<TITLE: Regeneration of the Pancreatic Beta Cells
ACADEMIC DOMAIN: medicine
DISCIPLINE: neurology
EVENT TYPE: lecture
FILE ID: ULEC150
NOTES: continuation of and continued in ULECD080

RECORDING DURATION: 45 min 37 sec

RECORDING DATE: 8.3.2007

NUMBER OF PARTICIPANTS: 15

NUMBER OF SPEAKERS: 1

S2: NATIVE-SPEAKER STATUS: Dutch; ACADEMIC ROLE: senior staff; GENDER: male; AGE: 31-50>


<S2> er thank you <NAME S1> , erm well i just planned to give a general talk about er how we see this er this field and er what we are doing in it , er so introduction we can see there (are) many others diabetes erm as a disease of the beta cell mass at least for a large part and erm only in the case of type one diabetes where it is of course known that er very important decrease in beta cell numbers is responsible for the clinical symptoms to appear but also in type two diabetes er there is very good evidence now that there is also eventually an important decrease in the beta cell er mass so this seems to be a common root for all types of diabetes that's the idea and an important thing is that <COUGH> for er , what can be planned as new er therapeutic options is that the proof of (concept) has been given that when your restore a bet- a functional beta cell mass by transplantation you can also restore their normal metabolic er functioning at least in long-standing type one diabetes patients this has been shown not only by pancreas transplantations but more recently also by a beta cell transplantation , as you know this er kind of therapy is seriously er hampered by a number of problems , well a major problem maybe is er the sh- the important shortage of organ donors there are many more er newly diagnosed cases every year than there are organs donors available erm another problem is as it has been shown recently that after a wrong follow-up that erm the metabolic control is not sustainable in 90 per cent of the successfully transplanted patients so after more than a year they require again er at least some insulin treatment , and so a new possibility that is now er under investigation in many labs and i think it will gain more and more of importance in the coming years is the idea to try to regenerate beta cells from endogenous sources , erm there are some studies that have er sparked these ideals so in er like the one by meier et al who claim that the observation of er continuing apoptotic deaths death of beta cells even in patients with that have got type one diabetes for many years that this would be indication that there is continuing regeneration in their pancreas and so that er regeneration is something that takes place but that is not er sufficient and m- could be helped in a therapeutic er way that's the idea so er is it possible to come to a regenerative therapy by stimulating beta cells that remain in the pancreas or their progenitors (that's the angle) and stimulating them with pharma- pharmacological er agents , er er- first important question (of it) has not been er fully solved at the moment is what is the capacity for expansion of the post natal beta cell mass what is er and how far can it expand and can we discontinue the expanse so can we really stimulate it and and and sufficiently what we know already for a long time there are many studies and we have done that already more than ten years ago that the beta cell mass in rodents adult rodents or in well young adult rodents can still compensate er for growing er cell (xx) demand for instance here you see that the beta cell mass which represents in fact the total number of beta cells in the pancreas of male and female er rats normal rats so from three weeks of age to 20 weeks in this study you see that there is a constant growth of the beta cell mass so- so- sometimes it has been thought that you are born with the number of beta cells for the rest of your life this is of course not true and the beta cell mass can compensate also to for instance er periods of er hypoglycaemia or females pregnancy for instance is a condition where there can be again more growth to the beta cell mass , er in this case the difference is because male rats er gain more weight than the females and <COUGH> , in this case what we have er investigated in this study ten years ago was the question er how important is replication proliferation of beta cells for this compensatory growth er and we did it in an in an indirect way by measuring the BRDU er pulse labelling index which wou- is in fact representing the percentage or fraction of beta cells which are in S phase of the cell cycle and er by counting this er we tried to calculate how important this er observed proliferation rate is this can be done by er , formulas equations which have been er wha- er different formulas that you can use which have been developed by er mainly er oncologists in the past so the cell birth rate resulting from this observed BRDU incorporation (effect) is in in a simplest equation is the ratio of this labelling index the fraction of cells in S phase divided by the duration of the S phase which has been estimated in our studies so you can er calculate the cell birth rate and new beta cells that can arise in 24 hours from this observed er incorporation of BRDU you can further calculate from that er the the beta cell number and which is er taking into account the the beta cell mass at the certain points and then you multiply this by the cell birth rate so we did these e- expi- e- experiments with our normal male and female er rats and what you see here in the white bars is the predicted or calculated beta cell number resulting from the observed replication in white and the black bars is the the measures the total absolute er a- amount of beta cells at these different time points weekly time points the (factual) conclusion from our study was that the er beta cell replications observed in normal animals er can explain the observed growth compensatory growth of the beta cell mass with exception of this time point , that is er as you know also the conclusion that has been gained more recently with transgenic mice by yuval dor in in melton's group with a a much er more er er elegant approach but er this confirms that in normal compensatory growth at least in rodents beta cell replication is probably er sufficient and i think there is no evidence whatsoever at this moment that another mechanism is operating under these conditions so normal physiological growth now what about regeneration that is a different story i think erm because if you end up with practically no beta cells anymore like in type one diabetes it is difficult to er expect that this very low number of beta cells may be able to regenerate er er functional beta cell mass er it doesn't occur in nature at least erm , so you want to study regeneration and how this can be stimulated then er we are limited to studies in rodents again and i think there are three important erm models i think it's very important to take into account that there are different models because the mechanisms that are operating er can be very different between different models that is my idea so the classical models are ablation b- of the beta cells or killing the beta cells by toxins like alloxan or streptozotocin and more recently there are now genetic ablation models that start to become available very e- elegant models like er the one by er from pedro herrera in geneva , er another type model is partial pancreatectomy which can be done in different ways 50 per cents two thirds or 90 per cent or more although 90 per cent or more in a mouse i think is er practically impossible to do from technical point of view and then there's a model that er we have liked very much to work with which is the partial duct ligation where you ligate part of the pancreas this part er the the distant part er is seriously affected by ligating which is in fact closing the ducts not closing the (xx) but closing the ducts whereas the unligated portion remains er totally normal . when you compare the growth rate as we did in the previous study during compensatory growth for the different er experimental models that i've just <COUGH> , mentioned or some of them so i've made some rough calculations from our studies and studies of other er authors so the beta cell growth rate in per cent per day er i showed you some data before er in male rats we have an average of 2.5 per cent per day in females 0.8 per cent per day when you compare that in the partial pancreatectomy model this is roughly in the same range is in fact and that is also my experience these are data from others but er also in in our experience this partial pancreatectomy only leads to very moderate er growth stimulation in fact there is no er (re- or vegeneration) of the normal beta cell mass i consider this so as as not a very interesting model 90 per cent pancreatectomy on the other hand seems to be very different this has been done with rats and there the growth rate is very er much higher than what is observed in these conditions and it's very comparable to what we have seen in the duct ligation also a much more important growth of the beta cell mass , er in an er another model that we have described the 2002 (rough) duct ligation er plus gastrin treatment so er in- intravenous infusion of gastrin in the ligated duct gives even a much more rapid growth and the comparable growth rates er we have found in mice made diabetic with alloxan and then treated with a combination of gastrin and EFG so i just want to say that the that in different experimental models you can have a very different outcome at least with respect to the capacity or the rate at with which the beta cell mass grows , and i would like to pres- th- this is the the working hypothesis that we have is that you could roughly divide er slow models of beta cell mass expansion on one hand and rapid or fast models of beta cell mass expansion on the other hand so compensatory growth is a relatively slow growth model two third pancreatectomy also also it seems already that partial beta cell ablation even (in new) genetic way er where you leave er about 30 per cent of the beta cell mass er still functional viable is also a slow model erm not growth growth roughly of er rate of between one and three per cent per day and so partial duct ligation with or without gastrin 90 per cent pancreatectomy are models where you have a much more important growth rate and i think these models are more interesting to study <P:05> erm in the duct ligation model we also try to estimate as i told you in the first slide what could be the quantitative importance of the observed proliferation of beta cells so in the same way with the same equations er we find er limited incorporation of BRDU by beta cells it's more than in normal rat okay but even from er that er er BRDU labelling index E er growth of the beta cell mass that can follow from that is er only limited and according to our calculations is at most 20 per cent of the observed growth what you see here is the beta cell mass or beta cell volume er of in the tail part of which is normally a ligated part of untreated animals here is three days after the ligation and seven days after ligation when er a maximum is reached in (a duct) in the heads there is no so er unligated parts of the beta cell mass remain stable so it's a nearly a doubling or approximately a doubling of the beta cell mass in one week time which can absolutely not be explained by this low growth (or birth) rate of beta cells so that's the indirect way with which we claim that in this condition this model there is another er mechanism which is more important which we have called neogenesis , the same goes for the mouse which i think is now emerging er and very interesting model to study beta cell regeneration or er expansion it's not real regeneration because the mass expands er following surgery so this is er mice that were er ligated and you see in the the beta cell mass here in white is the sham-operated tails of the control er in red their heads so their non-ligated parts of the , of the animals where the tail has been ligated and this is the ligated part of the tail where you see three days seven days so there is an increase this is er roughly a doubling of the beta cell mass in these parts in seven days time then it remains stable where there is no significant increase in the control , also in this case in the mice there's er an increase in the total insulin content of the pancreas which is significant after seven days er following ligation , again BRDU labelling index of beta cells in these conditions is very low it's even lower than in the rat and the potential growth rate that can follow from this is estimated with our equations as er two per cent per day which absolutely insufficient of course to explain a doubling in seven days , so this is er i think a very interesting model and er several of my colleagues <NAME> and his student <NAME> are er i think further investigating this mo- model by the use of transgenic mice erm er where you can do image tracing or you can do tracing of engine three positive precursor cells and er this is going to give very interesting results in the near-by future <COUGH> so my interpretation my working hypothesis is that self-replication of beta cells may be sufficient for compensatory growth in normal animals but is largely insufficient at least in the model of duct ligation and that neogenesis from progenitor cells is absolutely er active in this er model so if you want to study er neogenesis of beta cells from progenitor cells i think that duct ligated mouse is a very interesting model . erm , of course the big question still remains which are the beta cell progenitor cells in post-natal adults' pancreas and you know there are er many hypotheses and maybe they're all true maybe they're all false but we don't know at the moment and just to summarise er so as far as the neogenesis is concerned the oldest er hypo- no not the olde- i think the oldest hypothesis in fact is that acinar cells can act as progenitor cells this has already been er proposed many years ago that acinar cells could be able to trans-differentiate into beta cells there is the model erm amongst others bonne- susan bonner-weir is advocating that er duct cells either the differentiated duct cells can trans-differentiate to beta cells or that the duct lining contains a small population of undifferentiated cells stem cells which can neo-differentiate to beta cells there is evidence indirect evidence for all these er hypotheses i think it has not really been sorted out at the moment er i i have no doubt that this will be sorted out rather soon now thanks to the use of genetic er tracing lineage tracing using the transgenic mice you all know where er specific promoters specific for a given cell type er drives er (cre- recombinates) and so leads to expression of report which remains expressed in the cells even when they stop expressing this er cell type specific promoter so you can follow the fate of the cells and examples of this are er to use er as a promote amylase or elastase if you want to er trace acinar cells in the progeny carbonic anhydrase a part from the model that bonner-weir is using now or HNF one beta model that has been developed by er jorge ferrer in barcelona which we are now er using in collaboration with him and another very interesting as i already mentioned is the NGN three er (report) (xx) which we are using in collaboration with <NAME> in our er university , er our results are too preliminary so er even to show here i i just show you an illustration of the elastase cre-E-R-T er mouse that we use where we have a relatively good efficiency of labelling acinar cells so this can now be used to follow to trace the fate of these cells and to find out if acinar cells can contribute to beta cell neogenesis this is an example from the HNF one beta cre-E-R-T mouse where there is also a rather good efficiency about 60 per cent of er duct cells that can be labelled you see both in rather large interglobular the ducts in the intra- (xx) in the ligated ducts and even central acinar cells er <COUGH> you can er label them , so i hope that this will in the near-by future er answer the question and er especially when this is tested in different experimental models like duct ligation and others what is the contribution of these cell types to beta cell regeneration . er just to show you one more slide this is what we find as a good indication of new ge- what we consider at the moment as a good indication of ongoing neogenesis that is re-expression of NGN three normally er when we measure NGN three by a quantitative RTPCR in er in normal mice we have practically no er expression we can pra- you know detect it but after duct ligation this is in duct ligated pancreas part we have a rapid and very important upregulation of this embryonic er master switch for er pancreatic endocrine cells . i think that this is one of the good indications that can be used to demonstrate that er neogenesis is active in a given model , now this is in fact more the intentions that we have or preliminary results and what we are doing at the moment with er in vivo er beta cell regeneration studies but another option er and i want to more elaborate on that today is er to use not an in vivo model but an in vitro model ex vitro model to study beta cell regeneration or neogenesis er in more controlled conditions in your petri dish or whatever <COUGH> this is er a model that we have er devel- developed er during the last few years based on adult rat pancreas for which islets are isolated that's for other people we get the the rubbish in fact which is the er mainly the exocrine acini but these can be puri- further purified by (centrifugalisation) , er the everybody knows that when you work with er er isolations er of of er fresh cells or priming cultures you always have contaminated cells that you would n- don't do not want to have in this case what we do not want to have is contaminated beta cells because that makes it more difficult to evaluate whether we are generating new ones so what we do now er (xx) in this model is depletes the contaminating beta cells in the beginning of the culture with a rather high concentration of alloxan , er our model which has been er published in two papers is based on mono-layer cultures so first we allow the acinar cells to de-differentiate they lose their acinar characteristics they resemble a bit more like duct cells and then we allow them to make mono-layers and these mono-layers are treated with two factors EGF epidermal growth factor and LIF leukaemia inhibitory factor during three days and in these three days we see a very important regeneration in vitro of beta cells , in fact we have compared this with many other er factors or combinations of which only part is given here but it's only with this combination of EGF plus LIF or with a very similar factor to LIF maybe er ciliary neurotrophic factor CNTF that we have an effect on the in this case rough first test (xx) counting the er percentage of insulin positive cells after three days of culture with er the growth factors the differentiation factors , so there are many ways in which you can quantify that here is just an absolute count of insulin positive cells in a micro wells er at the end of the eight day er period here in control medium without EGF and LIF and here is the presence of EGF plus LIF where where we have very important increase of beta cells and in these conditions there are very very few beta cells that er incorporate BRDU , so this very important er generation of beta cells can certainly not be explained by a replication of very few remaining cells , we have shown that these er newly formed beta cells function there is a great insulin response to glucose in vitro but they also can normalise diabetic mice when transplanted onto the kidney capsule and the animals are normalised and when their graft is removed with with the kidney they revert to to hypoglycaemic <P:06> er th- the major question of course is remains where do they come from er our working hypothesis is that they come from the acinar cells that there is trans-differentiation of acinar cells under these conditions to beta cells we tried to prove that by <COUGH> genetic tracing but we worked with rats which is historical because we worked with er rat beta cells for physiological reasons this is a better model than than mouse but it's a problem if you want to to do genetic tracing studies er we have done already a lot of work still doing by using er lentiviral vectors or adenoviral vectors to transfuse the cells in vitro , er with er for instance er amylase-cre er reporter system but there are many technical problems here first adenoviruses kill the acinar cells they are very sensitive to adenovirus lentivirus works but has a lower efficiency 20 per cent approximately another problem is by the time that the lentiviral er construct has been er incorporated that it could be expressed the promoter is already switched off because the acinar cells de-differentiate very rapidly i- in during the first day of culture their expression of amylase drops to to practically zero so that's another technical problem we have to solve , but er in the meanwhile we have developed another er method which i think is very helpful at least in rats that is tracing of cells with er lectins in fact that's a very old er way but it seems to work very we er can use different er lectins that are (incorporated to) (xx) and which are taken up some of them are taken up specifically by er certain cell types like wheat germ agglutinin WGA is taken up by acinar cells whereas another one which is more known DBA is taken up by by beta cells it binds to the cells and then it's incorporated , you can also use them for in vivo labelling in fact erm at least for some time (xx) i i showed you two illustrations of WGA that has been injected directly in the pancreas of a rat and the the WGA is green it's inside acinar cells which are er labelled in the reds er for lipase acinar marker (and you) see an islet which is ab- absolutely negative does not contain er the lectin here's another example where islet has been stained for synaptophysin and where you see that the lectin is indeed not present in any of the islets but it's present in many of the acinar cells you can note here a duct which is also completely negative , so V- er WGA is a good er tracer for erm acinar cells <P:05> er so this is an illustration of the freshly isolated acinar cells which here are er cytospinned so they are clustered and they are labelled this er this er WGA which initially binds to the membrane and then will become incorporated this is seen eight days later the cells still contain the lectin but not inside the cell the isosomes here they are stained for er another marker (xx) , so what we find when we compare these two lectins acinar and the ductal lectin is that erm the beta cells at the end of our culture the insulin positive cells do contain the acinar lectin but we have never found a beta cell containing the lectin with which we bind which we with which we label duct cells in fact duct cells in our preparations are a contamination (that are) less than five per cent so there are not many duct cells but even these do not contribute to the beta cell neogenesis whereas WGA does so we consider this as a good proof that the acinar cells really trans-differentiate to the beta cells now i hear you (coming) or er we had the same question there are many studies er suggesting that beta cells could de-differentiate in the cultures and maybe these cells then er could re-differentiate and we could miss them well to check for that we also used lectins , and from the same rats <COUGH> when we purify on the one hand the acinars the acini on the un- ot- other hand (either) beta cells we labelled them in a different way so the acinar cells were labelled with V- WGA in this case this experiment a red (xx) and the beta cells we labelled with another lectin concanavalin A which in fact er will be taken up by many cells in this case we added it er with another fluorochrome to the purified beta cells then we again mixed the cells and we cultured them in exactly the same way er alloxan treatment mono-layer formation and treatment with EGF plus LIF and then we do insulin immunostaining and all the erm so this is the mixing all the insulin positive cells that we do find in mono-layers at the end of the culture well they contained the acinar lectin and we could not find any of the insulin positive cells containing FITC in fact this experiment shows that you really kill the beta cells with alloxan <P:07> another interesting thing is er so we considered NGN three as a good indicator of er neogenesis er it is in fact er s- similar to the generation of the first endocrine cells during (real) development which is dependent on NGN three expression by the progenitor cells and we found in our model that there is also a re-expression of NGN three this is at RNA level so normally we cannot detect NGN three in er in er freshly i- isolated <COUGH> er pancreatic cells erm this is at different time points i- in the mono-layer the mono-layers a- are treated during three days with EGF plus LIF you see that there is a rapid up-regulation of NGN three and then it goes down again and it's followed by insulin which is initially also practically undetectable then insulin er comes up this is in the state of er RNA level and as a protein level we have er measured these by immunocytochemistry in this case it's the absolute number of cells in the micro wells that's er express in the NGN three by immunocytochemistry or insulin and you see there is rapid up-regulation and then again it goes down to zero whereas insulin remains high so there is a transient expression of NGN three in this in this model it doesn't prove that NGN three is required for the er neogenesis but this we could prove by using er small interfering small hairpin RNA deviced er we can knock down NGN three er expression in in the cells so we have published this last year with these RNA-E experiments we could show that the beta cell neogenesis in this in vitro model is absolutely dependent on NGN three expression so in the cells where NGN three expression is blocked by the small hairpin RNA er there is no insulin expression <P:05> erm we also showed by er inhibitors pharmacological inhibitors of er possible relations specific inhibitors that the er model is entirely that the neogenesis is dependent on the jak-two-stat-three pathway which is the pathway that is known to be activated by er er LIF or CNTF that can also be activated by EGF and we think that in fact the we absolutely need the the two compounds we think that EGF and LIF are reinforcing the same er pathway so the jak-two-stat-three pathway and that this leads to NGN three expression although that is not entirely clear at the moment how that is done , and another interesting point is that in these cultures of course n- not all the acinar cells start to express insulin it's only about ten per cent of the cells that become insulin positive so we would like to know why not all the cells behave in the same way is there heterogeneity intrinsic to the acinar cells are there different subpopulations of cells could be another possibility is that there is a similar s- er situation as in the embryonic development er where NGN three expression is associated with lateral inhibition so you know this means that as it is also the case in neural development cells that er express NGN three that will er start to differentiate into endocrine cells in this case <COUGH> will also express er ligands for er the notch-receptor which will then er have an effect on the neighbouring cells that express a notch-receptor and the effect will be er a suppression of NGN three expression inhibition of endocrine differentiation so these cells will not become endocrine cells . it is believed that this is a very important mechanism that regulates the number of endocrine cells during embryogenesis so is this system reactivated also in our er adult model yes at least notch-signalling is clearly er becoming active this is er at RNA level you see that notch one is day one day two day three day four day five of the first five days of culture of our acinar cells notch one becomes expressed so like strongly up-regulated er notch two also then the target (xx) targets of notch-signalling are also expressed er from day two or three on namely hes-1 and also hey-1 and hey-2 and two ligands are also expressed to some extent in our cultures this is delta-like one and jagged one . erm also the protein level we can er stain for these this is a stain for hes-1 in the i- initial acinar cell culture day one we cannot detect it but er four days later day five most of the cells have strong hes-1 expression in th- in the nucleus so demonstrating that indeed in these cells notch-signalling is active , another proof of that is staining for the intracellular domain of notch one an antibody against the NICD part which is cleaved off when the receptor is activated shows you no staining at day one but many cells are stained mainly in nucleus er day five or in the mono-layers seen in (xx) they are the same cells as here , so notch-signalling is active and could well be the cause that er cells are erm inhibited from endocrine differentiation and that can be demonstrated in (these) by adding in fact the ligands and this experiment what we did was we coated the plastic of our mono-layer cultures before we made the mono-layers with the proteins the er (very) common proteins delta-like four which is a notch ligand or jagged one which is another notch ligand and you see that the percentage of insulin positive cells in these cultures normally around ten per cent is er very significantly er decreased when these notch ligands are coated so confirming that these ligands by activating notch are inhibiting the beta cell differentiation . so we hope that doing the ot- the other way around namely inhibiting notch-signalling could increase er the efficiency of our system and that is indeed what we have now found er so when we inhibit notch-signalling er with the notch-inhibitors we find , i hope that i have some data from that many more insulin positive cells in these cultures <P:09> well i have not the data but er i mean not erm the figures but we have on average there not 10 per cent beta cells but in these cultures we have an average er 30 to even 35 per cent beta cells which is er quite nice and which we hope now to even further improve by using other inhibitors or using S-I-RNA against er notch one for instance <P:07> and then er er this very small last part that it is also published er data but an interesting question of course er i think all this in my opinion ma- may lead to pharmacological interventions that is a hope at least that in the future we nee- we may be able to control the beta cell mass or th- the size of the beta cell mass by pharmacological interventions and er i just want to <COUGH> very shortly summarise the the model that we have er worked out for that which which works very well which has also been used by others namely where beta cells are er killed by alloxan treatment you can kill 90 95 per cent of the beta cells er by alloxan in- injection this is a such an islets one day later where most cells are dead and surviving cells in the islets are in fact (xx) beta cells and what we did was we treated the animals with combination of gastrin plus epidermal growth factor with mini alzet pumps one day later or we also tried at later time points er but normally we used one day after alloxan and then after seven days the pump is removed to be sure there is no further (pedigree) of the er factors and and then this model er seems to work very well so the glycaemia er just on the right the data here is er er going up to very high in the alloxan treated animals but it's rapidly very rapidly restored by the gastrin EGF treatment w- well we don't understand completely how it works but er there is certainly er an important er effect on the beta cell mass , this is the relative insulin positive tissue area which in the alloxan treated animals goes down very rapidly to to very low but which is partially restored not entirely but partially restored in the gastrin EGF treated animals so it er about 40 per cent of the normal beta cell mass is is is restored on the other hand the fact that insulin content the total er protein insulin content is restored to normal rate or even somewhat more than normal in these animals <P:05> so there is an expansion of the beta cell mass during this er one week of treatment which can be measured er it's about er , four fold increase or something in seven days in these between these ti- two time points so which is a radid very rapid growth and again in these particular conditions we do not see an effect on er beta cell er proliferation so the BRDU labelling index of beta cells is no- is not changed by this treatment . er although i know from <NAME S1>'s work that beta cells er express EGF receptor but in this case we do not see an effect on beta cell er proliferation , although we do see an effect on for instance duct cells <P:05> so i would like to conclude i hope you , got enough of this erm there exist experimental models in which beta cell mass expands slowly and possibly this slow growth can be explained by beta cell self-replication but there are other models where there's a very rapid expansion and where neogenesis is required at least that is concluded indirectly from the er evaluating the importance of the beta cell proliferation and as i said i hope that now the er genetic tracing models will be able to further er quantify the importance of neogenesis , er it can be stimulated or induced in vivo neogenesis by growth factors like gastrin and EGF or in vitro by EGF and LIF but i must add that this is only true in injured pancreatic tissue when we treat normal animals with a combination of gastrin and EGF we cannot increase the beta cell mass . and thirdly in vitro at least beta cell neogenesis from acinar cell trans-differentiation which i don't think is is necessarily the only way that er beta cells can er er regenerate from but it is i think certainly one of the possibilities (xx) we at the moment we even do not know if it occurs in vivo but in vitro you can induce it and there it is depe- dependent on NGN three expression jak-stat-three an- and notch-signalling <P:05> so our perspectives is to further unravel the mechanism and regulation of the trans-differentiation that er i described and of course a major aim is to try to transpose that to human pancreas cultures . from er the er er available er donors the largest mass of the er pancreas is in fact discarded not used for transplantation because of (exocrine) cells erm genetic linear tracing to characterise the beta cell progenitors i hope this will finally answer the the question in near-by future , and erm what we would like to continue in the future also is to find improved protocols to induce beta cell regeneration in vivo by pharmacological agents of which i hope one day they will be useful to treat diabetes , so i'd like to thank er my er students and collaborators in my er research group and also my colleague <NAME> and er his collaborators <NAME> and <NAME> thank you </S2>
<APPLAUSE>
