Fourth Annual Friedreich's Ataxia Symposium

Transcript: Research Topics and Discussion

Transcripts in This Video Series
Neurological Features of Friedreich's Ataxia Diabetes in Friedreich's Ataxia Biochemical Mechanisms and Pathways Questions and Answers Session - Clinical Panel Research Topics and Discussion

Printer Friendly Version

David Lynch: And we will shift gears to welcome some research topics for discussion. We have, courtesy of FARA, many individuals who are in town to talk about cell based approaches to the study of Friedreich's ataxia, so we have truly the best scientists in the world on this project here, and we've arranged for three of them to give, four of them to give their 15 minute presentations on what they do and how it's applicable to you. First, we'll have Mirella Dottori from Australia come up and speak for us.

Mirella Dottori: Thank you. [applause] Thank you and hello. First, I'd like to thank the organizers for the opportunity for me to be here and meet with you all. It's really a pleasure to be here and hear about all the developments and especially meet with the patients. So from a scientist's point of view, when we think about developing treatments and cures for a disease, first thing we think about is, what are the specific features of that disease that we need to target? And, first I need to work out how to move this by remote. Oh great, that's it, thank you. Okay.

And as you're all aware, the unique features of Friedreich's ataxia is that within the frataxin gene there are mutations. Most commonly, there's this GAA repeat expansion, and this causes very, as you all heard about today, very low levels of this frataxin protein being made. And later on, you'll be hearing about the function of frataxin protein from Helene. But the other curious feature about Friedreich's ataxia is that frataxin's expressed in every cell type in the body. However, for some reason these specific cell types tend to be far more sensitive to these low levels of frataxin, in particular cells within the heart and within the nervous system. And that's really important for us, because that's basically the target cells that we need to think specifically to cure and treat.

So how do we do this? So basically, as scientists, what we really want or need to have is human disease cells in a dish, so that we can study these. We can study the pathology. We can also look at different drugs to treat this disease that you'll hear from Joel and Marek as well. And, so far the type of human cells that we've had in a dish have been blood cells or fibroblast cells, which perhaps some of you have already donated to research, which we're grateful for. But ideally, we'd like to have the cell types that are specifically affected in the disease, so cardiac tissue and nervous tissue.

So how do we do this? That's where stem cells come into it. Now, there's various different types of stem cells. You would have all heard about stem cells in the media and so forth. And the most common types that you hear about are usually adult derived stem cells, or cord blood stem cells. And the way we define different types of stem cells is according to the potential of what they can become. So adult stem cells, as well as cord blood stem cells, are what we call multipotent, so they can become different cell types, but it's usually of that cell type of the tissue of origin, that they can't become other cell types. So for example, a blood stem cell can't become a nerve cell, and vice versa.

Now, one particular type of stem cell is called a pluripotent stem cell, and the unique feature of that is that it can become all cell types of the body. Now, up until a few years ago, the only way in which we could derive pluripotent stem cells was from embryos. So this, in the case of human embryos, they were derived from excess IVF embryos that were generated only for the purpose of IVF, and then they were no longer needed and so they were donated to research. They extracted out these embryonic stem cells and they kept them in culture for several years. And these cells have the potential to become all cell types of the body.

Now, there's two main issues with embryonic stem cells. Firstly, it's an ethically sensitive issue, and secondly, they're not derived from the patient. Now, a few years ago, these two issues were addressed with a big medical, or breakthrough, in the stem cell research world with the derivation of what we called induced pluripotent stem cells. Now groups, this was initially led by a Japanese group, Yamanaka and colleagues, where they were able to define, this is pretty amazing, the four or five main proteins that are involved in making a stem cell a pluripotent stem cell.

So what they did was that they isolated fibroblasts, initially from mouse tails, and then they even did it from humans, and they introduced these four or five proteins into these fibroblast cells. And essentially, they converted, or what we say, reprogrammed, these cells to become pluripotent. So they have the same characteristics as embryonic stem cells.

In the case of the mouse tissue, they were able to actually create a whole new mouse with it. So you can imagine this technology has rapidly infiltrated a lot of different areas into medical research, because now we have way in which we can derive stem cells from patients. So essentially we take skin biopsy samples, we introduce these four or five different transcription factors, or proteins, and that reprograms or converts these cells to take on embryonic stem cell characteristics. And, that is highly useful because from becoming stem cells we can then push them into the lineages that are specifically degenerating in this disease.

And this is extremely useful for drug development, because we have these diseased cells in a dish, but also, ultimately for transplantation. Now, one important thing I want you to keep in mind is, at this very stage, IPS cells are not suitable to be used in the clinic. Even though they're very, very close to ES cells, or embryonic stem cells, they're not equivalent. But the technology is moving so rapidly forward that they are already developing ways in which to derive these cells that are therapeutically relevant, but there's still some important hurdles that we have to get through.

So in the Friedreich's research world, there have been some groups, including from our lab, Joel's lab, Helene's, Marek, actually all of us who are here at the moment have all generated IPS cell lines from Friedreich's ataxia patients. And yesterday, we all got together and had a workshop and pooled our resources and heads together in order to move this forward as quickly as we can. It was very productive. And, for each of our groups we're using these stem cells for our interests, for our main purpose in the expertise in the lab.

So in our lab, we generated IPS cells from two biopsy samples taken from Friedreich's ataxia patients, and just to show you here, this actually a clone, a colony, sorry, of stem cells. In this case, it's derived from Friedreich's ataxia skin samples. This is from a control skin sample, and this is a human embryonic stem cell colony. So in each of these colonies, there's about 20,000 cells, and we culture them on fibroblast layer. So they all look fairly equivalent.

The important feature is, do they retain the characteristics of this disease? And the answer is yes. So we see that in the stem cell lines, they still retain the expansions on both frataxin alleles, and importantly, in these stem cells, they still have very low levels of frataxin protein, which is what we want because we need them as a model. But now, the main issue is, can we push them, or we say differentiate them, into the cell types that are specifically degenerating in this disease? And that's where the challenge lies.

So in the case for the nervous system, we're interested in pushing them more towards cerebellum neurons. As you heard earlier today, the cerebellum is particularly affected in this disease, and also into cell types of the peripheral nervous system, which is the dorsal root ganglia, which David explained to you earlier this morning, the large sensory neurons which is one of the first cell types that tend to degenerate. And also, of course, we're interested in pushing them towards cardiac cells.

So how do we do this? So our lab's been working with stem cells for several years now. We have pretty good protocols in place from going from a pluripotent stem cell to an end nerve cell type. And we adapted these protocols to these Friedreich's ataxia IPS cells. So basically, from these pluripotent stem cells we can push them down the lineage to get derived neural stem cells, and from there, we can also derive them into more mature nerves, and also other cell types of the nervous system, such as glial cells, which are supporting cells. Also, we have protocols in the lab to push them more towards the peripheral nervous system, so they're all the nerves through outside of the spinal cord and brain, and now we're trying to derive methods not just to the peripheral, but specifically of those of the dorsal root ganglia, of the large sensory neurons.

So with these Friedreich's ataxia cells, if we just look at them down the microscope, they look fine. They even make beautiful neurons. You can see these beautiful bipolar neurons here, which are mimicking, like, sensory neurons, but what's actually wrong with them? We need to find a pathology for them to be used as a model. So if we look deeper and we use microscopes that are called electron microscopes so they can amplify seeing something even 100,000 times fold, so you can actually look inside the cell to see what's going on, so you can see the nucleus, you can see the mitochondria, and other components.

And when we look at these electron microscope images of the Friedreich's ataxia neurons, you can see evidence of cellular stress. So you see these large inclusion bodies, these lipid accumulations and vacuoles. So that's quite good for us and quite promising, because we want a pathology in these cells so that we have something to work with and to actually form treatments for drug discoveries specifically in these cell types. So while we're moving forward in this, it's very important to have our controls. So we need to show is this reversed if we introduce Friedreich's, frataxin back in? And that's what we've done.

We've developed these Friedreich's ataxia IPS cells, and you can do a lot with stem cells. You can genetically engineer them to remove mutations, you can put in proteins, you can put in colors, you can make them blue, green, red, whatever you like. So here's a colony of Friedreich's ataxia cells. Again, there's about 20,000 cells, and every green cell represents a cell in which frataxin protein has been introduced. So we're in the process of purifying these cells, and now we can use them as a comparison to say, "Okay, is the pathology reversed in these cell types?" So that's where we're at at the moment.

And while all of this is going on in the background, we're also looking at other systems in which to investigate these cells. And transplantation is a big one. Again, to transplant these cells, firstly we need to develop tools. We've genetically modified them so that they express green fluorescent protein that's showing here, so that we can track them when we put them in animals in vivo and follow them through. From there, we make them into neural stem cells, and then we transplant them in-- specifically into the regions that we know are predominant problems in Friedreich's ataxia, in the cerebellum.

And one thing with a lot of animals and humans is that, amazingly, a lot of regions are highly conserved, especially in the cerebellum. The cerebellum's the region down here, I often say it looks like a cauliflower, and in the mouse, here it is around there. So we've transplanted some of these Friedreich's ataxia IPS cells into the cerebellar regions of these rodents in the adults, and all these dark regions here represent the human Friedreich's ataxia IPS cells. And here's a cerebellum, as I was saying it looks like a cauliflower, so you can see this, this is the structure here. And quite nicely, when we look under high magnification we can see that they form beautiful neurons in this region.

So again, now we're looking at certain things about this aspect. Are they functional? And then we do the same sort of experiments with the corrected cells and with frataxin back in there to look for differences. Now this is quite exciting for us. One word of caution, that I often get emails from a lot of patients saying, "Can we transplant these stem cells in us? You know, please, we'll be there, we'll fly to Australia, we'll get these things going." And there is a lot of what we call stem cell, sort of, like, hype, in certain countries where they transplant those stem cells.

One huge word of caution is that we're not there yet. We're pushing forward very, very hard for transplantation, but there's a lot of hurdles. There's a lot of dangers associated with that, and there's a lot, a lot, of unknowns. And, believe me, if we could do it, we would do it right now, tomorrow. But, we're working on it and, at the moment, we need to do it in animal models to see what actually happens. So my colleague, Alice Pebay, who group with the O'Brien Institute, is looking at ways to push these stem cells into cardiac tissue.

I won't go into the data at the moment, but just to show you that they can do this. And so, these Friedreich's ataxia IPS cells that they push them into early cardiac progenitors, and you can see down in a culture dish they are already starting to form beating, like, heart-like tissues, which is pretty cool. Then again, they're looking further into the pathology in these cells. So basically, that is the huge strength in this technology. The fact that we can now derive stem cells from patient samples is fantastic for us for various different purposes. That we can have these cells in a culture dish to look at how a drug specifically acts on that cell type, which you'll hear from Joel, as opposed to a blood cell or a fibroblast cell, and also, ultimately for transplantation.

And, like everything, we don't work alone. This is my lab in Australia, a small group but extremely hard working, but we collaborate with many other groups in this, in both generating the cells and characterizing it. It's a real team effort on an international level. So, and thank you for your attention.

[applause]

David Lynch: We'll take questions from everyone at the end of our four presentations. So next I get to introduce my friend Joel Gottesfeld from the Scripps Institute, who's going to talk about something. Whatever you want to talk about today, boss.

Joel Gottesfeld: Thank you, Dave. And I want to thank the organizers for bringing the group of scientists together such that we could have very productive discussions yesterday and to have the opportunity to meet all of you and share the status of our research with you today.

So my lab has been focused on developing therapeutics for Friedreich's based on trying to understand the mechanism of gene silencing in the disease. And as previous speakers, Dave, Rob Wilson, Massimo, and everyone has described, the preponderance of Friedreich's patients have these GAA repeat expansions in the first intron of the frataxin gene. Early studies suggested that those GAA repeats block the ability of RNA polymerase, that is, the enzyme that copies genes into messenger RNA, which is what is copied into proteins.

So the GAA repeats somehow block the passage of RNA polymerase through the gene. Now, as you've heard, since the repeats are in an intron, there's nothing the matter with the coding potential of the frataxin gene on pathogenic alleles, that is, the copies of the gene in Friedreich's patients. So one therapeutic approach, as Rob Wilson described, would be to increase the level of transcription that is copying the frataxin gene into mRNA, and hence, you should get higher levels of frataxin protein, and hopefully restore some aspects of normal physiology in patients by virtue of increasing levels of frataxin protein.

Okay. Now, to develop therapeutics along these lines, you have to understand how it is that the GAA repeats silence the gene, how it is that they block this enzyme, RNA polymerase, from getting through the gene. Well, studies in a number of labs have shown that these GAA repeats form unusual DNA structures, and that could be the heart of it. But what my lab showed several years ago, and published it about 2006, is that the repeats cause the normally active frataxin gene to go into a form in the chromosomes called heterochromatin.

So in this cartoon here at the bottom, you see a transition from an active frataxin gene through several steps to a coiled up, inactive frataxin gene. The steps are removing little chemical tags on the histone proteins. The histone proteins package DNA in the cell nucleus. And a series of enzymes are responsible for regulating this pathway between active chromatin structures that are able to be copied into messenger RNA and inactive, or heterochromatin, structures that are repressed. Some of these enzymes have to do with the post synthetic modification states of these histone proteins. One of these modifications is a little group called an acetate group that goes on the tails of the histones.

There are enzymes called histone acetyltransferases, or HATs, that put those acetate groups on the histone proteins, and another class of enzymes, called histone deacetylaces that remove them. Okay? So if you were able to shift the balance between the HATs and the HDACs, you might be able to shift the balance between an inactive gene and an active gene, and hence, relieve repression of the frataxin gene in Friedreich's patients. So what my lab set out to do back in 2005, 2006, was to identify small molecules that might be able to act as inhibitors of these histone deacetylaces, or HDACs, and we came across a class of molecules, here's one shown, that would actually do this in white blood cells taken from Friedreich's patients. To convince you that this really works, we use a technique called Western Blotting that Massimo introduced earlier today, where we look at both the level of histone acetylation in the cell, and the level of frataxin protein in the cell.

On the top set of gels, these are protein gels to analyze the levels of histone acetylation and frataxin protein, we see that this molecule, synthesized in our lab called 106, does increase the level of histone acetylation as seen by this difference between this lane marked zero and those lanes, the bright, dark spot is the acetylated histones, this is total histones. And then, a few hours after we see this acetylation, we see increases in frataxin protein.

Now, much to our surprise, another HDAC inhibitor, which we show as very active in cells, has absolutely no activity on the frataxin gene. This is actually a FDA approved cancer therapeutic called SAHA. It's a very potent HDAC inhibitor, but it doesn't work on the frataxin gene. We've done a lot of chemistry to understand why it is that this molecule works and why it is that this molecule doesn't work. But to suffice it to say, that these molecules really do work in patient's white blood cells. They also work in a mouse model, work done in collaboration with Massimo's lab, and more recently, in collaboration with another lab in London, Mark Cooks lab, has shown that these molecules are able to cross the blood-brain barrier and increase frataxin messenger RNA and frataxin protein in these mouse models.

Well, that's all well and good, but what about human neurons? So I'm going to follow on with what Mirella talked about, and that is, derivation of a human neuronal cell model for Friedreich's through the same technology that Mirella talked about, that is, induced pluripotent stem cells. So as she mentioned, Shinya Yamanaka in 2007 published a landmark paper where he identified four cellular proteins, that when introduced through genes into skin fibroblasts, you can turn those skin fibroblasts into stem cell-like cells, or induced pluripotent stem cells.

We've characterized these cells quite fully with respect to their pluripotency. They do mimic all the properties of authentic ES cells. They also mimic the Friedreich's condition. That is, they have reduced levels of frataxin mRNA and frataxin protein, and as Mirella indicated, they retain the GAA repeats. In fact, as one propagates these cells, the repeats tend to expand, and this allows us to have a good cellular model to try to understand the mechanism whereby the repeats are actually expanding. But that's a subject for another day.

Okay. So, can we take these cells, now, and derive neurons? Well, with Mirella's help and a publication from her lab several years ago, we followed her protocols to go from the IPS cells through induction of a structure called a neurosphere with a small molecule called noggin, and then through various cell culture techniques, we can plate the dispersed cells from these structures called neurospheres onto a surface stained for a neuronal cell marker, and you can see this bright green staining, very similar to what Mirella said. We can use a technique called FACS analysis: fluorescence activated cell sorting analysis. And we can estimate the purity of these cells.

Now, I'm a biochemist, not a cell biologist, so for the experiments we do in our lab, we like to have pure populations of cells. By using Mirella's techniques and a little tweaking of them, we can now get greater than 90% and quite often 95% of positive staining cells. So this is now adequate for us to do biochemical experiments to try to understand the mechanisms of gene silencing in a relevant cell type, that is, neurons, derived from Friedreich's ataxia patients, and also test our compounds. Okay. How do I go back?

Do these neuronal cells also mimic the Friedreich's phenotype? The answer is yes, they have reduced levels of frataxin mRNA, so if we set the level of frataxin mRNA in neurons from an unaffected individual's IPS cells to a value of one, the Friedreich's neurons have a value of about 25% or so, just like you would see in lymphocytes or other cells, fibroblasts, from a patient. And when one does this technique Western Blotting again to look at the levels of frataxin protein, you can clearly see that the Friedreich's neurons have reduced levels of frataxin protein compared to neurons from an unaffected individual.

When we look at the levels of histone acetylation, remember that mark that determines active chromatin structure versus inactive chromatin structure, again, we see that the Friedreich's neurons have lower levels of acetylation near the GAA repeats compared to this frataxin gene, the same regions of the frataxin gene in neurons derived from unaffected individuals' IPS cells. Okay. So now the most important question to us, very satisfying result, was would our compounds actually increase frataxin mRNA and frataxin protein in this human neuronal cell model? And the answer is yes.

These blue bars here on this graph on the left show the relative levels of frataxin mRNA measured in, here in IPS cells from an unaffected individual, again setting the value to one. The Friedreich's IPS cells, Friedreich's neurons, retain the same level of silencing as the IPS cells. Then when we treat these neurons with an active HDAC inhibitor, we see large increases in frataxin mRNA and when we do that technique, Western Blotting again, we see increases in frataxin protein that parallel the increases of frataxin mRNA. And, much to our satisfaction, when we use a molecule that is structurally similar but we know to be inactive, so this is sort of a chemical control experiment, it has no affect.

Okay. What did we do to the frataxin gene? Well, again, we go into the frataxin gene and probe the level of histone acetylation on the regions immediately adjacent to the repeats, and what we see is that we've increased acetylation through our molecules right on the gene. That says that we've, that the molecules are working directly to change the chromatin structure, the chromatin post-synthetic modification state on the frataxin gene, thereby relieving repression.

So, now we've heard a lot today that Friedreich's is a mitochondrial disease. A graduate student in my lab, Sherman Ku, has done a lot of work to try to find a biochemical signature of mitochondrial dysfunction in these Friedreich's neurons. And what he found was that he could use a dye, a very standard dye called tetramethylrhodamine, that measures the electrical potential of the mitochondrial membrane. And it turns out, when he does this assay, this assay has been used in the past in Friedreich's fibroblasts, for example, when he does this assay on the Friedreich's neurons and compares the value of this mitochondrial membrane potential in the Friedreich's neurons versus neurons from an unaffected individual's IPS cells, he does indeed see a statistically significant difference. This photograph here just shows you where mitochondria are. So the green stain is the nucleus DNA, staining in green. The red are mitochondria. Those are not neurons, those are actually lymphocytes using a related dye. Anyway, the important question was, "Now if you can increase frataxin mRNA and frataxin protein with our small molecules, can you restore this measure of mitochondrial function?" And, indeed, you can.

So treating the Friedreich's neurons or the unaffected neurons with our drugs, the histone deacetylase inhibitor, you can clearly see that we go from this reduced mitochondrial membrane potential back to the same value as you see in the neurons from an unaffected individual. It took three days to get there, but at least we got there. Okay? By that I mean, we had to keep the neurons in culture with this drug for three days before we saw this effect. Whereas, we would actually see the increases in frataxin mRNA after about 12 hours, increases in frataxin protein after about 24 hours, but we have to wait three days to see the membrane potential to come back.

That means we have had to make new frataxin protein and the mitochondria had to become healthy. Okay. So will these compounds lead to drugs? I'm sure that's the foremost question on most of your minds, not the basic science. Back in 2007, our compounds were licensed by the Scripps Research Institute where I work to a small biotech company in suburban Boston. The company is called Repligen Corporation. The first thing they did was to test our molecules as to whether they are drug-like. They are. They had good, fairly good pharmacological properties, but not perfect.

So what Repligen did was they synthesized a library of about 200 derivatives, assayed them for activity as HDAC inhibitors, sent a set of those molecules to my lab, some to Massimo's lab. We tested them in both cell-based assays for increase in frataxin, and in Massimo's mouse model, and after several years of testing, Repligen carried out full pre-clinical studies on a molecule called 109, which is the molecule that I used in all those previous studies. They're now calling it RG2833.

Back in May or June of last year, Repligen filed an investigational new drug application with the U.S. Food and Drug Administration to try to take this molecule into human clinical trials. However, the molecule was, that application was placed on clinical hold because in the pre-clinical studies, Repligen found that at approximately 100 times the therapeutic dose that you would imagine that we would use in humans, there was reproductive toxicity in rats. And so, because of that, Repligen had to go back and do a long series of additional studies to resolve this issue of toxicity. The good news is that they have now completed these studies, and a full clinical development plan will be submitted to the USFDA by the end of the year.

In the meantime, Repligen went forward and submitted an application to the European Medicines Agency in London. That was approved, and they are now pending approval of the Italian Regulatory Agency. As Massimo described, you would need European Union approval, that's the EMA, and then individual country approval to go into clinical trials. So we should, we're hoping for approval this month to initiate a Phase 1B trial in patients at San Luigi Hospital in Turin, Italy. This will be a crossover study. That is, all the patients, there will be 20 to 25 patients, they'll all be given either placebo or drug initially, and then a few weeks later, crossover those who are given placebo first will then be given drug, and vice versa.

What do we hope to learn? We hope to learn whether the compound is safe and well tolerated first of all, and secondly, because we have a biomarker for this disease, that is, frataxin mRNA and protein, we hope to see whether, when we draw blood from the treated individuals, whether their lymphocytes now have increased levels of frataxin mRNA and protein relative to their baseline levels measured pre-treatment. So we're very optimistic that this will go forward by the end of the year.

Now, I'm probably talking over my time, but I'll just finish-- it's okay? The boss, Jen Farmer, says it's okay. Okay. Well, no drug is perfect, and what I've learned in this journey in therapeutic development, something I've never done before in my life, is that you never know when you have the perfect molecule. There's always another molecule you can make. Seriously, you really know, you never know when you're done. Now, the molecule that Repligen has, RG2833, is really good, but it has limitations.

It doesn't have perfect brain penetration, and it's not perfectly stable in animal studies, or even ex vivo studies in human serum, for example. So Repligen has spent quite a lot of effort in collaboration with my lab to improve on these features. And, they've done so by doing a little tinkering with the molecule, and now we have backup compounds that have improved brain penetration. That is, when you inject a mouse with this molecule, measure the amount of molecule in the serum and in the brain, they're equivalent. Whereas, the earlier molecules only had about 15% brain concentration relative to serum.

These molecules now are also much more metabolically stable. They don't break down in the way the previous molecules. And they're equally active. Here's an assay for one of these new molecules in the human neuronal cell model comparing it to RG2833. Okay. So what I've shown you is that Friedreich's is a chromatin disease. That is, somehow the GAA repeats silence the gene through chromatin structure. We've identified molecules that will reverse this silencing. We've taken a path towards clinical development, and hopefully, next time we get together we'll have results from a clinical study to share.

Of course, I didn't do any of this. I have the lab, but all of these individuals, both in my lab and collaborators at other locations were responsible for the work. And, of course, we couldn't do this without support from FARA, National Institutes of Health, and other organizations, and I thank you for your kind-- [applause]

David Lynch: So I knew you were going to talk about something, Joel. So next is Marek Napierala, another of our collaborators who hopefully will not remind me that I owe him some fibroblast cultures.

Marke Napierala: I'm not going to. Okay. So I will do some neck exercise first and I'll start to use the right screen. Okay, this is working. So what I'm going to talk about today is something which I call, "the looking for a needle in a haystack." But using the right tools. I'm going to talk about our research towards identifying new drugs for Friedreich's via high-throughput screen. And I'm going to try to explain what high-throughput screen is and how we design it and how we actually conduct it. Okay.

So the outline is, I'll talk very briefly about the drug discovery. I'll talk on the one, two slides on mechanisms of Friedreich's ataxia. We heard it today a few times already, and then I will talk about high-throughput screen to identify the new drugs in Friedreich's.
So the first question is where our common drugs are coming from. Where most of our drugs come actually from something which we call the medicinal chemistry folklore. So they are coming from the past, something which used to be a poison, now it's a drug. About 25% of drugs which are actually U.S. and in U.K. contain compounds which are derived from plants, and they were known long, long ago. The other thing is, our drugs come also from quite accidental drug discoveries. That's quite a lot of drugs.
What you heard a few minutes ago from Dr. Gottesfeld's group, nowadays, the major impact is on so-called rational drug discovery, so Joel's group knew the mechanism, knew at least half of the mechanism, and knew how to design the drug to attack this mechanism of the disease. What I'm going to talk about is also part of some sort of rational drug design. We know the mechanism, but we're going to go to search for the drug between hundreds of thousands of compounds using a high-throughput screening strategy. So how to actually find the drug between so large a number of compounds?

So scientists believe right now that there is in nature approximately 10 to the 40th, which I don't even know how to say it in English, I don't know the number, 10 to the 40th possible drug molecules. So this is how, when you write down all zeros, that's how many is this, this is a comparison to U.S. data so it's much more than the total U.S. data. And, thus far the register of chemicals are on the approximately 30 millions. So it's a small, small number. However, the chemists also believe that if we have a library of approximately 50 to 100,000 compounds, and assuming this library is fairly diverse, the diversity within this library should be good enough to discover new drugs candidates, though they are likely to not be drugs, they will be drug candidates, which can be then further modified to be a drug.

So now, switching a little bit to something which you've seen today and heard about a few times, but this is the basis of our high-throughput screen, so I would like to repeat it. The basic of molecular pathogenesis of Friedreich's ataxia is the expansion of the GAA repeats, and while unaffected individuals have usually under 40 GAAs in the intron line of Friedreich's ataxia gene, the affected individuals, patients, have much more, sometimes even up to almost 2,000 GAA repeats. And what you've seen on presentation of Dr. Gottesfeld, this GAA repeat expansion in the Friedreich's ataxia gene causes the reduction of the transcription production of the mRNA, which then leads to the reduced level of frataxin protein in a patient.
So, Friedreich's ataxia patients have approximately 5 to 50% of the level of frataxin which can be found in unaffected controls. So one of the obvious moments of the therapeutic intervention is to try to increase this level to at least the level which is found in the carriers, the symptomatic carriers. So that was our goal when we were designing the high-throughput screening. But the question is how to actually analyze a million or more compounds. It's kind of like looking for a needle in a haystack. Well, we can look for a needle in the haystack if we have the right tools.

So what do we have? We have a metal detector, and this metal detector is the high-throughput screening assay. So with the emphasis on the assay, and this assay is specific for Friedreich's ataxia. So what is the assay? And why do we need an assay? So, if you look at the Friedreich's ataxia gene, you have an exon 1, exon 2, and those of the GAA repeats in the intron 1 to detect the changes in expression of the gene, the changes in the RNA or the protein, it is quite difficult to run those gels which Dr. Gottesfeld showed you, the Western Blot takes time, up to hours, sometimes even days. So it's impossible to analyze a million compounds using this technology.

We definitely need to replace the Friedreich's ataxia gene with something much simpler. So we designed the reporter gene, which is similar and I emphasize the word similar, it is not an identical gene, it's similar to the Friedreich's ataxia gene and closely resembles the mechanism of genetic defect in Friedreich's ataxia. In other words, we have also two exons and we have an intron and the GAA repeats in between. And we can observe silencing of the gene in the same way as the silencing is observed in the Friedreich's ataxia gene.

But, the reporter gene is different. It's not a Friedreich's ataxia gene. We designed this gene so the detection of its expression is very easy using either light or a color, something which we could do actually in milliseconds. And by doing that, we are able to analyze hundreds of thousand compounds a day. So how this is actually done, and this is done, I need to say, we don't do it in our lab, we collaborate with a so-called high-throughput screening facility, which is a part of the Scripps Research Institute in Florida, and all the high-throughput or the design of the assay is done in our lab and all the post-assay analyses are done in our lab, but the high-throughput screen is actually done in the Scripps Florida facility.

So in the traditional experiments, we'll have the patient cells on the plate, we'll have to isolate RNA protein. All this stuff is done practically manually, and it takes hours. How many compounds can we analyze? We can probably analyze 5, 10, if somebody really works hard, maybe analyze 50 compounds a day, but that's the maximum throughput of the system. When we have a high-throughput screening assay, we use completely different equipment. There is no manual labor. We use reporter cells, so not Friedreich's ataxia cells, and we analyze them straight away without any isolation using a robotics setup.

So we are able to analyze anywhere from 100,000 even 200,000 compounds a day. So just to show you an example how this is possible and how it's actually economically possible, so in a traditional assay, we frequently will use in the laboratory something which is called a six-well plate to grow the cells. So we get all the cells inside those little vessels, the little wells on the surface of those, and we need so-called media to keep those cells growing. Approximately in one well, you need about sixty drops of the media. So this is the same size, the six-well plate as this well, although this well, this plate has 1536 wells inside. And each of the wells has not sixty drops of the media, has one-twentieth of the drop.

So the cells are growing practically in the volumes, which are almost invisible, and cannot be added to those plates using manual pipetting. It has to be done by robotics. So it's a miniaturization which makes it possible, which makes the speed and also makes it economical. So how this strange strategy look like? So we start with a library of compounds, and the libraries could be as big as 5,000-- as small as 5,000 compounds, and as big as a million compounds. Usually we conduct a set of screens. The first one, which is usually salting the most molecules out, it's the primary screen, and usually reduces the number of so-called hits, or potential drugs, to a few thousands.

So-called confirmatory screen, or counter screen usually reduces it to few hundreds, and then we have a set of secondary screens which should narrow down the number of interesting compounds to about 10, 20, 50. And then, of course, from that, we'll take those compounds through the lead optimization and then pre-clinical, clinical trials up to potentially drug formation.

So how does it look like in our assay? How does the Friedreich's ataxia assay look? So we designed those cell lines, which, the model cell lines, the reporter gene and the long GAA repeats. As I mentioned to you, those long GAA repeats cause those cell lines do not produce as much of the light, because in this case we detect the light, as the cell lines with the short repeats, so the transcription of this reporter gene is reduced in the same way as the transcription is reduced in the Friedreich's ataxia gene.
Well, then, we plate those cells on those small, tiny little plates, and we up the library of compounds. And we're looking for the compounds which will cause that those cells will start glowing. It means that the expression of the reporter gene is increased, and that's what we would like to see, then, in the Friedreich's ataxia gene. The Friedreich's ataxia, the expression of the Friedreich's ataxia gene should also increase. So the goal is to find, to restore the production of the reporter gene.

So where we are right now, what's the common start of them? This is an update on, October is actually about seven days, well, maybe five days ago, so it's a very recent update. We used our system to screen for the National Institutes of Health library of 360,000 compounds, which is deposited in that library. The primary screen which was conducted on our reporter cell line, this reporter cell line is similar to the Friedreich's ataxia patient cell line, and this contains 850 GAA repeats in the intron. Narrow down the molecules from 360,000 to 19, to about 2,000.

So those are the molecules which increase the production of the reporter gene, that those 2,000 out of 360,000. Now, we want those molecules to increase the production of the reporter gene which has a lot of repeats, but we don't want those compounds to work on the cell line which has a little of the repeats. We don't want a nonspecific effect, as we call it. So we test those molecules in a confirmatory screen on a secondary reporter cell line which has only 30 repeats, and that's an equivalent of the unaffected person.

So any molecules which will stimulate expressions similar to production of the reporter gene here will be thrown out as nonspecifics. So after this counter screen, we end up with approximately 654 molecules. I'm sorry, 634 molecules. And 250 out of those molecules are currently, as of last week, analyzed for so-called dose dependence. So they will be analyzed for different concentrations to check if the effect is really concentration dependence, and that's crucial for the activity. So that's exactly where we are. We are at the stage of narrowing down from 360,000 to 250 compounds.

Next step, we'll be taking the cells, the cells which Mirella Dottori told you, and Joel Gottesfeld today, taking neuronal cells from Friedreich's ataxia patients and testing the compounds, which will come positive in a dose dependence test, testing them if they're actually increasing Friedreich's ataxia gene production, not the reporter gene. So can this strategy work? Well, it did work in several other instances. You can see, this is the list of the drugs which were actually produced or were in production by different companies, which originated from screening. Those are drugs which were not directly derived from the screens, they originated from the screen and then were modified by medicinal chemistry. So it can work.

Well, why then everybody is not doing high-throughput screening? Obviously, high-throughput screens have their limitations. And our first limitation is, as I told you before, reporter gene, it's similar to the Friedreich's ataxia gene. This is not an exact Friedreich's ataxia gene, so we cannot put here an equilibrium between them. It's as much similar as we could have made it, but it's not identical. So there is a possibility that our search for the needle in the haystack will produce just a bag of nails and other scrap metal, but it's much easier to go through the small bags with nails and scrap metal and find this needle in this small bag then go through the needle in the haystack. So that's the idea of the high-throughput screen. You want to narrow down your potential hits to the small number of compounds.

And, with this I wanted to finish with this slide, which you've seen today at least twice. We hope that the next, probably more than 12 to 18 months, we'll be able to add some potential compounds here on the bottom to that pipeline and start developing something which will come out from our high-throughput screen. So with this, I would like to acknowledge people who did the work. In my lab, the two people, collaboration with people from MD Anderson Cancer Center, they've collaborated with Joel's group, and the people who did high-throughput screen at the Scripps Florida. And the funding from FARA and NDS, and the National ataxia Foundation and our institutional funding. Thank you.
[applause]

David Lynch: And, last but certainly not least for now, Dr. Helene Puccio from Strasbourg to talk about one of her many avenues, but I don't know which one.

Helene Puccio: So, Good afternoon. Let me just get my coordination between the different things going here. So I'll go through this screen also. So I first want to thank FARA and all of you to be here and to have invited us to be here, and FARA for its inspiration throughout the many years I've researched, that my lab has been going through for Friedreich's, but that many of the labs have been going through. So Jen asked me to talk to you about more basic science and giving you new insights into what we know about frataxin function, because there's been a lot going in the past two years by many different labs around the world.

So before I go into those details, I do want to just talk to you a little bit about what we're doing in the lab, because we're doing many different approaches. The lab is mostly based on generating cell models and mouse models that are being used by many other researchers around the world to try to develop drugs, to understand pathophysiological mechanism, but also to understand the function of frataxin. So this has been extremely hard but useful to be generating these mouse and cell models. But as you've heard from Mirella, there's actually many other labs who are doing the same things.

So I think this is basically how we all look at what we do in lab for basic science is we try to understand both fundamental questions as molecule function of a protein, pathophysiological mechanism with the overall goal at the end of looking at therapeutic approaches. Here just to acknowledge the people in the lab, in my lab who are doing this. And so, this is the nine people in the lab who are working 100% on FA on these different aspects, and they're very hard-working people, and they actually are doing a great job.

So why are we so interested in cell and animal models for Friedreich's ataxia? Well, as I said, they help us understand the physiopathology of the disease. They are useful for therapeutic approaches, both large-scale screens that we just heard from, from Marek, or test specific compounds, as has been done in Massimo Pandolfo's lab for the HDAC inhibitors that Joel talked about earlier. But, these animal models and cell models are also important to understand the basic function of frataxin.

And, although I'm going to be presenting work from my lab, I just want to acknowledge the fact that there's many, many different labs around the world. This is a big international collaboration, a competition in research. That there are many, many cell models out there and animal models out there that have been helping us understand all of this, and actually each one of these cell models and animal models have a specificity and have their advantage and disadvantage. And so I think this is extremely important for you to understand when we develop models, none of them are perfect and some of them can answer some questions and some of them can answer other questions.

And so, for example, if I just take the animal models, there are at least four different mouse models, or mouse model approaches that have been developed. There are what we call the conditional mouse models developed in my lab, which I'll talk a little bit about it. And these are models in which we delete frataxin completely. So you heard that in patients, in Friedreich's patients, there's this GAA expansion and there's still a little bit of frataxin in the cell. And so, these models don't mimic what happens in the patient, but they are helpful to understand some of the pathophysiological pathways downstream. Massimo Pandolfo, Mark Cook, and Joe Sarsero, they've developed the GAA based models, and so these models are actually very good to be testing, to be understanding what the GAA does and how does the GAA silence the frataxin gene, and also for testing some of the compounds they are looking at increasing frataxin, such as what Joel and Marek have been talking about.

But, these models are less severe in terms of the phenotype, and they've been more disappointing, I think, in getting, trying to understand what the disease is. And so, when you hear the researchers say, you know, "We're still trying to develop other models," and I, you know, the patients all, that I meet in France often tell me, "Well, you guys have been developing models for ages and you need to have them." We have models, but they're not perfect, and we still need to develop more models. And so, our efforts are still going towards developing better models, and so this is important to understand.

So with that said, all these models, whether it's the cell models or the animal models have helped us really understand more of what is going on in FA, and these are the models and the inputs that help us understand, also, what goes on in getting towards therapeutic compounds. So frataxin. You heard already this morning a little bit about frataxin. It's a small protein that's highly conserved from the bacteria to the human. So this is actually a very, very conserved protein. What the protein does in the bacteria is the same thing that what the protein does in humans. So studying in the bacteria what the protein does actually helps us understand what happens in humans.

This protein, even though it's highly conserved, it's unique. There's no other protein that looks like this protein, so unfortunately, we can't learn from other proteins what this protein does. We need to work on this one particularly. In the mammalian, (unintelligible) the mammalian system, it's a mitochondrial protein, and so even though Massimo did, I think, a really good job of explaining what a mitochondria is, I just want to go back to some basics to make sure that we all know what we're talking about. So here's a cell, a eukaryotic cell is represented here, and there's many things in the eukaryotic cells. All I want you to look at and concentrate on is the cell has a membrane, which is the wall of the cell, and then, there's the nucleus here, and the nucleus is kind of like the central unit of the cell. And, this is where the chromosomes, the DNA, the chromosomes are.

And, a very important organelle in this cell are these blue organelles, which are the mitochondria. And, if we go and look at, by electron microscopy as Mirella showed, this is what a mitochondria looks like with the membrane, the internal membrane that Massimo Pandolfo was talking about this morning that has, that is here. We can see the drawing of this internal membrane. And, as Massimo explained to you, mitochondria are important for many different things. When you hear most about mitochondria being important for the energy of the cell by the generation of ATP, but I do want to really push, as Massimo did, it's not only ATP, not only energy. Mitochondria, we now know, is extremely important for all the iron handling of the cell, and mitochondria is also extremely important for lipid metabolism.

And so, this is really three of the many functions of mitochondria, but three functions that I think of our interest for FA. So frataxin is a mitochondrial protein, and so what we know now from many different studies in many different cell models, biochemical models, is that it's closely involved in the mitochondrial iron homeostasis or metabolism. And this is the way we often propose what frataxin is doing. So it's purpose functions-- here's a drawing of what a mitochondria is.

They have iron outside the mitochondria, it has to go in, and you have frataxin here in blue. And, it's involved in iron storage, maybe, iron donation for heme biosynthesis, so this is in the mitochondria that we make the last type of heme for hemoglobin. And, it's involved as a regulator, an iron donor for iron-sulfur cluster biosynthesis that you've already heard about. And I'll go a little bit more into that later on. So this is how we draw, or this is what we propose about frataxin function. So, as I said, we were developing mouse models to try to understand the physiopathology of the disease, but also to understand the function of frataxin.

So we're using a conditional approach, so this is an organ specific deletion of the frataxin gene in the mouse, and this is because, as Massimo told you this morning, when we completely delete frataxin in the mouse, it does not go through embryonic development. So here we have our mouse, and we have genetic tricks to be able to take away frataxin in their, for example, in the heart, and we generate a cardiac model. Or, in the neurons, represented here in blue, and we generate neuronal models. But we can also delete frataxin in other organs that are not specifically touched in FA, but for scientific reason will help us answer some questions.

And so, we've also developed models where we delete frataxin specifically in the liver and in the skeletal muscle. And what these models have given us, the neuronal and cardiac model actually reproduce most of the characteristics of the disease, most of the biochemical features and a lot of the clinical features, as much as a mouse can have the same clinical phenotype as a patient. But all of these models have helped us elucidate the primary function of frataxin. And so, the next two slides are very busy slides, and I don't expect you to see everything on it. It's just to illustrate what we have.

So this is a cardiac model, and it's a very severe cardiac model. It developed a cardiomyopathy in 10 weeks. And from this model, we can see here, this is electron microscopy of the mitochondria in these animals, and you can see that the mitochondria don't look as well as the mitochondria I showed you earlier. And so, they have iron deposit, there are crits that are collapsed. They are bigger. They are empty, also. So we've been able to get a lot of information by being able to dissect this model out on a regular basis by taking the two weeks, three weeks, four weeks, five week time points. So these are the advantages of animal models.

The liver mouse model, surprisingly also, is extremely severe, and surprisingly, really gives us the same type of phenotype. And liver's interesting when we look at iron, because liver is one of the main organs of iron regulation, body iron organ regulation, so this was extremely important for us to do this. And, again, we have a dysfunction, and we have the mitochondria are not doing well, and we actually have also this iron accumulation within the mitochondria. But what comes out of these models, really, is that the first thing that is affected is what here is in red, is iron-sulfur cluster enzyme prior to the iron accumulation.

So, this pointed out to us very early on that frataxin must be involved in iron-sulfur cluster biosynthesis. And then, we have other steps that occurred to deficiency on iron-sulfur cluster biosynthesis. So, you've heard a little bit about iron-sulfur clusters this morning by Massimo, so I will repeat probably a lot of what he said. Iron-sulfur clusters are these cofactors that are made up of iron and sulfur. They come in a variety of shapes and numbers of iron and sulfur. The most common ones are the ones I've represented here are two iron, two sulfur, and four iron, four sulfur.

And these cofactors are these, this element is actually essential for many proteins in the body. And so, they're what we call cofactors, so they're inserted into proteins. When the protein doesn't have them, it's nonfunctional. And when we insert this cofactor into the protein, it becomes functional. So this is a regulatory mechanism of protein. And, what we know about iron-sulfur cluster protein, there's a growing list of them, there's not many of them that were known, but there's a growing list of them.

And, we know that they are important, so here again, I represent the cell. That's a square cell. The mitochondria. We have iron-sulfur cluster proteins for energy metabolism, so in the respiratory chain. These are what Rob Wilson explained as transporting electrons and making ATP, so there's a lot of iron-sulfur cluster proteins there. There's a lot of iron-sulfur cluster proteins, but there's actually one main iron-sulfur cluster protein involved in iron metabolism, so regulating the cellular iron metabolism. And there's also iron-sulfur cluster proteins in the nucleus that are regulating the RNA, the DNA metabolism and are important for DNA repair, or translation of the DNA into, transcription DNA into RNA, and then translation of the RNA into proteins.

So this is an extremely fundamental pathway that's extremely important for all cellular processes. And this is where frataxin is probably involved in. So, why is it important to understand frataxin function? I know you don't really want to hear about frataxin function probably right now. But, for us as scientists, it's important to understand it because it's a fundamental pathway important for all cell function. So that's our fundamental reason to want to understand this. But, as I said, both for you and for us, it's also because this will help us, maybe, go towards therapeutic strategies. And so, I think we've seen this several times. Each one of us illustrates it differently, about how we think about FA.

We have a genetic mutation. This genetic mutation, which is this GAA expansion, leads to decreased frataxin, and then we have a number of steps. I've put three here, but there's a number of steps that we don't maybe all know about that leads to increased cellular dysfunction and disease. And each one of these steps where there's a genetic mutation, decrease of frataxin, or this dysfunction are possible places where we can think about therapeutic approaches, and so this is really what we've been hearing about today.

Joel Gottesfeld and Marek talked about therapeutic approaches for directly going at the genetic mutation, and Rob Wilson gave us different approaches for looking at this downstream effect into the dysfunction, so the antioxidants, the iron chelators. So if we understand what frataxin does, we can also think about frataxin mimetics, or replacing frataxin with something that would do the same thing. And so, this is why it's important to understand frataxin function.

So now, I want you to bear with me, because I'm going to go into a little bit of details on how we do this and how we've done this in the lab, but there's other people in other labs doing similar things. To try to understand how a protein works in the cell, you need to first know that a protein never works alone. So in the cell, we have here frataxin, so that's represented here in blue. And we know that frataxin probably interacts with one or two or more proteins, and that gives its function. And so, this is what we need to try to understand is what we called the protein interaction network, and that will give us an idea of what frataxin does.

So there's a lot of things that have already been shown about who frataxin interacts with. We have data in the literature that shows that frataxin interacts with enzymes and proteins involved in metabolism and energy metabolism. And there were also data showing that frataxin interact with these proteins that are also involved in the making of iron-sulfur cluster proteins. And so, what our lab attempted a few years ago is to try to get an approach of trying to understand who this frataxin interact with. And we wanted to try to take an approach that was a non-biased approach. So we took this non-biased approach, which is called co-immunoprecipitation, and I'm going to walk you through this quickly.

So if we want to be able to fetch out who frataxin interacts with within a cell, we can do this as if we're going fishing. We have our frataxin, we put a little tag, and this little tag just enables us to hook it out of the cell. We let it interact with who it interacts, and then, we take it out, we hook up onto this tag and we take it out of the cell, and we take it out in the situation where the proteins that it's interacting with stick on the frataxin stay there. They don't go away, and we go through different washing steps, we can elute, and then we can go and check out who is this x? We can identify this x by mass spectrometry. So it's a way to be able to identify proteins that we don't know anything about, or proteins that we don't know the sequence, but we can identify who they are. So that was, this is what it's called: co-immunoprecipitation. And this is one of the strategies that we took in the lab.

The second strategy that we took was, once we've identified proteins that interacted with frataxin, we needed to test to see whether these interactions are true. And so, we then express, we commonly, we were able to express in the bacteria the frataxin in the protein of interest. And then, again, we go through extraction purification, and we were able to check whether this interaction exists. So this is what Stephane Schmucker, a Calvin graduate student did in my lab. And, here is a Western Blot showing you the results. And, again, a Western Blot is, we're able to detect the proteins. This is a way to detect proteins.

And so, in this experiment, Stephane had the frataxin flagged here, so this is, we know what we weren't fishing out for, and then he looked to see if these different proteins that I talked to you about were into that. And, actually, you can see here by Western Blot, that he was able to find NFS1, this protein here where you can see a blob here, ISCU here you can see a blob, but other proteins that had been described, we can't see them in the experiment, even though they were here at the beginning. So this is the input, so we know they were there in the cell, but we can't fish them out with frataxin.

So this is what it showed us. It showed us that we were able to find frataxin interaction with these three proteins, but not with the other three proteins. And, actually, in this fishing out, we found that frataxin interacted with no other protein under our condition. Using the second approach, Stephane was able to show that frataxin is, can interact with these three proteins, but only when all three of these proteins are present. It doesn't individually interact with anybody, and what this tells us is that frataxin interacts in a complex. And this is the complex, this is the complex involved in iron-sulfur cluster biosynthesis. This is the early complex's first step of the complex of making iron-sulfur cluster biosynthesis.

So the next question is, we found this. Okay. Frataxin interacts with this, but is this essential for the cell? And so, again, we go back to cell models that we've generated, and these are cell models that we've generated from our conditional mouse mutations. And so we have cells from the mouse, but these are fibroblasts from the mouse in which they are genetically modified on the frataxin gene. We have tricks that we're able to delete frataxin. And what happens in a cell when we delete frataxin, the cell is not viable in culture. We were able to put that human frataxin in these cells, and then delete the mouse frataxin, and these cells are completely happy. They have a completely normal phenotype.

And what we've been able to do is when we put back frataxin, human frataxin with different mutations, so either mutations that do not affect the interaction with the complex, mutation that partially affects interaction, or mutations that abolish interaction, then we can put them into the cell, delete frataxin and ask what happens. And this enables us to decide, or to see if this interaction with this complex is essential, and if this is an essential function of frataxin. And so, when we do that, what we find is that if we have a frataxin mutation that does not prevent interaction with the complex, then we have normal cells. It's as if we have well-type cells. If we have a frataxin mutation that prevents completely the interaction, then we're in the situation we get no cell living, showing that this interaction is important for living. And then, when we have as a frataxin protein that only affects partially the interaction, then we get an FA-like phenotype in the cells with mitochondrial degeneration, iron accumulation, and iron-sulfur cluster deficit. So these results suggested to us that this function of frataxin in this complex is the essential function in the cell.

So animal models, interaction studies, and in cellular, so in-cell studies, all point out to the fact that the primary essential function of frataxin is in building these iron-sulfur clusters. These clusters are important for many different functions in the cells. There are many other labs working on this subject, and I've put in here four, probably the four leading labs who are working on trying to really understand the function of frataxin within this complex. And, what we have learned, and I'm not going to go, have you go through all the biochemistry, but what we have learned is, we know a little bit more about what frataxin does into this complex.

And so, I've adapted a scheme that was in a paper from David Barondeau, and I've adapted for you to have the same colors as before. So here's a complex without frataxin. This is the complex that makes iron-sulfur clusters. We have sulfur, we need to put iron. When frataxin's not there, this complex seems to be off. When we add frataxin and we add iron, or when the cell adds frataxin and iron, what happens is that this complex is on. We make iron-sulfur clusters, and then through a whole set of other proteins, this iron-sulfur cluster is able to be put into the protein of interest to make a functional protein. And so, what frataxin seems to be doing, it seems to be regulating the complex to make more iron-sulfur cluster proteins, and so this is important.
And so, if we want to try to find mimetics of frataxin, this is a way where we can work on, we can try to work on making this complex more active without frataxin. So with that, I hope I haven't lost you completely. I know it's a little-- lot's of details, but it's important. Again, I want to thank everybody, especially FARA for having us here and being here, and all the people in my lab. All my collaborators, which I didn't put up here, I just thank all of them and you for your attention. Thanks.

[applause]

David Lynch: Now, we have a few minutes for all of our scientists to come back up here for questions if they'd like to take them. Even if they wouldn't like to take them. So, if anyone would like to step to the mics and ask away for Joel, Marek, Mirella, and Helene. Sit anywhere you like, guys. Questions, anyone? Oh, in back, Sandy.

Sandy: (Inaudible).

Joel Gottesfeld: First question is, it's our hope that you only have to go to carrier level, and since most patients are in the 20% or so of unaffected level, by and large you would only need a 2.5 fold increase to get back to carrier status.
How will you know how much to dose the patients? Well, from animal studies, we have a good idea, and you can relate, there is such a background in pharmacology of what an effective dose is in an animal model and how you translate that into a human dose, that pharmacologists know these sorts of things. I don't, but pharmacologists do.

David Lynch: That's basically correct, yeah. And there are also rules on extrapolation of animal data to humans and where you can start, how high you can go to go up. Thus, the importance of that safety Phase 1 data for any drug. Because if you don't know what the safe dose is, you don't know how high you can go in the efficacy.
Next question, down front, Andy.

Andy: What's the general time frame for the HDAC inhibitor to receive FA trials in patients?

Joel Gottesfeld: Oh, the time frame. We're hoping by the end of the year, or very early next year to have the drug in patients at the Italian site, and then sometime later on next year in the U.S. if all should go well.

David Lynch: Question?

Audience: Okay. Dr. Puccio, this question is on turning the functions of the mitochondria, you were saying there were three parts: iron metabolism, energy, and lipid metabolism. You didn't really focus on the lipid metabolism, and my curiosity is, is there anything FA patients can do right now to work on improving lipid metabolism, taking lots of fish oil? I'm on Wikipedia, and I just don't know, but it is—

Helene Puccio: So, I think, actually, there's not much known about lipid metabolism in FA. What's coming out now is, I guess, more towards the diabetes. Lipid metabolism is closely related to diabetes, also. So I know that Massimo Pandolfo's group is working hard on their animal model to try to tease out what's going on in the lipid metabolism, and how it is affected. So to answer to your question, we know, what can FA patients do? I think that, I think it's Massimo who answered this this morning, or this afternoon, it's really on the diabetes part, I think you just need to be, like all of us need to try to stay fit, and not eat too much. But--

Audience: Has it been suggested to not eat as much fatty food?

Helen Puccio: I'll let him answer that.

Massimo Pandolfo: So, in fact, (inaudible)-- you cannot metabolize your fats you risk to accelerate the process of becoming diabetic. Now, not all Friedreich's patients become diabetic. That is to be clear. I mean it's just a minority that become clearly diabetics. So, although the trend and the risk exist, you have to behave like any good person should behave with a minimum of good sense. That is, try not to follow my example and not gain too much weight. Avoid eating too much fat, and try to stay fit as much as you can. All these things favor mitochondrial biogenesis, favor lipid, correct lipid utilization, and prevent better cell stress. So that's what you should do. No other miracle treatments.

David Lynch: Yes. On that note, obviously, exercise is a very important component of that. That's why I hope you're all in Ride Ataxia tomorrow. Help in raising money for everything. And, I would note that Massimo and I had talked about, we will ride. I will be there. And next year, we've decided to have a joint fundraiser, which is—

Massimo Pandolfo: -- a bunch of new mitochondria--

David Lynch: New mitochondria. Next year, we will have joint fundraiser, which is called weight loss ataxia, where the two of us will compete and people will sponsor us to lose weight. That's getting more applause than anything else. Question over here.

Audience: Good answer. I have one question just about the, does your FDA accept the European data? Like, for the Italian Phase 1B trial, or will that have to be replicated here?

Joel Gottesfeld: I know you have to report studies done in Europe, or anywhere, to the FDA, but I'm not directly involved in the clinical development. That's being done at Repligen, so, you know.

Audience: So we're not sure if it's going to have to be a repeat?

Joel Gottesfeld: I think Massimo knows the answer.

Massimo Pandolfo: My understanding is that the FDA will judge the trial on its own merits.

Audience: So it's—

Massimo Pandolfo: not because it's been done in Europe, but how it was done and what the outcomes were.

Audience: So it may be able to advance to--

Massimo Pandolfo: So it may be able to advance to the, also with the FDA, of course. Yeah. That's exactly correct.

Audience: And I do have one other question for Mirella. I was just wondering if she could expand a little bit on some of the dangers associated with the transplant of the stem cells.

Mirella Dottori: Yes. So if you transplant a stem cell, like an embryonic stem cell, it will form tumors. That's one of the massive dangers with stem cell transplants. A lot of the unregulated stem cell transplants going on in certain parts of the world, A. We don't even know what type of stem cells they're transplanting. It's not regulated at all in terms of, you know, where it's going in, and when you hear these success stories, it's often, you know, huge amounts of physiotherapy associated after the transplants, and we just don't know. That's the bottom line. We just don't know.

So at the moment, it's the tumor formation and especially with IPS cells at the moment, like some of them are produced using viruses or using some of the genes that could have oncogenic potential, but scientists are working towards making them a lot more therapeutically viable. Also, other types of stem cells, when we transplant, do tend to have, like, hyperplasia. So even if it's not forming a tumor, it's just a massive growth. So there's a lot of things that we need to address. We're also making the right cell types, there's a lot of other cells in there that you need to actually get rid of. So we're trying to purify the right cell types and do the right thing in the right numbers at the right place. That's basically it.

David Lynch: Question over here.

Audience: Yeah. I'm interested in the age of onset of this disease. I find it interesting that, you know, it's variable, but we generally don't see it during infancy or early childhood. Do you guys hypothesize that it's simply connected to the number of repeats that somebody has, or is there something that's protective in children that, is it prevented, do they respond better to oxidative stress? Or, what is it about the presentation? Can you hypothesize about that?

David Lynch: That's a very interesting question, and maybe we'll all have a-- The data are that if you, by regression analysis, the age of onset can be explained, roughly, the R-squared values, about .5 correlation coefficient between .7 and .8, so about .5. So you can explain about 50% of the variance of age of onset with the GAA repeat alone, so that is the primary determinant.

Some of the rest of that variance would be simple chemical variability in how we interpret it. There will be other genetic modifiers, which, for example, might superimpose upon the GAA and how much natural frataxin a person makes. And then the question of, in all neurodegenerative diseases, you develop a progressive course. So there certainly is an unknown. Are there protective factors earlier, or in the case of the nervous system, remember for the first few years of life, you are actually forming your neurological circuits. So that whether it can even come into play in that time is an interesting question. Other thoughts, including Massimo?

Massimo Pandolfo: One thing that happens in the degenerative diseases, is that the disease process has already begun and progressed to a certain point before symptoms appeared. And, in fact, you can find, you know, it happened to me to find a little child who was a sib with Friedreich's who has no reflexes, for instance. And that's, you know, that's like some genetic tests. I mean, otherwise this kid is fine, but is probably going to be, very likely going to develop the disease. So the disease process starts, the clinical onset comes when the disease has already progressed to a certain point, and that's probably one explanation.

David Lynch: Yeah. And I would note in other nervous system diseases, for example, Parkinson's disease, we say that you show up at a neurologist's office when you're down to 20% of your particular type of neurons. So that process started a long, long time earlier.

Jen: There's a compensation that's probably happening.

David Lynch: Yes. Yeah. As Jen notes, there's compensation. Our bodies are smart. You give it a toxic insult they try to do everything possible they can to get around it before it eventually fails. Other questions? Yes. Up front, Linda.

Linda: Just on a very high level, it seems like all these labs around the world are working together? Is that true of other diseases, or is this just--

Joel Gottesfeld: Let me say something about this. Friedreich's and, the Friedreich's research community, that man right there, Ron Bartek, Ron has created a research environment that is so highly cooperative and collaborative. I'm also involved in research in other diseases, and related diseases, and I won't name it because just the nastiness involved, infighting and no collaboration. This is remarkable. Yes.

David Lynch: Yes. Question there, Pam.

Pam: Yes. Dr. Gottesfeld, as far as the second generation compound that Repligen has, are they going to continue developing that, especially since they're trying to outsource, out license this to some partner? Are they going to continue to develop the second compound?

Joel Gottesfeld: Oh, absolutely. I mean, that's work ongoing in my lab in collaboration with Repligen. We have a grant application pending with the NIH right now to do the final development of the backup compound, so the answer is definitely yes.

Pam: And, have they started doing the six-month dual animal studies for that second generation compound?

Joel Gottesfeld: You know more than I do.

David Lynch: I wanted to hear the answer to that.

Joel Gottesfeld: I don't know.

David Lynch: Other questions looking around? Yes. Over there, Kathleen.

Kathleen: Hi. Just wanted to know if there are clinical trials that are ongoing or will be starting soon in Italy, Milan, or if we're from North America, are we still possibly eligible to go?

David Lynch: Is that directed to anyone specifically?

Kathleen: Any trial that would happen?

David Lynch: I would note, I can't answer what they say in Italy. I can answer what goes into our IRB approvals. We normally do not put restrictions on where a person may come from. Individuals in Idebenone Phase III included individuals from Canada. Now there may be practical aspects. If your visits are every week for the first month, you're going to probably, and you would need to come to Philadelphia, you're going to have to find a way to get here. If you live in Italy, that might be a little difficult or a little expensive. Travel, that would be a situation where they might specifically regulate the amount of travel which are paid for.

That said, we write these to be as inclusive as possible, unless there is a good scientific reason why a person should be excluded. Occasionally that will appear. Our IRBs will usually read that a person has to read the English language in our initial IRB, that if we needed to go back, we can amend that to translate the consent form into other languages, which is a systematic process you have to go through, which takes a while. So, yes it can be done. At least, coming this way. I can't say the other way, because I don't know.

Massimo Pandolfo: Well, the question is, of course, the amount of travel. We take patients from north of France and from the Netherlands, but the amount of travel is limited, and we have patient information in consent forms in French and in Dutch.

David Lynch: Right.

Massimo Pandolfo: That's the other point. I cannot take a Greek patient if I don't have consent forms in Greek. So that's--

David Lynch: Yeah. So the, yeah, it becomes somewhat practical and somewhat how much you need to enroll as many people as possible. Next question.

Audience: For children, for a lot of the, I guess, trial drugs, it's not really geared towards children. I mean, earlier you said that they should be the ones approaching you. And, at what age group, I mean, you said between 18 and 5. You, I don't know, how likely you would have a five-year-old coming up to you asking to be in a trial group.

David Lynch: Well, the five-year-old probably won't ask me, but their parents might be. I use that as the extremes of the example because it gets highly variable in that middle range how much you can increment down. The point is this, the question is, "Why are the trials essentially targeted to adults?" You go where your safety data is initially, and then you can modify over time. Routinely, in this country, if your safety data are in eighteen-year-olds, in your initial trial you can move to sixteen-year-olds, so let's just start at that level.

If some people who are sixteen-year-olds have done well, they may let you go down as you move to the next trial, or sometimes in a late form of that present trial. So you increment yourself down. An exception: if Deferiprone were to have a trial right now, that's a drug for which there are massive amounts of safety data in children. We know what the effects are. So that can go immediately to a pediatric trial if that would be the aim of someone.

Audience: Would that be the parents' authorization at that point again?

David Lynch: Well, it's the principal investigator and the company as to whether a drug enters in an experimental clinical trial, and then you devise your rules based on the safety data. Now, each individual parent has the-- or each individual has the right if they would like to participate in a trial, they can read the consent form and decide whether it's appropriate for them and whether they'd like to participate. It's an individual decision for each person, but the overall is determined by the sponsor and the principal investigator. And this might be a good time--