For a while, Rienhoff wondered whether an intestinal parasite was gobbling up Beatrice's food before she could digest it. Later he suspected that she might have a mitochondrial disease. He started her on a regimen of coenzyme Q10
, an over-the-counter treatment. There was no improvement.
Rather than visit yet another Bay Area physician, Rienhoff emailed an acquaintance from his Johns Hopkins days, David Valle
, director of the Institute of Genetic Medicine at the university. Valle agreed to see Beatrice, so father and daughter flew to Baltimore in March 2005.
Beatrice was examined by Valle and a bow-tied colleague named Tyler Reimschisel
. To Rienhoff's puzzlement, the two geneticists spent an inordinately long time peering down Beatrice's throat. Eventually they called for a third doctor, a Belgian named Bart Loeys
, to join them.
Loeys seemed gravely concerned by what he saw in Beatrice's throat. "I have a pretty good idea what this might be," he said. "We need to do an echocardiogram right away—today, this afternoon."
Loeys explained that Beatrice was likely suffering from a disorder named after himself and Harry Dietz
, yet another Johns Hopkins geneticist. Loeys-Dietz syndrome
is characterized by many of the same symptoms as Marfan and Beals syndromes, but the mutation is not found in the FBN1
genes but rather in two receptor genes for transforming growth factor beta, or TGF-▀
. (A receptor is a structure on a cellular wall that provides a binding site for molecules.) All of these genes play vital roles in the same metabolic pathway, the TGF-▀
signaling pathway, which regulates the growth and proliferation of cells.
One telltale sign of Loeys-Dietz is a forked uvula
—the cone-shaped blob of tissue that hangs at the entry to the throat. Other examiners had noted that there was nothing peculiar about Beatrice's uvula, but the Johns Hopkins team discovered otherwise.
By coincidence, Rienhoff had already scheduled an echocardiogram back in San Francisco for later in the week. He promised to keep Loeys apprised of the results; Loeys, in turn, gave Rienhoff a paper on the syndrome that had recently appeared in the scientific journal Nature Genetics.
Most of the plane ride home was spent playing games with 15-month-old Beatrice, whose illness has had no effect on her intellectual development. But toward the end of the flight, Rienhoff finally managed to scan the Loeys-Dietz paper. It was a horrifying read: The syndrome, which warps the aorta and twists the arteries, is far deadlier than Marfan. The average patient dies before the age of 27. "Even though I'm not a religious guy," Rienhoff says, "I was praying that she didn't have it."
She did not. The echocardiogram showed that Beatrice's aorta was fine. Follow-up genetics tests confirmed she didn't have the specific TGF-▀
receptor mutations characteristic of the disease, either.
But after the Loeys-Dietz scare, Rienhoff became convinced that the TGF-▀
signaling pathway must be involved in Beatrice's case. She now had verified symptoms from three diseases linked to abnormalities in the pathway—the Marfanoid feet, the clenched fingers from Beals, and the bifurcated uvula from Loeys-Dietz. But the problem most affecting her day-to-day quality of life was simple weakness. Rienhoff was determined to figure out whether there was any connection between skeletal muscle and TGF-▀
He turned his attention to myostatin, a protein that prevents muscles from growing too wildly and which is a chemical cousin of TGF-▀
. Mice stripped of myostatin become rodent versions of Arnold Schwarzenegger circa Pumping Iron. Perhaps Beatrice was suffering from a genetic defect that was affecting her myostatin production.
As it happens, Rienhoff knew the man who discovered myostatin, a Johns Hopkins molecular biologist named Se-Jin Lee
. The two had first crossed paths during Rienhoff's residency. "He called to describe his daughter and her circumstances," Lee says, "and told me that he was thinking about the role of myostatin in the TGF-▀
pathway." Lee sent Rienhoff a journal article on the protein; he was amazed at Rienhoff's voracious appetite for the complex topic.
Convinced his daughter's illness was connected to the protein myostatin, Hugh Rienhoff decided to examine the genes that affect myostatin production himself.
Rienhoff took vials of Beatrice's blood to a Stanford lab, where he extracted the girl's DNA by spinning the blood in centrifuges.
Back home, Rienhoff ran the DNA through a polymerase chain reaction machine. It starts by heating the samples, splitting the double-stranded DNA into two single strands.
Rienhoff applied chemical primers to the split DNA. These primers, which he designed and purchased online, flagged the genes Rienhoff was targeting. The PCR machine then used a type of enzyme called a DNA polymerase to amplify just the marked segments.
Rienhoff sent 50-microliter tubes of the amplified DNA to a contract lab, which posted the sequence data on a secure server.
Rienhoff printed out the dataŚ20,000 base pairs of DNAŚand compared it with a reference genome stored on Ensembl
, an online British database.
In the spring of 2006, the Rienhoffs made another trip to Baltimore to meet with Harry Dietz, who remained intrigued by Beatrice's case despite her normal echocardiogram and genetic tests ruling out Loeys-Dietz disease. Dietz still worried that the girl might have a fatal heart defect, and he asked Rienhoff to schedule a CT scan of her cardiovascular system.
Rienhoff balked; CT scans use x-rays, and Rienhoff was reluctant to expose Beatrice to a large dose of radiation. The two men compromised: In lieu of a CT scan, they scheduled another MRI. And they deferred the appointment until the following January.
But Rienhoff didn't like the MRI plan either—because of her age, Beatrice would have to undergo general anesthesia to ensure she remained still in the machine. He had about eight months to come up with a good reason to cancel the appointment. He focused more keenly on myostatin.
There are three receptors, known as activin receptors, that are thought to be critical to myostatin regulation: ACVR1B
, ACVR2, and ACVR2B
. Rienhoff wondered whether one of these genes was flawed in Beatrice, causing her body's myostatin system to go haywire. There was one way to find out: Examine her DNA. Activin receptors are fairly well-understood genes, so a problematic alteration would likely stand out.
Rienhoff first asked Lee to do the DNA sequencing. But Lee declined on bureaucratic grounds. Professors who wish to work with human genetic material need clearance from their university's institutional review board. This onerous approval processes can take months, and there was no guarantee Lee would get the OK.
Rienhoff was also turned down by several other doctors, including some acquaintances, who were frankly troubled by the request. They pointed out that myostatin had never been associated with human disease, so Rienhoff's hypothesis was far-fetched, bordering on foolish. The sequencing venture struck them as futile. "There were a few people who didn't want to get involved," Rienhoff says. "They think it's peripheral or wacky science."
Some skeptics were also disturbed by the notion of a father conducting, in essence, an experiment on his child: "When you start telling people it's your daughter, it weirds them out," Rienhoff says. "They're thinking that maybe you're going over the line a bit in your zealousness, maybe you're sacrificing your commitment to good science. "
By late 2006, it was clear to Rienhoff that if he wanted to get Beatrice's activin receptors sequenced, he'd have to do it himself.
As any fan of CSI knows
, polymerase chain reaction is a method for replicating a snippet of DNA, amplifying it over and over until there's enough genetic material to be sequenced. By making inquiries with local surplus brokers, Rienhoff discovered he could buy a secondhand PCR machine for less than a MacBook. He ended up purchasing a full working model for just $750.
Obtaining additional supplies, like the PCR reagents, for his experiment was tougher. Some chemical companies didn't want to ship to a private address, so Rienhoff pretended his house was the headquarters of the fictional Institute for Future Study.
Rienhoff went to friend's lab at Stanford and used the centrifuges there to extract DNA from a sample of Beatrice's blood. He then took the genetic material home and put his PCR equipment to work. The machine first blasted the genetic material with heat, splitting the double-stranded DNA into two separate strands. Then chemical primers, which Rienhoff had designed and purchased on Integrated DNA Technologies' publicly accessible PrimerQuest Web site, were added to the mix. They were coded specifically to amplify the genes that Rienhoff was focused on—the three activin receptors. This process, repeated many times over, created millions of copies of Beatrice's activin receptor genes, giving him a sample large enough for reliable sequencing.
While Rienhoff could spring for his own PCR machine, a used gene sequencer (assuming he could find one) would cost around $100,000. So he found a university lab (which he declines to identify) that would sequence the genes he had amplified, for $3.50 per 50-microliter sample. In spring 2007, Rienhoff mailed in more than 200 samples.
The sequencing results came back organized into charts resembling seismographs of a minor earthquake, with each sharp spike indicating an individual nucleotide. Rienhoff printed out everything and stored the sheets in white three-ring binders. He combed through these pages while logged on to Ensembl, a public database jointly funded by the UK's Wellcome Trust Sanger Institute and the European Bioinformatics Institute. Launched in 2000, Ensembl contains complete genomes for about 50 species, from alpacas to zebrafish. The Homo sapiens section is based on a full assembly of the genome compiled by the National Center for Biotechnology Information
; this data is supplemented by annotations submitted by researchers, who identify genetic variants suspected of causing diseases. Rienhoff compared Beatrice's DNA with the information on Ensembl, looking for any base-pair variants that hadn't been previously recorded on Ensembl. He was operating on the assumption that Beatrice's genetic blip was completely unknown, which would explain why she'd been so hard to diagnose.
The job was daunting: The printouts contained data for approximately 20,000 base pairs, and there was no feasible way to automate the hunt for variants. After putting in a full day of consulting and then getting the kids to bed, Rienhoff would retire to his attic and spend hours checking Beatrice's adenines (A), thymines (T), guanines (G), and cytosines (C) against the Ensembl reference genome. Sometimes he would fall asleep on the floor, only to awaken at 1 am and go right back to work. He would imagine Beatrice sitting by his side, patiently observing as he sorted through the data that defined her.
Progress was slow, but after a decade in the boardroom, Rienhoff enjoyed doing pure science again. Mostly, though, it felt good to be an active seeker rather than an anxious parent at the mercy of an overburdened doctor. The process of combing through Beatrice's DNA printouts became a form of meditation, a way for Rienhoff to steel himself against the sadness of seeing his little girl struggle up and down the stairs each day. It began to feel quasi- spiritual—he started referring to his work as "my journey."
Confident that he was making headway, Rienhoff canceled Beatrice's MRI. Two months later, in March 2007, he completed his study. He had identified about 20 places where the DNA for Beatrice's activin receptors didn't match the reference genome. Of those, only one, in the ACVR1B
gene, was not mentioned anywhere in the genetic literature. As far as Rienhoff could tell, the variant was unique—an adenine on one chromosome and a guanine on the other, whereas only two adenines had ever been reported.
But the variant was far removed from the area most obviously involved with myostatin. It seemed doubtful that such an oddly placed hiccup could cause Beatrice's serious disorder. Rienhoff decided to look at his own ACVR1B
gene to see how it compared. He later repeated the process with his blood, from centrifuge to sequencing; when the results came back, he discovered he also had an A and a G in that position. This ran contrary to Rienhoff's theory—a variant in that location couldn't cause an illness as severe as Beatrice's, if he was in perfect health.
Though his homebrew genetic analysis had been a bust so far, Rienhoff still believed that Beatrice's disorder was linked to TGF-▀
. And there was a piece of good news: During his research, he read a study in which Marfan mice were given losartan
, a common blood-pressure medication. The drug was intended to prevent aortas from dilating, but an Italian group had discovered that it might strengthen skeletal muscle, too. Rienhoff persuaded Beatrice's cardiologist to start her on losartan, figuring that they had little to lose—the potential side effects were minor; the potential upside huge.
Beatrice soon started to move with more confidence. It's difficult to say exactly why the losartan worked, or even if it was, indeed, the reason for her newfound grace. But Beatrice's physical therapist, unaware of the treatment and surprised by her improvement, pulled Rienhoff aside after one session and asked, "Has anything changed?"
In July 2007
, a few months after wrapping up his analysis of Beatrice's activin receptors, Rienhoff attended the National Marfan Foundation's annual conference
in Palo Alto. His main interest was a workshop dedicated to people like himself, parents whose children exhibited Marfanoid symptoms but did not have the disease.
The parents took turns introducing themselves and describing their various frustrations—insurance companies that wouldn't pay for genetic tests, doctors who prescribed unsuitable treatments. When Rienhoff's turn came, he spoke about his saga, from the early days of the Beals syndrome hypothesis to his sequencing of Beatrice's three activin receptors.
The attendees were both amazed and eager to hear more about Rienhoff's genetic exploration. But the doctor moderating the session, Dianna Milewicz
, was apprehensive. Director of the division of medical genetics at the University of Texas Medical School at Houston, Milewicz couldn't fathom why Rienhoff would pursue such a long-shot experiment. For starters, his focus was on genes that had never been shown to play a role in Marfan-like diseases, rather than more probable culprits such as fibrillin-2. And even if he succeeded in identifying a genuinely novel variant, that was only a tiny first step in a long diagnostic process.
"Then you have to prove the variant you're seeing is truly causing the disease," Milewicz says. "Understanding how a variant relates to a disease is often very difficult. And if you're sequencing genes that we don't even know cause disease, that's where it becomes a nightmare, even for those of us who have dedicated our lives to this."
As soon as Rienhoff wrapped up his presentation, Milewicz cautioned the workshop's attendees against following his do-it-yourself lead. "I said, 'This is not a good idea to recommend to parents,'" she recalls. "They're not trained to analyze diagnostic tests."
Rienhoff bristled at Milewicz's dismissive tone. "I remember thinking, 'Who the **** is this person?' I have never been in a situation where it was so obvious that a doctor had contempt for the curiosity of her patients. It was striking how insensitive she was to their dilemma."
Such a *****ly reaction is out of character for Rienhoff, a soft-spoken man who normally exudes an easy calm. But he has developed a cynical streak about doctors, especially those who are quick to dismiss inquisitive parents as nuisances. "Medicine in general is a slightly paternalistic activity," he says. "You hear these stories about patients bringing in all sorts of information from the Internet and doctors being exasperated. And part of that is because there is so much they don't know, and they're supposed to be omniscient."
Far from taking Milewicz's warning to heart, Rienhoff was preparing to do something even more radical than his activin experiment. In summer 2008, he launched a second, more ambitious phase in his effort to diagnose Beatrice—he started sequencing her transcriptome
The transcriptome, as Rienhoff puts it, is "the business end of the genome." In order to produce proteins, a small portion—about 1 percent—of the DNA in a particular gene is transcribed into messenger RNA. This mRNA, also known as a gene transcript
, then passes along instructions to ribosomes, the cellular machines that make proteins.
Analyzing a transcriptome is essentially a low-cost alternative to sequencing a person's entire genome. It provides a glimpse at small yet vital scraps of thousands of genes. And since the transcriptome represents only a cross-section of a person's DNA, there shouldn't be too many variants to sift through. A typical human genome will have between 100 and 300 mutations that weren't present in the subject's parents; a transcriptome should have five or fewer.
Rienhoff asked a lab to extract mRNA from Beatrice's white blood cells, in which more than half of human genes are expressed. Then a reverse transcriptase enzyme was used to convert this single-stranded RNA into double-stranded DNA, suitable for sequencing. The procedure yielded fragmentary snapshots of as many as 15,000 of Beatrice's genes. (A full human genome contains 20,000 to 25,000 genes.) As a last step, Rienhoff repeated the process for himself and his wife; he figured that the mutation in Beatrice's genome must have arisen spontaneously, rather than being inherited from either parent.
"Hardcore scientists would say this is a bad experiment," Rienhoff admits. "You're sampling only a portion of the genes—you're not casting your net over whole oceans; you're just casting it over the Pacific Ocean. But from my point of view, that's a pretty damn good start."
Rienhoff is still in the midst of this project, which he dubbed his triad transcriptome experiment—the "triad" consisting of mother, father, and offspring. He has tentatively identified three genetic variants so far that Beatrice alone possesses. To his consternation, none are directly related to the TGF-▀
signaling pathway: One is in a gene that codes transport proteins; one is in CRNKL1
, a gene that's barely been studied in humans; and the third is in a gene so obscure that it has yet to be named.
One of Rienhoff's heroes is Borgny Egeland
, a Norwegian mother of two mentally retarded children. In 1934, she contacted an Oslo doctor, Asbjörn Fölling
, and informed him that her children's urine was emitting a powerfully musty odor. At Fölling's request, she collected urine samples from the children every other day for two months straight—more than 5 gallons in all. Fölling found that these samples contained high levels of phenylpyruvic acid, a substance that was subsequently found in the urine of many other patients suffering from retardation. It turned out that a genetic disorder was preventing these children from properly metabolizing phenylpyruvic acid, and the buildup causes brain damage. The discovery led to the most effective treatment for the disease (now called phenylketonuria
)—a diet low in phenylalanine, which the body converts to phenylpyruvic acid. To Rienhoff, the story of PKU's discovery is a perfect example of how a concerned parent can press doctors to solve seemingly intractable problems. Countless afflicted children now live normal lives thanks to Egeland and Fölling's teamwork.
To encourage this kind of collaboration, Rienhoff created MyDaughtersDNA.org
. The Web site, launched in October 2007, invites parents to post the clinical histories of their undiagnosed children. The hope is that geneticists will also frequent the site, to help identify rare disorders and use the case studies to further their own research.
One of the first people to post to the site was a Bulgarian man named Stefan Petkov
, who wrote about his 12-year-old daughter. The girl had weak limbs and speech problems; she also lacked the ability to shed tears. Doctors in Bulgaria were stumped, and DNA tests done in both Belgian and Bulgarian laboratories didn't point to any known genetic conditions.
, a geneticist at Saint Louis University School of Medicine in Missouri, came across Petkov's saga after reading about MyDaughtersDNA.org in the journal Nature. The girl's inability to cry made him think of a disorder known as Triple A syndrome. He suggested that Petkov have his daughter checked for the disease.
Less than a month later, in March 2008, Petkov wrote again
. His daughter had finally been diagnosed with two of the pillars of Triple A syndrome, the most notable being Addison's disease, which limits production of the hormone cortisol. Shortly thereafter, Petkov's daughter began receiving medicine to replace the hormone, which should help build her muscle strength.
Rienhoff realizes that most parents lack the resources to operate PCR machines or jet around the country in search of a diagnosis. But he points to Petkov's story as proof that they can still play a role in their children's care by seeking answers outside traditional medical channels—embarking, in a sense, on their own Rienhoff-like journeys.
Rienhoff's personal quest has yet to produce any meaningful insight into Beatrice's condition, but his efforts have not been totally in vain. At the very least, he claims to have now conducted more work on human transcriptomes than anyone else. According to Rienhoff, there have been only two peer-reviewed papers on human transcriptome sequencing, and both have their shortcomings. "Certainly, no one has done the triad," he says, "which is the Holy Trinity of genetics." A genomics lab at UC Berkeley wants access to Rienhoff's data and is in the process of obtaining review board approval to examine Beatrice's transcriptome.
Rienhoff hopes that someone will eventually commercialize his triad transcriptome experiment, so that parents like Petkov will have a quick if crude tool for peering inside their children's genetic code. But such a diagnostic product may soon be moot: Complete Genomics
, a Mountain View, California, company, recently announced that it plans to start sequencing entire human genomes for $5,000. Just a few years ago, such a massive undertaking couldn't be done for less than $1 million.
But even when labs can sequence a person's genome for less than $1,000, it will still likely be years before such tools help children suffering from complex syndromes. It's one thing to process vast amounts of DNA data; quite another to turn that data into insight about an illness—let alone use it to develop therapies. "Certainly, our science cannot help Beatrice anytime soon," says Marc Vidal
, a Rienhoff confidant and director of the Center for Cancer Systems Biology at the Dana-Farber Cancer Institute.
Deep down, Hugh Rienhoff knows this. And he realizes that his DIY search for a diagnosis may never pan out. "I never pretend to know what's going on, because I don't," he says. "You can see—anybody can see if they look closely enough—that I don't have the answer to the story."
But he keeps hacking his way through Beatrice's genome, mostly because it's the only way he has of feeling some measure of control over an uncontrollable situation. Rienhoff compares his work on Beatrice's DNA to journalist Peter Matthiessen's search for the snow leopard
, documented in an award-winning book. Matthiessen never found the big cat in the mountains of Nepal, but his futile quest helped him come to terms with his wife's death from cancer.
The altered nucleotide at the root of Beatrice's problems "is as elusive and mysterious as the snow leopard," Rienhoff says. "And like Matthiessen, I may not find it. Curiously, it may not be important at all."
The journey, in other words, is its own reward. Of course, Rienhoff knows he can afford to be philosophical because Beatrice's condition isn't dire, at least not at the moment. "I am not watching my child melt away, which is the most desperate situation to be in," he says. "I can't even imagine that—I'd be working 24 hours a day."
Contributing editor Brendan I. Koerner
) is the author of
Now the Hell Will Start: One Soldier's Flight From the Greatest Manhunt of World War II.
http://feeds2.feedburner.com/~f/wire...eadlines?d=131 http://feeds2.feedburner.com/~f/wire...nes?i=MbeOl7WI http://feeds2.feedburner.com/~f/wire...nes?i=mZv0fm1i http://feeds2.feedburner.com/~f/wired/topheadlines?d=41