28 – Modulating gene regulation to treat gene dosage-associated diseases

Event Time

April 1, 2021 at 5:00 pm


In this video Dr Nadav Ahituv presents Modulating Gene Regulation to Treat Gene Dosage-Associated Diseases

Prof. Ahituv shared his labs work with us on April 1st. Prof. Ahituv is an advisor to two companies working to cure diseases like SYNGAP1: Encoded Therapeutics & Regel Therapeutics. His lab is also part of a team from across UCSF, UC Berkeley & the University of Washington that recently secured a grant from the Weill Neurohub. This grant was to look a haploinsufficencies and SYNGAP1 is one of the explicit targets.

Dr. Ahituv is a Professor in the Department of Bioengineering and Therapeutic Sciences and the Institute for Human Genetics at the University of California, San Francisco. He received his Ph.D. in human genetics from Tel-Aviv University working on hereditary hearing loss. He then did his postdoc, in functional genomics, at the Lawrence Berkeley National Laboratory and the Department of Energy Joint Genome Institute. 

The Ahituv lab is focused on understanding the role of regulatory sequences in human biology and disease. Through a combination of comparative genomic strategies, biochemical assays, regulatory element analysis, human patient samples, mouse and fish genetic engineering technologies, and massively parallel reporter assays they are working to elucidate mechanisms whereby genetic variation within these sequences leads to changes in human phenotypes.


00:00 Welcome and Introduction to the work of Dr Nadav Ahituv

03:10 Overview of talk

04:27 Haploinsufficiency: When one functional copy of a gene is not enough 

09:31 CRISPRa as a therapeutic for haploinsufficiency?

11:13 Sim1, an important obesity gene

14:44 CRISPRa for either Sim1 promotor or enhancer rescues the obesity phenotype

22:23 Can we develop it as a therapeutic?

23:35 Many genes cannot get packaged in an AAV

28:23 Summary

43:43 Q&A


0:05Okay then, hello everyone and welcome to today’s  session. My name is Marta Dahiya. I’m one of the  

0:12Syngap parents. I’m part of the SynGAP Research  Fund. We are excited again to have another webinar  

0:20in the webinar series. The goal of the series  is getting you closer to the science of Syngap,  

0:26make you aware of the research that is being  done and the opportunities to participate,   and empower your communications with clinicians as  you get clearer knowledge of Syngap. We also want to  

0:38plug you into the next presentation that is An  Integrated Approach to investigating GI and  

0:46neurological co-morbidities caused by Syngap. This  will take place on April 22nd at 1pm eastern time  

0:54and is with Dr Jimenez, Dr Dalman, and Dr Mosher.  Our talk today is Modulating Gene Regulation to  

1:03Treat Gene Dose Associated Diseases. I have the  pleasure to introduce today’s speaker Dr Nadav Ahituv.

1:13Dr Ahituv is a professor in the department of bioengineering  and therapeutic science and the institute for  

1:21human genetics at the University of California  San Francisco. He received his PhD in human  

1:27genetics from Tel Aviv University working on  hereditary hearing loss he then did his post doc  

1:36in functional genomics at the lawrence Berkeley  National laboratory and the department of energy  

1:43joint genome institute. the itube lab is focused on  understanding the role of regulatory sequencing in  

1:52human biology and disease through a combination  of comparative genomic strategies biology  

2:00chemical assays regulatory element analysis human  patient samples mouse and fish genetic engineering  

2:08technologies and massive parallel reporters  assays they are working to elucidate mechanisms  

2:16whereby genetic variation between these  sequences lead to changes in human phenotypes.  

2:22After this brief introduction i want to let  you know a recorded version of the webinar  

2:28will be available on the SRF website and  on our YouTube channel. We also are live in  

2:36Facebook right now. By the end of the  presentation you will have the opportunity to  

2:42get your questions answered. We love to hear  from you. Please write your questions in the chat.  

2:49For those who just joined us, welcome and I wanna  let you know again the name of the talk today is  

2:56Modulating gene regulation to treating dosage-  associate diseases with Dr Nadav Ahituv and  

3:05now I want to give him a welcome go ahead.  Thank you thanks so much for having me here and  Overview of talk

3:13it’s really a pleasure and hope to show you what  we’re doing and also a project that we’ve just  

3:20started actually on Syngap in the end and let me  start my share screen. So it’s really important  

3:28for me that people understand, and I don’t know  at what level some people are, so I apologize  

3:34that sometimes it’s too simple or or vice versa  but the most important thing is please stop  

3:40me at any time and ask questions via the chat  or any other way that you can through this but  

3:46really it’s more important that you understand  than anything else so so stop me at any time   don’t be shy there’s no such thing as a stupid  question so really please and i’m i love it  

3:56when people stop me in the middle so all right so  i’ll start with that so and i’ll start first with  

4:03some disclosures in terms of the companies  i’m involved in and then I’ll move into  

4:11the acknowledgements. And this work really  started by an amazing and very talented  

4:16postdoc in the lab Navneet Matharu who’s now  a researcher and assistant professor at UCSF  

4:21and where she wanted to tackle these diseases  that are haploinsufficient similar to some of  Haploinsufficiency: When one functional copy of a gene is not enough

4:29the Syngap mutations. And so the idea here is that  we have two copies of our genes, one from our mom,  

4:37and let me see if I can do a laser pointer here…  okay, can you see my… um? We can see your  

4:44mouse on the screen. Perfect! Okay great, so we have two copies  of a gene, one from our mom and one from our dad,  

4:51and when we have two copies that usually leads to  normal gene levels – everything’s fine and so forth.  

4:58But every so often we could have a loss-of-function mutation where only one copy  

5:04is working versus another copy here. And then the  problem with that is that we get 50% and basically  

5:17(in terms of functional protein) and sorry  i have to do something here for somebody  

5:24quickly which will take me one second i apologize

5:29i don’t do this i’ll be in trouble there we go.  Okay yeah, and so and if we have one functional  

5:40copy we get 50% of a gene and for some genes this  is perfectly fine having 50% but for some genes  

5:48having 50% is sort of like running on empty, and  that leads to human disease and it’s estimated  

5:54that over 660 genes actually cause human disease  due to what’s called haploinsufficiency when  

6:02you have one gene that’s loss of function and  Syngap some cases some mutations of course is  

6:07one of these examples where you just have one  copy and you have 50% of the protein being made.  

6:15And the number is even estimated to be even higher  so if people look at a whole genome sequencing  

6:22data sets or whole exome where, you  know, thousands of genomes have been sequenced  

6:29and there’s over 3,000 genes where you never see  this. You never see a loss of function mutation in  

6:36one copy and so it could be even greater than  the 660. 660 so far are known to lead to human  

6:43disease. And so the current treatment for for these  haploinsufficient genes, for most of them, is by  

6:50drugs and the limitation there it’s very timely to  make drugs, it’s costly, and it’s usually to treat  

6:57the symptoms not to actually fix or cure the  disease. Another approach that you probably may  

7:03have heard of is a gene therapy where basically  you can give, for example, extra copies of the gene  

7:13to the patient and then that might be able  to fix the disease. The problem there is sometimes  

7:21dosage. It’s hard to control the amount. The tissue  specificity: the gene can go into other tissues  

7:27and with AAV (adeno-associated virus) which we’ll  talk about quite a lot later on, there’s a limit  

7:34to the amount of DNA that you can put into an  AAV and that could be problematic for big  

7:41genes. And Syngap for example is 4,200 base  pairs which could be a problem in terms of  

7:46fitting it into this virus. Another option  which I’m assuming a lot of you have heard of  

7:53is CRISPR where you can go and try to fix the  actual mutation but the problem there is that the  

8:01success rate of this is currently not that great  and so it’s hard to fix in all the cells and also  

8:08if people have different mutations and you  need to sort of custom tailor, each sort of  

8:15fix would be different between patients and that  could be sometimes problematic in various ways.  

8:22So about seven years ago this cool technology  came out of UCSF from Jonathan Weissman, Wendell Lim,  

8:29Stanley Qi and Luke Gilbert where they  manipulated the Cas9 so it won’t be  

8:37able to cut. So it’s sort of like, a it’s called a  “dead-Cas9” basically. So it can bind to the DNA but  

8:46it has a mutation in the scissors and so it cannot  cut the DNA but it can bind. So what that allows  

8:53you is to use it similar to a delivery truck to  bring something specifically to a certain zip code  

9:00in the genome to a certain location in the genome  and you can connect stuff to it. So for example you  

9:06can connect to it something that will force the  gene to make less RNA and less protein and that’s  

9:14been called “CRISPRi” for CRISPR interference  or you can connect for example an activator to  

9:20it that will force the gene to make more RNA and  that’s been called “CRISPRa”: CRISPR activation.  

9:29So if there’s one take-home slide that  I hope you’ll get today is this one and  CRISPRa as a therapeutic for haploinsufficiency?

9:35so when this came out and we thought “hey,  this could be a great sort of a therapeutic  

9:40for haploinsufficiency for these diseases” because  here you still have this existing copy that’s  

9:48perfectly fine it’s just giving you fifty percent.  And so if we could target this existing copy for  

9:55example with CRISPRa, which as I mentioned before,  can force a gene to make more RNA and more protein,  

10:05we might be able to increase these levels  and fix the disease. I hope that’s clear for  

10:12everybody. Again, this is the sort of the main  take-home message that I hope you get.  

10:18And so basically the idea is you still have  that perfectly normal gene it’s just giving you 50%  

10:27and so if we can again force it and squeeze out  more RNA we might be able to fix the disease. And  

10:32so when we started this about seven years ago we  thought hard about what sort of disease we want  

10:39to go after that will be simple and will be  a good proof of principle for this and you know   some phenotype that we can easily look at and so  we came up with obesity as sort of a phenotype  

10:51because the idea is even if we up regulate maybe  just a little bit and we’re just here we might  

10:57be able to see a nice quantitative change you  know that the mouse is less obese just a little   bit or so forth and that would be a very helpful  phenotype to see if this idea could actually work.  

11:10And so we chose two genes. I’ll tell you  about one of them here. Both of them are part  Sim1, an important obesity gene

11:17of the leptin pathway and again this is not as  important. The main thing to understand here is  

11:23that in adipose tissue we have this peptide,  this hormone that gets secreted leptin when we eat  

11:31and through a signaling cascade it  tells us that we’re not hungry anymore   and a lot of these mutations in almost all of  these genes right here lead to severe obesity  

11:42in humans and in mice also. And the gene that  we chose to focus on here is SIM1 where there’s  

11:50tons of mutations in obese individuals. If you take  sort of the top one percentile of the population  

11:56that is obese, and the highest body mass index,  about two to three percent of them will have  

12:02mutations in this gene in SIM1 and  mutations in it are haploinsufficient. So if  

12:10you have two copies of the gene you’re perfectly  fine. If you have one copy of the gene and one, you  

12:17know, is loss of function, so haploinsufficient,  you have people who are extremely very obese  

12:24and then no copies is actually lethal. So  there’s nobody walking around with it because  

12:31it’s very important for development of  the hypothalamus and other organs in the body.  

12:37So the idea is that we go here into these mice  that are very similar that have basically one  

12:45copy and are extremely obese and try to rescue  their phenotype, their obesity, with this approach.  

12:52And so when you have as I said one copy of  this gene, one is loss of function, you have 50%  

12:59protein and you have very obese mice. And so  what we did was target a regulatory element,  

13:06switches, that turn on this gene to force it  to make more RNA. So we target the promoter  

13:14which is right next to the gene and tells it  to turn on with this CRISPR activation system  

13:21and that hopefully should make the existing  normal copy make more RNA and more protein.  

13:29And we target another switch that’s called an  enhancer. That could be very far away from a gene  

13:35and in this case the enhancer is very specific to  the hypothalamus where this gene is expressed  

13:42and the idea is hopefully also here when we target  this enhancer we force the gene to make this  

13:49existing sort of normal copy to make more RNA and  more protein and that will lead to more protein  

13:55and no obesity in these mice. And so we got these  mice and from a collaborator in Montréal that  

14:05become extremely obese after 16 weeks. We made  mice that have the Cas9 then the activator  

14:17the thing that forces them to be to basically  make more RNA and we cross them to these mice  

14:25that become obese. And with Cas9 you need another  component with CRISPR called the guide and so we  

14:33made mice with another guide right here for the  promoter or for the other switch the enhancer and  

14:39our hope is that these guys become less obese  than all of these guys here. And so this is a  CRISPRa for either Sim1 promotor or enhancer rescues the obesity phenotype

14:46sort of the summary of this, where we target the  promoter the switch that’s right next to the gene   and force it to make more. And you can see here  graphs of the body weight of these mice. So you  

14:58can see mice here that are wild type that after 16  weeks weigh about you know 23 grams right here and  

15:07these are the heterozygous mice the mice that are  haploinsufficient that only have one functional   copy and you can see they become extremely obese  weighing about almost 50 grams after 16 days and  

15:19then in yellow are the ones that we rescue with  this CRISPRa and you can see that they look in  

15:24terms of their body weight over 16 weeks just like  normal mice. And the same for the other switch that  

15:32enhancer that was very far away when we target it  you can see that in green here the mice over 16  

15:40weeks weigh exactly like the wild-type  mice and not like the heterozygous mice  

15:48and it’s always nice to show a picture so here you  can see a mouse that genetically has only one copy  

15:55so only one copy of this gene the other one is  loss of function and you can see it’s extremely   extremely obese for a mouse and here to the right  you can see exactly the same mice in terms of the  

16:07genotype which should be as obese as this mouse  right here on the left but you can see it looks  

16:14much more like a normal mouse because we managed  to squeeze out more RNA and more protein from  

16:21that existing normal copy in that mouse. So I’m  happy to stop here if there’s any questions it’s  

16:30important for me that, you know, this part will  be understood and I see one question in the  

16:38chat. “For haploinsufficiencies do you think that  reducing the expression of one or more proteins  

16:44direct protein-protein interactors or downstream  could be treatment approach so total synthetic   dosage not increase but the dosage of SynGAP and  interactors would be more similar?” and so, yeah I  

16:56mean, that’s definitely a potential approach  and one that we’re actually trying here so if you  

17:03know if i go back here to the cascade like can  we rescue things here… so as I said mutations in  

17:12any one of these genes leads to obesity this gene  that we targeted is way downstream it’s it’s after  

17:19all these genes and what we’re trying to do now  is see if we can up regulate it does it help for  

17:25example mutations in this gene like this one is  way you know down sorry upstream and so if you  

17:32have mutations in here you’re extremely obese but  maybe if we up regulate this guy we can rescue it  

17:38because the signaling pathway has to go through  here so hopefully that answers the question.  

17:45And another question in the chat from Janie is  “Remind me, when you treated mice with CRISPRa, was  

17:52it over different time points?” and that’s a great  question. I’ll show you later on when we treat for  

17:58another gene at later time points but with CRISPRa and and i’ll show you later with av we treated  

18:07and i’ll show you an a v in a minute but we  we treated only at four weeks and you can see   that this is before they be this is the mice that  are heterozygous that are haploinsufficient you  

18:17can see they become extremely obese after that  so we sort of treat it before they become obese  

18:22and so we’re really interested also in treating  you know somewhere around here to see whether we  

18:27can rescue it or whether there’s some kind  of a pathway that’s already sort of fixed   and it’s too late to treat and so that’s some of  the stuff that we’re doing now is trying to treat  

18:38later in time to see can we rescue if  there’s any other questions let me know

18:50okay so if there’s no other questions I’ll keep on  moving on and again feel free to ask at any time  

18:59and so this slide i’ll walk you through it it’s   a little bit complicated but what I’m trying  to show here is that the way we rescue  

19:11is actually very tissue specific so the problem  that I mentioned in the beginning with AAV is  

19:16that it sometimes can go to other tissues and  upregulate a gene in tissues where you don’t want  

19:24one thing that we saw that’s advantageous to  this approach also is the tissue specificity  

19:29so the the gene that we targeted SIM1 is  expressing the hypothalamus in the kidney  

19:35but our CRISPR activation approach is expressed  in all the tissues so i’m just showing you  

19:41four tissues here but you can see that cast  nine are all sort of CRISPRa is expressed in  

19:47the hypothalamus in the kidney and the  lung and the liver and all the tissues   so technically we could be upregulating this  gene in all the tissues which we don’t want and  

19:59and so we looked are we up regulating in all  the tissues so here I’m showing you just the   expression of of this gene called SIM1 in normal  mice and wild-type mice and it’s only expressed in  

20:11the hypothalamus in the kidney but not in the lung  in the liver this means below detectable levels so  

20:16you know just like I said it’s only a hypothalamus  in the kidney and we plot this as one and then we   compare everything to the normal mice and so the  the mice that are heterozygous that just have  

20:28one copy of this gene have half expression as you  would expect for haploinsufficient and they’re  

20:34only expressing the hypothalamus in the kidney  the mice that we up regulate and with the promoter  

20:42with the switch that’s right next to the gene  you can see that we’re only up regulating in   the hypothalamus this gene and the kidney but  not in the lung in the liver even though it’s  

20:51expressed in the lung in the liver like all our  CRISPR components are expressed in every tissue  

20:57and so we see that it’s very tissue specific  and then when we target this enhancer which is a  

21:05switch that’s specific just to the hypothalamus  we see that we’re only up regulating in the  

21:10hypothalamus but not in the kidney and we  think that it’s because when when you have  

21:17this regulatory element the switch the DNA  is much more open and when the switch is active  

21:23and that allows the cas9 to bind to it and so  for example here the the switch sort of is only  

21:31open in the hypothalamus cas9 can only bind  when it’s open in the hypothalamus and that  

21:38and it only works there and so that provides  you tissue specificity so by targeting specific  

21:43switches that could provide us a tissue  specific activation which is pretty good for  

21:49for a therapeutic and I apologize if this wasn’t  clear don’t don’t worry about it is there is  

21:56there concern over over expression in something  like the kidney here yes there is and and I’ll  

22:02show you everything here is transgenic i’ll show  you with AAV we could sort of modulate the over   expression but there is concern we didn’t see  any phenotypes here in the kidney at least not  

22:12nothing observable the mice seemed fine but yeah  it’s definitely a concern where you’re getting   you know three threefold more than you should  be getting so yeah okay so everything i’ve shown  Can we develop it as a therapeutic?

22:25you was a transgenic approach where basically  the mouse has the CRISPRa system from birth  

22:32but we really wanted to go the distance and  develop it as a therapeutic and I’m assuming a lot  

22:38of you have heard of adeno-associated virus  which is now used in a lot of clinical trials and  

22:44the advantage of the adeno-associated virus  is that it goes into the nucleus of the  

22:51the cell but it doesn’t integrate into the  genome and so it’s a sort of a more safer  

22:56approach for a gene therapy and also it’s less  immunogenic so it’s been used for a lot of  

23:03clinical trials and some approved gene therapies  have been done with adeno-associated virus.  

23:10The issue that a adeno-associated virus has  is that it it can only package a small sort of  

23:17amount of DNA so only 4,700 base pairs or  nucleotides is what it can package inside  

23:24and you also have to take into account other DNA sequences that you need to get your transgene  

23:30your DNA to be expressed and so if we go back to  the genes that cause haploinsufficiency and  Many genes cannot get packaged in an AAV

23:41660 or the 3,000 that are predicted that I  mentioned before and we draw sort of our a red  

23:48line in terms of their size on what would nicely  fit into this adeno-associated virus you can see  

23:55that there’s a lot of genes that lead to disease  due to haploinsufficiency that will not fit into  

24:01an AAV and Syngap is one example unfortunately  that’s about 4,000 base pairs here so  

24:09it’s on the cusp it may be but it’s pretty big to  fit into an AAV along with all the other additional  

24:16things that you need to make a strong sort of expression for this so another  

24:25approach and that could be used is our CRISPRa approach for some of these long, large genes  

24:31and so the idea is here that we made, at least  for the obesity work, a virus that has the  

24:41CRISPR system that activates that makes things and  that increases expression and for it it needs this  

24:48a guide that allows CRISPR to know where to bind  and we made a guide for the promoter the switch  

24:54that’s right next to the gene and a guide for  the enhancer the switch that’s right next to  

24:59the other farther away sort of switch and the idea  is that we inject these two viruses into mice  

25:07or these two viruses into the mouse and hope that  they rescue their obesity phenotype via now again  

25:15an adeno-associated virus approach and so we  injected this into the hypothalamus of the mouse  

25:24the mouse that’s has only one copy of SIM1  and becomes extremely obese and here we’re  

25:30just showing you that she’s injecting it  into the right place, into the hypothalamus and  

25:37now we measure the levels of this gene SIM1 and  again we plot a wild type the normal mice is one  

25:46and the mice that have one copy you can see  show half we had a sort of a negative control  

25:53where we just inject the activation system but  without the guide that tells it where to work  

26:00and so here you can see that if we just inject the  activation system without the guide that tells it   where to bind in the genome we get also a half the  expression but now when we add the promoter guide  

26:14that tells it to bind right next to this genes  in one or the enhancer guide that tells it to bind  

26:19right next to this faraway switch of this gene in  the hypothalamus, we’re now able to get expression  

26:26of this gene that’s very similar to normal levels.  So we’re managing to up regulated expression with  

26:32AAV and get it to normal levels what about the  phenotype here so we took these mice and again  

26:38we inject at four weeks when before they become  obese we wait a seven weeks and then we measure  

26:46how much weight they gained in those in this time  point and so you can see that the normal mice  

26:54gain about five grams the mice that only have one  copy of this gene gain quite a lot about 11 grams  

27:02and our negative control where we just have the  activator system but without the guide that tells  

27:07it where to bind again similar to the heterozygous  but are the ones that have the guide for the  

27:14promoter or the enhancer the switch gained very  similar to the normal mice so you can see that um  

27:23we managed to basically rescue their obesity  phenotype another thing we did was take these  

27:28mice all the way to nine months from that  single injection so we only inject it at  

27:34four weeks of age and the advantage of AAV is  that it can express the transgenes for the life  

27:40of the mouse and so when we inject  it at four weeks and then wait nine months  

27:48we can see that our mice still look very normal  in terms of their body weight so these are the  

27:54normal mice these are the ones that have just one  copy and you can see that they weigh quite a lot  

28:01and these are again the negative control which is  the activator but the ones where we injected both  

28:06the promoter and the enhancer look just like  wild-type mice in terms of their body weight  

28:12with 30 grams so from that single injection at  four weeks we have mice now at nine months that  

28:19have the same body weight as the wild-type  mice as the ones that have two copies of the gene  Summary

28:25and so this looks like it could be a long-term  therapeutic and for this obesity and gene  

28:37so just to summarize this part when we did  the transgenic and we saw that we could rescue  

28:44this phenotype and what it allowed us since  the transgenic approach we basically could  

28:50up regulate in all the tissues what we saw  uh was that we only upregulate the gene  

28:57based on the regulatory element that we  target so the switch sort of determines  

29:02where we can up regulate or not only where  it’s open and active and we up regulate  

29:08which was good to see for tissue specificity  and we have these mice and these vectors and  

29:14we are using them for other diseases as I’ll show  you in a minute and for me the most important was  

29:20that the adeno-associated virus approach  actually worked and rescued their obesity  

29:26phenotype suggesting that this could be a  viable therapeutic for a lot of these haploinsufficient  

29:34diseases. So what I thought I’d show  you today is how we’re following this up and a few  

29:41other in one more gene but I’m happy to take  stop here and take some questions about this.  

29:47Yes there are. I was going to say it looks  like there is one question in the chat  

29:52So Pavel asked how do you gauge the frequency  of this approach on nonsense versus missense  

29:59mutations and what are your thoughts on allele-specific targeting possibilities? Great question  

30:04Pavel so we have to make sure that the  the other the allele is really loss of  

30:11function because if it’s not I will also  up regulate it and that’s not good. So  

30:17your approach, at least for therapeutics, is the  first thing is really make sure that the mutation  

30:24is causing a loss of function mutation because  yeah we don’t want to up regulate something   that’s not loss of function so there it would  not work and definitely allele-specific targeting  

30:33is something that we’re thinking about also if  for example if the switch that we’re targeting  

30:39has a variant, a snip, that is only in the mutant  but not in the normal allele or vice versa  

30:49we could just target the normal allele with  the guide and not the other one. So that’s that’s   definitely something that we’re thinking  about and I think and a lot of other people  

30:59are thinking about in various other ways so it’s  definitely approach that we would want to use.

31:06What about male mouse? Yeah I only showed so Roman  asked what about the male mice I only showed the  

31:11female but we had the same results with the males  so exactly similar the nine months and a lot of  

31:18these mice we had to take out and sacrifice to  do a lot of these assays that I showed you like  

31:23the up regulation or other stuff so we only had  female mice that we took for the nine months but   it worked everything else worked beautifully for  the male mice exactly the same just that I didn’t  

31:33want to overwhelm you with too many slides and  and then Janie asked at cellular level what is  

31:39the percent of cells that have AAV? That’s a  great question. We we didn’t look we didn’t do  

31:47any single cell on the hypothalamus to see what  cells got the AAV and what didn’t and overall  

31:54we saw broad as i showed if i go back here sorry  and here you can see that overall we get a broad  

32:04distribution of our AAV from our injections and  it’s in the location in the pvn where we want in  

32:10the hypothalamus and that being said we we didn’t  look in a single cell manner and another thing is  

32:17that i mentioned we injected two different  viruses so we hope that both viruses go into   every cell but it may not be the case in some  cases but overall our sort of bulk up regulation  

32:28as you can see here was enough to get levels  at least on on the region of the hypothalamus  

32:34that we looked at in bulk not on a single cell  level of something that looks like a normal and  

32:41like a normal mouse a mouse  that has two copies of this gene

32:48okay no other questions so i’ll move into um another  story that’s that’s I think more relevant for  

32:57for Syngap and this is beautiful work by  Serena Tamura who’s an amazing grad student in my  

33:06lab who’s on the call now also today and will be  much better to answer questions than me and then  

33:12Perry Spratt was a grad student in Kevin Bender’s  lab and along with many other people including  

33:19Stephen Sanders here and this is also through  the SCN2A families foundation and also through the  

33:29SFARI the simons foundation autism research  initiative where we got funded for this project  

33:36and and the gene that we targeted here is SCN2A  and it’s a sodium channel and if you look at  

33:44this graph is showing you all the mutations  that lead to autism and if you are more here  

33:52to the right it’s much more common and SCN2A  is probably one of the most common mutations  

33:58and they lead to ASD but of course also Syngap  as you can see is also here CHD8 and others and so  

34:07we chose to go after this gene to see whether we  can use the same approach that I just showed you  

34:14to rescue some of the phenotypes that  are known to be in this gene in mice.

34:23And in terms of SCN2A, mutations in this gene  are associated with three different disorders so  

34:31if you have for example increased function of the  sodium channels then you have infantile epileptic  

34:38encephalopathy and if you have a certain some  de novo mutations lead to BIS and then if you  

34:46have loss of function haploinsufficiency  like I showed you for the vcdg, that leads  

34:52more to ASD and intellectual disability  and so we targeted of course the  

34:59loss of function since that’s you know similar to  the approach that we did before in terms of this  

35:05sodium channel it’s a very important early in  in development and in terms of action potential  

35:13and so in the first year of life it has the main  function in terms of action potential and in mice  

35:20this is about one week in human it’s one year  and then after that uh SCN8A takes over and then  

35:27this sodium channel has an important role in back  propagation and maintaining velocity as I’ll show  

35:33you in a minute so this is beautiful work by Perry  sprout and kevin bender’s lab who characterized  

35:41mice who have only one copy of this gene and  what you can see that early in life again like  

35:48i mentioned there’s there’s differences in the  action potential between the mice that have one  

35:53copy versus two copies but then later on it seems  to be fine but you have issues with velocity here  

36:01where uh basically you see differences between  the nor the sort of the ones that have two copies  

36:06versus the ones that have one copy right here  and these could be plotted as these nice squiggly  

36:13lines showing differences here in velocity and so  what we wanted to do is see can we rescue these  

36:20phenotypes by upregulating the extra copy and um  another thing that perry saw is that um the and  

36:31neurons themselves were less mature in the mice  that had one copy and that could be measured by  

36:38nmda ratios where you can see that they’re much  lower in mice that have only one copy versus mice  

36:44that have two copies right here and we also try  to see if we can rescue those phenotypes so serena  

36:51worked hard to generate um an adeno-associated  virus that targets the promoter of this gene and  

36:58similar to the previous work that i showed you we  used activator a virus and we inject it into mice  

37:06and serena injects into one hemisphere and  uses the other one as a control she waits for  

37:13four weeks for the viruses to express and then  she takes out the prefrontal cortex checks that  

37:20she’s injecting and gets expression there checks  that she’s up regulating and then along with kevin  

37:26bender’s lab does electrophysiology to see are we  rescuing the phenotype and this is just showing  

37:32you that she’s injecting into the right place and  that she manages to upregulate this gene with the  

37:40crispr a approach and here both in wild type mice  and these are in mice that have just one copy  

37:49so what about phenotypes so as i mentioned  the velocity and this is all in a sorry i  

37:55forgot to mention this is all in adult mice  so we did not inject in um in early mice yet  

38:02we’re working on that but everything we’ve  done is an adult mice where they’ve already  

38:08started if i go back already past this page and  now have just issues with velocity right here  

38:15and so um if we go back here and what happens when  we up regulate so you can see that we’re rescuing  

38:24the velocity phenotype and here this is each one  of each dot here is individual neurons and you can  

38:30see that this looks more like mice that have two  copies versus mice that have one copy right here  

38:38and then what about amp nmda ratios which tells  you about maturity of the neurons and you can see  

38:44here that we are rescuing that phenotype also here  and it’s looking much more like wild-type another  

38:53concern that we have here and going back to the  question before by peter is what happens if we  

39:00up regulate too much and the problem that we  have here that i mentioned in the beginning is   if the channel works too well that could also lead  to severe epilepsies and so we’re worried that if  

39:12we up regulate too much that might be a problem  and so to check that we took mice that have two  

39:18copies of the gene and then we up regulate them  with crispr a and to have twice as much of the  

39:27gene itself and then we checked are we having do  we have any phenotypes from this so this is again  

39:33showing you the normal mice with two copies and  one copy ones that we rescue and then in the mice  

39:41that have two copies even if we up regulate  we see that we’re getting similar velocities  

39:48and also you can see it um throughout the  different neurons that we check even though we’re  

39:54upping it by twofold like twice as much and we  think the reason for this is because it’s a sodium  

40:01channel it has other subunits that basically  bring it to the membrane and associate with it and  

40:09even if we have too much of this sort of gene  and this channel the other ones are at normal  

40:15sort of normal wild-type levels and so because  of that we don’t get as much of that channel  

40:23in the membrane and that’s why we sort of get even  if we up regulate it twice as much we still get a  

40:30sort of a normal and velocity phenotype here due  to that so just to summarize this part i’ve shown  

40:38you that we can up regulate SCN2A in mice and  i’ve shown you that we can improve the velocity  

40:46in these mice with this upregulation and also the  APMPA NMDA using this approach and as I showed  

40:53you at the end in the mice that have two copies  if we up regulate it too much we still see normal  

41:01function of this channel and we think that’s  due to just because the other subunits are sort  

41:07of lower levels and so I think I’ll stop here for another question oh no question okay sorry  

41:16and I’ll end with what we’re doing now for Syngap. So we were actually just awarded  

41:26a grant from the Weill Neurohub, you  can feel free to check out their website. This is a  

41:34joint program between Berkeley UCSF and  University of Washington and basically  

41:42and we’re working with two other with a few other  groups and both in Berkeley and the University of  

41:48Washington in berkeley both with dan feldman and  Helen Bateup and in the university of washington  

41:54with um jay shandur’s lab and at UCSF  with kevin bender and Stephan Sanders  

42:00and we are awarded to do this to use exactly these  similar approaches but for three different genes  

42:07and one of them is actually Syngap and  we’re actually now in the process of and  

42:12finding guides and and optimizing CRISPRa  conditions for Syngap we this just started so I  

42:19sadly don’t have any information or any data  yet to show you but we are now in the process  

42:25of trying to optimize these conditions for CRISPRa and then it will go into the mice and see  

42:31can we rescue some of these phenotypes and  in particular Helen Bateup in Berkeley has  

42:38characterized a lot of these in mice and and also  in cell lines and a lot of the phenotypes in  

42:43Syngap so it’s it’s really amazing to be able  to work with her and and see can we basically  

42:49rescue a lot of the phenotypes that we  see in these mice and so that’s sort of the  

42:54the goal of this again it just started a  month ago and again the idea is we generate  

43:02basically viruses specifically for Syngap and  we’re going to inject and do very similar things  

43:07that I’ve shown you for SCN2A but for Syngap  and then I’ll end with the most important slide  

43:15is just thanking members of my amazing labs  our collaborators in particular kevin bender  

43:23and stephen sanders on all these projects and of  course dan feldman Helen Bateup labs and jason  

43:30dewer labs and university of washington for this  new neurohub project that we’re really excited  

43:35to start with and of course our funding  for the wild norah safari and all the other  

43:41funding that we have in the lab so i’ll end  there and very happy to answer more questionsQ&A

43:51and and then one question from janie could  you say again who would you that will you work   with on Syngap? Yeah so we will work with uh Jay  Shendure uh on Syngap. Jay’s a genomicist and with him  

44:03we’re what we’re trying to do is actually develop…  so for every one of these genes when we try to  

44:10find guides that upregulate the genes or  try to find the best conditions for CRISPRa it  

44:18actually takes quite a long time and so with Jay  we’re working on developing technologies that can  

44:24allow us to do it in one experiment that we can  test for example 6,000 guides in one experiment and  

44:31and that will save us a lot of time so then we’ll  have already a whole list of guides for all the  

44:36genes that we want to upregulate and we don’t  need to optimize the conditions every once in a  

44:41while again which could take like three months  or so. So that’s what we’re doing with jake

44:48Thank you so much. That was great. Very good  talk. Very, very nice. I have a question about the –  

44:56if you are injecting with the CRISPRa you  are injecting the hippocampus that is so deep in  

45:03the brain and I didn’t saw that you measure  in the cortex that is where most of the people  

45:10measure this type of experiment and what is  your explanation that for example the kidney  

45:16was high or that is tough like because  it’s very deep in the brain very localized  

45:23down then… yeah so I apologize if I  confuse you so the the first part  

45:28that I showed until the AAV slide was we we we  just did everything with a transgenic so the mice  

45:35we made mice that have all the CRISPRa inside  so there was no injections whatsoever the mice   from the one cell stage already had all the CRISPRa and so that’s why the kidney was upregulated  

45:48after that we we turned to then adeno-associated virus  and just injected in the hypothalamus and  

45:53we only stop regulation the hypothalamus and  because it’s localized and same for SCN2A  

45:59and SYNGAP1 that will do and so forth but the  the initial work that we did with obesity  

46:07was transgenic on and that allowed us  to see the tissue specificity because the   the CRISPRa was in all the cells because  we basically made transgenic mice which  

46:19basically to make transgenic mice you inject  the construct in a one cell stage embryo  

46:25and then it, you know, the the whole mouse  has the gene in every possible tissue   and so in every possible tissue you could  up regulate but we did not see that so that sort  

46:36of allowed us to see the tissue specificity which  we couldn’t see with AAV because it’s localized

46:45i hope that’s clear. Yeah. Yeah   and going forward we’re just doing AAV. That  was just because it was a proof of principle.  

46:54We wanted to see if there’s tissue specific  if there’s an advantage for this approach in   terms of tissue specificity if there is any other  brain tissue that is high like the hippocampus or

47:07so for now depending on the gene we’re going  to inject to you know the the best target in the  

47:13brain and we’re also playing with different  stereotypes so for example I’m assuming you heard  

47:20about this php stereotype it’s great in mice it  covers most of the brain actually and you get  

47:26very nice expression but it only works in mice and  only in certain types of my c-57 black mice and  

47:33but there are others that are coming out and  now there’s one article on bioarchive and it  

47:38looks like a very good stereotype at least in  macaques that is similar to this one that covers  

47:45the brain much more you can envision injections  in other locations not not directly but through  

47:50other ways and so a lot of that will also  be hopefully a trying to see can we improve  

47:57and in terms of the the phenotype and we first  want to get first the you know the pressure a  

48:03conditions well that we know that we’re up  regulating well and then we could play with   various ways of delivery to see can we upregulate  in more places or in the places that we want and  

48:15and another thing we could do is is upregulate in  certain locations versus others and see where are   we rescuing where is the important region in  the brain to rescue some of these phenotypes.  

48:25Thank you. Dr. Ahituv I had just one quick  question for you. You I think if I understood  

48:32correctly earlier on you had mentioned with  the CRISPR technology that it’s going to be   mutation specific in terms of the actual  treatment was that correct? So with this  

48:44the standard CRISPR approach where you’re  fixing the mutation and you have to basically  

48:54if you’re using a standard approach there’s  what’s called a donor plasmid that sort of   goes in instead of the and fixes the mutation but  you have to make that for each one there are other  

49:05sort of flavors of CRISPRs like um base editing  or CRISPR prime that you may have heard that can  

49:10specifically fix a certain location and which  which could be great the problem is for each  

49:17mutation you’ll have to generate a different  sort of guide and a different fix and and  

49:23another concern that you could have with those  is “DNA scar” what we call DNA “scars” that they  

49:30they could fix the mutation but they  could also cause mutations in other places   yeah and that’s i mean again being  optimized and worked and improved and but  

49:44the advantage of this technique is that the the  scissors of the cas9 are broken so it doesn’t  

49:49cut the DNA whatsoever it just binds and and makes  more copies so it could also bind in other places  

49:55and lead to up regulation but we feel that you  know and again I’m not as objective but we feel  

50:03that that might be less severe usually than  causing DNA scars but sort of an advantage and  

50:09the main advantages you could have many mutations  but for all of them you could just use this right  

50:15if they’re loss of function you can use just  this one moety and you don’t have to make  

50:21taylor for each one the different ones  so that that’s sort of the difference  

50:26i see quite a lot of questions in the chat  and from Pavel in your paper you mentioned  

50:34nanoparticles as a delivery alternative to  AAV. How realistic do you think this approach is  

50:40versus AAV and what needs to happen to advance it  and how long yeah so again great question Pavel  

50:46we don’t have to deliver it with AAV we could do  it with nanoparticles or any other thing that that   people try to deliver and and you know DNA and  a lot of work is beautiful work has been done  

50:59on that you know even now with the COVID vaccine  right or others and to deliver RNA and so that  

51:05definitely could be a great approach and you know  if people develop good delivery to the brain using  

51:12these nano particles that that would be great.  Another sort of worry I always have with AAV  

51:18is that it’s going to be there all  the time and if it up regulates   like you mentioned before Peter and others  too much that that could be a problem and so by  

51:30giving it for example in the nanoparticle  we could control and we could you know give   it every so often and so forth so it could  be a better way sort to control expression  

51:39and so and so that might be so definitely you know  i think that would be a good approach for this  

51:47and janie asked one comment on the prevalence  of Syngap it is way under diagnosed and we  

51:54trend toward only those with strong symptoms  of epilepsy getting diagnosed stay tuned for   its actual prevalence when more extensive whole  exome and genome yeah sequencing will be available  

52:05and that would be great and again I should  mention that our approach what I showed you  

52:10works well for the loss of function and you  got to make sure that it’s loss of function   and Pavel asked this before and it’s a  problem if it’s not a loss of function we  

52:20we may be up regulating the bad allele also right  and so and and so we got to be careful about that  

52:28also but for loss of function and through  these hopefully genome sequencing and exome  

52:34sequencing we may know better about what these  mutations are doing and then Heather I can take  

52:41Dr Mefford off of mute real quickly if she wants  to ask her question give me just one moment here

52:48dr Mefford yeah hi can you hear me? great nice work  Nadav we actually just had Helen Bateup here to  

52:56give a seminar earlier this week so it’s fun to  hear about the new grant you got to the neurohub  

53:02i was just curious how many genes you’re going  to try to target through that grant? Is this like   a large, you know, all neurodevelopmental genes  screen or have you picked out a few favorites?  

53:13Can you share that? sure sure so with  the first approach of the grant first aim is  

53:20to try to find them a lot of them sort of the  reagents the guides like I mentioned it takes us a  

53:26lot of times to optimize and so together with jay  we’re we’re going to target about I forget the  

53:33number of at least i think 300 genes are that lead  to disease and in a high frequent manner just to  

53:39find guides for them so that’s sort of the first  approach and the the second name is really drill  

53:46down on three genes and and try to use these  approaches for them and the genes are SCN2A which  

53:52I just showed you but doing more phenotyping  there. SYNGAP1 which I mentioned and the third one  

53:59is anchoring so those three genes are the ones  that we’re going to drill down a bit more do   much more extensive phenotyping, optimize our  CRISPR reagents and injections and then similar  

54:09to what I’ve showed you with SCN2A do the same  for these other two genes that I mentioned

54:16perfect I think mike has a couple of questions as  well mike yeah thank you can you guys hear me yeah  

54:23yeah great so thanks Professor Ahituv that  was awesome and I think just to calibrate for the  

54:31parents on the phone who aren’t like basically  what you’re presenting is something that could   really worked because of both the specificity and  the ability to up regulate so it’s super  

54:42exciting and generally after webinars like this  that are super exciting some parent calls me and   says when will it be ready so and i obviously i  have no clue so i i just a path for i mean so the  

54:53question to you in a few parts is like and without  raising expectations or making any promises like  

54:59how long do you think this neurohub work  is going to i mean i’m sure it’s years   and then at what point do you think the Syngap  work will be something that could go out to  

55:08a company to actually start working on delivery  and when would we see you know clinical trials  

55:15and when could we hope just in terms of managing  people’s expectations around this stuff and then  

55:20after that hard question the other one is do you  how will you think about the age of intervention  

55:26and how how old will be too old or is there hope  based on the the mouse work that maybe we could  

55:33do rescue later in life i mean obviously these  are next to possible questions but I thought I’d  

55:38throw them at you since you’re  clearly good at impossible things   thanks mike and yeah so for the first question as  you all probably know and sadly these things take  

55:49time and and they won’t come from academia and  and then you know they have to come from industry  

55:55and and so on our end we we formed a company  called regal therapeutics that that is sort of  

56:01working on trying to take these into the clinic  there are two other companies also that are  

56:07are doing this one that i’ve been advising for  a while and that one for for one disease and  

56:14i think they’re close to being in clinical trials  in a year from now and not not Syngap but but  

56:21another one and so i think that will probably be  the first clinical trial for this kind of approach  

56:28and so hopefully in a year from now we’ll  we’ll be able to see some results and and and  

56:35you know with our company also we’re just gearing  up and starting up so it will take time and  

56:41sadly as these things do but again i think  it’s only from industry you know in academia   there’s only so much we could do and that  will we’ll be able to move this forward  

56:52and in terms of the second hard question in  terms of what time to rescue and what age and  

57:00i mean at least looking at the SCN2A it was very  promising to see that we could we could rescue the  

57:05phenotypes that we see in in the adult mice and we  are going now to the earlier stages and we’ll see  

57:12what you see there but one thing that really you  know one of the things that made me so happy you  

57:19know with coed and everything and all this here  is i don’t know if you follow the the bio marin   sorry not the by marine the ultragenics story  about angelman syndrome with the the asos where  

57:31where children were given at a much older age  asos and and they see a very nice improvements  

57:39in in the phenotype in their clinical and so so  again this is a completely different disease and a  

57:47completely different gene but but for me this was  you know one of the rays of hope and and sunshine  

57:54that i had this whole COVID year so there could  be yeah yeah no I agree I follow fast and i’m  

58:04i’m actually i listen i listen to everything  that yale and ultra genics say because i think   they’re amazing and i think the quote that  came out of that one of the fast studies was  

58:11that’s that i like to use is one of the doctors  said we had no idea how plastic the brain was so  

58:16you know as parents that that gives us  a lot of hope i guess another question   and i should have i’m just throwing hard questions  at you because i know you can answer them but  

58:24I hope i said we’re so  grateful that that you and your colleagues are   doing this work. It’s really exciting.  You mentioned up front there are  

58:35hundreds of haploinsufficiencies  and yet you chose Syngap next and i’m  

58:40i’m just curious like what was the thought process  there was it the availability of science was it   the fact that Helen and Stephan Sanders and  others were already working on it? Or was it  

58:47i mean what how did Syngap get lucky enough to  be one of the three genes that are in this grant?  

58:54yeah no great question i think for  us it was in the availability of  

59:01mouse models that we could you know in  the end you want to go into the organism i  

59:06i there’s so much i can do in a cell line but  but you really want to if you can show it in  

59:14on an organism level and the availability of cell  lines and and characterize them and might sorry  

59:20of mice and characterization of them so there’s  actually a phenotype that we can go in and see  

59:26that we can rescue and was very important when  in serve our selection for this and then two of  

59:32course was the the collaboration i i you know i  should mention that this this wild neurohub grant  

59:38you had to have three from the three  sites so both from ucsf berkeley and um  

59:44and then uw and and sort of you know and i  kevin has been working a lot both with helen  

59:52and was dance for dan feldman’s  first phd student so it served was an  

59:58easy survey tie with them to work  on this and and helen’s been working   and doing beautiful work on Syngap so i think  that’s how what brought it together to Syngap.

1:00:11great can i get can i ask one more question  sorry it’s too much fun um so so you’re working  

1:00:17on you’re working on characterized mice which  are which are which are knockouts or you know   some of some of the fancier flocks i don’t even  understand what those mice do but i don’t think  

1:00:25we have that many patient or we don’t have enough  in my opinion and i i don’t know enough science  

1:00:30to really have an opinion but what i’m learning  from talking to different companies is it would   be great if we had more patient-derived cell  lines and potentially patient-derived mice  

1:00:40so that as things progress with various  approaches those can those can be tested  

1:00:47it it just in general like as a general thought  does that does that make sense to you or do you   think that might might be less true for this  approach no i’d say that would be very helpful  

1:00:58and again i i for me i’m you know my training  is mouse so i’ve always want to see things in  

1:01:04an organism but i think i mean even for scn2a what  we’re doing now is we have two cell lines that are  

1:01:10a lot loss of function of scn2a and and we are  basically differentiating them into neurons and  

1:01:17then test testing our CRISPRa approach on them and  and that’s worked by Serena in lab to see can we  

1:01:24rescue and so that you know could one you know if  you want to get it to to you know therapeutic in  

1:01:30the clinic one you wanna the guys are the sequence  is very different for these switches between mice  

1:01:36and humans right so you want to optimize you’ll  need to optimize the guides on on human cell lines  

1:01:42and two with this approach we can go you know all  the way to at least some kind of a phenotype that  

1:01:47we can see that we can rescue in these neurons  and we are we are doing electrophysiology on   them to see you know and we get them to a stage  where they they fire and see electric you know  

1:01:57we could see electrophysiology and and we could  see differences between the and sort of the ones  

1:02:04that have two copies versus one copy and now we’re  trying to see can we rescue at least you know in  

1:02:09in this cell culture system i think you know more  and more there’s advance beautiful advances in in  

1:02:15making brain organoids and and other sort of them  tissues and in the lab from cells and so i think  

1:02:23by having them that will be very helpful and some  some things that usually come up when people do  

1:02:30this research is um sometimes people prefer to  have the same cell line but with the mutation so  

1:02:37they make the mutation on the cell line because  it provides a similar genetic background and   then there’s no other backgrounds and and so so  that’s sometimes also what what people like to  

1:02:48do and with CRISPR now it’s much easier to do  and so so another approach could be instead of  

1:02:55you know having ipscs from from individuals  you could also just generate mutations in  

1:03:03ipscs and then you know start working  with those that have the mutation   the advantage there would be the  genetic background is the same  

1:03:17sorry i was muted there janie is a Syngap  parent and geneticist i know that she’s been   asking a lot of good questions here so i just  pulled her off of mute to ask her next question  

1:03:28hi uh thank you so much for all your work  i was thinking about the phenotypic assays so  

1:03:36do you already have phenotypic phenotypic assays  for the drill down work on Syngap and you know the  

1:03:42idea is that treatments the eventual treatments  that seem to be coming pretty fast we’ll need  

1:03:47biomarkers in order to tell if they are working or  not and so development of biomarkers is something  

1:03:53i’d love to understand and um even if you’re  even if your assays are all lab specific you  

1:04:00know wouldn’t really work in patient populations  i was just wondering what you think the the   assays will be and so and i think you may have  already answered it since you said characterized  

1:04:08mice right like if are you gonna use the  same the same phenotypes that have already   been characterized like the going back and  forth in the y maze and stuff like that  

1:04:18yeah so um so we are limited in you know both the  funding and the time that we have for this grant  

1:04:24but um our goal and as i said we just started and  but our goal is first to optimize and and get the  

1:04:31CRISPRa working nicely for Syngap  and then once we do that we’ll pass it on to  

1:04:38Helen who will do all the phenotyping but you know  once we have all these reagents in place we’re  

1:04:43happy to share them with with anyone who wants to  use them for for any types of phenotyping you know  

1:04:49in mice or that the ones that we’ll optimize  will be mice but we’re you know we could also   work quickly to do a newman man so he could  be used in human cell lines or so forth and  

1:04:59very happy to share in terms of or test  any other biomarkers or anything that   we can look at either in the mice or later  on in in other cell types and but you know  

1:05:09one once we have these in hand you know it’s  very easy for us to to share and send around   and give to to check for various other phenotypes  and we’d love to collaborate with people to

1:05:23with this approach and then uh one last question  that i didn’t put in the chat is with haploin  

1:05:30sufficiencies have you have you considered  or have you already been doing whole genome   sequencing to see what the like say say you have  a haploinsufficiency in five people and they  

1:05:42have very different a very different range of  effects right very different range of symptoms   um can you do whole genome sequencing or some  other kind of assay to see what the dosages of  

1:05:53the protein and protein interactors so you know  like kind of seeing are some kind of rescued by  

1:05:59having less of something that it interacts  with or or some worse because they have more

1:06:06yeah we we have not done it would be interesting  to do sort of a phenotype genotype correlation   on on the actual mutation in the expression and  but uh or some kind of the rna-seq right to see  

1:06:20expression levels you would need the specific  tissue which of course is going to be hard   impossible to get but and i i think that that  would be very interesting we we are envisioning  

1:06:32of course doing um you know for for mutations  that are not clear loss of function or or even  

1:06:37you know there could be some kind of a transcript  is to really make sure that they are loss of   function we can model them in some kind of a cell  type first to see the expression and so that that  

1:06:47would be important and and i think they’re also  maybe get like mike said getting ips from from  

1:06:53various patients would be very important to  look at the differences in expression we could   for example differentiate them into and neurons  and and see the difference is an expression for  

1:07:03each mutation make sure that it is a loss of  function because as i mentioned our approach   won’t work if it’s not a loss of function and  but now we’re not doing that but i think yeah  

1:07:13that’s definitely important to figure out and and  on the industry side i think that that’s key also  

1:07:20is really making sure that it’s a you know loss  of function before you target anything because you  

1:07:26could be doing more worse than harm and harm if  you are targeting an allele that is expressed and  

1:07:34should be still doing something that’s not that’s  deleterious right but once you get five that you  

1:07:42really know or loss of function it’s it would  be interesting to have i you know human derived   psc lines that you could then switch the  mutations in and out and see you know is  

1:07:53this line better with no matter what loss of  function right and is this second line worse  

1:07:58no matter which laws of function and then you can  start interrogating them for what is up and down  

1:08:03yeah you know what not what non-syn gap genes are  up and down in there yeah yeah i think definitely  

1:08:09if i’m trying to figure out also what what are  the you know the other genes that interact what   are we rescuing that with sim one we i didn’t  show that just the amount of data but we did  

1:08:18a lot of that also it’s a transcription factor  in this case so it turns on a lot of genes and   we looked at when we up regulate you know how it  changes expression of a lot of other genes also  

1:08:29and then are we are we sort of getting it closer  to the expression of the oven sort of two copies  

1:08:34of this gene so and yeah i think i definitely  you know going back to mike’s question i i  

1:08:40definitely think it would be very helpful to have  and these um and cell lines because we you know  

1:08:47you could test all that and look at differences  in these mutations and what they’re doing and also  

1:08:53as preliminary sort of trying to see  can we up regulate not only rescuing   and you know even the little phenotypes  that we could do in cells would be helpful

1:09:09sorry i don’t see any other questions on  the chat any last-minute questions

1:09:27this is a lot of science and for those of us  who’ve been tracking you and the biohub it’s   it’s it’s it’s truly incredible but i think it’s  easy to miss how incredible it is because you  

1:09:37you’ve covered so much i can i just ask  you in 60 seconds or less to kind of  

1:09:44in your own words because i don’t want to because  60 seconds about how long i have when i’m talking   to different families about this stuff like could  you could you try to just sum up like what makes  

1:09:53this approach different in terms of everything  from from the delivery to the specificity  

1:09:59to the up regulating the haplo and then when i  heard you answer my difficult question about time  

1:10:04i know you can’t put a number on it but i think  in terms of orders of magnitude we’re thinking   years not decades so i guess the question is  could you give us a quick summary and just  

1:10:15something that i can quote or memorize  to to try to describe this this project

1:10:24yeah so so i think i mean the the advantage of  this approach is that um it does not cut the dna  

1:10:31it just up regulates it can be used universally  for any mutation that is a loss of function but  

1:10:38you have to be sure that it’s loss of function and  because and i i didn’t show too much data of this  

1:10:44but because we’re targeting the the regulatory  machinery the switches of the genes themselves  

1:10:51and we think there’s an advantage in terms of  tissue specificity and also sometimes we see  

1:10:59like i showed you with scn2a that even if we up  regulate too much it’s not and not too much of  

1:11:06an issue is saying we saw for for this gene sim  one which i didn’t go into detail too much but  

1:11:13we think there’s some advantage  also for ectopic expression   and in terms of delivery there’s still a lot to  do and and my worry with this approach is with av  

1:11:24is that it’s going to be there all the time and so  you need to have probably some kind of a kill neck  

1:11:30switch if if you can if it’s doing something bad  or use other things like pavel mentioned you know  

1:11:38lipo proteins or sorry nanoparticles or other  ways to deliver and this that could be used and  

1:11:46i i feel and you know it’s still very  preliminary very academic i’d say and then  

1:11:54and you know i i think we’ll we’ll be much  smarter in a few years from the work that  

1:11:59industry is doing and as i mentioned  there’s other than the company that we   formed there’s two other companies so there’s  three companies now there that i know of that  

1:12:07are using this approach and and one in a year  will hopefully be including full trials so  

1:12:13and we could see does it actually work and but  and you know until then i i have no idea and  

1:12:21but you know i think it has it has a lot of  potential and but there’s a lot of fears that  

1:12:27i have with it as i mentioned is that with  av in particular it’s there all the time and

1:12:35you know and if you mess something up it’s it for  me it’s very frightening so and so i think having  

1:12:43kill switches or having some kind of a safe way  to control expression i think is really important

1:12:53that’s amazing thank you like i said before i’m  i’m so grateful that you’re the thing gap’s on  

1:12:58your radar and that you’re working on it so really  appreciate your time today and sharing all this  

1:13:05uh one one last question uh you mentioned that  uh you guys saw the same change for nine months  

1:13:11after the application to are you expecting that  with time is the it will decrease the effect or  

1:13:20so we saw we measured the levels in those mice  of the cast of the crispr a and also of the gene  

1:13:27the sim one that we up regulated and they were  exactly the same as in you know in the beginning  

1:13:32when we injected early and one advantage for  neurons and this is the case also would be for  

1:13:39syngap is that they’re non-dividing so av is  there for the lifetime in dividing cells like  

1:13:44um when people use av for liver and there  is a an issue that you may at some point  

1:13:51have to re-inject because the levels will go down  because the cells divide and the av gets diluted  

1:13:56and but with neurons that that’s an advantage for  av again that it’s it’s there for the lifetime  

1:14:04they don’t divide but again as i mentioned my  biggest fear is it’s also a disadvantage where  

1:14:09if something’s wrong it’s still there for the  lifetime so so that’s my biggest concern for that

1:14:19any other questions just a standing ovation thank you so much  yes we second that this is an excellent  

1:14:30presentation really really appreciate it uh  fascinating information covered thank you so much  

1:14:37um we’ll go ahead and wrap if there are any  other questions but thank you so much dr ahutuf  

1:14:43and i think that’ll be it thank you so much  thank you thanks for having me thanks everyone