Victor Dubowitz

Transcription

[applause] Kenneth Fishbeck: But I'm really -- I'm really glad to be here. I work just across the street, on the other side of Old Georgetown Road in that glass building, the Porter Neuroscience Center, Building 35, and come over here usually when there's some kind of medical emergency in our -- in our group, or our family or neighbors. To start off with, I should say the title is a little different from what you see on the -- as advertised. I will be touching on diagnosis and management of these diseases, but the main focus of the talk is on developing treatment, because I -- in my mind, that's the most interesting aspect of converting the fruits of the Human Genome Project into clinical practice. The other thing I wanted to mention at the beginning, I realized when I was putting this together, is that kind of an updated version of a talk I gave across the street over in Masur for the Astute Clinicians Series. So those of you who were are here for that -- or there for that talk, there will be some overlap. Let's see here. The other thing is I have nothing that I have to disclose. As an NIH employee, I'm not paid for outside activities, but there are some volunteer activities that I thought would be good to mention verbally to keep in mind as I go through this talk. I serve as an unpaid member of advisory boards for a number of voluntary -- for patient foundations, voluntary organizations. And these are both within the United States and abroad. So the Muscular Dystrophy Association, the Spinal Muscular Atrophy Foundation, and the French Muscular Dystrophy Association, the AFM; and I'm also an unpaid member of advisory boards for a couple of companies, Biogen Idec and Prosensa. And one thing that I also wanted to mention is that I've just finished a sabbatical, a six-month sabbatical in industry. In government speak this is a training experience, and I was at Novartis, Novartis Institutes for Biomedical Research in Cambridge, Massachusetts for six months, and I think that is something that gives a kind of perspective on what I'm going to be talking about, but it's also good to bear in mind. Where I was in Cambridge, on the east side of the city of Cambridge of industrial -- formerly industrial area around the MIT campus, Kendall Square and Central Square, there's an amazing burst of activity in biotech and pharmaceutical companies. A number of different companies have sprouted up in that area on, I think, land that was owned and managed by MIT. And where I was was a former candy factory, a Necco candy factory, and across the street from where I was working in Cambridge was a park that was developed along with the industry development in that area, and there was a -- engraved in stone in the monument in the park was this quote from Henry Thoreau, from "Walden," that was published in 1854. And I think it captures the sense of being up there on sabbatical is, "What recommends commerce to me is its enterprise and bravery. Commerce is unexpectedly confident and serene, alert, adventurous, and unwearied." And I think this captures the spirit of these companies in that area, with academic collaborators who are trying -- working to develop treatment to reduce the burden of disease that I'll be talking about. Now, this is a slide -- this next slide is one that I've been using now for over almost 10 years; it originated from Maynard Olson and Francis Collins, and I think that it is good at showing, in context, the work that we do. "We" meaning all the people working on hereditary diseases using the Human Genome Project to identify the causes of these diseases and to use that information for diagnosis and development of treatment. Now what you can see here is a process that begins and ends with patients. So clinicians who see patients have to start by characterizing the disease, the phenotype, the clinical manifestations of the disease; collect samples from patients and family members to do -- to map the disease gene and identify the disease gene. Now this has become a lot easier over the last few years with high-throughput sequencing methods. We can sequence through all the genes in the genome, the coding regions for all the genes, and identity the specific mutations that track with the disease in these families, sequence variants that are specific to the disease, and that tells us what the cause of the disease is. It's -- it gives us a very accurate diagnostic test that can be used to identify, see who's got the disease and who doesn't. And for the diseases I'm talking about today, these diagnostic tests have been available now for over 10 years, 10, 15, 20 years for the two diseases I'm going to focus on: Duchenne muscular dystrophy and spinal muscular atrophy. But with the disease gene identified, we can characterize the disease gene, we can see what is the normal function of that gene, and how do the mutations affect that normal function. And then we can identify or develop animals -- mice, flies, worms -- that have the animal equivalent of the disease by putting the same kind of alteration into the animals. And we can develop cell culture systems to study; use these animals and cell culture systems to work out the disease mechanism, and start to identify targets for therapeutic intervention. We can take the cell culture systems and look -- screen through hundreds of thousands of chemical compounds to find compounds that correct or mitigate the disease manifestations in cell culture, test them in the animals. Or we can develop biological approaches to treatment. That's what I'll be talking about. But we get down -- we can -- so we can use these cell culture systems as assays to test and screen for potential treatments. But that brings us back to the -- what I say is the most important and the most challenging part of this curve, is to take treatments, which may be very effective in the animal models, to take them into patients, and to use that information to develop a treatment that's safe and effective at reducing the disease burden in the patients. And that's where we're stuck. Boy, I don't know if this is a good analogy. When I came back from sabbatical, our kitchen sink was clogged from when we were away, and we had to get the plumber in to root it out. I think this process is -- with hereditary diseases has kind of come around to this stage, and it's beginning to back up. We're -- this is where we really need to focus our efforts, and this is where the enterprise and bravery of companies comes into play. And I think -- I think it's important for us to engage these companies, biotech and pharmaceutical companies, to bring -- to bear their resources and expertise to solve this problem, so that we can really fulfill the promise of this curve and come back to patients with effective treatment. Now, as I said, I'm talking about two diseases this morning, muscular dystrophy and spinal muscular atrophy, but I really want to focus on where we've come with these disease and where we're going, and where the work is that still needs to be done, which, again, involves clinicians, people who see these patients, who connect the patients to the right kind of clinical trials and develop the clinical trials that are needed to bring home an effective treatment. So that's what I'm going to be talking about. First, for Duchenne muscular dystrophy, this is the most common severe form of muscular dystrophy. In children it causes progressive proximal weakness; so muscle weakness in the shoulders and the hips to start off with. The -- it's an X-linked disease, and the -- so only boys are fully affected by the disease; they start to become weak at usually around age three or four. And the weakness is gradually progressive, to the point where they become wheelchair-bound around age 10 or 12. And then with progressive weakness, they -- of the extremities and the respiratory muscles, they also get a cardiomyopathy, and they die. They die, usually without any kind of therapeutic intervention, by their late teens or early 20s. And with aggressive respiratory support and cardiac management, they can -- they can live now into their -- you know, up to age 30. Now, the underlying pathology here is, if you look at the muscle, there's a problem with the muscle, and it causes the muscle weakness and respiratory and cardiac manifestations. And if you look at the muscle pathologically, what you see is degeneration and regeneration. Over on the left here are necrotic muscle fibers; the muscle fibers fall apart, their inflammatory cells go in, macrophages gobble up the debris. And on the right here are regenerating muscle fibers. Muscle has a regenerative capacity. If any of us injure our muscle, the muscle will regenerate by activation of what are called satellite cells around each muscle fiber, and the -- you know, it can make new muscle fibers. But that regenerative capacity is limited, and with the ongoing degeneration of the muscle fibers, eventually, the regenerative capacity gets depleted, and by the late stages of the disease the muscle is replaced by fat and connective tissue; there's really very little muscle left by the time the patients die from the disease. Now, as I said, this is an X-linked. It's been -- long been known to be an X-linked recessive disorder. The clinical manifestations were first described back in the 1840s, so the disease had been well-characterized clinically. And then, just about 25 years ago, in the late '80s, the gene was identified. And this was really one of the first, probably the first gene to be identified by this process of positional cloning, to find the defect on the X chromosome, which was in a gene that was -- the product of the gene was given the name "dystrophin." So a protein that is missing because of mutations in patients with this disease. The mutations cause a loss of dystrophin, and that's the cause of the muscle breakdown in this disease. Now the dystrophin gene is still one of the -- this made it easy to find, I guess -- it's still one of the largest genes known; it has 79 exons, these coding portions of the gene, that are spread over 2.3 or more billion base pairs of DNA on the X chromosome. So it takes up more than 1 percent of the X chromosome, 0.1 percent of the whole genome. There is another disease -- for a while it was thought be a separate -- different disease -- called Becker muscular dystrophy, also X-linked, but a milder disease. And it's caused by mutations in the same disease -- in the same gene, the dystrophin gene, but it has less severe manifestations. These patients can have onset later in life, may be quite normal into their -- into early adulthood. They have elevated creatine kinase, a sign of muscle breakdown, and they variably develop weakness later in life. Now, dystrophin. What is this protein dystrophin that's missing in Duchenne dystrophy and deficient in Becker dystrophy? Dystrophin is now known to be a structural protein. It underlies the muscle plasma membrane, and its job, really, is to hold the muscle plasma membrane together as the muscle contracts and relaxes. There's a lot of tension on muscle when it contracts and relaxes, and dystrophin forms the key structural link between the interior side of the skeleton, the actin side of the skeleton, within the muscle fiber through a group -- a cluster of proteins in the -- in the plasma membrane to the basal lamina, the sheath [spelled phonetically] that's outside the muscle fiber. So it's an important structural link in the integrity of the muscle plasma membrane. And loss of dystrophin leads to instability of the muscle plasma membrane. The muscle membrane breaks when the muscle contracts and the contents of the muscle leak out. That leads to a very high creatine kinase level, and calcium enters into the muscle fiber, activates proteases, and that leads to degeneration of the muscle fiber. So loss of dystrophin causes Duchenne dystrophy. Mutations in these other proteins also lead to muscular dystrophy, usually called limb-girdle dystrophies; can be very similar to Duchenne or Becker's, but has a different pattern of inheritance, autosomal recessive rather than X-linked, because the genes are located on other chromosomes. Now, since the gene was identified back in the 1980s, people have known that the distribution -- looked at the distribution of mutations in the gene. And most of the mutations in this -- the cause of the disease are deletions of one or more exons, so internal deletions. And they're distributed in such a way that most of them are near the middle of the gene. This is one end of the gene to the other, and where the -- where these deletions are distributed is mostly in the middle of the gene, and some at the 5' end of the gene, the left side of the gene here. These deletions are mostly around exons 45 to 55 out of 79. And the effect of these internal partial deletions is to shift the translational reading frame. I'll show you that a little bit later. They -- you know, as the ribosome comes along and reads codons, three nucleotides for each amino acid, the deletions that cause Duchenne dystrophy throw the reading frame off; they take out -- they take out an odd number of nucleotides so that the message downstream from the point of the mutation is altered. And this leads to a truncated protein that is lacking the C-terminal end, which is encoded by the 3' end of the gene here. And this part of the protein is important for the protein stability, it's important for the interaction with the other structural proteins, the -- you know, that help dystrophin to do its job. And so disruption of the C terminal end leads to an unstable protein which is rapidly degraded. And when you look for dystrophin in muscle, you really can't see it very much at all. Now, the -- knowing this, knowing that the cause of the disease is loss of this protein and the result of the loss of the protein is muscle degeneration, there are number of different approaches to treatment that come to mind and have been actively pursued over the 25 years since this gene was identified. One is to block muscle degeneration. As I said, the degeneration is due to activation of proteases, and one approach is to use protease inhibitors, like calpain. Another is to stimulate muscle regeneration, and it's been figured out what factors are involved in muscle regeneration, like insulin-like growth factor one and follistatin, and, in particular myostatin. Myostatin is a hormone, a peptide hormone that prevents muscle overgrowth. And if you inhibit myostatin, that can stimulate muscle regeneration. And that's another approach that's been taken. Now, there is another gene called -- it's been discovered, called utrophin. And utrophin is very similar to dystrophin, it's encoded elsewhere, so the patients have utrophin. Another approach to treatment is to stimulate the production of utrophin to compensate for the loss of dystrophin. And yet another approach has been to look at those patients who have other mutations in the gene that lead to premature truncation. These are called nonsense mutations; they lead to a premature stop signal. And there are drugs that have been identified that lead to -- cause a read-through of premature stop signals, nonsense suppression, and amino -- commonly used aminoglycosides like gentamicin have this effect at low levels, and there've been trials with gentamicin treatment. And a drug specifically designed by a company in New Jersey, PTC Therapeutics, to have this effect, which is more potent -- or said to be more potent than gentamicin at having this effect of suppressing nonsense mutations, called ataluren, has been in clinical trials. Now -- and then, you know, just replacing the gene is another approach. The problem here is that the gene is so big that it's hard to replace it, but there are small versions of the gene that are -- that are still quite functional that have been developed, and there have been experiments to directly inject the gene into the muscle, the truncated mini-dystrophin constructs into the muscle, or to put them into a virus. First, this was tried with adenovirus. The problem with adenovirus is it's very immunogenic, and you get a lot of inflammation, not very much expression. And then more recently with adeno-associated virus, a smaller, less immunogenic virus; but here, too, all of these approaches have been tried in animal models and in -- some of them in patients, with, I should say so far, limited success. There's one approach that -- left here on this list that I would like to, you know, spend a little bit more time talking about, because I think it's the one that bears the best promise right now for treatment. Yeah? Male Speaker: Just before you get to that. The first two approaches, the protease inhibitors and the myostatin inhibitors, would that be for, like, any muscle degeneration? Kenneth Fishbeck: Yeah, you know, that's a good point. And I think that's the appeal of those approaches, is, you know, as you get down the list, you're getting more and more specific to Duchenne muscular dystrophy, but that's a good point. The top two here are -- would work for any muscle disease, right? Or any muscle degeneration, any muscular dystrophy or muscle degeneration disease. And I think that that's -- particularly when it comes to myostatin, that's attracted the interest of the pharmaceutical industry, Wyeth, Pfizer, and Novartis, to develop drugs that work on that. Then it would work not just for Duchenne dystrophy and Becker, but would work for polymiositis or other kinds of muscular dystrophy, or even, perhaps, I think this is in the back of their mind, for age-related muscle weakness, or what's called sarcopenia, that we all get as we get older. It starts in our 30s, and then gets to be more of a problem as you get into your 60s, 70s, and 80s, this age-related muscle weakness. So some of the treatments that are being -- have been and will -- are being tested for Duchenne muscular dystrophy may have broader applications. But the sense is that the more effective treatments are further down on the list; they're targeting the mutation, going right to the source of the problem. So you have more general approaches that are less likely to be specific, and then more specific approaches that are -- that aren't going to work generally. That was a very good question. So I'd like to focus on this exon skipping idea, I guess because it has genomic relevance, and it's something that's being looked at in a variety of other diseases as well, genetic diseases, so explain a little bit about how that works. So, again, Duchenne muscular dystrophy is usually caused by gene deletions that have this effect of shifting the translational reading frame. And what's been developed recently are oligonucleotides. These are short stretches of nucleotides, 15, 20 nucleotides, that can be used to promote the skipping of the exon -- can promote a skipping of the mutant exon, or, more commonly, downstream exons, to restore the reading frame. And this has been an idea for a long time; I think what's led it to take off here in the last few years is that people have worked out chemical modifications that enhance the stability of oligonucleotides, so that with a single injection you can get an effect that lasts for months. They're very stable; in some ways, kinds of frighteningly stable. It's better to have a drug that turns over, that's not going to be staying in the system for so long, but it does lead to a potential for a very effective treatment. And the results in mouse models have been very, very good. Now, there's a good -- I think a very good mouse model for Duchenne dystrophy called the mdx mouse that was identified some years ago. The mice are less -- a lot less severely affected than patients, but they have a clearer phenotype and lack dystrophin. And when you -- when you inject these oligonucleotides -- here the dystrophin is stained in red around each muscle fiber, and you can see without injection there's no dystrophin; and then with injection at four weeks or 24 weeks of age, you get a nice dystrophin expression. I've tried to diagram here how this exon skipping works. Here are the exons in the gene, and each three nucleotides encodes an amino acid, so, you know, we have, like, glutamine, arginine, tryptophan, lysine, phenylalanine. And when there is a loss of an exon, one or more exons, it can throw the reading frame off so that the amino acids that are encoded are the wrong amino acids because the reading frame has been shifted. And you lead -- and you get a premature stop signal. Now with this oligonucleotide, it binds to the messenger RNA near the splice site, and alters the splicing such that this exon is skipped. And when this exon is skipped, the exon with the mutation, you get -- you bring back the downstream amino acids. You may be missing part of the gene. Here I guess you're just missing the two amino acids between the arginine and the phenylalanine, but, you know, dystrophin is such a long protein that in the middle of the protein you can get away without a few amino acids and you still have a very functional protein. That's the idea behind this approach. Now, moving into the clinic. First, well, over 10 years ago now, a group in Holland, in Leiden, did experiments in cell culture with this oligonucleotide-induced exon skipping. And here you see the messenger RNA without the skipping; and here you see that a shorter messenger RNA is made with the oligonucleotide; and here you see a muscle fiber in cell culture, a myotube in cell culture, and you see that the dystrophin here, stained in green, is nicely present. And then they went on -- this paper was published about five years ago in the New England Journal of Medicine, to inject the oligonucleotide directly into patient muscle. Here they had four different patients. They biopsied the muscle after the injection, and here you see what a Duchenne patient looks like without the treatment. And you have no dystrophin, except for occasional muscle fibers where there's a spontaneous mutation that reverses the effect of the mutation, so-called "revertant fibers." Here's a control patient -- or control -- healthy control for comparison. See the green around each muscle fiber and cross-section. And here are the four patients who were injected. And you see, you know, very nice, you know, levels of dystrophin after the injection. So the protein is clearly made when you inject the oligonucleotide into the muscle. And then they went on, more recently -- this was also published in New England Journal in 2012, with a multicenter study, a small Phase 2 proof-of-concept multicenter study with a subcutaneous delivery. And this was remarkable. So it makes sense that if you inject directly into the muscle it works, but if you inject it subcutaneously it works, generally, to -- the oligonucleotide gets into muscle throughout the body and corrects the splicing defect. Now, part of that might be because the muscle fiber membrane is so leaky in this disease that the oligonucleotide can get in easily, but it worked remarkably work. So a biotech company named Prosensa in Leiden picked this up and a larger pharmaceutical company, GSK, has picked it up from Prosensa and carried this into full development mode. A remarkable investment from a large company in a relatively rare disease, so Duchenne dystrophy affects about one in 3,000 boys. And the treatment is exon-specific, so it's only a subset of those patients who will respond to this kind of treatment. Now -- so in this multicenter study, they saw an increase in dystrophin in the muscle, and it was an uncontrolled, open label study. They -- it looked like maybe they were having an effect on the -- on the walking. So here's the effect on the six-minute timed walk after a year. And, again, this is uncontrolled; they're looking at -- comparing to untreated patients, you know, that weren't in the same study, so it's not really a fair comparison. But it shows the typical course of the disease as these patients get older, here six, eight, 10 years. They get better -- they do better in terms of the walking, gradually, like any child would, but when they get to a certain point, their walking just really stops, you know? When they -- as I said, they get wheelchair-bound around age 10 to 12. And with this treatment, the patients that they followed, after the treatment, a majority of them after one year were doing okay, and they've now looked out two, even three years, and they're still -- most of them are still doing pretty well. So it was encouraging enough for GSK to get involved, and now there is -- well, nearing conclusion, I guess, the enrollment's finished for the Phase 2 U.S. study: 14 sites, 54 patients, and a large Phase 3 placebo-controlled study in 30 sites, 20 different countries around the world, 180 patients. We should know within the next year or two whether this treatment is effective. GSK is putting a major investment into this treatment, and it'll tell us whether this -- whether this approach works, whether it -- whether it restores dystrophin and whether it has a clinical effect. They're using this timed walk as the primary outcome measure. Now, there's another company -- it used to be called AVI and is now called Sarepta. It's a start-up biotech company that had been based in Washington State, but just recently moved to Cambridge. And Sarepta is using a different kind of chemistry in the oligonucleotides; they're called morpholinos. And the results suggest that morpholinos may be just as effective, but with less toxicity than the oligonucleotides developed by Prosensa. And here's a similar experiment, where they injected -- this was a study done in England. They injected directly into the muscle and looked at the untreated muscle and treated muscle, and you see no dystrophin here, and then here in black, again, the ring around each fiber showing nice correction of the dystrophin deficiency. And as with the Prosensa oligonucleotide, they see a very nice -- relatively nice expression with subcutaneous or systemic injection, as well as with direct injection. Now, just recently -- it's funny, it's a small company, their stock went down when they got some negative results, and then came way up again when they had some positive results in October -- they reported at meetings in October. But it's -- a small study, just a handful of patients, they claim to see a dramatic effect on the six-minute timed walk, and that did wonders for their stock. They, too, are now planning to into a larger scale trial. So both of these, you know, therapeutic development efforts are actively underway. And I think we should get results here before too long, as to whether -- that'll indicate whether this approach will work. And a lot of people in the field are optimistic about it. So what are the issues? A research nurse in our group, Angela -- I don't think she's here -- is always talking about issues, and I realized that when she talks about issues, she's talking about problems. You know, the potential problems are safety, first of all. So, in the Prosensa/GSK study, there -- a lot of the patients get injection-site reactions, and a lot of the patients get proteinuria. And there's a concern about these and about whether there might be more serious inflammatory reactions in some of the patients, but it's a large enough study we'll get a handle on that by the end. There's a problem with delivery and -- efficiency and stability of delivery. As I said, the oligonucleotides are very effectively distributed in Duchenne muscle. Whether this approach will work in other muscle diseases like we talked about is an open question. It's still an unanswered question is how much dystrophin do you need to mitigate the disease manifestations? These treatments and studies that have been done so far can get up to 20 to 30 percent normal levels, and it's a question of whether that's enough. Probably it is, but the trails will answer that question. And then the other interesting question in terms of genomic medicine is where we go with this. So what's the path ahead? The problem here is that every patient has a different mutation, and the treatment is targeted to the mutation, the specific mutation. There are many exons that need to be targeted, and you have to know exactly what the mutation is in order to know what exon to target with the -- with the oligonucleotide therapy. So each of these oligonucleotides need to be developed and tested. The two companies are both working on the same exon, exon 51, and each -- but there are many other exons that have to be addressed, and each will need evidence of safety and efficacy in order to get regulatory approval, and really just to know whether they work. Let's see. So here the companies picked exon 51 because that's' the one that is, you know, best targeted in the largest percentage of patients. But of the deletions, only about 18 percent can be predicted to be corrected by exon 51 skipping. So, of total patients, about two-thirds have deletions. So it's only about 12 percent of patients will have a -- will be amenable to this treatment with this particular exon. And, you know, both companies are looking down the list: exon 45, 44, 53. Each exon you target after that -- so this is a different therapeutic that has to be targeted based on the precise diagnosis of that patient, to know that this -- skipping this exon would be likely to correct their defect. It gets to smaller and smaller percentages of patients. If you get up to 12 different oligonucleotides, picking the right oligonucleotide for each one, that would cover three-quarters of the -- of the -- of the deletions, so half of the patients. So this is really the beginnings of what we're seeing as personalized medicine. It's a kind of medicine which is specifically targeted not just to the gene but the mutation in the gene that needs to be corrected. And that's a challenge, but it's a challenge that I think is going to be overcome bit by bit. It has a lot of issues with regard to safety, efficacy, FDA approval, and so on, that are being worked out here in Bethesda and elsewhere around the world. Now, one thing that would help with this is to have a good biomarker to give an early read as to whether a particular oligonucleotide is working before you see clinical manifestations -- clinical effects of the treatment. So a non-invasive biomarker that will give an early indication of biological effects. And Ami Mankodi here in the audience is working in our group and others on an imaging study here at the NIH doing cardiac and skeletal muscle imaging to see if we can pick up -- or if she can pick up [laughs] -- we can pick up an effect. It's riding on the Phase 2 multicenter U.S. trial of oligonucleotide therapy by GSK targeting exon 51. And so looking at cardiac and skeletal muscle MRI and ultrasound, here is a image from University -- a group at the University of Florida who's also involved in this kind of work with Sarepta. And you can see that there are a lot of changes in the muscle MRI. But the question is whether the drugs -- the oligonucleotides correct these changes in a way that can be seen early in the course of the treatment. Okay, so I'd like go on now in the time that's remaining to talk about another disease that's, I guess I say, close to my heart. It's one -- it's an important disease. The -- said to be the most common severe hereditary disease of infancy and early childhood, and that's spinal muscular atrophy. Now here it's a autosomal recessive disease, and it affects about 1 in 8,000 to 10,000 babies. And so the carrier frequency is about 1 in 40. Here in this room there would be two or three people who would be carrying the mutation, would be at risk of having children or grandchildren affected by the disease. It causes early onset progressive symmetrical weakness and muscle atrophy due to motor neuron loss and muscle denervation. So the problem's not directly in the muscle here, but a loss of enervation of the muscle, because the motor neurons are lost. And here the gene that was identified back in the '90s was given the name "survival motor neuron," or SMN. So it's a loss of SMN, a relative loss of SMN that leads to this clinical phenotype. Here's a picture from a -- from Victor Dubowitz from some time ago, showing the severe early onset form of the disease. The babies are kind of floppy, like little rag dolls, and really don't do very well; the survival is limited. Now -- a pretty bad disease. Now, there are milder forms of the disease. Here is a patient I saw in a trip to Dominican Republic a few years ago, and chose a patient with milder, we call Type 2, form of the disease. And there are still milder forms with long-term survival, pretty -- still pretty severe weakness, but not to the point where it limits survival in the way the severe Type 1, or Werdnig Hoffman form of the disease does. Now, here the gene that was identified back in the '90s, the SMN gene, it's interesting in that it's present in multiple copies and two different varieties. There's SMN1, which is lost in the disease, and SMN2, which is kind of a backup gene; it's very similar to SMN1, differs at only a -- there's only one nucleotide difference which really accounts for the difference in the -- in the function of these two genes; the coding sequence is the same, but there's a single nucleotide difference that leads to an effect on exon splicing, such that a particular exon, exon number 7, is missing in the messenger RNA transcript that is encoded by SMN2. So you get full-length SMN transcript from SMN1, and you get some full-length transcript from SMN2, but the majority of transcript messenger RNA from SMN2 is lacking exon 7. And, again, is -- you know, it's an unstable protein, it's rapidly degraded, and it's less functional than the full-length form. Now, the disease is usually caused by deletions of SMN1, just like the deletions in dystrophin, although here, oftentimes, the whole gene is missing. But the patients still have SMN2, so they don't have a complete loss of SMN; they have a relative depletion of SMN. And that's what leads to the disease manifestations. Now, what is SMN? SMN is expressed everywhere, it's not just in the muscle or motor neurons; it's in every cell. And it plays an important role, ironically, in splicing. The normal function of SMN, the best established normal function of SMN, is to put together this complex of proteins called the spliceosome, which is responsible for splicing introns out of messenger RNA to lead to a mature messenger RNA transcript. In addition, SMN likely has a role in the axonal transport of messenger RNA. So here's a -- here's a neuron in culture with an axon that extends out from the cell body; here's a growth cone. And you can see SMN in the cytoplasm, the cell body, in the nucleus, but you also see it distributed along the axon, in particular at the growth cone. Now there is protein synthesis occurring out here, at a distance from the cell body, and SMN, in promoting the transport of messenger RNA, may play an important role in supporting protein synthesis at a distance from the cell body. Now, why are motor neurons particularly vulnerable? I mean, this is a protein that's in every cell; it plays an important role in RNA splicing. It maybe because of this shape and dimensions of modern neurons; they have long axons, multiple muscle fibers that they innervate and large nerve terminals, and they may be particularly vulnerable to splicing defects in the messenger RNA, and/or motor neurons may be particularly dependent on this axonal transport of messenger RNA. Now, since the gene was identified, SMA has been reproduced in a variety of model systems: cell-free biochemical assays, cell culture, and a whole menagerie of animals. We have worms, flies, zebrafish, and mice that are all deficient in MSN, and they all develop some variety of motor weakness. You know, the worms don't crawl very well; the fish, like our goldfish at home, sometimes go belly-up because they just can't swim very well; and the flies, at the larval stage, they don't -- they don't crawl around very well; and the mice, the mice look a lot like the patients, I think, a mouse version of the patient. So here we see a wild-type mouse, and here a untreated mouse with SMN deficiency. This mouse model was developed at Ohio State. And you see the mouse very weak at an early age; it doesn't grow very well, it doesn't feed very well, and dies within -- in the mouse within about 15 days. Now, there are drugs that have been tested in the mice, and the drugs actually can do -- can have a significant effect; in this case, even after the onset of the disease. So it comes on around three or four days; by five days they're starting to get pretty weak. Here is a drug that -- called trichostatin that inhibits histone deacetylase, and that has the effect of opening up the chromatin around the SMN gene, allowing more transcription of the SMN -- the remaining SMN to -- gene to increase SMN protein levels. And here you can see that it has an effect, a significant effect, but it only increases the survival by about three or four days. Now, you know, it's an effect, but it's not -- it's not that much of one. Now -- whoops. Oh, yeah, that's the -- my special effect there showing the arrow. And then work in our lab has shown that you can also block the degradation of SMN. So you can stimulate its production and block its degradation with a drug called bortezomib. This is a drug that is a proteasome inhibitor, so it blocks the complex of proteins that's responsible for degradation of proteins like SMN, increases the levels, and improves the motor behavior in spinal cord pathology in the mice, and this effect of bortezomib is synergistic with trichostatin. Now, there's work done Deborah Kwon and Barrington Burnett in our lab; Deborah and Barrington have identified a specific enzyme that targets SMN for degradation called mindbomb 1, or Mib1. And this might be a more selective -- rather than inhibiting proteasomes in general, the proteases, this might be a more specific target, and that's something that we are working on now -- or the people in our lab are working on now. Now, another interesting finding, Heather Narver is a veterinarian, a mouse veterinarian across the street in our building, kind of took -- I think she kind of took pity on these little mice and had the idea of giving them some additional nutritional support, gave them infant formula once or twice a day. And that had a dramatic effect; it did not have an effect on its own, but it had a dramatic effect on enhancing the response to treatment with this drug trichostatin, so that rather than a three- or four-day effect, it was up to -- it increased the survival two- or three-fold. And this is something I'll get back to at the end, the importance of nutrition. Now, there have been efforts, screening assays and efforts to develop drugs based on a variety of different assays: just stimulating the SMN promoter to increase the levels of the transcript; increasing the retention of this exon 7, and -- with oligonucleotides and drugs; and increasing the protein levels by blocking its degradation; working further downstream to enhance its function and -- or to increase the survival of SMN-deficient cells. And there are drugs that are at various stages of development now. A couple that are in clinical trials, a drug from Trophos in France, a drug that's -- it's been a while, over 12 years now, in development, but it's finally getting into the clinic. A drug championed by a patient organization, Families of SMA, taken up by a biotech company, Repligen. And just yesterday, a press release that Pfizer has been encouraged by the Phase 1 results with this drug and Quinazoline, that blocks the degradation of the messenger RNA for SMN. And the -- so Pfizer's put it -- just like GSK with Prosensa and the oligonucleotides for Duchenne dystrophy, Pfizer's putting a big effort into supporting the development of this drug. And then tetracycline derivatives from a biotech company called Paratek to enhance the splicing. And here at the NIH an NCATS screening was done that identified a class of compounds that increase SMN levels called arylpiperidines. And there's also a similar effort that's underway at Novartis up in Cambridge. Now, again, as with Duchenne dystrophy, you can take more general approaches, but working close to the cause of the problem, working at the level -- at the level of the gene or the mutant transcript gives a more specific and potentially more effective approach to treatment. Now -- so that can be done by gene therapy, gene replacement, or, again, by oligonucleotide therapy. Now, here, not promoting exon skipping, but promoting exon retention. And I'll just describe that briefly here. Now, first with gene replacement, you can surprisingly get the SMN gene -- this is a small gene compared to dystrophin -- you can get this gene into the spinal cord, into motor neurons, with peripheral injection in mice. So intravenous injection, the right kind of adeno-associated virus gets taken up into the mouse spinal cord and into motor neurons. And here's another approach with intramuscular injection, a group in Lozan [spelled phonetically], showing, in a monkey, uptake into motor neurons in the -- in the spinal cord. And, I want to say a couple of years ago now -- well, in 2010, there were three different groups reported effective -- remarkably effective treatment with gene replacement in SMN-deficient mice. Here, a group at Ohio State, Ryan Kaspar's group was one of those three papers, shows, again, in the mice -- so these mice, again, die at 15 days, but with the treatment, they can show that they get the SMN into the motor neurons in the spinal cord, they have an effect on the behaviors, and a remarkable effect on survival. So with just one or two injections in the first few days of life, these mice, which normally die at 15 days, live out a normal lifespan. You can cure the disease in these animals with just one or two injections of AAV carrying the SMN gene; it's pretty remarkable. If we had -- I like to say if we had mice as patients we'd be all set; unfortunately, we've got to figure out how to get this to work in patients. Now, the other approach that I mentioned, oligonucleotides, was developed by -- first by Adrian Krainer's group at Cold Spring Harbor. They screened for oligonucleotides that bind to the messenger RNA around exon 7 that will -- and some of these will promote skipping of the exon, some of them will promote retention of the exon, some -- so depending upon exactly where their binding and whether they're affecting -- how they're affecting the splice site. And the -- what they found is that with -- injecting this into the -- into the cerebral ventricles that get long-term expression and benefits. So here exon 7 inclusion, which is at a low level in the untreated animals. Here for after a single injection these oligonucleotides, just one injection, you get -- whoops -- you get, you know, effects that last up to six months -- you know, virtually normal expression of the protein after six months. And then, more recently, they showed -- marked survival effect with -- also with subcutaneous injection. So you wonder how that works. Some of it gets into the central nervous system, but some of it might be working peripherally also. And they came up with an idea that it might be having this effect through -- by correcting a deficiency of insulin-like growth factor one, which is -- has -- is a neurotrophic factor for motor neurons. And it's not an effect directly on IGF-1 but on a binding protein, IGF ALS, acid label subunit, that correct the deficiency of this binding protein that corrects the levels of IGF-1, and that has a beneficial effect on the mice. But here again -- so the mice without treatment die within 15 days -- whoops, [unintelligible] -- and the mice with treatment, increasing doses of the treatment, have, you know, survival that goes up to a normal mouse life expectancy of, you know, one to two years. Now, this is a news article in Nature Medicine that just came out. If you want to read more about it, it's a good place to get caught up on where therapeutics development efforts are now. But just to illustrate here what a remarkable effect you can have on these mice. So here's a mouse without treatment, dies as a newborn, and here's a mouse with the treatment who looks like a normal mouse. So call in the backup, I mean, it's getting -- you know, activating SMN2, stabilizing the transcript, and enhancing the retention of exon 7, and this is the effect. So, again, you know, very effective treatment in mice. How do we -- how do we get that to work in patients? Isis Pharmaceuticals is taking this into patients in Phase 1 studies, intrathecal injection, and they have a lot of support from Biogen Idec -- financial support from Biogen Idec. But it's a challenge -- it's a challenge to get this to work in patients. We found this now with several diseases: People are much more heterogeneous than animals. The laboratory animals we use are pure bred, they have a very predictable phenotype. As I mentioned, the phenotype varies in patients, from very severe to less severe. So, you know, dealing with the heterogeneity is an issue. Also this disease, after the early onset, those patients that survive have a very stable course, very slow disease progression after onset. So showing an effect in the survivors on the progression is hard. And most -- or an important aspect of this issue to address is the need for the treatment to start early. In the mice, you have to give the treatment within the first few days. And the challenge -- the clinical challenge we have to face is how do we identify patients with this disease early enough to have it have an effect? One way to do that -- you know, we can do genetic testing. It's very easy and generally available kind of genetic test; not as cheap as you would like, but there are efforts to reduce the cost to do genetic testing at birth, in newborns. But it's hard -- there's a chicken and egg kind of situation of justifying broad newborn screening to pick up a disease that affects 1 in 8,000 to 10,000 without having the treatment. And it may be hard to develop the treatment without having newborn screening. So I think -- I think this is an overcome-able problem. It's something we've had discussions in NINDS and the Genome Institute, is how to -- how to deal with this. The Child Health, NICHD, has supported development of newborn screening for SMA with the idea that this would be something that would spur therapeutics development, and -- as well as providing the information to families about the diagnosis. But this is an issue that needs to be addressed, I think, for this approach to be successful. Now, what did I learn from my time at Novartis? The company, remarkably, like other companies, has taken an interest in rare diseases like this, rare genetic diseases. There are about 40 different rare diseases that Novartis is working on now. So it was fun, as an academic investigator, to be up there and see that work, see the company using the expertise that they have and resources they have for therapeutics development to take on diseases like SMA. How do they choose the diseases and how do they -- how do they develop what we call a clinical proof of concept, so to show that a treatment works in early-phase clinical trials. They select diseases based on unmet medical need, so clearly we have that with SMA. A strong biological rationale for the therapeutic approach, increasing SMN levels is -- clearly works in animal models and would be expected to work in patients as well. Favorable risk-benefit ratio, to have the right kind of safety profile on -- for the potential therapeutic agent to match the disease manifestations. You don't want to use cancer chemotherapy-type drug for headaches, for example; you want to match the risk to the benefit. Every drug had side effects, and, you know, the issue here is -- the issue here is getting this on. Let me see if I can get this done; get a little klutzy as I get older here. Maybe -- let's see; maybe you have to do "end show" -- oh, I know. Yeah, okay. So the -- and then this is an important thing. Having a biomarker that tells you that you've engaged the target. So just to know -- it's called a pharmacodynamic measure -- to know that the drug, the potential drug, has the biological effect in a patient that you want it to have; to know whether the drug's working biologically, and then to know whether it's working clinically. And then clinical trial readiness, to know that patients are available to have reliable clinical outcome measures and natural history data to know that you're having an impact on the -- on the clinical manifestations of the disease. Now we have each of these with SMA. We have unmet need, strong biological rationale, favorable risk-benefit ratio with the drugs that are being considered. This is the hardest one: to know that in the central nervous system we're having an effect with the treatment that's designed to increase SMN levels, but there's a lot that's been done recently with developing good tests for SMN levels and to get things ready for clinical trials. There have been clinical trials -- there are clinical trials ongoing, but to make sure that the patients are available and willing -- and interested in getting involved in the clinical trials -- the clinicians are interested in referring patients for clinical trials; and that there are good outcome measures in natural history data. So we're getting close. Now one thing to say just at the end here, I've mentioned before in other talks, is an interesting thing that's been going on with Duchenne dystrophy and spinal muscular atrophy, is that the patients are doing better. I mean, all the while we're trying to develop drugs, small molecules, gene therapy, oligonucleotide therapy, the patient's seem to be doing better; the survival has improved, as I said, with Duchenne dystrophy, from 20 to 30 years. So a 10-year increase in survival over the last few years -- over the last decade. And also with spinal muscular atrophy, here's a study done by Petra Kaufmann, who's now in our institute, did this when she was at Columbia in New York, showed -- just reviewed the history of the severe Type 1 SMA patients, here back in the '80s and early '90s. Most of them died within a few years, really within two to three years of diagnosis. But then looking at patients in the late '90s into the aughts here, the current century, the survival is much better. And likely -- this is probably due -- there's a follow-up study that's being done now, but it's probably due to a better respiratory nutritional support. You can offer these patients a lot with feeding tubes and external non-invasive ventilator support, good, you know, aggressive respiratory management. And the quality of life is -- if you ask the patients and the families, it's not that bad. It's, you know, it's pretty remarkable what can be done just by looking to see what do we currently have to offer these patients in terms of symptomatic support, respiratory nutritional support, and how can we make that better. So just to close, this is our group over across the street. There are few extra people in this picture, but this is a neurogenetics branch; it's our lab. And, in particular, I wanted to talk -- point out Carsten Bonnemann who's -- has a group that's working on early onset neuromuscular diseases; Craig Blackstone; Ricardo Rodas [spelled phonetically] in that group working on spastic paraplegia. I wanted to highlight Ami Mankodi in the -- in the back there, who's doing the Duchenne muscle imagining study that I mentioned. And a nice group in our lab working on spinal muscular atrophy, including Barrington Burnett, Deborah Kwon, Katherine Birchenough, and others. So this -- it's been said some time ago, when I say "I," I mean "we," and when I say "we," I mean "them." So the work that's been done has really been done by that group over across the street. Thank you. [applause] Male Speaker: No comments or questions [unintelligible]? Male Speaker: I have one. Male Speaker: Go. Male Speaker: About how large the oligonucleotides are, and whether they get into other cells? Kenneth Fishbeck: Yeah, yeah. So the question was how large are the oligonucleotides? I think they're about 15 to 20 nucleotides long. And for the Duchenne treatment they get into the muscle very well; they're pretty widely distributed. But it's -- you know, it's really in the muscle where they're needed, and they get into that. It -- you know, there's less entrance into the central nervous system. You can do that in newborn period, but -- in animals, at least. But to get into central nervous system, it helps to give it intrathecally or intro the cerebral ventricles. So the clinical trials are using intrathecal delivery for this disease. Male Speaker: And so you're not really terribly worried about toxicity -- Kenneth Fishbeck: Yeah, we're always worried about toxicity. So the question is, are we worried about toxicity, and, you know, there are some, you know, concerns based on the animals studies and the clinical studies that have been done with oligonucleotides. I don't know, I think it maybe goes back to Hippocrates or somebody that every drug is toxic, it's just a matter of dose. And so to get the dose right, I think, is the challenge here. Male Speaker: [unintelligible]? Kenneth Fishbeck: Yeah, exactly. Yeah? Male Speaker: Would it be reasonable to assume that the SMN therapy, at least that therapy, could be applied to the adult form of motor neuron diseases [spelled phonetically]? Kenneth Fishbeck: Oh, you know -- ALS or Lou Gehrig's disease. You know, there's been some evidence that there's an association with SMN copy number, or SMN levels with ALS, and it's been back and forth. I think the results are -- the results are really mixed on that. It's not a strong correlation. I mean, I think that's the hope, that there might be some way to get -- that SMN might be important in other motor neuron diseases, and it's something that we're investigating, too, but it's not clear that that will happen. It might. Male Speaker: Please join me in thanking Dr. Fishbeck. Kenneth Fishbeck: Thank you. Thank you. [applause]