This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
The overall objective of this thesis is to investigate the regulation of beta cell mass in adult mice in vivo. We employ a â€˜switchable transgenic mouse model pIns-c-MycERTAM , in which controlled beta cell ablation and regeneration can be initiated and followed accurately across time. This strain has been crossed with different mice carrying gene knockouts and or other transgenes in order to conduct functional studies of beta cell mass regulation and regeneration in vivo. Specifically, to explore the role of IGF-II in beta cell regeneration pIns-c-MycERTAM mice were crossed with IGF-II knock outs. To investigate the role of PML in beta cell apoptosis, replication and tumour formation, pIns-c-MycERTAM and pIns-c-MycERTAM/Bcl-xL double transgenic mice were crossed with PML knock outs. Key results from these studies and limitations are discussed in the following sections.
6.1.1 Impact of IGF-II Loss in Regulating Beta Cell Mass
There are multiple mechanisms involved in beta cell renewal. Many efforts have been devoted to identify those growth factors which can potentially promote and support beta cell regeneration. However, the key signals triggering beta cell replication and regeneration still remain to be discovered. The IGF family has drawn a lot of attention in the recent decades as has been discussed in detail in the introduction. IGF-II is widely expressed during murine embryonic development whereas the expression level drops dramatically postnatally (D'Ercole, Applewhite et al. 1980; Moses, Nissley et al. 1980; Brown, Graham et al. 1986; DeChiara, Robertson et al. 1991). In the adult, Igf2 loss of imprinting (LOI) and reactivation have been reported in human cancer (Kaneda and Feinberg 2005; Cui 2007), and in beta cell derived cancers in mice (Christofori, Naik et al. 1994). Taken together one can intimate that IGF-II is important in regulating pancreatic beta cell mass during ontogeny, might be usefully deployed therapeutically in the adult, and is functionally important in beta cell neoplasia (Christofori, Naik et al. 1995). However, whether endogenous IGF-II plays any role in regeneration of murine beta cells in the normal adult is not known yet and forms the basis of this study. Here for the first time we exploit one of the new conditional beta cell ablation models to address this question in vivo.
In this study we investigated IGF-II re-expression in pancreas. Numerous studies in different model systems highlight the similarities between regeneration, neoplasia and ontogeny. By activating c-Myc, beta cells are killed, raising the possibility that such injury may result in the reactivation of a developmental program which could in turn support regeneration subsequently. We have shown, for the first time that IGF-II is re-expressed after beta cell injury in vivo. We chose to examine this at an early time point (after 48 hours of Myc activation), as clear beta cell injury and apoptosis is present but islet mass is still moderately maintained, maximizing the chance of detecting changes in low abundance mRNA for IGF-II. We detected IGF-II mRNA during 24hr and 48hr beta cell injury in IGF-II wildtypes but not in IGF-II KO mice (Figure 3.3.2d). Quantitative real-time PCR was performed to confirm and quantify the IGF-II re-expression from RT-PCR data. The results confirmed reexpression of IGF-II mRNA in IGF-II wildtype pancreas after 24hr c-Myc activation (Figure 3.3.2d). In IGF-II KO mice no expression was detected after either 24hr or 48hr beta cell injury.
In order to localize the origin of newly re-expressed IGF-II, we extracted RNA from islets isolated following Myc activation in both wildtype and IGF-II null mice following beta cell injury. Again early time points were needed as it has proven difficult to isolate Myc-activated islets after 24 hours because of friability through loss of cell adhesion molecule expression. IGF-II expression was not detected in islet mRNA. It is possible that the time point chosen was too early to detect IGF-II re-expression. However, it is equally possible that IGF-II re-expression takes place in the exocrine pancreas or other site without the islets. Unfortunately, neither IHC nor Western blotting has proved successful, due to the lack of decent effective antibodies against IGF-II. Future studies could address this question by means of in situ hybridization.
As mentioned in Chapter 1.2.2, IGF-II can bind to IGF-IR/IR to activate IRS/PI3K/Akt as well as Ras/Raf/MAPK pathway to trigger cell proliferation, growth and provide survival signals (Baserga et al., 1997; Franke et al., 1997; Kulik et al., 1997; Parrizas et al., 1997). Given the re-expression of IGF-II after injury in adult pancreas, we hypothesized that IGF-II might be a critical factor in the successful recovery of beta cell mass in adult mouse islets after injury.
To investigate the role of IGF-II in regulating beta cell mass, we activated c-Myc in IGF-II wildtypes and KO mice for 11 days. Allowing the mice to then recover for 4 days or 3 months, we examined the beta cell mass and numbers. Previous studies have suggested that IGF-II KO mice have 40% less body weight than normal mice at birth (DeChiara, Robertson et al. 1991). We can confirm these findings here: IGF-II KO mice have fewer beta cells and a lower beta cell mass compared with IGF-II WT mice. Most of this difference can be accounted for by the small body/pancreas weight/size of the IGF-II KO mice, because, the cross sectional area alone, (which is not multiplied by pancreas/body weight) are very similar at day 0 (before c-Myc activation) in both strains (See Appendix 2). As expected, c-Myc activation leads to a large degree of beta cell ablation in both IGF-II wildtypes as well as in KO pancreas. However, beta cell mass quantification suggested IGF-II wildtypes may suffered more beta cell ablation as we observed similar beta cell mass after ablation in both strains. To check the details of beta cell regeneration after 4 days or 3 months c-Myc deactivation in IGF-II WT and KO mice, beta cell mass/numbers at the ablation time point (c-Myc activation for 11 days) was set as a starting line for comparison. After 4 days recovery both beta cell number and beta cell mass have increased significantly in the IGF-II wildtype mice, whereas this is not seen in IGF-II knockout mice. Moreover, after 3 months following Myc deactivation, although both strains achieved recovery, IGF-II wildtypes consistently increased more beta cell mass/numbers than IGF-II KO mice. And in both strains such beta cell mass/numbers recovery was sufficient to control blood glucose even during an IPGTT.
These findings suggested the re-expression of IGF-II following c-Myc-induced beta cell injury is not sufficient to prevent c-Myc induced apoptosis. Both strains suffered large number of beta cell loss after c-Myc activation. However, the re-expression of IGF-II may be important for the early recovery of beta cell mass, as it has been clearly shown impaired beta cell regeneration at 4 days c-Myc deactivation in IGF-II KO mice. Upon long term c-Myc deactivation, data suggested that re-expression of IGF-II may continue to affect beta cell regeneration, as we found IGF-II WT consistently increased more beta cell mass/numbers than KO mice. However such effect may be less likely to be essential in beta cell regeneration in long term recovery as we also observed some degree of beta cell mass recovery in IGF-II KO mice. The results suggested other factors/mechanisms may also be activated cross time. This could also be a response to beta cell loss or a compensation of loss of IGF-II.
In the adult beta cell numbers can be renewed by differentiation of stem cells (Bock 2004; Hao, Tyrberg et al. 2006) or by replication of existing differentiated cells, of the same (Dor, Brown et al. 2004) or of a different lineage (Bouwens and Rooman 2005). An intriguing study showed that alpha cells can transdifferentiate to beta cells after extreme beta-cell loss (Thorel, Nepote et al. 2010). Also recent work by Dr. Abouna from our group has shown that beta cells may arise from a non-beta cell source, such as a stem cell progenitor, during pregnancy (Abouna, Old et al. 2010). We have also therefore determined the rate of beta cell proliferation in IGF-II wildtypes and KO mice after 3 months recovery. However, no difference was detected between the two strains. Hence loss of IGF-II may not affect beta cell proliferation in long term recovery. However, given observed more beta cell increase in IGF-II WT mice, one possible explanation could be that IGF-II expression promoted beta cell neogenesis by triggering other sources, other than proliferation.
6.1.2 Impact of PML Loss in Regulating Beta Cell Mass
The protein PML is a tumour suppressor firstly identified in acute promyelocytic leukemia. PML and its nuclear macromolecular structure PML-NB are largely involved in cell apoptosis. Research from pml null mice and cells indicated that PML can induce apoptosis in a p53 dependent or independent manner. Continual evidences suggests PML is important in inducing growth arrest and tumor suppression (Gottifredi and Prives 2001; Pearson and Pelicci 2001; Bernardi and Pandolfi 2003; Ito, Bernardi et al. 2009). However, relative few studies have been done to link PML and c-Myc. The only direct evidence was done by Lawrences group showed that Max, an important c-Myc partner, is concentrate within PML bodies, strengthening the relationship between PML bodies and Myc/Max (Smith, Byron et al. 2004). Whether PML is playing a key role in pancreatic beta cell has not been well described yet. Therefore, we aimed to determine if PML is involved in c-Myc induced apoptosis.
Firstly, we investigated the role of PML in regulating beta cell mass in the pIns-c-MycERTAM/PML+/+ (PML wildtype) and pIns-c-MycERTAM/ PML-/- (PML KO) mice. Before c-Myc activation islet morphology in PML KO mice was indistinguishable from that of PML wildtype mice. No spontaneous tumours were found in PML KO pancreatic islets at day 0 (before c-Myc activation). As previously described c-Myc activation in pancreas can induce early beta cell proliferation and afterwards large number of beta cell ablation (Pelengaris, Khan et al. 2002; Radziszewska, Choi et al. 2009). After 4hr or 24hr c-Myc activation we found a similar increase of beta cell proliferation in both PML KO and WT mice. Following up to 3 weeks of Myc activation, we observed a similar extent of islet ablation in both PML wildtypes and KO mice. These results indicated that with the loss of PML in beta cells and concomitant c-Myc activation, the beta cells do undergo early proliferation. However, the loss of PML in beta cells is not able to override the predominant apoptotic signal conferred by sustained c-Myc activation in beta cells.
Moreover, unlike the Myc-induced cancer formation that occurs, when Myc is activated in pIns-c-MycERTAM mice that have been crossed with those expressing anti-apoptotic lesions such as Bcl-xL, or where p19ARF or p53 are knocked out, even after 3 weeks c-Myc activation no evidence of islet tumour development was observed in the crosses with PML null mice. The results indicated that the c-Myc induced apoptotic signal is not modified by loss of PML. Finally, we tested the function of PML in our tumor model RIP7-Bcl-xL/pIns-c-MycERTAM by crossing with PML KO strain. As previously mentioned c-Myc activation in RIP7-Bcl-xL/pIns-c-MycERTAM mice can induce beta cell tumorigenesis (Pelengaris, Khan et al. 2002). Hence in this study we investigated whether loss of PML accelerates or aggravates beta cell tumor tumorigenesis. After 2 weeks c-Myc activation in both RIP7-Bcl-xL/pIns-c-MycERTAM/PML+/+ (RIP/PML wildtype) and RIP7-Bcl-xL/pIns-c-MycERTAM/PML-/- (RIP/PML KO) mice, there was no visible difference in the extent, nature or spread of islet tumours between mice with or without PML.
Both c-Myc and PML seem to have a strong relationship with p53 in their apoptotic functions. Both of them can stabilize p53 via regulation of MDM2, either through the formation of trimeric complex or promoting p19ARF (Zindy, Eischen et al. 1998; Kurki, Latonen et al. 2003; Louria-Hayon, Grossman et al. 2003; Bernardi, Scaglioni et al. 2004; de Stanchina, Querido et al. 2004; Qi, Gregory et al. 2004; Alsheich-Bartok, Haupt et al. 2008). Hence the deletion of PML is expected to reduce the stability of p53 and prevent cell from undergoing apoptosis. However, in our study loss of PML didnt protect beta cells from c-Myc induced apoptosis. Also loss of PML didnt accelerate c-Myc induced tumorigenesis, which suggests PML may not be involved in the c-Myc related apoptosis pathway, at least in beta cells in vivo. A novel PML/PTEN/Akt/mTOR/FoxO signaling network was described recently (Ito, Bernardi et al. 2009), suggesting an interesting relationship between PML and PTEN. The PTEN (phosphatase with tensin homology) is an important tumor suppressor. PTEN loss or mutant is frequently found in cancer progression (Li, Yen et al. 1997; Liaw, Marsh et al. 1997; Zhou, Loukola et al. 2002). PTEN could dephosphorylate PIP3 to PIP2 and acts as a potent negative regulator of the PI3K/Akt signaling pathway. By utilizing the same transgenic model pIns-MycERTAM the effects of the deletion of PTEN has been studied in regulating beta cell mass (Radziszewska, Choi et al. 2009). Research found PTEN deletion does not lead to tumor formation in beta cells. Moreover, surprisingly, they found PTEN loss did not protect beta cells from the overwhelming apoptosis and development of diabetes induced by c-Myc activation. And the conclusion was drawn that there was no evidence of tumour development with sustained c-Myc activation despite of the combined effect of loss of a tumour suppressor and activation of an oncogene in beta cells. Evidences to date suggest that loss of either PML or PTEN does not strong enough to inhibit c-Myc induced apoptosis or lead to tumour development.
6.1.3 Quantification of Beta Cell Mass and Numbers
Quantification of beta cell mass/number is of great importance in diabetes research. However, most conclusions were drawn basing on the measurements of cross sectional area or a relative volume (beta cell mass). Although the methods for beta cell mass quantification differ, to determine a percentage or proportion of the tissue in this study we first obtained the insulin cross sectional area (insulin positive area divided by pancreas area). And in this thesis to estimate the beta cell mass pancreatic weight was multiplied by insulin cross sectional area results. To avoid any bias caused by overestimated pancreatic weight, the parameter of body weight was considered during normalization. Furthermore, beta cell counting was performed both by human and software which provided the relative total beta cell number for each animal. Thus the information of beta cell mass/number was achieved to investigate the role of IGF-II in beta cell regeneration.
Bioimage information is now a novel branch of the rapidly approaching field. Due to developments in the field of bioimaging, an increasing amount of data needs to be analyzed. Data volume increases in different dimensions: higher resolutions, faster preparation and imaging, time lapse imaging, multispectral imaging. As a result, manual evaluation is no longer feasible without proper software support. The beta cell number quantification by human hand labeling, as mentioned above, is proven to be very time consuming and it can limit the analysis of data volume. In this study more than 2000 islet images were hand labeled for beta cell number counting and also a machine learning based system was trained and employed achieving the same results (Herold, Zhou et al. 2009). As described in Chapter 5 such system was validated by human labeling results. The comparison data suggested the software is reliable and counting beta cell number unbiased. We can foresee that the softwares, TIScover and Pankreas Analyzer, will largely save time and offer researchers the opportunity to achieve even more accurate information in diabetes study.
6.1.4 Limitations of Thesis and Directions for Future Work
To date we have not been able to determine the exact cellular source of IGF-II in the pancreas. Our data demonstrated that mRNA for IGF-II is found within the pancreas, but we have not been able to successfully or reliably localize peptide expression to any given cell type using immunohistochemistry, further compounded by the unavailability of an antibody that works with Western blots either, which has previously also been noted by other groups. It is hoped that in situ hybridization may help resolve this issue and we are pursuing this matter further at the present time.
To quantify beta cell mass/numbers experiments were set up with n=3 at each time point from IGF-II KO and WT strains. And we did noticed animal variation during the data analysis which probably can be corrected by introducing more replicates. However, limited by animal maintenance and time of analysis, n=3 for each time point was the maximum. Hence it is recommended to use more replicates in future beta cell mass quantification study.
It is however certainly that these studies reveal the utility of conditional beta cell ablation models in the functional studies of beta cell regeneration. It is anticipated that many more such will follow so that a more complete picture can be drawn of those factors essential for beta cell regeneration in vivo. Such factors, are of obvious therapeutic interest as they may be exploitable as new therapeutic agents in the treatment or even prevention of diabetes.
Accumulated evidences point out that PML is involved in many apoptotic pathways as well as inhibition of survival signals. The deletion of PML, a crucial tumor suppressor, was found to not affect c-Myc induced apoptosis or provide protection of diabetes development. However, Here in this study not only does loss of PML in beta cells not lead to tumorigenesis, but even with a concomitant induction of an oncogene, a situation which should create an environment that is favorable for tumor formation, we observed no tumorigenesis. Hence PML may not be a good candidate for therapeutic use in curing diabetes.
To quantify beta cell mass/numbers is a difficult and time consuming work. Here with the help of software counting system we are able to achieve a more accurate cell number data in a faster way. However, the estimation of total beta cell mass/numbers is still restricted by examining limited pancreas levels. Recently ApErio and others have developed automatic slide scanning system, ScanScope, which with suitable imaging software can be used to measure islet and beta cell area in an automated fashion, thus dramatically reducing time needed for such studies. It is reasonable to believe that more and more novel instruments will be developed to offer a much rapid, more precise and reliable information on beta cell mass/number and would be a useful tool for all diabetes researchers.
In this thesis we demonstrated for the first time that IGF-II is reactivated after beta cell injury. We detected IGF-II mRNA during 24hr and 48hr beta cell injury in IGF-II wildtype mice but as we expected not in IGF-II KO mice. However the re-expression of IGF-II didnt prevent IGF-II wildtype mice from c-Myc induced beta cell ablation. After 4 days recovery beta cell mass and numbers in IGF-II wildtype mice were significantly increased, whereas in IGF-II KO mice there was no recovery at all. After 3 months recovery both strains achieved some degree of beta cell mass and numbers which was sufficient to control blood glucose even during an IPGTT. However, results suggested consistently more beta cell regeneration in IGF-II wildtypes cross to 3 moths recovery. Data to date support the view that IGF-II is important for beta cell regeneration.
In a related study we investigated the role of PML in c-Myc induced both beta cell apoptosis and proliferation. We demonstrated that PML loss does not prevent c-Myc induced diabetes. After 3 weeks c-Myc activation both PML wildtypes and PML KO mice developed hyperglycaemia in 4 days and stayed high blood glucose in the whole procedure. Results from histological analysis confirmed islets from both strains were largely ablated after c-Myc activation, which matches the previous observation of hyperglycaemia.
The effects of PML loss in c-Myc induced proliferation was examined by activating c-Myc for 4hr or 24hr. The results indicated that with the loss of PML in beta cells and concomitant c-Myc activation, the beta cells do undergo early proliferation. However, the loss of PML in beta cells is not able to override the predominant apoptotic signal conferred by sustained c-Myc activation in beta cells. Also after 2 weeks c-Myc activation we didnt observe any severe tumorigenesis condition when PML is absent in our tumour mouse model RIP7-Bcl-XL/p-Ins-c-MycERTAM.
Finally in this study we introduced a novel machine based counting system for beta cell number identification and quantification. By analyzing more than 4000 images from 30 pancreata the validation of software was successfully achieved with the support from human counting data. Moreover results indicated when sampling islets with small or medium size (containing less than 250 beta cells) the software is counting very similar results as humans. But when sampling big/huge islets (containing more than 250 beta cells) the software is consistently counting more beta cells than human. The validation suggested software counting is reliable in either counting small, medium or big islets as it is reasonable to believe that when facing too many beta cells researchers can easily get lost or distracted. Such useful software based counting would provide a much quicker, more accurate and efficient research tool in diabetes study in the future.