June Fireside Chat Recap
The focus of the latest fireside chat was the UPD, ICD, and Mosaic genotypes of Angelman syndrome (AS), with questions led by parents Nicole Meade (ICD) and Vlad Korotkov (UPD). We were joined by FAST's Chief Science Officer Allyson Berent, and Dr. Yong-Hui Jiang of Yale, who joined so that both genetics and therapeutic questions could be answered by someone who sees patients in the clinic as a clinical geneticist and understands the nuances of genotypes better than most.
Please note that none of the answers to the questions relating to clinical trials are specifically from any of the companies. These are general answers to questions based on public disclosures and do not reflect the opinion or statements made by any specific company.
QUESTIONS
- What exactly is UPD isodisomy?
- At a high level, UPD is uniparental disomy. Isodisomy is a form of uniparental disomy (UPD) in which both copies of a chromosome, or parts of it, are inherited from the same parent. In the case of UPD for AS this is when there is no maternal copy but instead 2 paternal copies, both silenced by the UBE3A-antisense (UBE3A-AS) transcript. This results in both copies not expressing UBE3A in neurons, and therefore resulting in the symptoms of AS.
- At a high level, UPD is uniparental disomy. Isodisomy is a form of uniparental disomy (UPD) in which both copies of a chromosome, or parts of it, are inherited from the same parent. In the case of UPD for AS this is when there is no maternal copy but instead 2 paternal copies, both silenced by the UBE3A-antisense (UBE3A-AS) transcript. This results in both copies not expressing UBE3A in neurons, and therefore resulting in the symptoms of AS.
- Do any of the ongoing trials include UPD/ICD?
- Currently, there are 4 clinical trials active and 3 use investigational ASOs, or antisense oligonucleotides. These ASOs in early stage clinical trials using paternal activation approaches and focus on safety and early evaluation of possible benefits. These early-phase ASO studies are not including UPD/ICD for a few reasons:
1) treating the most homogenous (meaning most similar) group of patients to best understand dose effect and safety;
2) not introducing any theoretical risk of double activation (meaning unsilencing 2 copies of the Dad’s UBE3A gene) until it is understood what the target dose is for single allelic activation (meaning unsilencing one UBE3A gene - in this case from the Dad in patients who have a maternal deletion or mutation genotype);
3) establishing a consistent dose for >90% of the population from the start and then once understood work in parallel on the other genotypes <10%. - The 4th study is a downstream targeting approach and UPD/ICD is included in that study, this is Neuren Pharmaceuticals study.
- Currently, there are 4 clinical trials active and 3 use investigational ASOs, or antisense oligonucleotides. These ASOs in early stage clinical trials using paternal activation approaches and focus on safety and early evaluation of possible benefits. These early-phase ASO studies are not including UPD/ICD for a few reasons:
- Do they know if the clinical trial in Australia by Neuren Pharmaceuticals Limited, is testing UPD cases?
- According to clinicaltrials.gov, Neuren has included all genotypes of Angelman syndrome except for mosaicism.
- According to clinicaltrials.gov, Neuren has included all genotypes of Angelman syndrome except for mosaicism.
- Are there any novel solutions for people with UPD?
- The main cause of AS in individuals with a UPD/ICD genotype are the same as that of the other 90% of individuals affected by Angelman syndrome--that is, deficiency of UBE3A. So as shown in our pillars, maternal replacement, paternal activation, etc. are all applicable to UPD/ICD genotypes, but it is important to remember that paternal activation approaches could maybe lead to increased expression of UBE3A compared to the other genotypes because a treatment might “turn on” UBE3A production from both copies of the Dad’s gene. All maternal replacement strategies apply to all genotypes. Research teams are also working to understand what impact the expression of additional Small nucleolar RNAs (snoRNAs) could have specifically on this population of patients, and if knocking those down individually, which are in excess in UPD/ICD, affects the phenotype or symptoms of UPD/ICD neurons. This is being assessed with the landing pad and organoid work being done at NCSU with Dr. Albert Keung.
- The main cause of AS in individuals with a UPD/ICD genotype are the same as that of the other 90% of individuals affected by Angelman syndrome--that is, deficiency of UBE3A. So as shown in our pillars, maternal replacement, paternal activation, etc. are all applicable to UPD/ICD genotypes, but it is important to remember that paternal activation approaches could maybe lead to increased expression of UBE3A compared to the other genotypes because a treatment might “turn on” UBE3A production from both copies of the Dad’s gene. All maternal replacement strategies apply to all genotypes. Research teams are also working to understand what impact the expression of additional Small nucleolar RNAs (snoRNAs) could have specifically on this population of patients, and if knocking those down individually, which are in excess in UPD/ICD, affects the phenotype or symptoms of UPD/ICD neurons. This is being assessed with the landing pad and organoid work being done at NCSU with Dr. Albert Keung.
- Is UBE3A only functionally important in the brain? Angelman syndrome is characterized as having external phenotypic features like scoliosis, lordosis, strabismus, widely spaced teeth etc. which clearly can't be due to UBE3A levels in the brain. Do we really know how important UBE3A is outside of the brain?
- We don’t believe anyone feels UBE3A is only functionally important in the brain, it is just that neurons of the brain have no UBE3A, or very little, being expressed in those living with Angelman syndrome due to the imprinting phenomenon where the maternal allele is either missing, mutated or silenced, and the paternal is naturally silenced in these cells. Whereas, in all other cells of the body the paternal copy is not silenced so there is a functional paternal copy that is expressed in everyone with Angelman syndrome. This means that in all cells of the body, except for neurons, there is half the amount of UBE3A when compared to a person without Angelman syndrome, but in the neurons of the brain there is none. With UPD/ICD outside of the brain there is 100% of UBE3A expression because they have 2 copies and the silencing doesn’t happen outside of neurons. Therefore, they should not have any impact of a loss of UBE3A outside of the brain like the other genotypes. Because individuals with UPD/ICD still have symptoms of AS like individuals with other genotypes it supports the idea that symptoms are likely centrally mediated (meaning that these other symptoms are caused by UBE3A deficiency in the brain).
- Many of the symptoms above, like scoliosis, lordosis, strabismus etc are likely due to the mixed tone of the muscles that control movement, which are directly influenced by the neurons not expressing UBE3A, so this can be “centrally driven” (from the central nervous system or “CNS”). This can result in the eye muscles and the core muscles being weak, which can have consequences like scoliosis, lordosis, and strabismus. The distinct facial features could be due to having only half of UBE3A in cells of the rest of the body, like bone, ligaments, etc, in the non UPD/ICD genotypes. But in the case of UPD and ICD individuals, who have full expression of UBE3A outside the brain (and not 50% like other genotypes), any peripheral symptoms they have (meaning outside of the CNS) are likely either driven by lack of UBE3A in the brain or due to central excess of the SnoRNAs. . The fact is, UPD/ICD individuals still have some of the non-CNS features of AS, but they are often less severe than those individuals with mutation and deletion genotypes, therefore this would support that effects from missing UBE3A in the brain plays a role in symptoms in other body systems outside the CNS.
- It has previously been hypothesized that imprinting of UBE3A is ‘leaky’, i.e., not leading to a 100% silencing, such that UBE3A in UPD/ICD could have additional residual expression from two incompletely silenced copies compared with the residual expression from one incompletely silenced copy in the other genotypes. If the ‘leaking hypothesis’ was true, UPD/ICD should be less affected than mutation, but one study found the phenotype of mutation was less severe than that of UPD/ICD. Does this support that residual expression of silenced UBE3A (i.e., ‘leaking’) has less relevance for the overall severity of the phenotype than overexpression of maternally silenced genes in UPD or other genetic factors in the paternally duplicated region?
- It is important to acknowledge that the papers comparing UPD/ICD to single base pair mutations in the Mom’s gene are comparing very small numbers of individuals and the overlap of the severity and symptoms is large. The differences between these 2 genotypes are very small, so this must be carefully interpreted. The mutations can be missense mutations where 100% of the UBE3A protein is being made but it is misfolded and therefore has some function, but not typical function because a protein needs to be folded correctly to work the way it needs to. Other mutations result in no protein being made at all. Those different types of mutations often present differently clinically, but in these publications are often grouped in the same categories. To try and dissect that out and say that mutation individuals have a less severe presentation than those with UPD/ICD is likely a reach and more prospective detailed assessments would need to be done to truly conclude this. What was clear is that non-deletion genotypes generally present differently from deletion genotypes, but that too is quite variable and there are individuals with large deletions that are on the same range of severity as those with non-deletions (mutation/UPD/ICD) and those with non-deletion genotypes that experience the same degree of severity as those with deletions. In addition, we must remember that we have over 20,000 genes in our genome and we likely all have other variants that impact the way we function. The impact other genes can have on UBE3A deficiency should not be ignored, and helps to explain some of the significant variability that we see from person to person.
- It is important to acknowledge that the papers comparing UPD/ICD to single base pair mutations in the Mom’s gene are comparing very small numbers of individuals and the overlap of the severity and symptoms is large. The differences between these 2 genotypes are very small, so this must be carefully interpreted. The mutations can be missense mutations where 100% of the UBE3A protein is being made but it is misfolded and therefore has some function, but not typical function because a protein needs to be folded correctly to work the way it needs to. Other mutations result in no protein being made at all. Those different types of mutations often present differently clinically, but in these publications are often grouped in the same categories. To try and dissect that out and say that mutation individuals have a less severe presentation than those with UPD/ICD is likely a reach and more prospective detailed assessments would need to be done to truly conclude this. What was clear is that non-deletion genotypes generally present differently from deletion genotypes, but that too is quite variable and there are individuals with large deletions that are on the same range of severity as those with non-deletions (mutation/UPD/ICD) and those with non-deletion genotypes that experience the same degree of severity as those with deletions. In addition, we must remember that we have over 20,000 genes in our genome and we likely all have other variants that impact the way we function. The impact other genes can have on UBE3A deficiency should not be ignored, and helps to explain some of the significant variability that we see from person to person.
- Mosaicism: I’ve heard of mosaic ICD and mosaic UPD, is there a mosaic deletion/mutation too?
- Mosaicism likely does exist for all genotypes but the largest reports are on ICD, followed by UPD. There are deletion cases that have been seen, which are generally very small deletions and somewhat rare overall.
- Mosaicism likely does exist for all genotypes but the largest reports are on ICD, followed by UPD. There are deletion cases that have been seen, which are generally very small deletions and somewhat rare overall.
- What patterns of mosaicism exist in terms of numbers of differentially expressed cells in the body and how that looks in different organs compared to the brain?
- It is not exactly understood how the pattern of expression looks in the brain compared to other organs, as it is really assessed by looking at the pattern of UBE3A expression in blood cells. Because we diagnose mosaicism with blood testing we have an idea of % blood cells that are expressing UBE3A but that doesn’t necessarily translate into the exact amount of expression in the brain and the patchwork that has some cells expressing and some not. This could only be assessed with an evaluation of neurons directly from the brain, which can’t be done without a brain biopsy. Therefore, we make assumptions. Interestingly, a higher percentage of maternal UBE3A being expressed in the blood has been associated with less severe clinical symptoms, so we believe that it is translatable to what we see in the brain (more UBE3A expression leads to stronger function). We also know that most are between 5-15% mosaicism which means that 85-95% of cells are NOT expressing maternal UBE3A in those living with mosaicism.
- It is not exactly understood how the pattern of expression looks in the brain compared to other organs, as it is really assessed by looking at the pattern of UBE3A expression in blood cells. Because we diagnose mosaicism with blood testing we have an idea of % blood cells that are expressing UBE3A but that doesn’t necessarily translate into the exact amount of expression in the brain and the patchwork that has some cells expressing and some not. This could only be assessed with an evaluation of neurons directly from the brain, which can’t be done without a brain biopsy. Therefore, we make assumptions. Interestingly, a higher percentage of maternal UBE3A being expressed in the blood has been associated with less severe clinical symptoms, so we believe that it is translatable to what we see in the brain (more UBE3A expression leads to stronger function). We also know that most are between 5-15% mosaicism which means that 85-95% of cells are NOT expressing maternal UBE3A in those living with mosaicism.
- What is the difference between Prader Willi Syndrome (PWS) and Angelman syndrome (AS) and why can’t we consider using the same medications to address them both since it is just the paternal allele affected instead of the maternal allele?
- PWS and AS are separate disorders even though they occur in a similar region of Chromosome 15. Angelman syndrome results from the lack of maternal UBE3A being expressed in neurons. This can occur from either deletion or mutation of maternal UBE3A, having 2 paternal copies that are silenced by the UBE3A-AS (UPD), or an imprinting center defect resulting in maternal silencing, all culminating in the loss of functional maternal UBE3A expression. There is no loss of the paternally expressed RNAs, which is what you get with PWS. PWS results from the loss of paternally expressed Small nucleolar RNAs (snoRNAs), due to various genotypes as well. This condition is not a result of the loss of paternal UBE3A, since UBE3A is not expressed paternally due to the silencing. So, one disorder (AS) is really due to the lack of UBE3A maternally, and the other (PWS) is due to the lack of expression of various genes/mRNA from the paternal allele that are not usually expressed on the maternal copy. Therefore, the underlying causes of both diseases are very different, and the targeted treatments are therefore very different.
- PWS and AS are separate disorders even though they occur in a similar region of Chromosome 15. Angelman syndrome results from the lack of maternal UBE3A being expressed in neurons. This can occur from either deletion or mutation of maternal UBE3A, having 2 paternal copies that are silenced by the UBE3A-AS (UPD), or an imprinting center defect resulting in maternal silencing, all culminating in the loss of functional maternal UBE3A expression. There is no loss of the paternally expressed RNAs, which is what you get with PWS. PWS results from the loss of paternally expressed Small nucleolar RNAs (snoRNAs), due to various genotypes as well. This condition is not a result of the loss of paternal UBE3A, since UBE3A is not expressed paternally due to the silencing. So, one disorder (AS) is really due to the lack of UBE3A maternally, and the other (PWS) is due to the lack of expression of various genes/mRNA from the paternal allele that are not usually expressed on the maternal copy. Therefore, the underlying causes of both diseases are very different, and the targeted treatments are therefore very different.
- Question about current clinical trials: 6 months ago we heard from Ultragenyx and Ionis at the Science Summit on the topic of UPD/ICD inclusion. One company said there was no further research needed and that they are going to include UPD (and ICD) into the trial, while the other company said that more clinical data is required, referring to measuring protein levels using a newly developed assay tool on CSF (biomarker) and targeting a certain protein level by adjusting the dose. Neither of the companies gave an indication of any timelines but both felt they did not need more work to be done on animal models to support including these genotypes. Could you provide the current snapshot of the position of all active ASO trials?
- We engage these companies regularly about this topic and their opinions have not changed. UGX and Ionis have communicated their plans for UPD/ICD in various forums. Prior to the recent news by Roche deciding not to conduct any more clinical trials of their ASO, Roche had not shared much related to this topic. Ionis and UGX have told the advocacy groups, and shared publicly, that they do not feel they will learn anything more from additional animal studies in UPD/ICD and are not advocating for that. They made this very clear to our community at the FAST Science Summit, and when specifically asked why they don’t feel we should put more work into rodent studies for UPD/ICD the answer was that more rodent studies will not model a large animal brain as far as expression, behavior, and biodistribution of a drug. To get the dose precise, if you can measure UBE3A in the CSF, one can try to determine an ideal dose to get target protein expression in the CSF. This can potentially be used to titrate the right dose for all genotypes, including UPD/ICD. Ionis shared a promising test method that they hope to use to figure out the ideal dose for deletion and mutation patients and then target that amount for UPD/ICD. They have not shared the details of how this test is working post-ASO treatment in their trial. If it is working as they hope, then this is a nice way to try to know the target dose. In addition, UGX stated that they feel that dosing UPD and ICD patients are in their near-term plans based on determining the ideal dose in deletion patients that are currently in their trial.
- We engage these companies regularly about this topic and their opinions have not changed. UGX and Ionis have communicated their plans for UPD/ICD in various forums. Prior to the recent news by Roche deciding not to conduct any more clinical trials of their ASO, Roche had not shared much related to this topic. Ionis and UGX have told the advocacy groups, and shared publicly, that they do not feel they will learn anything more from additional animal studies in UPD/ICD and are not advocating for that. They made this very clear to our community at the FAST Science Summit, and when specifically asked why they don’t feel we should put more work into rodent studies for UPD/ICD the answer was that more rodent studies will not model a large animal brain as far as expression, behavior, and biodistribution of a drug. To get the dose precise, if you can measure UBE3A in the CSF, one can try to determine an ideal dose to get target protein expression in the CSF. This can potentially be used to titrate the right dose for all genotypes, including UPD/ICD. Ionis shared a promising test method that they hope to use to figure out the ideal dose for deletion and mutation patients and then target that amount for UPD/ICD. They have not shared the details of how this test is working post-ASO treatment in their trial. If it is working as they hope, then this is a nice way to try to know the target dose. In addition, UGX stated that they feel that dosing UPD and ICD patients are in their near-term plans based on determining the ideal dose in deletion patients that are currently in their trial.
- ASO trials started 2-3+ years ago and we expect that by now companies have a good idea about the optimal dose for deletion. Has there been any tangible progress made about the inclusion of UPD and ICD?
- We only have feedback from Ionis and UGX and nothing has changed. Once they establish their target doses for deletion/mutation they have stated that they will then move to UPD/ICD. Neither have finished their Phase 1/2 trials so we have no further information from the companies to share at this time.
- We only have feedback from Ionis and UGX and nothing has changed. Once they establish their target doses for deletion/mutation they have stated that they will then move to UPD/ICD. Neither have finished their Phase 1/2 trials so we have no further information from the companies to share at this time.
- Is there a risk that UPD and ICD are not going to be included in the ongoing trials and instead will be pushed out to separate satellite trials? If separate trials - how does it work - is there a precedent we could use as a reference?
- No company has shared their plans with us to be able to answer this question precisely. In general, studies will have different cohorts and many trials will include different populations in different cohorts. What we know is that both Ionis and UGX have said they do not require UPD specific models to include patients in clinical studies and that they will use their early data from the ongoing studies to consider next steps.
- No company has shared their plans with us to be able to answer this question precisely. In general, studies will have different cohorts and many trials will include different populations in different cohorts. What we know is that both Ionis and UGX have said they do not require UPD specific models to include patients in clinical studies and that they will use their early data from the ongoing studies to consider next steps.
- If UPD/ICD are included as part of Phase 3 of the currently running trials, would they also be mixed with placebo? Or will they be dosed with the drug due to the rarity of the genotype?
- That will be up to each company, and we can’t answer this question as we have not been given any information on Phase 3 plans.
- That will be up to each company, and we can’t answer this question as we have not been given any information on Phase 3 plans.
- The “theoretical risk of overexpression” has been a term that all UPD/ICD parents who are into science dread to hear mentioned every time a pharma or academia representative is asked about the topic. What is the latest on this - e.g. what do we actually still need to establish and what is known already? Would be great to have a collection of study results that directly or indirectly address this.
- The answers to this were carefully and thoroughly laid out in a blog that FAST put out last summer with all recent literature on this and the interview on the topic by each pharma. Nothing has changed per each pharma. The only modality where there is a “theoretical” risk of overexpression is paternal activation (right now this is the strategy is being used on the ASO trials). This is not the case for any gene replacement therapy (AAV, HSC-LVV), enzyme replacement therapy, or downstream targeted approach.
- Each ASO company has been interviewed on this topic and specifically, Ionis and Ultragenyx commented explicitly on this, as discussed above. They both stated very clearly and succinctly that they do not feel any more research is needed in animal models to support including individuals with UPD/ICD in clinical trials. They need more human data on the deletion/mutation genotypes. Experience of dosing in the mutation and deletion will drive safety decisions and selection of dose levels that would expect to be potentially safe and effective. Once that is established, they will then look at UPD/ICD genotypes. If there were a clear biomarker allowing them to measure UBE3A expression in the CSF that would also help more easily drive the dosing decisions needed in this population but at the time of their presentation they did not yet have the method ready for use.
- Most do not feel that it is “unsafe” to treat individuals with UPD and ICD. We know in large brains (monkeys and therefore assumed the human) that an ASO does not get to 100% of brain cells. Therefore, you will not get 100% UBE3A expression in the brain (so realistically it will be much less than 100% because the ASO will not get into every single neuron in the brain). In short, it means the risk of overexpressing UBE3A in individuals with UPD/ICD is probably low because you would not turn on both paternal genes to make UBE3A at 100% each (meaning they get 200% UBE3A) if the ASO does not get into every cell.
- The answers to this were carefully and thoroughly laid out in a blog that FAST put out last summer with all recent literature on this and the interview on the topic by each pharma. Nothing has changed per each pharma. The only modality where there is a “theoretical” risk of overexpression is paternal activation (right now this is the strategy is being used on the ASO trials). This is not the case for any gene replacement therapy (AAV, HSC-LVV), enzyme replacement therapy, or downstream targeted approach.