Editing Humanity: The CRISPR Revolution and the New Era of Genome Editing

CHAPTER 10 THE RISE AND FALL OF GENE THERAPY

It took more than twenty years for French Anderson’s worthwhile speculation to become a clinical reality. But the first tentative steps, albeit misguided, began shortly after his spurned manifesto. Stanfield Rogers, a physician at the Oak Ridge National Laboratory in Tennessee, had long been advocating the use of viruses to transmit genetic information. He had found that researchers handling the Shope rabbit papillomavirus^I had lower levels of arginine than normal people, suggesting they were picking up the virus and supplemental activity of the viral enzyme called arginase, which breaks down the amino acid. Rogers had reported high levels of arginase in warts on the skin of rabbits infected with the virus, and speculated that the virus was a therapeutic agent in search of a disease. “The possibility of tying specific synthetic DNA information on to the genome of passenger viruses, thereby using viruses as a vector, could prove to be a useful technique,” he suggested.¹

Rogers got his chance after reading a report in the Lancet about a pair of young, mentally retarded German sisters who had a rare inherited disease called argininemia—excess arginine in the blood caused by arginase deficiency. Believing he could supplement the missing enzyme using the Shope virus, Rogers persuaded the girls’ pediatrician to let him try a ludicrously premature experimental procedure—the first human genetic engineering experiment. In 1970, Rogers flew to Germany and injected small doses of the virus into the two girls, hoping to boost levels of the enzyme. There was no response. Later a third sibling received the virus, only to develop an allergic reaction. His reckless gene therapy adventure over, Rogers went back to studying plant viruses.

Two years later, Theodore (Ted) Friedmann and Richard Roblin published a commentary in Science entitled “Gene Therapy for Genetic Disease?”² Friedmann, a physician, is widely credited with coining the term “gene therapy.” Friedmann was born in Vienna but fled with his family to the United States in 1938 to escape the Nazis. At the University of Pennsylvania, he attended lectures by Colin MacLeod, who with Oswald Avery had proven in 1944 that DNA was the genetic material. He later trained with Fred Sanger in Cambridge before joining the NIH.

Friedmann worked on Lesch-Nyhan syndrome, a debilitating, sex-linked genetic disorder in which affected boys suffer retardation, abnormal movements, and self-mutilation. Friedmann was able to correct cells from Lesch-Nyhan patients using gene transfer by replacing the DNA that codes for the key enzyme. The experiment was terribly inefficient—only about one cell in a million was corrected—because Friedmann was using a full genome’s worth of DNA (this was years before the ability to isolate specific genes).

Friedmann admired the work of Renato Dulbecco, who had just discovered that a tumor virus did exactly what gene therapists wanted to do, “taking a foreign piece of genetic information, a foreign DNA, and inserting it into a cell and forever changing that cell.”³ Viruses could indeed be used to ferry normal copies of genes into cells carrying a broken version of the same gene. Friedmann helped popularize the concept of using modified viruses for gene therapy, while warning of the ethical dangers of pushing ahead too quickly. “Gene therapy may ameliorate some human genetic diseases in the future,” he wrote. The idea of gene replacement therapy using viral vectors had just received a major shot in the arm.

This approach was exciting but why be content to just add a healthy gene, papering over the cracks in the genome as it were, rather than actually trying to repair the broken sequence? In 1978, the same year as he won the Nobel Prize, David Baltimore offered one approach to this medical milestone. A patient with a blood disease like hemophilia or sickle-cell could be treated by transferring a normal gene into the patient’s bone marrow stem cells that ultimately give rise to blood cells. This would allow a normal protein to replace (or be made alongside) the faulty protein and cure the patient’s disease. “It is likely to be the first type of genetic engineering tried on human beings, and might be tried within the next five years.”⁴

“The concept of repairing a defective gene such as the sickle-globin gene is appealing,” wrote one physician, “however, existing technology cannot direct a site-specific recombinational event. Therefore, the concept of gene-repair in a genome as complex as that of man is for the moment impractical.”⁵ The author of those words was a UCLA hematologist named Martin Cline.

In 1979, Cline proposed treating patients with beta thalassemia with gene therapy, but a UCLA review committee insisted on additional animal experiments. Frustrated, Cline looked overseas, and in June treated two young women—a twenty-one-year-old at Hadassah Hospital in Jerusalem and a sixteen-year-old in Naples, Italy, a few days later. The process involved extracting some of the patient’s bone marrow, transfecting the cells with the beta-globin gene, and infusing about 1 billion treated cells back to the patient following irradiation of their femur. Cline told the women that the chances of success were slim, but he felt compelled to try. “When do you consider animal experiments adequate?” Cline asked. “When do you feel ready for a transition [to man]? Here’s a patient who has a life-threatening disease with a limited life expectancy and no options with modern treatment. Is now the time to try an experimental treatment?”⁶

In the opinion of the NIH and most of Cline’s peers, the answer was emphatically no. Cline had taken it upon himself to conduct the first recombinant genetic engineering experiments on humans. Following censure by the NIH, UCLA’s dean of medicine accepted Cline’s resignation as chief of the oncology department in February 1981. He chastised Cline for conducting an unprecedented experiment on two patients without the necessary institutional approval. Although no medical harm had been done, he continued, “the freedom to conduct experiments of benefit to mankind is jeopardized by failure to act in accord with the relevant regulations.”⁷ Hematologist Ernest Beutler called the Cline episode tragic “because it interrupted the efforts of a highly talented, productive scientist who was in too much of a hurry to see patients benefit from the marvels of modern molecular biology.”⁸ Beutler softened his criticism when he judged that the patients were probably more at risk from the three hundred rads of ionizing radiation they received than the therapy itself.

Two years after the Cline affair, Bob Williamson, a prominent gene hunter in London at the time (and my PhD supervisor) argued in Nature that gene therapy, while not on the immediate horizon, “will be possible in the future, and it should be considered now, before the headlines break on us all.”⁹ But whereas Williamson preached caution, Anderson felt it would be unethical not to embark on human trials once issues of safety had been met. In 1984, he wrote:

[A]rguments that genetic engineering might someday be misused do not justify the needless perpetuation of human suffering that would result from an unnecessary delay in the clinical application of this potentially powerful therapeutic procedure.¹⁰

Despite increasingly vocal opposition from anti-biotech campaigners led by Jeremy Rifkin, the realization of gene therapy had turned from a matter of if to when. “Egos and expertise will clash like cymbals as the technology of gene splicing keeps racing along so fast that it laps ethical debates about what it all means,” wrote Jeff Lyon and Peter Gorner in Altered Fates.¹¹

The year 1990 was an annus mirabilis for human genetics, notably the launch of a fifteen-year, $3 billion enterprise known as the Human Genome Project under the leadership of Jim Watson to build the definitive roadmap to identify the locations and identities of all genes, including those underlying the most devastating genetic diseases. In October 1990, I savored a taste of things to come as thousands of scientists traveled to Cincinnati, Ohio, for the annual conference of human geneticists. Late one evening, with the World Series featuring the hometown Cincinnati Reds blaring in every hotel bar, UC Berkeley’s Mary-Claire King hit a walk-off home run. In front of a standing-room-only crowd, she announced that she had mapped an errant gene, BRCA1, that when mutated increased a woman’s risk of breast cancer.^II

Identifying human disease genes was big news, and the genetic detectives commanding those search expeditions like King and Collins and Lander became scientific celebrities. Identifying the genes mutated in CF or DMD or hereditary breast cancer marked a transformation in medical diagnostics and immediately raised hopes for a successful drug or gene therapy. The genomics gold rush spread to Wall Street, but identifying a disease gene is just the start: it takes on average a decade and $1.3 billion to bring a drug to market, and even then success is not guaranteed.¹² The molecular basis of sickle-cell disease was identified in the 1950s but the disease is still incurable sixty years later. It took more than twenty-five years since the discovery of the CF gene in 1989 for Vertex Pharmaceuticals to develop a drug that successfully treated a subset of patients with a particular gene mutation.

Shortly before King’s walk off in Cincinnati, a group of clinicians at the NIH took a major step on the path to gene therapy by treating a young girl with a rare genetic disease. If a disease is caused by a typo in the genetic code, then the most logical way to treat that disease is to introduce a normal copy of the same genetic sequence. If the disease was caused by a faulty nonfunctional gene, then why not just replace the gene? A gene transplant, if you will. After two decades of false starts and deliberation, gene therapy was about to get real. But as generations of gene therapists can attest, it is anything but easy.

After three years of fierce debate, French Anderson finally won approval to start the first official gene therapy trial in the United States. Doctors and nurses gathered in the pediatric intensive care unit of the NIH Warren Grant Magnuson Clinical Center. It took all of twenty-eight minutes for the first history-making infusion to take place after the final paperwork was completed earlier that morning.

On September 14, 1990—12:52 P.M. to be precise—Kenneth Culver took a small syringe and injected some liquid into the left arm of four-year-old Ashanthi de Silva from Cleveland. Wearing a white top and turquoise trousers, Ashanthi was a model patient, distractedly sticking cartoon stickers on her doctors’ lab coats. About one billion genetically modified T cells slowly flowed into her body. “She was wonderful, a lot calmer than I was,” said lead investigator Michael Blaese.¹³ Ashanthi suffered from a rare form of severe combined immune deficiency syndrome (SCID), caused by a deficiency of the enzyme adenine deaminase (ADA). Only about a dozen children are born with this recessive disorder each year in the United States. Ashanthi had been sick for most of her life—about as long as the bureaucratic wrangling over the trial.

This was Anderson’s medical Mount Everest, a scientific summit he had wanted to conquer since his audacious manifesto was dismissed by the medical establishment two decades earlier. As a student, inspired by the double helix and Roger Bannister’s four-minute mile, Anderson had two goals in life: “I was going to be in the Olympics and I was going to cure defective molecules.”¹⁴ His first goal had already come true: he was a team physician for the U.S. Olympic team in Seoul in 1988.¹⁵ As Anderson, Blaese, and Culver exhaled after Ashanthi’s treatment, Anderson could reflect on twenty-five years of hopes and dreams. “At long last, the great adventure has started,” he said.¹⁶

Four months later, Anderson’s team began treating another pioneer, ten-year-old Cynthia Cutshall. In order not to treat the two girls as guinea pigs, Blaese and Anderson continued to administer the standard treatment, PEG-ADA. Did the gene therapy work? Well, yes and no. A twelve-year follow-up on Ashanthi declared that about 20 percent of her T cells were producing the ADA enzyme.¹⁷ More than twenty-five years after the trial began, Blaese said the amount of ADA produced in the treated cells was only about 15 percent of what had been expected. But the two girls are “both beautiful young ladies”¹⁸ who happily invited Blaese to their weddings. Anderson was elated. “I eat and sleep and breathe gene therapy 24 hours a day,” he told a New York Times reporter.

There was a giddy euphoria in gene therapy circles in the years following the NIH trial as more disease genes were identified, many amenable to genetic therapy. As the founding editor of Nature Genetics, I was down for the ride. More and more manuscripts arrived in our Washington, DC, office touting advances in gene therapy. Ted Friedmann, whom Horace Freeland Judson called “gene therapy’s most ardent advocate,” wrote a review I entitled “A Brief History of Gene Therapy.” Friedmann declared that the first phase of human gene therapy—the emergence and acceptance of the general concept—was over. “We are now in an explosive second phase—one of technical implementation.”¹⁹

In March 1994, I took the train up to Philadelphia to attend a press conference led by a rising star in the field. James Wilson of the University of Pennsylvania walked onto a makeshift stage in his pristine lab coat, flanked by two colleagues, to discuss his latest study that we were publishing the next day. Wilson’s team had treated a patient with an inherited form of coronary artery disease by removing part of her liver, treating it with a recombinant virus, and restoring the liver. The woman’s LDL cholesterol levels fell by 25 percent. Among the reporters in attendance was New York Times Pulitzer Prize–winner Natalie Angier. Her front-page story said Wilson’s study was “the first to report any therapeutic benefits of human gene therapy.”²⁰ It happened to be April Fool’s Day.

The swashbuckling pace with which investigators were publishing results made some observers nervous. Harold Varmus, then director of the NIH, commissioned a report on gene therapy. “Today, the announcement of a [disease] gene being discovered is tantamount to the belief that gene therapy exists for the condition. We’ve seen an extrapolation from hope to hype. In the long run, this will be destructive to basic clinical science,” said Varmus sternly. The report criticized the rush to clinical trials without sufficient understanding of disease pathology, the low frequency of gene transfer, and overselling of results leading to “the mistaken and widespread perception that gene therapy is further developed and more successful than it actually is.”²¹

Nevertheless, by 1999, Friedmann and his fellow gene therapists were euphoric about the future. Despite the efforts of critics who had “fomented mistrust and misunderstanding of the goals and techniques of gene therapy,” the inevitability of the science had been established “long before it was able to provide truly believable clinical benefits.”²² Friedmann continued: “The success of the concept of gene therapy has been phenomenal and represents a truly epochal new direction for medicine.”

Even before the ink was dry on Friedmann’s essay, those words rang hollow.

In 1984, a few months shy of his third birthday, Jesse Gelsinger fell into a coma while watching Saturday morning cartoons on television. After a lengthy hospital stay, he was diagnosed with a rare X-linked genetic disorder, ornithine transcarbamylase (OTC) deficiency. A missing enzyme results in an inability to process nitrogen (found in all proteins and other biomolecules) and a toxic buildup of ammonia. About fifty OTC deficiency babies are born in the United States each year; only half will live past the age of five.

Jesse’s father, Paul, and his partner put Jesse on a low-protein diet to manage the disease. Fortunately he had a mild form of the disease, but if he forgot to take his pills (up to fifty per day), he could fall into another coma. In September 1998, Paul heard about an OTC deficiency clinical trial being conducted at Penn. But that Christmas, Jesse fell into a coma and almost died. Shortly after graduating from high school, Jesse volunteered to take part in the clinical trial as soon as he was eligible. The following June, the Gelsinger family flew to Philadelphia on Jesse’s eighteenth birthday. They visited the usual tourist landmarks, with Jesse posing next to the statue of Rocky, arms aloft in a Minnesota Twins T-shirt.

The Penn doctors explained that the purpose of a phase I trial was to test safety, not to expect any clinical benefit. There would likely be side effects such as flu-like symptoms as Jesse’s immune system ramped up to attack the virus. The doctors also explained that while OTC deficiency was rare, affecting one in 80,000 people, there were at least two dozen similar metabolic liver disorders, in total affecting one in five hundred people. Jesse would be a pioneer for thousands of other patients. Three months later, Paul took Jesse to the airport for his flight from Arizona to Philadelphia. “Words cannot express how proud I was of this kid,” Paul said. “Just eighteen, he was going off to help the world.”²³

The OTC trial was directed by James Wilson, the founding director of Penn’s Institute for Human Gene Therapy. Because Wilson had founded a biotech company, Genovo, he was not allowed to have direct contact with the patients. On Monday September 13, 1999, Jesse received his first infusion of a recombinant adenovirus containing a normal copy of the OTC gene. As planned, he received the highest dose among the eighteen volunteers. He soon developed a fever but that was not unexpected. Jesse spoke to his father by phone that evening. It was the last conversation they would have. The following morning, Jesse had developed jaundice and his ammonia levels were spiking. They informed Paul and Wilson. Over the next two days, Gelsinger’s kidneys and liver started to fail. He was put on an artificial lung to help his breathing. When Paul finally got to Jesse’s bedside, he didn’t recognize his son because Jesse’s face was swollen beyond recognition. He had suffered irreparable brain damage.

With seven of his siblings and spouses in attendance, Paul held a brief bedside ceremony for his son. At 2:30 P.M. on September 17, physician Steve Raper shut off the ventilator and pronounced Jesse dead. “Goodbye Jesse, we’ll figure this out,” he said. The grim news reached the press two weeks later. “Teen Dies Undergoing Experimental Gene Therapy” was the Washington Post headline.²⁴ Ironically, Wilson had never met Paul or Jesse Gelsinger.

In early November, Paul led about two dozen mourners on a hike to one of Jesse’s favorite places—Mount Wrightson, close to the Mexican border. After Paul shared some memories of his son, Raper read a poem by Thomas Gray:

Here rests his head upon the lap of earth

A youth to fortune and to fame unknown.

Fair Science frowned not on his humble birth,

And Melancholy marked him for her own.

Moments later, Paul and other mourners scattered Jesse’s ashes into the Arizona air.²⁵

A short time later, Wilson flew out to meet Paul Gelsinger for the first time. Sitting on Gelsinger’s back porch, Wilson shared Jesse’s autopsy results and told Paul that he was just an unpaid consultant to Genovo. Initially supportive, Gelsinger learned that the Penn team had seen fatalities in animals (albeit receiving much higher doses of virus) and adverse events in some patients before Jesse’s fateful enrollment. Gelsinger eventually filed a lawsuit against Wilson and his two senior colleagues that was settled out of court. Wilson’s gene therapy center was disbanded, and he was forbidden from running any clinical studies until 2010. In the formal scientific report published by Raper, Wilson, and colleagues, much of the blame was cast on the viral vectors that triggered Jesse’s fatal cytokine storm.²⁶

Writing shortly after the Gelsinger tragedy, molecular biologist Peter Little reached a grim diagnosis:

I suspect the judgment will be that we were arrogant and came extremely close to using humans as experimental animals; we knew too little and expected too much, and the expectation of success was used to roll back objections.²⁷

In his book The Gene, Siddhartha Mukherjee condemned the OTC trial as “nothing short of ugly—hurriedly designed, poorly planned, badly monitored, and abysmally delivered. It was made twice as hideous by the financial conflicts involved; the prophets were in it for profits.”²⁸ Wilson, when interviewed for The Gene documentary, said: “I’ll think about [the tragedy] frequently until the day I die. I don’t know what else to say.”²⁹

The Gelsinger tragedy was soon compounded by news from across the pond. In 2000, French physician Alain Fischer of the Necker Hospital in Paris held a press conference to mark the preliminary success of gene therapy in two infants with an X-linked form of SCID. Fischer was using a retrovirus vector to shuttle the healthy gene into the patients’ blood stem cells. But two years later, two boys were diagnosed with a form of leukemia that was traced back to the therapy itself, and one eventually died. Retroviruses work by integrating directly into the host genome, like hiding the joker in a pack of cards. In most cases, the integration event is harmless, but in rare instances it can trigger a cancerous event. Fischer’s viral vector was finding a comfortable landing spot in the host DNA, inadvertently switching on an adjacent oncogene with devastating results. In 2003, the FDA responded by halting the use of retroviruses in the United States.

As Judson observed, despite hundreds of millions of dollars lavished on hundreds of gene therapy trials involving thousands of patients and volunteers, “new hopes cyclically turned to ashes, dramatic claims to sad farce.”³⁰ Gene therapy itself was on life support.

With some sober reflection, many of the setbacks could be understood. After all, viruses did not evolve simply to be used at our beck and call as delivery drones. As one gene therapy expert said: “We underestimated the fact that it took billions of years for the viruses to learn to live in us—and we were hoping to do it in a five-year grant cycle.”³¹ There was also the complication of our immune system, which is designed to combat foreign agents such as viruses. The human body isn’t going to automatically give billions of recombinant viruses a pass just because they mean well.

It has been a long haul back to respectability and success for gene augmentation therapy. The roller-coaster ride follows the Gartner Hype Cycle: the inflated expectations of the 1990s, the trough—or abyss—of disillusionment at the turn of the century; followed by the slope of enlightenment. What’s been holding up the field is not a lack of suitable targets—we have an encyclopedic catalogue of thousands of eligible Mendelian genetic diseases—but the capability of delivering the therapeutic gene safely and effectively. Researchers had to go back to the drawing board, focusing on viral delivery and safety. Two new candidates emerged as reliable, adaptable, and effective delivery vehicles for a swath of gene therapy (and genome editing) indications: adeno-associated viruses (AAV) and lentiviruses.

AAV was discovered by accident in the mid-1960s as a contaminant of an adenovirus preparation. The attraction of AAV comes from its barebones structure—it is the tiny house of viruses, a protein shell that can carry a small gene cargo. The virus is very safe—about 90 percent of humans have been exposed and infected by AAV without knowing it. Wilson’s group, clutching a financial lifeline from GlaxoSmithKline, got to work. There were only six known varieties when Guangping Gao in Wilson’s team set out.³² But at the end of 2001, he presented Wilson with a bounty of novel AAVs he had isolated from monkeys. More than one hundred forms of AAV are currently known. “Penn’s viral vector center became the Amazon of AAV,” observed science journalist Ryan Cross.³³

Why so much interest in this tiny virus? AAV naturally carries just two genes, REP and CAP, which encode proteins that make up an icosahedral capsid coat. Fully clothed, the virus is just twenty-five nanometers in diameter, holding a payload of single-stranded DNA about 5,000 bases in length—enough for a small therapeutic gene. And unlike retroviruses, AAV does not integrate into the host genome; that means it will dilute out over time as the cells it infects divide.

Despite their popularity, there’s room for improvement, says UC Berkeley’s David Schaffer. “We need better viruses. Viruses did not evolve in nature to be used as human therapies.”³⁴ Treating spinal muscular atrophy (SMA) patients, for example, has required the highest dose of AAV used in a human to date. Schaffer’s team is evolving in the lab the amino acids on the AAV surface to create novel vehicles with improved targeting properties to the appropriate cells, such as the retina.³⁵ Jean Bennett was able to overcome the limitations of AAV2, which doesn’t travel through the vitreous of the eye, by delivering it directly through the retina. Schaffer’s group has evolved a new AAV that can penetrate the full surface of the retina by injecting into the vitreous.

Lentiviruses, the other emerging virus class, form a sub-family of retroviruses and includes HIV. The tenacity with which HIV can lurk in the T cells of a patient illustrates their potential value as a modified gene delivery vehicle. Lentiviruses can infect both dividing and non-dividing cells and have a cargo hold double that of AAV. The first clinical trial using a lentiviral vector was conducted in 2005.

A decade after Gelsinger’s death, an editorial in Nature signaled a renewed sense of optimism in gene therapy circles. “The pervading sense of disillusionment is misplaced,” Nature stated.³⁶ It was time for researchers and biotech “to consider its successes with as much intensity as its setbacks.” Wilson reflected on the lessons he had learned. “With what I know now, I wouldn’t have proceeded with the study,” he said. “We were drawn into the simplicity of the concept. You just put the gene in.”³⁷ Carl Zimmer penned a story for WIRED with a striking hero image of two viruses: on the left was the adenovirus that “laid waste” to Wilson’s career. On the right, the AAV, the bright new hope of gene therapy that could bring Wilson “redemption.”³⁸

In 2015, Friedmann and Fischer were awarded the Japan Prize for their contributions to the gene therapy field. Friedmann opened his award lecture in Tokyo by showing a picture of the serpent-entwined Rod of Asclepius, the ancient Greek symbol of medicine. Next he substituted the snake with the double helix—the repository of all genetic information. “We would like to think that knowledge of this molecule is going to markedly change the way we understand disease and the way we treat disease,” he said.³⁹ As an example, Friedmann paid tribute to a female physician in Philadelphia who was pioneering a method to deliver a gene therapy directly into the retina of patients with a rare genetic form of blindness. Those early results, Friedmann proclaimed, were of “biblical proportions.”

Indeed, they were little short of miraculous.

I. The virus was named after Richard Shope, a Rockefeller University pathologist who, studying a flu outbreak in pigs in 1918, helped prove that influenza was caused by a virus, not a bacterium. In 1933, Shope injected himself with the eponymous virus.

II. As I described in my first book, Breakthrough, King’s quest to isolate BRCA1 was thwarted by Myriad Genetics. Twenty years later, I served as the technical advisor for a film called Decoding Annie Parker, based on the true story of the first woman in North America to undergo BRCA1 genetic testing. Helen Hunt played Mary-Claire King. Sadly, few of my suggestions were incorporated—the writers told me we weren’t filming a Nova documentary. My name is buried at the end of the closing credits, right after Aaron Paul’s guitar coach.