Nature Medicine: “Call in the Backup” By Elie Dolgin

By Elie Dolgin

The most common genetic killer of infants, a disease known as spinal muscular atrophy, is caused by mutations in a single gene. The human genome contains its own backup system—near-identical copies of the defective gene—yet these secondary sequences rarely get used correctly. As Elie Dolgin finds out, drug companies hope to change that, with the first wave of targeted therapies that reboot the body’s backup system now entering clinical trials.

Vivienne is stacking toy blocks on the coffee table in her family’s living room in Millville, Massachusetts. She adds a sixth block to the growing column and carefully nudges the cube into place. “It’s Rapunzel’s tower,” she exclaims.

For a three-year-old with a hereditary neuromuscular disease, the fine motor skills in Vivienne’s fingers are surprisingly refined. And as she talks to her parents, switching effortlessly between English, Polish and German, it’s clear that she’s exceptionally bright. But Vivienne’s condition betrays itself as she attempts to pile the blocks higher. Her shoulders bunch up. Her arms strain under their own weight.

Vivienne has spinal muscular atrophy (SMA), a condition caused by mutations in the gene SMN1. Under normal conditions, this gene produces the ‘survival of motor neuron’ (SMN) protein, which, among its many functions, has a crucial role in maintaining the specialized nerve cells that control muscle movement. In individuals with SMA, however, working versions of SMN are in short supply, causing crucial motor neurons to progressively die off. This leaves people with poor muscle tone as well as difficulties with eating, moving and breathing. The resulting whole-body paralysis typically leads to early death.

Even though SMA affects around 1 in 10,000 babies born worldwide, there are currently no drug treatments that influence the course of the disease. Unless one is developed soon, Vivienne will probably die before she reaches her 30s. Many others won’t make it past infancy.

Vivienne’s parents Helena and Helge (who asked that Nature Medicine withhold their family’s last name) are both carriers for mutations in SMN1. They learned of their genetic status—and Vivienne’s diagnosis—in March 2011, when Helena was eight months pregnant with their second daughter Lara. Fortunately, Lara inherited only one faulty copy of the SMN1 gene, and she is now a healthy and active toddler. “We call her ‘hurricane category 5’,” says Helena, as she chases Lara around the living room picking up tossed-aside stuffed animals on a Saturday morning in mid-July. In contrast, Vivienne—her legs bowed, her knee joints extended—can barely take measured steps and, even then, only with the help of someone holding her hands or with something solid to lean on.

She feels confident enough to try to walk again, though, which wasn’t always the case. Thanks to an experimental drug from Isis Pharmaceuticals that Vivienne received six weeks earlier as part of a phase 1 clinical trial, her parents now say she has more strength, more poise and more self-assurance. She sits more upright on the horse at the equine therapy clinic she attends each Monday. She can now wade through the water on her own during her weekly aquatic therapy sessions. “She shows significant signs of improvement—improvement that shouldn’t be there in her type of disease and improvement that I believe can only be explained by the trial,” says her father Helge.

Isis’s drug, called SMNRx, works to boost levels of SMN protein. In most animals, mutations in the SMN1 gene cause embryos to die during development. But because of a duplication on chromosome 5 of the human genome, most people (and all those with SMA) have varying copy numbers of a nearly identical backup gene, dubbed SMN2, that can partially rescue SMN protein production.

SMN2 differs from the normal, functioning version of SMN1 by just a single nucleotide change: in SMN2, the sixth letter of the gene’s seventh coding region, known as exon 7, is a T rather than a C. That small difference has big consequences, though. The altered DNA letter usually causes the RNA processing machinery to splice out exon 7, leading to the creation of a shortened, less stable version of the SMN protein that rapidly degrades. Only around 5–10% of the time does the exon get left in place to yield a full-length, functional protein.

The more copies of SMN2 people with SMA carry, the more full-length protein they produce and the less severe their disease. One or two copies usually results in type 1 SMA, the most common and deadly form of the disease. Around 60% of all children with SMA are in this category. They never develop the ability to sit, stand or walk, and most will die under the age of 2. People with type 2 or type 3 SMA typically have at least three copies of SMN2, and, despite physical and respiratory disabilities, these individuals can live into early adulthood and beyond. Then there are rare individuals who harbor four or more copies of SMN2. Despite lacking a functional version of SMN1, these people have much milder forms of the disease and, in some cases, show practically no symptoms at all.

The dream for scientists, physicians and pharmaceutical companies alike is to bring more people with SMA effectively into this last category by helping affected children make the most of the SMN2 genes they already have. A number of different strategies are being pursued, including small-molecule drugs and antisense therapies, all with the same goal in mind: to therapeutically boost the levels of full-length SMN protein. “Nature gave us this second copy [of SMN] that can make a little bit of full-length protein,” says Basil Darras, director of clinical neurology at Boston Children’s Hospital and a site coordinator in the Isis trial. “And now everybody is trying to develop drugs that will induce more full-length SMN protein production from the SMN2 gene.”

With multiple therapeutic approaches now either in early clinical trials or nearing human testing, optimism is starting to mount that one of these approaches will ultimately prove a success. “There’s an excitement in the field,” says Karen Chen, chief scientific officer of the SMA Foundation, a New York–based research organization. “There’s now more hope for a treatment for SMA in the near future—more than at any other time.”

Keep the motor neuron running

SMNRx is the first therapy that specifically targets the underlying splicing defect behind SMA to be tested in children with the disease. In Isis’s first phase 1 trial, which launched in December 2011, doctors administered a single injection of SMNRx directly into the fluid-filled space at the base of the spinal cord in 28 children with type 2 or type 3 SMA. Vivienne, a type 2 with three copies of SMN2, received a six-milligram injection at Boston Children’s Hospital in May of this year.

The therapy is an 18-letter string of DNA that corrects aberrant splicing by binding a unique sequence on SMN2‘s messenger RNA located just downstream of the crucial seventh exon, thereby obstructing components of the RNA splicing machinery (see ‘Bring to one’s antisenses’). “We displace some negative regulatory proteins, which allows proper splicing to occur,” explains Frank Bennett, head of research at Isis, which is headquartered in Carlsbad, California. In mouse models in which the endogenous Smn1 gene has been knocked out and human versions of SMN2 have been swapped in, the Isis therapy—a so-called ‘antisense oligonucleotide’—delivered to the mouse central nervous system (CNS) increased the expression of full-length SMN protein in motor neurons, improved muscle strength in behavioral tests and extended the rodents’ median lifespan from 16 days to 26 days³. “We think we have a drug that will have a big impact on the disease,” Bennett says.

Isis, which is co-developing its drug together with Biogen Idec of Weston, Massachusetts, is hardly alone in its endeavor to correct SMN2 splicing—but it is the only one currently using an oligonucleotide approach. Other companies, including the Swiss drug giants Novartis and Roche, along with Boston’s Paratek Pharmaceuticals, are developing ingestible small-molecule drugs designed to increase the inclusion of SMN2‘s exon 7. Of note, splice-modulation therapy is also being pursued for other rare neuromuscular disorders. For example, the exon-skipping drug drisapersen, from the UK’s GlaxoSmithKline, is currently being tested worldwide in a 180-person, phase 3 study for the treatment of Duchenne muscular dystrophy, an X-linked disease that affects approximately 1 in every 3,500 boys.

Before adopting this splicing-correction strategy for SMA, many researchers tried increasing the amount of SMN protein in a nontargeted fashion. For example, by using histone deacetylase (HDAC) inhibitors, which, by remodeling chromatin structure, are thought to promote gene expression broadly, including for SMN2, numerous groups showed that they could increase SMN protein levels—both truncated and full-length—in mouse models of SMA as well as in patient-derived cell lines. And because several HDAC-blocking drugs are commonly used in other fields of medicine, including neurology, it proved relatively straightforward for physicians to take some of these drugs into clinical trials in patients with SMA. However, even though there are case reports of some children showing dramatic improvements on the drugs, no study has yet demonstrated a statistically significant overall benefit.

Kathryn Swoboda has repeatedly seen HDAC inhibitors disappoint. As director of the pediatric motor disorders research program at the University of Utah School of Medicine in Salt Lake City, she led the Project Cure SMA Investigators Network, a trial consortium that recently published the results of two placebo-controlled trials of the HDAC-blocking agent valproic acid, neither of which demonstrated much clinical advantage from the drug^{1, 2}. Looking ahead, Swoboda, now one of the four site directors for Isis’s phase 1 trial, expects more from the experimental splicing-directed therapy. “I’m more excited about this approach than anything I’ve been excited about in a very, very long time,” she says.

Fellow trial site director and pediatric neurologist Claudia Chiriboga, of the Columbia University Medical Center in New York, agrees. “I’m very optimistic that once we find the optimal timing and the optimal dosing we’re going to see results that are life altering,” she says. Isis’s drug, Chiriboga adds, “could change the face of SMA as we know it.”

Still, many investigators—and their big pharma backers—see potential in boosting SMN levels through mechanisms other than splicing modulation. These alternate strategies include using small-molecule drugs that affect RNA metabolism or protein stability, as well as administering modified viruses for therapeutic gene delivery (see ‘Getting a fix on SMA’).

“It’s one thing to only try to correct splicing events, but there are a lot more opportunities in terms of drug targets,” says Gideon Dreyfuss, a biochemist at the University of Pennsylvania Perelman School of Medicine in Philadelphia who is working with the New Jersey–based pharmaceutical firm Merck to discover new compounds. “Nothing’s a winner yet,” adds Elliot Androphy, a molecular biologist and dermatologist at the Indiana University School of Medicine in Indianapolis who has conducted many of his own chemical screening experiments. “The world is open to people getting more compounds and several companies have expressed interest in this space.”

One need not look further than the drug in late-stage preclinical development at the US National Institute of Neurological Disorders and Stroke (NINDS) in Bethesda, Maryland. Through its SMA Project, a $32 million contract therapeutics development program that ran from 2003 until earlier this year, the agency identified a lead compound, dubbed ALB-111, that induces the read-through of stop codons during protein production. This creates extended SMN proteins that, even though they still lack the seventh exon, may have some functionality or may help stabilize the small amounts of full-length protein. “We are discussing the merits of this compound, the [SMA] Project and its assets with a whole panel of external interested parties—foundations as well as companies—to see if we can establish a licensing agreement where this could be developed further,” says Rajan Ranganathan, head of the NINDS’s Office of Translational Research.

Capping it off

A separate $13 million nonprofit drug discovery program, this time funded by Families of SMA, an Illinois-based research and advocacy organization, yielded the compound being advanced by Repligen, a small biotech located in Waltham, Massachusetts. The licensed drug, now called RG3039, blocks an enzyme called DcpS that normally takes a protective cap structure off of messenger RNA transcripts and primes them for destruction. By inhibiting that process, researchers have shown that they can increase the levels of SMN transcripts—both truncated and full length—resulting in elevated overall protein production and improvements in behavior and survival among mouse models of the disease⁴. “We’re changing the steady-state metabolism of RNA,” says James Rusche, head of research at Repligen.

In mouse studies, RG3039 increased the expression of SMN in central nervous system tissue by only about 50%—a “pretty modest” amount, admits Rusche. But that may be enough to generate real therapeutic benefits. “The production of SMN protein doesn’t need to increase by much to go from almost no survival to almost no disease,” Rusche says. “That appears to be true in the natural history of patients, and it appears to be true in the animal models as well.”

Although Isis’s drug was the first SMN-targeting agent tested in children, RG3039 was actually the first SMN-directed drug discovery program to enter human trials, albeit in a healthy adult population. Repligen launched a single-dose, phase 1 trial involving 32 adult volunteers in May 2011. A year later, at the 2012 Annual Meeting of the American Academy of Neurology in New Orleans, the company announced that its oral liquid formulation of RG3039 proved safe, with molecular signs that the compound inhibited DcpS in peripheral blood cells. A multidose safety study among a second 32-person adult cohort kicked off in September of this year.

The drug’s effect in peripheral blood cells is not inconsequential. Studies are beginning to show that SMA might not simply be a disease of the motor neurons, but one borne out in the peripheral muscles and other tissues, too. If validated, these emerging—and, it should be said, somewhat controversial—findings could mean that SMN-targeted drugs will need to work throughout the body, not just in the CNS, to have maximal therapeutic effect.

By default, that’s already the case for small-molecule drugs such as the ones under development at Repligen and Roche. “If it gets in the brain, it gets in every cell in the body,” notes Luca Santarelli, head of neuroscience at Roche, which is co-developing its splicing-targeted drug with PTC Therapeutics of Piscataway, New Jersey. It’s trickier for oligonucleotide approaches, such as Isis’s antisense therapy, which, in its first phase 1 study, was administered only into the CNS.

Although researchers found that delivering the Isis drug directly into the brain nearly doubled survival time in a mouse model of SMA³, systemic administration in the rodents—which usually die within two weeks—has been shown to extend median survival by up to eight months. “The results unequivocally say that you need high levels of SMN somewhere in peripheral tissues to have a healthy mouse,” says study author Adrian Krainer, a molecular biologist at the Cold Spring Harbor Laboratory in Long Island, New York, who has been working with Isis since 2004.

Beyond location, there is also the question of timing. A recent preclinical study that involved an experimental oligonucleotide drug similar to SMNRx found that waiting to administer the therapy drastically reduced its effectiveness. Mice treated four days after birth had a median survival of 41 days; by comparison, most mice treated immediately after birth lived for 100 days or longer⁶, a finding that mirrors earlier gene therapy studies⁷. “The later you go, the worse the impact gets,” says study author Arthur Burghes, a molecular geneticist at Ohio State University in Columbus. “Where are those time points in a human individual with SMA? We don’t absolutely know.”

Getting there early

In its first safety tests, Isis’s drug is being tested only in children who are over the age of 2. (A second, phase 1 study involving multiple doses of SMNRx and a similar demographic of patients is scheduled to begin recruitment before the end of the year.) But as the company gears up to test for signs of efficacy, Basil Darras, Vivienne’s doctor at Boston Children’s Hospital, expects to see the most pronounced effects in the youngest cohorts. “Maybe we’ll be able to achieve cure if we intervene early—shortly after birth or before you develop symptoms,” he says. This consideration has led to increased interest in SMA screening for all newborns. For example, Kathryn Swoboda plans to launch a state-wide pilot screening program in Utah sometime early next year.

Cure is the ultimate goal, especially for the thousands of children born each year with type 1 SMA. But for people with type 2 or type 3 SMA, “the reality is that any sort of treatment would be a blessing,” says Jill Jarecki, research director of Families of SMA who, in a previous job, led the small-molecule screening effort that yielded RG3039, the Repligen compound. “Typically what the patients say is, ‘Of course we want a cure. Of course we want the slam dunk. But just give us something that will stop progression.’”

That’s all that Vivienne’s parents were realistically expecting at the onset of the trial. Yet, Vivienne’s renewed abilities have continued to defy expectations. In September, she went from lying down to a sitting position all on her own. She also took a single, unaided walking step. She hadn’t managed to perform either of those actions for well over a year.

“If that’s already the effect of one injection of just six milligrams,” says her father Helge, “the question that arises of course is, ‘What would more of the drug do?’” Vivienne and her family may soon know the answer to that question. As part of a planned follow-up study, Vivienne is scheduled to receive another dose of the drug sometime next year.