist who
discovered the APP mutation and who now runs a lab at the National Institutes of Health.
“Though maybe I’m delusional. But the fact is that if you have mutations in tau, you get
a disease that causes tangles. If you have mutations in either amyloid or the presinillins
you get Alzheimer’s disease, which, has both plaques and tangles. I think what that
shows is something very simple, and that is that there is an order to the pathology. The
analogy I use is a river: in Alzheimer’s disease you get on the Mississippi River near
Duluth and you end up in New Orleans with plaques, tangles and dementia. In Picks
disease, which is caused by mutations in tau, you get on the Mississippi River in St.
Louis and you still end up in New Orleans, but with tangles and dementia. My view is
that most people, very generally, agree with that. And that’s come out of the genetics.”
“So much of the work that we’ve done, going forward, is asking ‘how do these
genes cause disease?’ ‘What are the biological pathways involved?’ ‘What goes
wrong?’” Rudy Tanzi, John Hardy’s antagonist on most matters but this one, observed
the day we spoke. “What goes wrong is that you produce too much a-beta 42.
“What was controversial was whether the plaques, where a-beta 42 eventually
makes its home, are the cause of Alzheimer’s disease. And the answer is probably not.
Plaques are not good because they cause inflammation and inflammation can make things
worse. But the newest data says that we have to instead focus on the synapses. The main
place where a-beta 42 does its work is in the synapse. So every minute of the day an
Alzheimer patient is producing a-beta 42, for one reason or another, and it’s
accumulating in the brain, and where it’s accumulating is in the synapse. And it’s not
these big ropey fibrils that take years to form. Way, way before the plaques form you get
tiny little aggregates of a-beta 42. The peptides stick together and they get into the
synapse and they disrupt the most basic synaptic function for learning and memory. For
all we know plaques may be a beneficial attempt by the brain to sequester a-beta away so
you don’t have it in synapses any more. It’s the newly made a-beta 42 that is relentlessly
attacking the synapses and probably this is why an Alzheimer patient has trouble
remembering what happened five minutes ago. When you impair the synapse, when you
cause it to not function properly, eventually it starts to break down. And eventually it
goes away.”
This point, exactly, was illustrated last fall in a short article in the journal Nature
Neuroscience, though the piece was about mice, who don’t get Alzheimer’s disease
unless they are genetically altered to develop plaques and tangles, which these mice were.
The title of the paper, “Fibrillar Amyloid Deposition Leads to Local Synaptic
Abnormalities and Breakages of Neuronal Branches,” was not exactly sensational, but the
accompanying photographs—images of real mice brains in real time—were vivid, and
chilling. In the pictures, dendrites and axons, the parts of a neuron that carry information
to and from a cell, highlighted in a radiant shade of green, start out long and robust,
encounter the amyloid (which shows up as tomato red) and wither. The maw of amyloid
simply chews them up.
One of the authors of that paper, Karen Duff, was also the “author” of one of the
original transgenic Alzheimer mice. Duff, a precociously young geneticist who trained at
Cambridge, worked closely with John Hardy, and developed the first mouse model of the
presinillin mutation while a post-doc in his lab, continues to make mouse models of
neurodegenerative diseases at the Nathan Kline Institute in Orangeburg, New York. NKI
is about half an hour north of Manhattan, on the grounds of the Rockland Psychiatric
Center, a state-run facility for the mentally ill. It, too, looks institutional, in a scrubbed,
white-walled sort of way. Duff’s office, though, is cheerier, with a bobble-head doll of
James Watson on the desk, a cuckoo clock on the wall, and a stuffed mouse toy perched
on a bookshelf. The real mice were across the hall, where Duff, who recently received an
eight million dollar grant to work on tangle diseases, was raising multiple generations of
mice made to come down with something like Lou Gehrig’s disease.
To model a human disease in a mouse Duff had to micro-inject select bits of
human DNA into a mouse egg with a tiny needle. (“You’re going from the outside to the
inside and sometimes the eggs burst.” ) The introduced DNA bonded with the mouse
DNA and was passed along to offspring. The first mouse born with this engineered DNA
was called the founder mouse. Breeding the founders produced the actual “models”-real, live mice--which were then used to observe the progression of a disease or to test
theraputic interventions. With the Lou Gehrig mice, Duff was doing both.
“When I was in school I wanted to study physics, but I couldn’t do the math,”
Duff said as she picked up a paraplegic mouse and gently stroked its back. “Then, when
I was sixteen I went to a lecture about genes and learned that you could change one thing
in two million and have an effect on the whole organism and I said ‘that’s it.’ I wanted to
make that one change and see what it did, what pathways were involved, and then go on
to treat it. My post-doc was making transgenic mice. Making mice is very hard. You
have to be very specific with what you’ve done. If I have a demented mouse that can’t
get around a water maze, I have to know that that is because I changed one gene.
“I’m happy with mice. I think mice are a lot harder to work with than flies or
worms, and they are more accurate. They’re more like humans.” She ran her finger
along the sick mouse’s spine, then laid it back into its cage.
Though Duff had shifted her focus from making mice that mimicked human
disease processes, to mice that allowed her to test drugs and therapies that had the
potential to reverse or limit the progression of those diseases, she still supplied her old
established mouse models to other researchers, including Mayeux’s group at Columbia.
(Another feature of Mayeux’s public lectures was a movie clip of two mice negotiating a
water maze. The healthy mouse needed two tries before it could consistently find its way
to a submerged platform on the opposite side of the maze, where it could escape the cold
water. Meanwhile, the demented mouse, a relative of Duff’s original transgenic
“Alzheimer” mouse, could never remember where the platform was. In real--which is to
say human--Alzheimer’s disease, a loss of orientation—an inability to negotiate a world
of angles and depths and roads that may or may not lead to home—is one of the first
observable cognitive deficiencies.) Mayeux, too, was one of the few researchers for
whom Duff was still willing to create a whole new mouse model—once he had a whole
new Alzheimer’s gene to give her.
“It’s a symbiotic relationship,” Duff explained. “The geneticists want their
findings to be more than a gene on a piece of paper. They want to look at its
functionality, want to see it really does cause disease by putting it into animals. Richard
needs me to put the genes in the animals, and I need someone like him to give me the
genes to put into my animal models to see what they do. It’s a multi-part process.
Finding the gene is just the first bit.”
As tricky as finding this first bit had been, by the summer of 2004, six years into
the Dominican genetics study, the Columbia team, working closely with Peter St. George
Hyslop in Toronto, who was also collaborating with Dr. Lindsay Farrar at Boston
University, was pretty sure they were closing in on it. Having hunted through the entire
genome searching for places where shared patterns of DNA showed up in people with
Alzheimer’s disease, having pored over thousands upon thousands of genes, some four
spreadsheets long, they believed that they knew where the rogue gene resided and what it
did. Hyslop and his Toronto group had pointed the way. They had been studying a
promising gene that belonged to a family of genes called retromers that traffic proteins
around a cell, and proposed that the three research groups look at all the members of the
retromer family, which were scattered across ten different chromosomes. It was standard
work for Joe Lee, and he began the genetic close-work, applying micro-satellite markers
along the chromosomes at regular intervals and examining the space between them,
looking for variations in the genetic code, called alleles, that segregated with the disease.
“It’s like walking down the block between Broadway and 47th Street and Broadway and
48th, looking for a person with red hair,” Lee explained. “The micro-satellites are the
street signs.
“If more affected people tend to have certain alleles in a certain marker, then
that’s a candidate we should be looking at more carefully. If the affected people in the
family have the same allele, then that suggests that the disease gene must be close by.
Microsatellites give you coarse resolution to find which areas are interesting and once
you find the interesting part you go on to the next level of resolution, called snips, which
are single neucleotide changes, which give you a bigger resolution. You want to find the
snips that are really strongly associated, and then see if that single change in the
neucleotide changes the protein, and if that protein change is meaningful. For example,
in the case of AD, we try to see if it changes the level of beta amyloid.”
While Hyslop’s group in Toronto examined the DNA of a few hundred people of
European descent, and Farrar’s team looked at the genetics of sibling pairs—one with the
disease, one without it—collected in the United States, Canada and Vancouver, Lee
focused on the thousands of samples gathered from the Dominican Republic and
Washington Heights studies. By sheer size alone, Mayeux’s genetic library, with its
branching trees of large, extended families that often spanned two or more generations
offered Lee great power to see which alleles were segregating with the disease. But the
design and scope of the complementary Columbia studies, WHICAP in New York and
the family genetics survey in the DR, also gave Lee another advantage: where the large
families from the Dominican Republic enabled him to more easily find segregating
alleles, the large numbers of Dominicans in the Washington Heights study allowed him to
see if those alleles also showed up more generally in the Hispanic population.
Meanwhile, as Lee and the researchers in Toronto and Boston systematically
moved along the chromosomes—think of SWAT teams, going door-to-door--a funny
thing happened. A clear association had appeared on the same gene in each population,
but the alleles were different and isolated by race. The association for blacks was at one
end of the gene, the association for whites was at the other, and the association for
Hispanics was in the middle. Something else was in the middle, too, not far from the
Hispanic allele, a hyperactive segment of DNA that seemed to defy the rules of cell
division. Typically, genes that are close together on a chromosome are inherited
together, while genes that are far from each other are not because they break up during
reproduction and the fragments crossover and become something different and new.
Geneticists rely on the number of crossovers to tell them how close a marker is to a
disease gene. Generally, the rule is that if there are a lot of crossovers they are far from
it, and if there are none, they are near.
“It was a Saturday and I was looking at the data,” Mayeux recalled. “We couldn’t
figure out why we couldn’t get an overlapping haplotype—alleles that are inherited
together--across the ethnic groups. So I went to these big genomic databases that are in
the public domain and looked at some of the markers that are publicly available and
noticed there was a big gap between them. Why would adjacent markers not be inherited
together?” What the three teams had hit upon, Mayeux suspected, was a recombination
hotspot, where there was a lot of breaking and joining and statistically unpredictable
crossovers of genetic material.
“Joe Lee was away,” Mayeux continued, “so I asked Ron Chang, who works in
our lab, to run a genetics sequence called Merlin, a statistical package that can show
crossovers. Ron came back and said that in this segment there were way too many
crossovers. Something like thirty-eight in a small, small segment. And there should be
one. Or none.”
Mayeux called Hyslop, who worried that they had made a geneotyping error, and
suggested that the DNA be analyzed again. It took a month. The result was the same.
They really were looking at a site of unpredictable genetic activity, a red-herring that had
probably thrown-off previous researchers looking at the same gene. With so many crossovers, it never would have shown up on their analyses.
The three lead scientists, Mayeux, Hyslop and Farrar, were growing more
convinced that the gene they were looking at was one of the undiscovered late-onset riskfactor genes. The association was strong, and not just in one sample population, but five,
and not in one racial group, but three, scattered around the globe. Each scientist had
replicated the work of the others, and in their line of work replication was rare. By
nature, the three admitted to being skeptics—Lindsay Farrar liked to refer to himself as
Dr. No—but as the evidence mounted, their ability to explain it away deserted them.
“The ‘aha!’ moment happens to different people at different times and sometimes
it never really happens,” Hyslop said. “We are aware of little bits of data as they come
out that say ‘yes, it’s real,’ but not very strongly, so what you get is not really a eureka
moment, but something that is incremental. It starts out as “uh huh, but it’s probably a
fluke,’ to ‘maybe it’s not a fluke,’ to ‘this could be real, let me see what I can do to make
it go away,’ to ‘well, it seems pretty robust but there are still problems,’ to “we’ve taken
this as far as we can and we concede that there are many things to be done on this story,
but before we do too much more it needs to be put in the hands of some other people,
with totally different data sets and totally different ways of analyzing things and see if
they get the same results.’
“If we’re right about this gene, the next thing that will happen is that people will
say ‘okay, how does this fit in the metabolic pathway? You want to find which other
proteins it interacts with, whether you can make a drug that will affect that function,
whether you can put it into animal models to make a transgenic mouse.”
To get from the Taub Institute to the neurology mouse lab, you have to work your
way down two sets of stairs, pass through a double set of security doors, climb other
stairs, and turn a couple of corners until you end up in a long corridor of one of the upper
floors of the medical school. On the morning I visited, Ken Hess, the lab technician, was
just finishing taping a breathing mask over the snout of an inert black mouse; the mask
was delivering anesthetic gas. Once the mouse was knocked out, Hess wrapped it in
plastic and laid it on a tiny bed that fit into a long tube, and secured the mouse’s head
with tape. The mouse did not stir. Then Hess inserted the tube into a machine that
looked like a storage tank on a dairy farm but was actually a relatively petite magnetic
resonance scanner made specifically to look inside the bodies of small animals. Hess
turned to his computer. On the screen were images of the mouse’s brain, its beating
heart. The mouse looked peaceful in there, unperturbed by the percussive, hydraulic
noises issuing from the machine, noises that overwhelmed the room.
The mouse nursery, where this mouse was raised and lived, was one flight up.
There were white mice, black mice, wild mice, sick mice, genetically designed mice and,
though I did not know it then, a mouse that would, many months later, support the
finding that the particular gene on which Mayeux and his collaborators were closing in,
influenced the development of late-onset Alzheimer’s disease. The mouse was transgenic,
the descendant, many generations removed, of a founder mouse bred fourteen years
earlier for a different set of experiments on retromer dysfuction. Mouse models are
expensive to produce, and this one had been perpetuated for so many years with the idea
that someone in the future might find its particular collection of disabilities, including
memory deficits, to be of use.
That someone was Scott Small, the Herbert Irving assistant professor of
Neurology at Columbia who was still in medical school when the mice were being
developed. Small was a nimble scientist, able to jump effortlessly between cell biology,
physiology, electrical engineering and medicine. He was also a gifted clinician. I had
been in the examining room when he diagnosed an eighty-six year old women with
Alzheimer’s, a diagnosis that was more apparently devastating to her middle-aged son
than it was to her, and watched him talk to the son with tremendous compassion, then
slam his fist on the desk in anger as soon as the man and his mother were gone, furious—
with the universe, with science, with himself--that the best he could offer were free
samples of a drug that might help for a couple of months, if that. In the lab Small had
distinguished himself by developing high resolution functional magnetic resonance
imaging that let him see into specific regions of the brain at work. Pairing fMRI with a
technique called microarray, which let him examine all the cells in that region to see
which ones were contributing to disease, he had, independent of the genetics work
Mayeux was doing with Hyslop and Farrar, hit upon the retromer complex as a pathway
that lead to common, late-onset Alzheimer’s disease.
As was typical when I went to talk to Small in his artfully under-furnished work
space on the 18th floor of the hospital, it was not long before he had left his chair and
stood at the large chalkboard that dominated one wall, drawing diagrams. On this day,
which was in early August, he was trying to explain the retromer complex and how it
contributed to Alzheimer’s. The picture he was drawing had a lot of what looked like
Legos stacked top of each other, and a big circle in which there were four smaller circles,
two on the right, two on the left.
“God solved the problem of having different kinds of molecules that didn’t like
each other by compartmentalizing the cells. The garbage disposal is separate from the
refrigerator. You have a cell, which is the big circle, and you have different
compartments called organelles, which are the small circles. Once there is separation,
you need a way to take something from the refrigerator to the garbage disposal. That’s
what the trafficking molecules do. They’re like a shuttle bus.” He drew a line from one
organelle, the nucleus, to another, called the golgi, and then from the golgi out of the big
circle altogether, and from there back to one of the smaller circles, the endosome, and
from the endosome to the fourth circle, the lysosome. It was the Lego-stack, he
explained, that escorted the molecule back inside the cell after it had breached the cell
wall. It was called the “retromer” because it moved in reverse, and “complex” because it
contained a number of different molecules. One of those molecules, it turned out, was
the gene that had showed up so convincingly in Mayeux, Hyslop and Farrar’s data. Until
that gene appeared in the microarray analysis Small had done of cells taken from people
who had died of late-onset AD, he said, he had never heard of it before.
“So how does this tie into Alzheimer’s disease?” Small said. “If you cause
retromer dysfunction, you prevent the cargo from being removed. We think the cargo is
the amyloid precursor protein, APP--which would account for there being too much
amyloid in the brain. If our model is right we’ve uncovered something completely novel
that contributes to late-onset Alzheimer’s which, since finding APOE4 no one else has
done. All the other findings were genes that might have affected AD, but they are not
actually primary to the disease process. What we’ve found is primary to the disease
process. We’ve shown it in a Petri dish—when we turn down the retromer gene the level
of beta amyloid goes up--and now we’re working with the mouse, and the results look
very promising. I can’t say more than that right now.”
For months as I ducked in and out of Richard Mayeux’s office, the name of the
promising new gene was written on the whiteboard near the door, but I had no idea that
the word I was seeing was the gene we were talking about since no one would utter its
name in my presence. There was some concern that if I knew it I might inadvertently tip
off another research group, which could claim the finding as theirs. In science there is
only one winner, Karen Duff had said in the spring, describing how, during the race to
make the first presinillin mouse, she was afraid to spare the time to get her broken arm
set, for fear than she would lose too much ground to her competitors. And then there
were the cautionary words of Scott Small when he finished explaining the retromer
complex: “If tomorrow someone publishes the whole story I just told you,” he said, “I
can cry till I’m blue in the face but they will have published first.” It was only at the end
of the summer, when I was sitting with Mayeux in a small cabin on a lake in the
Adirondacks where he was vacationing with his wife, relaxing by taking forty and fifty
mile bike rides, that he said the name, and I realized that it had been in front me for half
the year.
There was also a matter of protocol. As real as a finding might seem in a lab, it
had to debut in a peer-reviewed scientific journal before it was accepted as real science.
Announcing it first in the popular press could queer that. For the longest time I didn’t
want to know the name of the gene, did not want to possess the ability to compromise
Mayeux’s work and the work of his collaborators. And the name seemed beside the
point, anyway. Names of genes are hardly descriptive—does UBQLN1 (Tanzi’s most
recent candidate gene) mean anything to you? More important is what the genes do, and
how they do it, and if they offer scientists a way to circumvent or override the disease
process.
“Five years,” Peter Hyslop told me when I asked when the first treatment for
Alzheimer’s disease, based on the genetics, would be available. “Five years,” said Karen
Duff. “Five years,” said Rudy Tanzi. All are actively working on novel, geneticallybased, Alzheimer therapies. And at Columbia, at the end of August, Larry Honig was
looking to sign up the first patient willing to participate in a clinical trial testing a method
of immunizing the body to produce antibodies that, in animals at least, reduced the
number of amyloid plaques in the brain.
“Ten years from now you’ll go to your doctor and she will look at your gene
profile and say ‘you know, you have three high risk genes for Alzheimer’s disease and I
would strongly suggest that when you turn fifty you get the Alzheimer’s vaccine,”
Mayeux said that day in the Adirondacks. “Right now you don’t know what the hell to
do. You don’t know whether you should take vitamins, whether you should take
ibuprofen, and if you do if you’ll get a stroke, whether you should take estrogen, and if
that will give you a stroke. People tell you to use your brain, to use your body, and those
are all well and good but you don’t know if it’s a lifetime of doing those things, you don’t
know if it’s starting to do crosswords when you’re ninety. If we can solve some of these
genetic puzzles, we’ll know how to treat the disease.”
It was, by then, three years since he, Hyslop, and Farrar, had begun their active
collaboration, and despite their successes—the statistically significant association of the
same gene in five independent Alzheimer populations and the functional studies that
showed that when the gene was knocked out of a cell more beta amyloid was produced—
Mayeux remained circumspect. There was still more work to be done: having found the
gene they now had to find the mutation, and they had to show how the mutation
influenced the course of the disease. It might be another few months of hard labor at the
very least before they could go public.
“Let’s say this is a real finding,” he said, avoiding the presumption of a
declarative sentence. “You can bet there will be a ton of work on how this particular gene
fits into the big picture. It’s really a jigsaw puzzle with 500 pieces. You can look at it
and see some of the key pieces and you can tell that there is a brain on a brown
background, but you can’t figure out where all the other pieces go. But then you get one
piece in there that fits and it helps you get a whole section together.
“We think we got a piece—it may be more—but you don’t know till you nail it.
What I feel best about is that the collections I’m making are going to be around for a
while. Collecting these Dominican families, putting the data together, having them very
well characterized, having the cell lines—if it’s not us who finds the gene, then someone
will find the genetic variant eventually and that will help.
“It’s a lot like the movie The Maltese Falcon. You look for, you look for, you
look for, and you find something--and then you realize it’s not the right thing, and by that
evening you’re booked on another ship to begin the next search. If this fizzles out, we’ll
be booked on that boat.”