“No,” she said calmly. “I don’t like shots. But … if you want some of my blood for DNA, that’s no problem.”32
A few days later, Beatrice turned five, though at twenty-eight pounds she could pass for much younger.
I had met father and daughter seven months earlier at Penn Station in Baltimore, an enormous and beautiful old Beaux-Arts edifice defiantly standing over a part of Charm City that had seen much better days. I stood just outside the massive revolving doors and watched as a good-looking fiftysomething man in a dress shirt and wire-rimmed glasses carried a blond child in a bright yellow raincoat in one arm, umbrella in the other, through a cold May downpour.
Hugh had brought Bea here to see Hal Dietz, her doctor at Johns Hopkins. Hugh knew the landscape well: he came from a family of Baltimore doctors. His father, Hugh Sr., broke the line by becoming a metallurgist and wanted his son to resist medicine as well; Hugh Jr. could not. In the early 1980s he was a genetics fellow at Hopkins under the tutelage of the father of twentieth-century medical genetics, Victor McKusick.33
When Bea was born, Hugh had long since traded in his full-time clinical and lab-bench vocations for biotech entrepreneurship and business consulting. “Working as a consultant is another way of saying you’re unemployed,” he told me. His current venture was called FerroKin Biosciences, a start-up developing a treatment for iron overload in anemia patients who had undergone multiple transfusions. He had kept his medical license current and for a while did pro bono work at an HIV clinic in San Francisco; now he volunteered at the city’s Department of Public Health. But since 2004 his passion—his obsession—had been trying to figure out what was wrong with his daughter.34
When Rienhoff’s wife, Lisa Hane, was pregnant with Bea, their third child, she was forty-two and at higher risk for having a baby with a chromosomal abnormality like Down syndrome. Her chromosomes—pictures of chromosomes are a fairly crude view of one’s DNA—looked normal. Lisa’s doctor looked elsewhere: one of the common ultrasound tests obstetricians run near the end of the first trimester is a nuchal scan, which measures the amount of fluid behind the neck of the fetus. More fluid is associated with a higher likelihood of a chromosomal problem and/or heart defects.35 Bea’s scan showed a high level of fluid. At nineteen weeks the couple got an echocardiogram to look for major heart problems; the doctors did not observe anything unwarranted on the ECG.
But with Bea’s emergence from the womb came the first moment of recognition for her father. “I saw her feet,” Hugh said. “Marfan syndrome flashed through my mind.”36 Marfan syndrome is a connective tissue disorder that affects multiple organs; patients may have enlarged aortas, severe nearsightedness, cataracts, and a swelling of the sac around the spinal cord, among many other features, including the long feet that Hugh noticed on Bea.37 A person with Marfan is typically tall and thin; there has been speculation that Abraham Lincoln might have had it.38 Without surgical aortic replacement, Marfan patients’ aortas may eventually weaken and rupture, leading to sudden death.39
Bea had other physical peculiarities, too, though they didn’t look like Marfan. She was floppy, she had a port wine stain (a large red or purple vascular birthmark) on her face, and her fingers were contracted. “I knew there was something going on,” Hugh said. “Were these things isolated or part of a syndrome? Was it genetic?” He tried not to think about it.40
Lisa was mostly oblivious until the three-month visit to the pediatrician, who told the couple that something was up. Bea was still floppy and she was not gaining weight. She would nurse but would consume only tiny amounts at one time. There was discussion about inserting a feeding tube. Lisa’s lowest moment came when a friend visited. “I said, ‘Do you think she’s okay?’ My friend said, ‘She’s just … really small.’ The way she said it made me think it was bad. When an ordinary person gave her true assessment of what she saw … Bea just looked so weak.”41
She did not appear to have classic Marfan syndrome: she lacked certain features, such as ocular problems. And she had others not typically associated with Marfan, such as severe muscle weakness, widely spaced eyes, and failure to thrive. The Rienhoffs went from specialist to specialist on the West Coast. Each one poked and prodded at Bea, ordered tests, and offered tentative diagnoses, none of which were satisfying to Hugh. One intriguing suggestion was that Bea had a form of Beals-Hecht syndrome, a rare Marfan-like condition characterized by crumpled ears, long fingers, contractures, and scoliosis.42 Hugh contacted Rodney Beals, who described the syndrome in 1971. But based on Bea’s hyperextensible limbs, he didn’t think she had it.43
Hugh reasoned like a clinician: his daughter needed a diagnosis. A diagnosis would suggest a management plan. A management plan would, he hoped, lead to weight gain for Bea and alert him and Lisa of what to be on the lookout for. Simply getting a muscle biopsy or ordering another round of tests was not a management plan. He was frustrated.
“I thought she had something in the Marfan family. She had long fingers, long feet, and a caved-in chest. That was a place to start. But I wanted her to have a good old-fashioned physical exam. She had to be seen by someone of the old school.”44
He brought Bea to medical geneticist Dave Valle, someone he knew slightly from his Johns Hopkins days. Unlike the other docs they’d seen so far, Valle gave Bea what Hugh called “a great exam. He knows his syndromes. He knows that if the patient has inverted nipples, then he’d better check the fat on the bottom of the butt. Things like that. Dr. McKusick called clinical geneticists ‘the last generalists in medicine,’ and I think that’s true. That’s Dave Valle.”45
Valle brought in two colleagues, Bart Loeys and Hal Dietz. Loeys was a genetics fellow; Dietz had been at Hopkins since his pediatrics residency twenty years earlier. He was part of a group that had identified fibrillin-1 as the causative gene in Marfan syndrome in the 1990s and probably knew more than anyone on the planet about the genetic basis of the disease.46 Like Valle, Loeys and Dietz examined Bea for clues to her condition. They looked at her uvula, the small piece of flesh that hangs from the palate. Hers was split: a bifid uvula, something that occurs in anywhere from 1.5 to 10 percent of newborns.47 They examined her widely spaced eyes and listened to her heart. When they were done they told Hugh that Bea needed an echocardiogram as soon as possible. As it happened, Hugh had already ordered one, thinking that perhaps a defect in blood flow between the chambers of the heart was causing Bea’s failure to thrive.48 But Loeys and Dietz wanted an echo for a different reason. They were worried about Bea’s aorta, the largest blood vessel in the body. If she had a Marfanoid aorta that was enlarged and weakened, then it could tear and kill her.
Okay … but if it were agreed that Bea didn’t have Marfan syndrome, then why the urgency? Loeys and Dietz gave Hugh a paper they had just published on manifestations of a Marfan-like disease described by and named after them: Loeys-Dietz syndrome. Loeys-Dietz patients had mutations in a gene in the same biochemical pathway as the mutations that caused Marfan syndrome, the transforming growth factor beta (TGF beta) pathway.49
I asked Hugh how he felt at that moment, knowing that a possible diagnosis for Bea had been found. “Terrible,” he said. “I was depressed. I read the paper and saw that the average age of death was twenty-six or twenty-seven. Loeys-Dietz was much worse than Marfan. I wasn’t expecting that Bea might have catastrophic vascular disease. Sometimes you want a diagnosis … but really you just want your daughter to be okay.”50 Hugh brought Bea home and nervously watched the cardiologist perform the echo. It was completely normal … The family could exhale. Meanwhile, the Johns Hopkins team sequenced Bea’s copies of the two TGF beta receptor genes that had been associated with Loeys-Dietz syndrome. They were clean. Hugh pondered these data: a normal echo, no mutations in the Loeys-Dietz genes, and the presence of extreme muscle weakness, a feature not seen in Marfan or Loeys-Dietz. In his mind, all of this added up to three strikes against these two syndromes. The best news was that Bea did not have any signs of life-threaten
ing vascular problems. The bad news was that Hugh was back to square one.
He went into hypothesis-generation mode. He read everything he could about the TGF beta pathway. Perhaps Bea’s phenotype was also the result of something gone awry in TGF beta signaling. But whatever it was, it would have to account for her muscle weakness. Bea could certainly walk, but anyone could see that climbing stairs was a challenge—she would use her hands to help propel herself (to descend, she often opted to slide down the stairs on her backside). Bea’s underdeveloped musculature led Hugh to a biochemical pathway related to TGF beta and centered around a protein known as myostatin. Myostatin’s normal function is to limit the growth of muscle.51 Mice that have been engineered without a myostatin gene develop extremely large muscles: they are known as “Schwarzenegger mice.”52 A few champion athletes have been found to carry myostatin mutations.53 Hugh reasoned that if Bea’s muscles were overproducing myostatin or something like it, then that might explain her weakness. Moreover, myostatin shared at least one receptor with TGF beta. “The more I looked at it,” he said, “the more I thought this pathway was a credible way to get what Bea has.”54
He began asking around for help. He talked at length to one of the people who cloned the myostatin gene in the 1990s.55 This scientist thought Hugh’s TGF beta hypothesis was reasonable, but he was reluctant to sequence Bea without Institutional Review Board approval—Bea was a human, after all, and entitled to the protections offered all human research subjects, even if Hugh was her dad. This meant that research results from a lab that was not clinically certified could not be returned to research participants or their families. Other requests from other laboratories were met with similar polite refusals.
“I thought, well, shit, I’ll have to do it myself,” said Hugh.56
He found surplus stores in the Bay Area that sold used lab equipment. He bought a PCR machine. He ordered primers and began setting up PCR experiments to amplify the relevant parts of the activin receptor genes that were known to interact with myostatin (activin receptors receive signals that tell cells to grow, divide, differentiate, and/or die). He was just about to buy a cheap DNA sequencer when a friend stopped him and said, “You’re crazy. Just send the samples to a core lab at a university. They’ll do it for four dollars per reaction.” He did.57
Within a few weeks sequence data began coming back. Hugh transferred it all to Word files on his computer. He then went to every genomic database to find all the DNA variants that were known in the activin receptors. Did Bea have any that had not been reported? If so, did Hugh and/or Lisa carry them? And if they did, then why were he and Lisa healthy?
After scouring the three genes he’d had sequenced, Hugh found a variant in Bea that had not been reported in other published human genomes (of which, admittedly, there were few). But he had still not had himself or Lisa sequenced. It seemed like an obvious experiment—so why not do it? And why not sequence the myostatin gene itself?
“Because by that time I’d kind of lost interest,” Hugh said. “I had a management plan. Bea would get regular echocardiograms. And for treatment we had losartan.”58 Losartan is a generic, FDA-approved drug that is commonly used to treat high blood pressure. For the last few years, it has been used to treat Marfan patients59 thanks to a fateful Google search performed by Hal Dietz.
When I suggested this version of events to him, Dietz laughed. “That’s not the whole story.” He came across as a mild-mannered and affable guy; he wore a rumpled sweater and small rimless glasses and had a receding hairline. When I said that I thought he had devoted his professional life to Marfan syndrome, he gently corrected me again. “Certainly Marfan is an important part of what I do, but it’s not the majority. I would categorize [my work as being dedicated to] anyone who has a problem in the first three centimeters of his or her aorta.”60
In 2001, Dietz’s lab found that mice with Marfan mutations developed emphysema, something that occurred in about 10 percent of Marfan patients. The conventional wisdom was that the emphysema was the result of deterioration of the patients’ lung tissue over time. But Dietz’s team noticed lung problems very early in the Marfan mice’s development, which meant there must be another explanation. They suspected that too much TGF beta was the culprit. Dietz’s hypothesis was that fibrillin-1's normal function is to keep TGF beta in check in the extracellular matrix, the scaffolding that supports our cells and is a defining feature of our connective tissue. In Marfan patients, the thinking went, fibrillin-1 is disabled and TGF beta runs wild, leading to enlarged aortas and other problems in places where fibrillin normally keeps TGF beta tamped down, such as in the lens of the eye. The Johns Hopkins group showed that antibodies to TGF beta could prevent heart valve problems and aortic aneurysms in Marfan mice.61
But could they also prevent lung disease? This was an obvious follow-up experiment, but Dietz wanted a more clinically relevant way to suppress TGF beta. Like, say, with a drug. One of his postdocs did some digging and learned that several blood pressure medications also inhibited the growth factor. Dietz then went to Google and typed in “TGF beta antagonist, FDA-approved.” Losartan was at the top of the list. At six months, Marfan mice given a placebo or a traditional beta-blocker had severe aortic aneurysms; they were in bad shape. The losartan-treated mice, on the other hand, looked completely healthy. The drug had actually reversed the aortic damage and had positive effects on lung tissue and skeletal muscle.62 “It was beyond anything I could have anticipated or hoped,” Dietz told Science.63
In some cases, discovering that a drug has profound effects in mice might be sufficient to get a grant funded or drive a stock price up, but rarely does it have immediate ramifications for patients. Losartan was different: it was already FDA-approved and had been used safely and effectively in millions of people, including children, for nearly two decades. On top of that, it was used to lower blood pressure, which was already known to be a good thing in Marfan patients. Dietz and his colleagues began offering losartan to a small number of Marfan kids with severe aortic problems who had not responded to other drugs. A preliminary study by the Dietz group on eighteen kids with Marfan looked promising: once on losartan, their rate of aortic enlargement slowed dramatically. A fullblown clinical trial began, but it would be years before definitive results were in; meanwhile the buzz about the drug was so strong in the Marfan community that some patients opted not to participate for fear that they would be assigned to the control group and not receive losartan. Some obtained losartan prescriptions from their own doctors.64
For the Rienhoffs, even though Bea did not have Marfan, losartan offered hope. “Two things were compelling to me,” Hugh said. “First, we had not ruled out that Bea was at risk for vascular disease. She will be getting echocardiograms for a long time to come. Losartan actually arrested the vascular disease of the Marfan mouse! Second, I looked for cases where people taking this family of drugs had an increase in skeletal muscle mass; I found a bunch of papers supporting that. It seemed to me that, with this drug, Bea could get prophylaxis [prevention of aortic problems] on one hand, and therapy [an increase in muscle growth] on the other.”65
Dietz, who tended to err on the side of caution, was somewhat lukewarm to the losartan-for-Bea idea, but he was prepared to go along with it given Hugh’s research and rationale. But it wouldn’t have mattered: Hugh had already made up his mind. “I didn’t ask for Hal’s permission. I only asked him for the correct dose.”66
I drove to Hugh and Lisa’s place about thirty miles south of the San Francisco airport in San Carlos, a bedroom community with Spanish street names and two-story houses on windy culs-de-sac. Their old farmhouse seemed more Virginia than California—hardwood floors, high ceilings, and a wraparound porch. They had three kids: Colston, then age ten; MacCallum (“Mac"; age seven); and Beatrice. When I showed up in December right before Bea’s fifth birthday, no one was home; a neighbor let me in with little to no suspicion. Inside, amid the artwork, children’s furniture, and tasteful decor, st
uff was strewn everywhere: books, toys, mail. I felt right at home.
An hour later Hugh showed up wearing a bow tie—his doctor outfit, I reckoned, though he only saw patients a couple of days a month. He greeted me warmly, kicked off his shoes, and insisted on carrying my bag upstairs. We retired to his carpeted office in the attic: the Bat Cave, he called it. Piles of paper, notebooks, journal articles, and books littered the floor. Pro-Obama and anti-Bush stickers adorned the wall; Lisa had been a socialist labor organizer when she and Hugh met in the 1990s. Another sticker said, “Evolution is God’s Intelligent Design.” The room featured two computers. One was for all of the mundane business of being a serial entrepreneur. The other was for “My Daughter’s DNA,” as Hugh called it, or what I will call, for the sake of brevity, “Project Bea.”67
Management plan or not, Hugh had not been able to let go of Project Bea. Since I’d last seen him he had resumed his molecular detective work in a massive way. Sequencing three genes was nothing. Now, loaded onto his computer, he had tens of thousands of gene sequences from himself, Lisa, and Bea. Each had given blood. From those samples, someone at Illumina, the market leader in the new DNA sequencing technology, had isolated white blood cells and from those extracted RNA and enzymatically converted it to its DNA form known as complementary DNA, or cDNA.* The resulting collection of genes from Bea and her parents were their three transcriptomes: the sets of all of the active genes (transcripts) produced in a particular type of cell, in this case white blood cells. Transcriptomes were easier to work with than whole genomes because 1) they represent only 1 to 2 percent of the genome, “only” 60 million plus base pairs instead of 3 billion; and 2) they include much if not most of the “business end” of the genome: the parts that instruct the cell what proteins to make.
Here Is a Human Being Page 20
