good afternoon everyone. we have a special speaker doctor, dr. douglas melton. we tried to get him here not quite a year ago, and it was one of those amazing moments that happens every three or four years when we had over 20 inches of snow and it just didn't
happen. even a guy if boston could not make it here under those circumstances. if he had, i think he would have spoken to an empty room. thank you to all of you people watching by video. dr. melton has dedicated over
the past 20 years developing an approach that would be therapeutic to those with diabetes. in 1993 he was studying frog development and his infant son was diagnosed with type one later on his daughter also diagnosed with that same
condition, and he did what a dad would do who had scientific skills, well, okay, what can i do about this, and converted his own research interests in that direction. and over the course of the past many years, he has been devoting himself to trying to understand
how one can actually differentiate cells as in for instance embryonic stem cells and therefore generate an artificial pancreas made up of human cells. you would think that that kind of process could be worked out by a dedicated group in a couple
of years of just adding the right cocktail. well, this is turning out to be an extremely complex process of trying to convince cells to go down developmental pathways, which they do in all of us. we have cells that manage to find their ways into the
appropriate differentiated states. but doug's group has really been the one that tried to sort out what those signals must be. all of us have been increasingly excited over the past couple of years to see papers coming from his group documenting that he
has been able to achieve this. he will tell us that it is still a finicky bit of recipe construction to be able to make this happen, but we are clearly much further along than i thought we could be five or ten years ago and most of this is because of this hard work of
douglas melton. not only does he run this laboratory at harvard, he founded a company. it comes from the saeuplz of his considers, sam andtha. he tries to take some of these developments and turn them into a rapid entity that could be
clinically applied for all of the individuals who would benefit from having an artificial pancreas. just his credentials, he did his undergraduate work at the university of illinois and got phd in molecular biology at the mlb, at cambridge university in
the uk but has been at harvard pretty much continuously since 1981 where he is now the associate member in the children's hospital, a biologist at mass general. he is investigator in howard hughes medical institute, he's a named to if he iser, that's
cabot professor, he's zander professor as well. he he's been elected at the national academy of science and the national academy of medicine and received numerous prizes including the jocelyn medal and happy to say he continues to be a grantee of the national
institutes of health where he has also been a faithful servant in many of the thing we ask your scientists to do. so i think we're in for a treat to hear from the guy whose the furthest along in trying to tackle a problem of enormous consequence for anybody who
cares about diabetes. please join me in welcoming dr. douglas melton. [ applause ] >> thank you, dr. collins, for that very generous introduction. what i'd like to talk about today, as you heard, the idea of making pancreatic beta cells for
diabetics. one is for transplantation and the other is to understand the disease, maybe use cells for drug screening. constructed the talk is relatively simple. i'm going to spend a few minutes giving you my perspective on the
disease recollections everyone has a different view of what the diabetes is and then going to relatively quickly summarize some published work and spend more than half of my talk on what i see as the challenges for the next couple of years. i hesitate to say a couple of
years, i thought it would only take a few more years. i'm going to take on the challenge of how do we get these we deal with auto immunity. my view of the world of diabetes is relatively simple, as i said. i think beta cell is a, if not the, problem.
the picture there, a diagram of a pancreas, above an islet. the blue cells are the insulin producing cells. in type two diabetes it's well known because of insulin resistance and other factor, there's insulin malfunction. type one diabetes beta cell
destruction by the immune system leads to the absence of insulin production and the requirement for injection. in my time of looking at this tsunami of tkaoeubz is people tended to think when i began that in type two diabetes the data cell might be involved.
and now because of the work of francis collins and his collins and others from genetic studies, it's become increasingly clear, the islet is important for both types of diabetics. i'm sure some might disagree with that. if you're interested in
diabetes, you have to pay attention to the blue cell in this picture. there i'm showing that cell going away in type one diabetes. again with my wiggish history of this subject, i would say for insulin dependent diabetics, injecting insulin has been the
solution since the 1920s. what we see here is park products, enough to treat about 10,000 people for one year. what advances have been made since then. i'll say not much. cloned the human gene. lily started making it, then
this were different versions of the insulin protein, fast-acting, slow-facting. as far as i know the idea was to inject it in a pen. this has really been the advance on diabetics treatments and i don't mean to minimize the importance of injecting insulin
but it is not in my opinion a cure for the disease. the last thing i would say in that regard is when i read these sort of like pieo tech news, they talk about how important it is to develop a billion dollar drug. they don't ever mention insulin
because it's past being a drug. it's now a commodity. these numbers are listed here are one that i can verify. they are from the cdc and other places but it's estimated that two years from now $32 billion of insulin will be injected into people.
that's a lot of insulin. it's not what you'd call a block buster drug. so what do i want to do? i want to describe, as francis indicated, a kind of san francisco -- transformative approach. had the idea that you should be
able to transplant islets into patients and relieve them of the daily blood glucose checks and insulin. what i want to show you as a result of this, again the work of other, is if you look at blood glucose control in a diabetic patient in the top
panel a, you see it's pretty much all over the place, following transplantation with the immuno suppression, you see the glucose control after that. this then becomes a very simple problem. other have discovered stem excels that can make any better
of the body, beta cells are missing. connect the dots, make beta cells. i like to show this slide because i've been showing this slide to my lab for an embarrassing long time. it's more than ten years and
less than 20. they now call it my skittle diagram and they are sick of seeing it. the idea it was relatively simple, let's understand how mice and humans make beta cells and then recapitulate that. can we learn enough about the
normal development of a beta cell and use that information to direct invitro the differentiation of different human stem cells. so this diagram shows that the first step is to begin to the stem cell, then telling it to make mid gut and pancreas and
then an upbs ins producing beta cell, a cell that see creates the right amount of insulin. the little purple dots are supposed to be the insulin coming ou. this is such an obvious idea, many groups have worked on this after about a decade everyone
got to this stage of being able to make a cell capable of making all of the parts of the pancreas. when that cell or that popping of cells was transplanted into the kidney capsule of a mouse, three months later possible to show human beta cell being made.
it was hr-rpblly stuck here about three or four years ago, that is stuck here. let's just call it the green cell space. you'll have to take my word for it. when i say we have a green cell, that means with anti-bodies and
trance description -l and know for example for example that cell has -- that cell is capable of making ductal cells. just take my word for it that we have genetic markers for these skittle diagrams. because that was so difficult, two of the companies working on
this one decided to move into a clinical trial which they are in the middle of using this population of cells, some of which can become pancreatic functional beta cells after three to four months in a mouse and we will soon determine whether or not when they put
that popping of cells into a device do they similarly mature inside humans. they've transplanted them subcutaneously. that data is now available. it hasn't been made publicly available. can we go all the way, if you
like? can we make mature beta cells. i'll come back to the use of the word mature later. but let's say that respond to good luck good luck and see creates insulin. this is the work i'm going to summarize which, as i said is
many years of work which shows you could get signals to mature those cells invitro. we then went on to do all kind of expensive transcription ass al analysis and then screening which is a fancy way of saying we couldn't figure ou the answer and had guesses.
we did that by setting up fluorescent marker from the cell and asked can we move them from one skittle balls to another. i'll just remind you that that's okay if you have two factors but if you have many factors, that turns out to be a difficult what we discovered was what you
might expect is that all of the signaling pathways known by developmental biology were important for telling sells what to do and than were sufficient either to stop or move cells into the right direction. so if you count the number signaling pathways there and
imagine you have to get the right concentration and right order, you see quickly it's a very big and tedious problem. after a very long period of time, let me just guess and say approximately a decade, this is the protocol we published two years ago now, the names of the
factors here don't really matter. it's all published but some of them you might recognize. you can look along bottom there and maybe see some growth factors of small molecules familiar to you. so what i'm going to show you
now is the evidence that we solve the problem of making functional cells but i want to pause here and emphasize that this protocol which makes 30 to 40 days is absolutely not unique. in other words, it is not the only way this could be done.
the only point i'm trying to emphasize is that we demonstrated that it was possible. some in the audience might be old enough to know that when we started on this, there were people who said you can't get full differentiation invite row,
-- invitro. you cannot study diseases in a dish bit such a simple-minded while those ideas are widely accepted, they weren't in the common mindset of the time and so we weren't sure we could do this. the second thing we set
ourselves for doing which may or may not have been a good idea -- some days i thought it was a bad idea. we didn't do this with mouse embryonic stem cells. if we figured it out for mouse, it wouldn't apply to the humans. we set out to do this with human
embryonic stem cell. francis collins and others here it the nih were very helpful and trying to enable that to happen at the time congress thought that wasn't such a good idea. that's old history now. the final part we set for ourself and this came from a
meeting we had with gordon weir. i'll paraphrase, he said i'm sick of looking at papers that show one image of one cell. get real, you need a billion cells it treat a patient. we set a goal of doing it with human cells at a scale that could be clinically relevant.
to do it at that scale we had to grow a large number of cells. we grow the cells in clusters figuring out how to grow them that are islet size clusters and this shows some of the stages stained with anti-bodies. c-peptide is a better marker there at the end.
this movie is made by an undergraduate, mike siegel who is an author on this paper. these are the esl beginning. that's after to hours of growth. then after one day we figured out how to grow them as islet size steers spheres. that flash can contains enough
functional beta cells to treat one person. an undergraduate at my labs wrote up his senior thesis which was entitled making human beta cells from embryonic cells. this is a fantastic thing to have as a title of your thesis so i pushed to win the prize.
it didn't come close. someone wrote like reading shakespeare upside down with rose-colored glasses or something really annoying to me. i like to show mikey' movie for that reason. what's the evidene we made functional cells.
this is published so i'm going to go over it quickly and concentrate on the challenges. the first one shown here is if we give multi--l group challenges to these clusters of stem cell derived beta cells, you see insulin on the y axis and the glucose challenges three
times on the x axis. the idea would be -- like you can joke about it, you give it three challenges of glue glue or depolarize the membrane with potassium which are i'd and see insulin see creation. this compares very favorly to human.
and these are the cells and others that have brought into clinical trials. but i'll -- that have gone into the clinical trial are mixed that express at least or have in them at least glucose and upbs ins and other hormones.
the cells we make are indistinguishable from an adull beta cell. what this looks like in real life and is my first chance to show you one of the problems or challenges is these are the clusters section and the first path is they are remarkably loam
-- homogenius. i might look different but if you think about it being on the edge like this of 16 and a half%, this must might look like more but that's just the way they were sectioned. up until now if you're sort of half paying attention,
everything i've said would imply that we were 100% efficient, we made all of the cells beta we did not. we show that following trance plan take two weeks later they reorganized themselves and formed these islet like structures.
the red cells here are the few alpha cells. the bright green are the insulin producing c-peptide positive these cells here are the kidney and the poly hormonal cells will make some insulin positive cells where you see a huge difference. this is comparing to our stem
cells in an immuno compromised mouse with the cells are transplanted under the kidney capsule. so if i'm rushing a bit it's because i'm telling you about published things. i want to summarize that to talk about unpublished and the
challenges. let's compare the stem cell derived. they are not indistinguishable but there's a number of small number of genes that are different between the two and none of them strike us as particularly important.
the stem cell derived cells as popular insulin granule. they respond perfectly to calcium fluxes. i also didn't show you that they make these unusual finesta. i did show you they have glucose system laying invitro and there's rampant rescue in mice.
the last point i might make an editorial comment for those of you that work in this area, there are lots of thing that can cure diabetes in a mouse. an insulin pellet will, transplants, so that's not such a high bar but it does work extreme will he well to make a
diabetic mouse non-diabetic. so i want to focus on sort of a bigger goal and tell you unpublished work where we're trying to get maybe i would call it complete mastery and dominion over this process. we want to control this process. i want to talk about the
accuracy of the glucose response, spend a lot of time on the efficiency and move to you could say it's phase two of the quest to cure pipe one diabetes. so let's first look at where we stand on glue glue responsiveness. on the top right is a panel
where in the yx, you're measuring the amount of human insulin produced. so a low glucose challenge is in blue, high in red and depolarization in green. again this sort of previous protocol, the one that's in clinical trials should change
after three to four months after incubation in the human. it certainly does change after incubation in the mouse. these are one of the best preps we've received. and this is the very best prep we have of our stem cell derived beta cells.
i want to make two points, one, you may not be able to see it but that number is three and that is one so they are see creating -- secreting less whether they are responding properly is an open question, and this comes back to that issue i referred to as
maturation. problem of maturation. can we make a fully functional beta cell invitro or did something important happen when you transplant it under the kidney capsule. we don't know the answer to that.
in fact, my thinking has changed about this a bit. it may not be an event. it may not be you go from immature to mature. i could ask the audience to raise your hand if you're mature, and if you are, when did you become mature.
it would be a slow process. we really don't understand this process. the other thing we want to do is say why should we stop at the beta cell? in type one diabetics that's the key cell for sure but mouse studies among others show alpha
cells are not only preventing you from low blood glucose but delta cells are clearly important for turning off the see creation of both. as far as i know, nobody really know what pancreatic peptide does. maybe i'm missing important
papers in the literature. as far as i know, at least for my own children, appetite is not a problem. let's just say could we make alpha and beta cells. i won't show you the data. we published it, so have others. we're about a year behind making
alpha cells as well as making as far as i am, no one has made these before. so by making the pancreatic progenerators and tweaking the cells we can make delta cell. when we use our favorite em, we again see that our stem cell derived stem cells look like
delta cells. what i wanted to show today was kind of a puzzle and someone can answer this in the question period or in the reception is what should these cells do in response to glucose. they secret less at high and they can be depolarized.
i can't figure out hat cells are supposed to do. no one seems to know this because it's been impossible to get a pure aid preparation of delta cells to know what the answer should be. i don't know if that's a good thing or bad thing but that's
what we find. i don't know if that's the right direction or the wrong. i think, and this may be a daring thing to say, that within a couple of years, i could come back and tell you we now know how to make beta cells, alpha cells and delta cells.
maybe equally important for the kind of transplantation that we like to do. so now let me move to our approach to a bigger puzzle which is not only how do you control the cell types but how come we can't get the efficiency up?
by efficiencyunder we making approximately 20% of the beta that's up with the other cells? are they not getting the message? what's their problem or are we not having the right factors at the right time. we spent a lot of time, bad
practice for me to say we've wasted a lot of nih money trying to solve that problem. we had not solved that problem by just changing the factors and tweaking the system. we cannot figure it out. it could be because -- i forgot toth tpasize there are 15
different factors at six different stages. so the first thing we started to do is could we get a clue by looking at the genes that are expressed. what i'm showing you here is a good example of an environment that was very difficult, cost a
lot of money and didn't really teach us much of anything. what this shows is looking at gene expression in 14 different flash cans offer a period of time marching through the different stages of the differentiation process, looking at just some key genes.
insulin at the bottom, obviously the famous tkpaoe tphoepl. what this shows you in a way is a sort of variability from flash can to florida can but equally important is you're dealing with populations and that became such a hard problem for us that we're looking at poppings.
we couldn't figure out which gene to use for marker. fortunately for me, a graduate student in the lab was one of the aupblers on what is now like all the craze, at least in boston is single kelsey consequencing. if you go to a restaurant in
boston, there's a 10% chance the table next to you will be talking about single cell frequent wednesdayie. here's a little movie from adrian where many of you will know this. we've now really gone into this in a big way by doing single
cell seek consequencing. for the purposes of my talk today, at every stage of the differentiation to try to get a handle on how the cell make a decision. so the data i'm going to show you is based on what you could say is 50,000 different
experiments. looking at transcripts in 50,000 different cells at different stages of development to try to understand how does a cell decide if it's going to become a so this is the sort of goal. the nice thing about this work which we and two other labs
published at the same time is that the cells that were discovered in the islet were known to endocrinologists for a long time so that's sort of good news. there weren't any big surprises. to me there was one big surprise.
but other than that, if you look at the cells in the adult islet, biologists knew what was there. that's your target. can we make this beta purple cells in the middle? we found interesting facts about type two diabetics and showing some differences.
this is what we want to do. if we start look at when did your process become inefficient? we're extremely inefficient at get to the green cells but then things start to go off the rail. so here's the facts showing that at that stage only 30% of the cells do what we want, 42% of
the cells didn't respond to the factor we added and 23% of the cell went and did something we didn't want them to do. if you take that to the last stage and you ask how many different kinds of cells are there? here adrian looked at
purposefully one of a good but not a great differentiation where nearly 20% of the cells are of we want. they express the transcription factor. so this is what they want. there are a number of ways you could pose a puzzle with the
other cells like what are they doing and how come they are not paying attention to what we want them to do. many of you will know now if you're reading the single cell papers. in this case taking 18 dimensions of data, each of the
dots you see here is single cell and the intensity of the color is the intensity of the gene expression and the only thing you need to know are those two cells have very similar gene expression. so when we look at this top left program here, you see that those
end cap cells are the ones we want but it's all those other cells starting to do thing we won't and we're now learning for example that it's expressed in this line of cells. what are they doing? they've become positive, turn off nk-681.
i think it's fair to say that the literature of the field is confused about whether or not those cells ever exist normally or whether they can ever be told to do just one thing. there are very strong opinions. mine will simply be there's no telling experiment but we don't
want them. i want to get all of the cells to pay attention and become beta so if we look at the last stage, the first important thing we find from this is the following. we find that if we look at the cell we want, the insulin positive cells, we found a new
trance membrane known to be on insulin beta cells. these are the cells that haven't yet become purple. let's say they are still blue. we also find these which are the one we don't want. it's a transcription factor. that comes on of cells are going
to become non-beta like alpha or so we don't want that. and then we find a really weird cell population that i'm virtually certain does not exist. it's a very stable population that divide at a low rate of represently indication.
we don't have very good data but i'll bet that that cell type doesn't exist. it's kind of an interesting finding is you can make a kind of cell which is not cancerous in the sense that it doesn't have any mutation. the bileologist in the lab find
it fascinating. how do you make a stable kind of gene network that doesn't normally exist. so there you can say is our we want this purple cell up at the top but we had these other i've already admitted to you that we've now tried enough
times by imperial tests to control the system that we're not doing much of it anymore. i forgot to say that this is extremely expensive when you do it the way we do it because we do it in flash -- flsaks: so this is a very expensive thing so we should stop doing that.
i'm going to stop wasting nih money on that approach. what i now want to do in collaboration with the broad institute is can we get a better handle on the genes by doing this experiment. this is the model for what happens and this is what we're
going to do. can we produce a designer cell line that can only make beta we want to close the door on all those other options. so even if you're as a cell inclined to become an alpha cell, that door is going to be closed to you.
we've got a virus pool which contains knock-out targeting constructs for every human gene. the only reason we can do this experiment is because we set up our cultures at scale. you have to have 500 million cells if you want to do this properly.
we infect this with a virus pool, cells which have gone into one direction or the other. we different them it the stage we want. we separate the populations by facts, the kind i showed you before, and them we do sequencing.
if you've gone in one direction or another, what genes are you we can also do this with over-expression. where are we in this? we're at the point where adrian has solved the problem with how to create the pool with the institute and we figured out how
to do the vie al infection with each of these stage. we got the multiplicity infection down and we've got the sequencing down to know what construct was put, what virus was put in that cell. so i wish i was going to now follow with the slide to tell
you the answer. i don't have the answer. i'm just telling you that's how we're going to solve the this is my dream that we will knock out key transcription factors. just make it impossible for cells not to go down the wrong
path. once you have that information, it should be easy. i say easy, to do the same thing for making an alpha cell or a delta cell. my goal could be would be to come back in a few year's time and say this is how you make a
this is how you make an alpha cell. this is how you milwaukee a then reconstruct islets for transplantation and studies. with your sort of forgiveness, i'm going to assume something that's not true which is we've solved the problem of how to
make beta cells. you already know that's not completely true. there must be some analogy of we're in the goal line. companies are working on that but i'm now convinced we now know how to make beta cells and we'll get better at it, more
efficient and make them more mature. i'm going to finish my talk with the second challenge which is a little bit for developmentologist. i want to talk about auto immunity and where i'd like to move my lab and the others,
particularly young people in the audience. autoimmunity is a fascinating problem i think no one has a very good handle on. i'd like to ask the primary cause of autoimmuneology. some people think it's fundamentally a problem in
t-cell education. others think it's a beta cell i'd contend we don't really know. one of the things we're doing with the help of the nih over some years now is to try to reconstruct human diabetes by taking ipsos from diabetics,
turning them into beta cells, turning them into blood and turning their ipsos into blood and see in this kind of circus trick way if we can put them in a competent mouse, can we watch the disease develop. in the shorter term what we've been concentrating on is a
different approach. so to simplify the problem, you can say there are two things you've got to do or one thing you've got to do. to transplant the wells we make into a human and have them survive and function so that's like goal one.
and in my dealings with many bio engineers in boston and elsewhere, i've learned many people first come up with a new material and ask how can i make use of this. i'd like to turn that around to say here's an inexhaustible amount of cells.
can you find the material that does that. i'd glad to say we have a number of collaborators who have taken on that club and i'm going to show you the results, very promising results with one of our collaborators. two ways we can think about
block the immune system. i'm come back to biological protection. dan anderson's lab has been working for years on chemical you cannily modified algenaeuts. what i'm going to show you is the rul where we pu our human stem-cell derived beta cell-like
clusters into these spheres that have a chemical modification that seems to work very well to prevent any kind of globing. these now go way past the 175. must be more than 200 days with no effect on their ability to control blood sugar or any demonstratable immune reaction
to these clusters. that's the good news. the bad news is everyone tells us that you won't be able to go into human clinical trials with this because you'll need to put in 100,000 of these capsules. until you know that they won't make tumors and/or that you
could vacuum them out, they are not going to allow that. and some groups are working heavily and actively and aggressively into which you can put these cells and have it retreatable to have it demand straight efficacy. my last slide, there's a cartoon
i like now replacing my skittle diagram. diabetics can't control their blood sugars because they have dead, missing or dysfunctional we can make beta cells, how do we make them into patients to control their blood sugars. i've described superficially the
state of an immuno protective device. i have been excited by advances into cancer and therapy and others. i like the simple minded idea of making use of what two kind of cells due, the first being tumor so let's wonder out loud what
would happen if you expressed 4 ig in these cells. what would happen? what would the t-cells do? now because everyone and their brother can do -- can do that. the other idea i like came from my colleague. who pointed out to me that
there's this puzzle which i never thought about before. why aren't fetal cells rejected by the mother since half of their genome is their forth's gee no, ma'am. we had can chad collins knocked out class bun and class two and now trying to express hla gog.
right now the problem is not doing the genetic moderate i can says and not making the beta it's how do you test it. is testing it in a model helpful? is that going to intpopl your thinking about auto immunity and can you get blood from a type
one diabetic, would an invitro blood reaction on granular release or activeing a of t-cells convince you that you should be able to put these genetically modified cells. i'm happy for advice about that. you can tell my thinking is embryonic but that's where i
want to move my activities for now. so i'm going to the custom of everyone to remind us all that i do very little of this experimental work. in the black are the students who have done the work i've described to you today and i've
enjoyed wonderful collaborations with gordon weir, now we're working mgh. some of our collaborators, and as you can imagine i need to be writing grants all the time. those are the funding agencies. but since it was nice to be indebted here.
i did want to give special thanks to a couple of people in the nih in several ways. first for my entire career, i had depended on nih for money. and i have all enjoyed particularly recently a change i've seen in the nih to the niddk in particular.
sheryl and olivia -- i'm embarrassed. i don't know what their title is. they have a wonderful way of making gentle suggestions and helping me find other collaborators without actually telling me what to do, and i
think that's been a change in my lifetime, then i just say no, i can't do this, didn't get a grant. so i don't know if they are hear but i'm grateful to this invisible hand approach. related to that is the formation of two con con consortium.
i think these are great inventions because they force people to work together and i'm delighted to be part of them. i wish they funded my entire lab. like having single-payer health insurance, i'd like to have a single grant.
anyway, let me stop there and thank you for your attention. i'm glad to take questions. >> thanks, doctor, for really informative and up to the minute description of what's happening in this fascinating space. we have time for questions and please use the microphones so
people listening by video can hear and we can start right over here. >> first the comment now that it's approved by the fda, their artificial pancreas which measures good good luck --glucose and also secrets insulin. we have islet cell itemmers or
other endocrin tumors -- then be able to put them into people with a little bit better results starting with stem cells themselves? >> those are both two good questions. thank you. let me see if i have a slide to
tell you why i don't think it's going to work. i'll just say the best evidence, if you look at the glucose exertion is that it's much better than a hormonal pump with and without the continuous glucose monster. still if you look at how much
time the patient's blood is kept out of the good range, it's competing with how many years of evolution. those pumps, those monitors in the best case every few minutes. i shouldn't ever say google won't be able to solve this problem because i would never
believe google would look you look on every street in america on google map but i'm going to vote on the natural solution. i'm delighted by the continuous good luck good luck monitor but their blood sugar are not controlled in the normal range. your second question was should
we have taken a different approach, maybe tried to slap an insulin on and get it to behave better. sort of related to the idea is why should we have to go through a 40 day procedure. why can't we begin with the cell and add eight genes and turn it
into a beta cell. i first asked someone what are the eight genes and how would you found them? we know there's about 1,000 genes turned off and about 800 that are turned on. so i don't disagree with your idea why didn't we try this
other thing. i guess i'd say life is short and you have to place a bet on something. i'm a developmental biologist. but if you want to do it, more power to you. go for it. >>> thank you.
>> hi, doug. thank for a great talk and kind words about nih administrators. i really like the last five minute of your talk, this business of escripting beta cells with defense mechanism that may not even exist. if we push that arm all the way,
you could theoretically use some sort of gene therapy approach to protect the beta cell master that's already there in diabetic so do you see a time when actually reimplanting cells in individual may not even be needed and what are your thoughts about using the
original capacity of the pancreas itself? >> that's a great point. i would say that first of all by the time some diabetics are diagnosed there are residual is luckily to be insufficient. you'd not only have to protect those cells, you'd have to get
them to replicate. i'd also remind the audience, i don't need to remind you, that i'm already asking a lot of the fda if i want to put in an ef cell that's been turned into a beta cell. you're saying let's do gene therapy in a person with
multiple genes. i'll just remind you then that the challenge there is off tissues in the body you don't want to touch. maybe the brain ranks number one, i would say the pancreas is next. i like a good example of what i
said, why stop at cell trance plan talks and then that brings us away from something i don't think about. i think about treatment but we should think about pure or prevention. >> one thing i'm curious about are all those other cell type in
the islet. you emphasize what you need is the beta. if, for example, if you have just the beta cells and say by sorting get rid of everything else, does that work as well as the mixture does? >> a great question.
we haven't been able to do that experiment. the reason one's focused on beta is from the work on many other over the year that in a type one diabetic that's the only cell they can say is missing or dysfunction functional. they change in number and they
sha rifle -- shrivel up like a raiseen. it's amazing that you can inject you don't have to inject it into your pancreas. can you inject it anywhere and it works. you may not have to put it anywhere special.
you may be able to put it anywhere. that's the argument for the initial focus on the beta cells. having said all that, my intuition is it won't work as well as an islet would. those weren't beta transplants. no one has ever put pure beta
cells actually into any animal and said is that sufficient to control blood sugar because no one's ever had pure beta cells. does that answer the question? >> yes. >> how critical is the portal circulation? you just said it probably
doesn't matter too much. don't we have some evidence that if you want completely normal logical situation, those beta cells ought to be in the circumstance laying or is that just an idea -- >> i think it's an idea more than there being strong
evidence. there's no doubt that the portal circulation is key and that works normally but whether that's required is you be clear. one of my favorite surgeons in miami likes the omeantum. how would you put it? insulin injections work.
they don't work well enough. we may not have to have perfection. my goal could be people wouldn't have to prick their fingers and inject themselves with insulin and use an insulin pump and importantly not worry anymore than you or i do about how much
we eight and how much exercise we got. >> it's a nice way to sum up and a worthy goal indeed it is. please, let's thank dr. douglas melton again.
0 comments:
Post a Comment