Introduction

Loss of pancreatic -cell function is one of the major causes of diabetes mellitus.¹ Transplantation of pancreatic islets is a potential therapeutic approach for diabetes. However, its clinical approach has several limitations, including immunological rejection and insufficient availability of cells for transplantation.²

Thus, the latest attention has focused on the possible therapeutic use of the controlled differentiation of adult pancreatic stem cells into insulin-producing cells. Recently, several studies showed that the prolonged culture of isolated ductal tissues from mice and humans resulted in the production of functional endocrine cells, suggesting that adult pancreatic stem cells are located at or near the ductal tissues.^3,4 Pdx-1, a pancreatic -cell-specific transcription factor essential for normal pancreatic development, is highly expressed in ductal cells and in islets during pancreatic regeneration following 90% pancreatectomy in rats.⁵ Islet neogenesis and an enhanced expression of pdx-1 in the proliferating ductal tissue was observed in transgenic mice that have an aberrant expression of interferon- in their pancreatic -cells.^6,7 These data suggest that transcriptional regulation involving pdx-1 is essential for endocrine neogenesis in the adult pancreas and also suggest that ectopic expression of pdx-1 in ductal cells might induce their differentiation into endocrine cells, which may provide a novel strategy for the in vivo induction of -cell neogenesis.

In the present study, we performed the adenoviral vector-(AdV-) mediated gene delivery of transcription factors, pdx-1 and neurogenin3, to the mouse pancreas by the retrograde intra common bile ductal (ICBD) injection. The results showed that the AdV-mediated expression of pdx-1 induced the proliferation of pancreatic ductal cells and the neogenesis of insulin-producing cells, providing a potential new approach for curing diabetes.

Results

Construction of adenoviral vectors

Full-length cDNAs of mouse pdx-1 and neurogenin3 were cloned, and AdV-pdx-1 and AdV-ngn3, AdVs expressing these cDNAs, were generated as described in Materials and methods.⁸ These vectors were designed to express enhanced green fluorescence protein (EGFP) in addition to pdx-1 or neurogenin3 from a single transcript (Figure 1a). To test the ability of these AdVs to express pdx-1 and neurogenin3, we used them to infect 293 cells. On day 3 after the infection, EGFP expression was detected in the 293 cells by fluorescence microscopy (data not shown). The cells were harvested and subjected to RT-PCR analysis and Western blotting. Expression of exogenous pdx-1 and neurogenin3 mRNA was detected by RT-PCR analysis (data not shown), and the expression of pdx-1 protein (45 kDa) was detected by Western blotting using an anti-mouse pdx-1 serum (Figure 1b).

Figure 1.

The structure of the expression cassette integrated into the adenoviral vector and AdV-mediated pdx-1 expression in vitro. (a) The expression cassette contained the CAG (human cytomegalovirus immediate-early enhancer-chicken -actin hybrid) promoter⁴⁷ or the AG (chicken -actin) promoter, mouse pdx-1 or neurogenin3 cDNA, followed by an internal ribosome entry site (IRES) and EGFP cDNA,⁴⁸ which led to the concomitant expression of EGFP, detectable by fluorescence microscopy. An AdV containing an expression cassette lacking transcription factor cDNA was used as a control. Red arrows indicate the RT-PCR primers for exogenous pdx-1 or ngn3 mRNA. (b) Expression of mouse pdx-1 protein (45 kDa) was detected by Western blotting using anti-mouse pdx-1 serum on day 3 after infection.

Full figure and legend (105K)

AdV-mediated gene expression by ICBD injection

The AdVs were introduced into mice by ICBD injection (see Materials and methods). To determine whether efficient gene delivery could be performed in vivo by ICBD injection, we examined AdV-mediated expression of -galactosidase in the liver, pancreas and duodenum on day 4 after the ICBD injection of AdV-lacZ. Expression of -galactosidase was clearly observed in both the liver (data not shown) and the pancreas (Figure 2a, b), whereas it was not observed in the duodenum (data not shown). In the pancreas, -galactosidase was mainly expressed in the exocrine cells surrounding major ducts, but not in the islets. These results suggested that most of the AdVs infected the exocrine cells and the peripheral ductal cells. Histological analysis also indicated hardly detectable inflammatory or mechanical lesions in the pancreas after ICBD injection. We conclude that ICBD injection is an efficient and safe method to introduce AdVs into the mouse pancreas, especially into exocrine and peripheral ductal cells. Thus, we used the ICBD injection method in the following experiments.

Figure 2.

AdV-mediated expression of -galactosidase. (a, b) High-level expression of -galactosidase was observed in the pancreas after ICBD injection of AdV-lacZ. -galactosidase was mainly expressed in the exocrine cells and the peripheral ducts surrounding major ducts. Scale bars represent 50 m.

Full figure and legend (528K)

Adenovirus-mediated gene delivery of transcription factors

To determine whether efficient gene delivery was performed in vivo by AdV-pdx-1 or AdV-ngn3, liver and pancreas were examined by fluorescence microscopy on day 7 after the injection. The ICBD injection induced EGFP expression in both the liver and the pancreas (Figure 3a–d), whereas intravenous injection induced EGFP expression only in the liver (data not shown). Expression of exogenous pdx-1 and neurogenin3 mRNA was also detected by the RT-PCR analysis (data not shown). Furthermore, double-immunohistochemical staining for pdx-1 and EGFP showed that the EGFP expressing cells also expressed pdx-1 (Figure 3e,f). To further increase the efficiency and specificity of gene delivery into the pancreas, we clamped the bile duct during injection. However, this procedure tended to cause severe pancreatitis later, probably due to excessive hydrostatic pressure (not shown). These data suggest that ICBD injection is an efficient and safe method of AdV-mediated gene delivery into pancreas.

Figure 3.

AdV-mediated gene expression in mice on days 3–7 after ICBD injection. (a, c) The appearance of the mouse pancreas (Pa), liver (Li), common bile duct (CBD), and spleen (Sp) were observed by stereoscopic light microscopy on day 7. (b, d) Expression of EGFP was observed in both liver and pancreas, but not in the spleen after ICBD injection. (e, f) Pancreatic sections containing ducts were immunohistochemically stained for EGFP (green) and pdx-1 (red) on day 3 after AdV-pdx-1 injection. Scale bars represent 3 mm (a–d), 50 m (e, f).

Full figure and legend (579K)

Ductal proliferation and newly induced insulin-positive cells in pancreas

Histological analysis on day 7 after the injection revealed that the administration of AdV-pdx-1 led to extensive ductal proliferation (Figure 4b). Proliferation of ductal cells was confirmed by BrdU incorporation (Figure 4c). The proliferated ductal cells showed immunohistochemical staining for cytokeratin 19 (CK19), a marker of ductal tissues (data not shown). In the area displaying ductal proliferation, we observed the appearance of insulin-positive cells or clusters (Figure 4 d–f). The administration of AdV-ngn3 led to more minor changes in the pancreas (data not shown) and that of control AdV did not lead to any detectable changes (Figure 4a). To quantify these histological observations, mice were allocated to five groups and each group (n = 4–6 per each group) was subjected to ICBD injection of lactated Ringer's solution without AdV (AdV(-)) or with AdV with or without pdx-1, ngn3, or both, as indicated in the Figure. More than four randomly selected sections of the pancreas from each mouse were examined. Thirty-one to 39 normal islets were found per section (Figure 5a). There were no statistically significant differences in the numbers of islets among the five groups. However, insulin-positive cells adjacent to the proliferated ductal cells were detected in 94.4% of the sections from mice treated with AdV-pdx-1 and in 22.2% of the sections from mice treated with AdV-ngn3. The average number of newly induced insulin-positive cells and islet-like cell clusters were 11.9 and 4.5 per section from mice treated with AdV-pdx-1, which was estimated to be about 5% of the total pancreatic endocrine cells, and 0.65 and 0.18 per section from mice treated with AdV-ngn3, respectively (Figure 5b). In contrast to AdV-pdx-1, neogenesis of insulin-positive cells and cell clusters adjacent to the proliferated ductal cells were not seen in the pancreas of control mice (AdV (-) or control AdV). These data clearly suggested that introduction of AdV-pdx-1 into mouse pancreas resulted in ductal proliferation and endocrine neogenesis.

Figure 4.

Pancreatic sections on day 7 after AdV-mediated gene expression and quantification of histological changes. (a) Pancreatic section from a mouse treated with control AdV. (b) Ductal proliferation induced by AdV-pdx-1. (c) Ductal cells showing BrdU incorporation (arrows) by immunohistochemistry. (d–f) Insulin immunohistochemical staining revealed the neogenesis of newly induced insulin-positive cells (brown) accompanied by prominent ductal proliferation. These newly induced insulin-positive cells formed various sizes of cell clusters that resemble pancreatic islets. Such changes were observed to a lesser degree in pancreata treated with AdV-ngn3, and no changes were observed in pancreata treated with control AdV (data not shown). Scale bars represent 50 m (a–c), 10 m (d–f).

Full figure and legend (843K)

Figure 5.

Quantification of histological changes (days7). Four to six random sections for each pancreas were stained with hematoxylin/eosin and with anti-insulin antibody and examined by light microscopy. Then, the islets and insulin-positive cells on each section were counted as described in Materials and methods. (a) The numbers of normal islets. No statistically significant differences were observed among the five groups. (b) The number of newly induced insulin-producing cells and islet-like cell clusters near or in the proliferated ductal tissues. Statistical analysis revealed that AdV-pdx-1 led to a significant increase in the number of newly induced insulin-positive cells or cell clusters in pancreata. Error bars represent standard deviation. #P < 0.05 compared with AdV-ngn3-treated mice.

Full figure and legend (40K)

Immunohistochemistry and quantitative RT-PCR for several markers

To investigate the molecular mechanism underlying these changes, we performed immunohistochemical analysis on day 7 after AdV injection. In the proliferated ductal area, the insulin-positive cells were observed, but EGFP-positive cells were hardly observed. It was probably because the EGFP expression by AdV had already been diluted out due to extensive proliferation of the infected cells by day 7. On the other hand, insulin-producing cells became detectable after transient AdV-mediated gene expression had decreased. Glucagon- or nestin-positive cells were not detected in this area (data not shown).

Quantitative RT-PCR analysis⁹ was performed for several marker genes. Endogenous pdx-1 mRNA increased 7.2-fold in the pancreata treated with AdV-pdx-1 and increased mildly in the pancreata treated with AdV-ngn3 (Figure 6a), but endogenous neurogenin3 was not detected in the pancreata treated with AdV-pdx-1or with AdV-ngn3 by RT-PCR (data not shown). CK19 mRNA, which is a marker of ductal tissues, increased significantly in pancreata treated with AdV-pdx-1, but not in the pancreata treated with AdV-ngn3 (Figure 6b). Nestin mRNA, a recently suggested marker of pancreatic stem cells in islets, tended to be higher in the pancreata treated with AdV-pdx-1 than in those of other groups (Figure 6c), but immunohistochemistry for nestin did not show the apparent induction of nestin in the pancreas (data not shown). These data suggested that endogenous pdx-1 was increased by the exogenous expression of pdx-1 and that this pdx-1 induction may be directly or indirectly related to ductal proliferation and -cell neogenesis.

Figure 6.

Quantitative RT-PCR analysis. Pancreata treated with AdVs were subjected to RT-PCR on day 7 after the injection. (a) Endogenous pdx-1 increased 7.2-fold in pancreata treated with AdV-pdx-1 and 3.9-fold in pancreata treated with AdV-ngn3. (b) Cytokeratine 19 (CK19) expression was increased significantly in pancreata after administration of AdV-pdx-1 (P < 0.05). (c) Nestin mRNA tended to be higher in the pancreata treated with AdV-pdx-1 than in other groups, but the difference was not statistically significant. Error bars represent standard deviation. *P < 0.05 compared with control AdV-treated mice. P < 0.05 compared with AdV-ngn3-treated mice.

Full figure and legend (31K)

Discussion

The purpose of this study was to investigate the possibility of the controlled induction of -cell neogenesis by AdV-mediated expression of a transcription factor. AdV is an attractive vector for delivering genes into various mammalian tissues.^10,11 In view of potential clinical applications, ICBD injection ¹² is an ideal method for AdV-mediated gene delivery into the pancreas because it is very similar to the well-established ERCP method in humans, but few attempts have been made to perform it in animals. In the present study, we successfully performed AdV-mediated gene delivery into the mouse pancreas by ICBD injection. The side effects, such as pancreatitis were not evident.

To induce -cell neogenesis in the pancreas, we focused on two transcription factors, pdx-1 and neurogenin3. Pdx-1 is a homeodomain transcriptional activator of a number of genes involved in endocrine function of the pancreas, such as the insulin, somatostatin, glucokinase, and GLUT2 genes.^{13,14,15,16,17,18,19,20} Moreover, pdx-1 is a crucial regulator of pancreatic development. During embryogenesis, pdx-1 is expressed in the ventral and dorsal walls of the primitive foregut where the pancreatic buds later form. As development proceeds, pdx-1 is expressed in the exocrine and endocrine progenitors, and in the duodenal epithelium. In adults, pdx-1 is mainly expressed in the pancreatic cells.^{5,21,22,23,24} The targeted disruption of the pdx-1 gene in mice and an inactivating mutation of the pdx-1 gene in a human infant both resulted in agenesis of the pancreas.^23,25

Neurogenin3 is another determinant factor for the development of endocrine cells. Neurogenin3 expression during endocrine cell genesis peaks at E15.5, but is greatly diminished at birth, and is largely absent from the mature pancreas. Mice lacking the neurogenin3 gene fail to generate any pancreatic endocrine cells and die postnatally from diabetes. In their developing pancreas, lack of expression of several essential regulators such as isl1, pax4, pax6, and NeuroD/BETA2 were observed.^{22,26,27,28,29,30,31} These data suggest that pdx-1 and neurogenin3 lie upstream of other transcription factors involved in endocrine development and play a key role in initiating endocrine cell specification and differentiation from progenitor cells.

Here we showed that induction with AdV-pdx-1 led to the proliferation of pancreatic ducts and -cell neogenesis. On the immunohistochemical analysis for EGFP and insulin on day 7 after AdV injection, insulin-producing cells were observed but EGFP expression had diminished in the proliferated ductal area. It was probably because the EGFP expression was diluted out due to the extensive proliferation of ductal cells (Figure 4b). Therefore, we could not conclusively tell that the insulin-positive cells were derived from the AdV-pdx-1-infected ductal cells. It would be necessary to establish the method to stably mark the AdV-infected cells. Regarding the origin of -cell neogenesis, two possibilities can be considered. The first is the transformation of non-insulin-producing cells into insulin-producing cells. Recently, several studies have shown the transformation of adult non-insulin-producing cells into insulin-producing cells. In vitro, the alphaTC1.6 cell line, which was derived from pancreatic -cells, was stably transfected with a pdx-1 gene and shown to express insulin and glucokinase in the presence of betacellulin.³² In vivo, AdV-mediated pdx-1 expression in mouse liver activates the insulin 1, insulin 2, and prohormone convertase 1/3 genes.³³ In the chicken embryo, pdx-1 is normally expressed in the duodenum, and duodenal cells never form pancreatic buds, but ectopic expression of pdx-1 using in ovo electroporation causes cells to bud out of the epithelium-like pancreatic progenitors.²² These reports suggest that the endodermal cells, when pdx-1 is expressed at high levels, lose their cellular stability or differentiated status, gain a transient pluripotency, and obtain endocrine properties. Sharma et al. hypothesized that the pdx-1-positive ductal cells, which can transiently regain their pluripotency, can be considered the true precursor cells of the adult pancreas in their rat model of pancreatectomy.⁵ In the present study, endogenous pdx-1 expression was also significantly elevated in the pancreas by AdV-pdx-1. Marshak et al.³⁴ showed that the proximal promoter region of the pdx-1 gene is important for its -cell-specific regulation and that the pdx-1 protein itself binds to this enhancer element.³⁵ These data together with our results suggest that the positive autoregulatory loop of pdx-1 was activated in the cells of pancreata treated with AdV-pdx-1, and that the ductal cells, when pdx-1 was highly expressed, lost their mature status and transformed to an undifferentiated proliferating status, some of these cells then redifferentiated into insulin-producing cells.

In the present study, CK 19 mRNA was also significantly elevated in the pancreas treated with AdV-pdx-1 (Figure 6b). CK 19 is normally expressed in the ductal cells and in adult islets. Recently, several studies revealed that CK19 is expressed more strongly in neonatal islets, which undergo significant islet growth, than in adult islets, and is also expressed in proliferating ductal cells during tissue regeneration after ductal ligation or partial pancreatectomy in rats.^36,37,38,39 These data suggest that the ductal proliferation and -cell neogenesis are reflected in the high level of CK19 expression in pancreata treated with AdV-pdx-1 in the present study.

The second possibility for the origin of newly induced insulin-producing cells is the differentiation of adult stem cells into insulin-producing cells. Numerous studies have suggested that adult pancreatic stem cells are located at or near the ductal tissues.^3,4 Recently, the existence of nestin-positive pancreatic stem cells within rat islets, ducts, and exocrine cells was demonstrated. Nestin has been thought to be a marker of neural stem cells.^40,41 The nestin-positive pancreatic stem cells derived from islets can proliferate and differentiate into cells with phenotypes of hepatic and pancreatic cells in vitro.⁴² In the present study, expression of nestin mRNA tended to be higher in pancreata treated with AdV-pdx-1 than in those of other groups, although nestin-positive cells could not be detected in the proliferating ductal area by immunohistochemistry (data not shown). It is possible that pdx-1 induced the differentiation of nestin-positive stem cells into insulin-producing cells, leading to the activation of self-renewing nestin-positive stem cells, which was reflected in the elevation of nestin mRNA, although further studies are needed to confirm the role of nestin-positive cells.

AdV-ngn3 showed a weak induction of insulin-producing cells in the pancreas (Figure 5b), which was not as prominent as AdV-pdx-1. During endocrine cell genesis, neurogenin3 is expressed transiently, and it is absent from mature -cells. Because it was possible that the effects of neurogenin3 were manifested after the disappearance of neurogenin3 expression, as is seen in endocrine cell genesis, we examined the pancreas 1 month after the injection of AdV-ngn3. However, no further histological changes were observed (data not shown). Consistent with our data, several studies have shown that neurogenin3 alone is not able to induce the differentiation of insulin-producing cells.^22,29,43 We also examined the effect of coexpressing pdx-1 and neurogenin3, but we could not detect any synergistic effect (Figure 5b). Further studies are needed to clarify the role of neurogenin3 in the regeneration of insulin-producing cells.

In the present study, the proportion of the newly induced insulin-producing cells in AdV-pdx-1-treaed mice was estimated to be about 5% of the total pancreatic endocrine cells. We tested this approach in the streptozotosin (STZ)-induced diabetic mice. All mice showed severe hyperglycemia 4 days after administration of STZ, and then we performed ICBD injection for them. However, significant anti-diabetic effect was not observed by AdV-pdx-1 treatment, although some mice showed improvement of their hyperglycemia (data not shown). We believe that further improvement of this method, such as simultaneous delivery of AdV expressing -cell growth factors or other transcription factors, should promise to increase the number of newly induced -cells.

In conclusion, we demonstrated that -cell neogenesis and ductal proliferation were induced by the AdV-mediated expression of pdx-1 with concomitant activation of the pdx-1 autoregulatory loop. This is the first demonstration that -cell neogenesis is induced by gene delivery into mouse pancreas. Because ERCP is established as a safe technique in humans, the successful introduction of transcription factors by ICBD injection has promising applicability to humans and provides a novel strategy for gene therapy to treat diabetes mellitus.

Materials and methods

Animals

9–12-week-old male C57BL/6J mice, 25–28 g in body weight, were used in our study. All experiments were conducted in accordance with the Osaka University Guidelines, which are based on the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Construction of AdV-producing mouse pdx-1 and neurogenin3

A full-length cDNA of mouse pdx-1 was cloned from the cDNA of MIN6, which is a mouse insulinoma cell line,⁴⁴ and a full-length cDNA of mouse neurogenin3 was cloned from mouse brain cDNA. After confirmation that their sequences were the same as those reported previously,^13,28 AdVs were generated to express these cDNAs as described previously.⁸ An AdV containing an expression cassette lacking transcription factor cDNA was used as a control.

Western blotting and RT-PCR

Total protein was extracted from 293 cells 3 days after they had been infected with AdV-pdx-1, and assessed by Western blotting and enhanced chemiluminescence (ECL kit; Amersham Corp., Arlington Heights, IL, USA) using rabbit anti-mouse pdx-1 serum (a gift from Dr Kajimoto) and horseradish peroxidase-conjugated goat anti-rabbit IgG antibody.

After AdV infection, total RNA was extracted from 293 cells on day 3 and from mice pancreata on day 7, by the acid guanidinium–phenol–chloroform (AGPC) method, and was subjected to cDNA synthesis using ReverTra Ace (TOYOBO, Tokyo, Japan). The cDNA was subjected to PCR with the following primers (Figure 1a): AdV-pdx-1 and AdV-ngn3 forward primer which hybridizes with the region of the -actin promoter, 5'-GCTGGTTATTGTGCTGTCTC-3'; AdV-pdx-1 backward primer, 5'-TCCTCTTGTTTTCCTCGGGTTC-3'; AdV-ngn3 backward primer, 5'-ACCCAGAGCCAGACAGGTCT -3'.

To determine the level of endogenous pdx-1, endogenous neurogenin3, CK19 and nestin mRNA, RT-PCR was performed with the following primers: endogenous pdx-1 forward primer, 5'-TGGATAAGGGAACTTAACCT-3'; backward primer, 5'-TTGGAACGCTCAA GTTTGTA-3'; endogenous neurogenin3 forward primer, 5'-TGGCACTCAGCAAACAGCGA -3'; backward primer, 5'-AGATGCTTGAGAGCCTCCAC-3'; CK19 forward primer, 5'-AAGACCATCGAGGACTTGCG-3'; backward primer, 5'-CTATGTCGGCACGCACGTCG-3'; nestin forward primer, 5'-GGAGAGTCGCTTAGAG GTGC-3'; backward primer, 5'-GTCAGGAAAGCCAAGAGAAG-3'; HPRT forward primer, 5'-CTCGAA GTGTTGGATACAGG-3'; backward primer, 5'-TG GCCTATAGGCTCATAGTG-3'. These primers were designed to encompass the intronic sequences to distinguish the appropriate PCR products from products amplified from contaminating genomic DNA. For the detection of endogenous pdx-1 and neurogenin3 mRNA, the backward primers were designed to hybridize with the 3' UTR regions of their sequences that were not included in the AdV expression cassettes. PCR was performed with taq DNA polymerase (Gene Taq, Wako, Tokyo, Japan), and within the log phase of the reaction (23–35 cycles). The mRNA levels were measured by the methods of non-radioactive RT-PCR and CCD imaging as described previously.⁹

Administration of AdVs

AdVs were purified by cesium chloride gradient.⁴⁵. Mice were subjected to laparotomy under general anesthesia with pentobarbital. The mice were then injected with AdV solution (1109 pfu in 250 l of lactated Ringer's solution) into the CBD adjacent to the duodenum, through a 29-G needle in 30 s. ICBD injection without a CBD clamp led to an appropriate intra-ductal pressure to allow the successful introduction of AdV into the mouse pancreas without any side effects, such as pancreatitis. Following confirmation of the absence of backward flow, the mouse abdomens were sutured. Another group of mice was treated with injection of the same amounts of AdVs into the tail vein. The mice were killed at various times after the injection.

Detection of -galactosidase and EGFP

On day 4 after injection of AdV-lacZ, a -galactosidase-expressing AdV, frozen sections (10-m thick) of the liver, pancreas and duodenum were fixed and stained with X-Gal as described previously.⁴⁶ These sections were then counterstained with eosin and examined by light microscopy. On day 7 after injection of EGFP expressing AdVs, mice were subjected to laparotomy under general anesthesia and the liver and pancreas were examined by stereoscopic light microscopy and fluorescence microscopy (Olympus, Tokyo, Japan) with a 460–490-nm band-pass excitation filter and a 510-nm long-pass emission filter.

Histochemical analysis

For paraffin sections, pancreatic tissue was fixed for 2 h at room temperature in freshly prepared 20% formaldehyde and processed for embedding. Sections of paraffin-embedded pancreatic tissue (5-m thick) were deparaffinized, dehydrated, and stained with hematoxylin/eosin. For insulin immunohistochemical staining, deparaffinized and dehydrated sections were incubated with 10% normal pig serum in PBS to block non-specific staining, then incubated with diluted guinea pig anti-human insulin antibody (Oriental Yeast Co., Tokyo, Japan). Negative controls were incubated with 10% normal pig serum without the first antibody. The sections were then incubated with swine polyclonal anti-rabbit immunoglobulin/biotin (Dako Japan Co., Kyoto, Japan), and then with diluted peroxidase-conjugated streptavidin (Dako Japan Co., Kyoto, Japan). After incubation with a DAB solution (20 mg of 3, 3'-diaminobenzidine tetra hydrochloride and 20 l of 30% hydrogen peroxidase in 100 ml PBS), the sections were dehydrated and examined.

For frozen sections, pancreatic tissue was embedded in OTC compound (Tissue-TEC, Miles, Elkhart, IN, USA) and frozen in liquid nitrogen. 10-m-thick frozen sections were cut with a cryostat and placed on slides, and fixed in cold acetone for 10 min. The sections were then rinsed in PBS, incubated for 5 min in 1% Triton X-100, and after a second rinse, incubated in blocking serum which derived from the same animal as second antibody. The sections were incubated for 60 min at room temperature with the first antibody, washed with PBS, incubated for 60 min at room temperature with the fluorescein-conjugated second antibody. The first antibodies used were rabbit anti-GFP antibody (MBL Co., Nagoya, Japan), guinea pig anti-porcine insulin antibody (Dako Japan Co., Kyoto, Japan), rabbit anti-rat pdx-1 serum (gifted by Dr Kajimoto, Osaka University), rabbit anti-human glucagon antibody (Dako Japan Co., Kyoto, Japan), mouse anti-human cytokeratin antibody (Nichirei, Tokyo, Japan), mouse anti-human nestin (Chemicon, CA, USA). As the second antibodies, Alexa Fluor 488-conjugated anti-rabbit IgG, Rhodamine Red-conjugated anti-rabbit IgG, and Alexa Fluor 594-conjugated anti-guinea pig IgG, Alexa Fluor 568-conjugated anti-mouse IgG (Molecular Probes, OR, USA) were used. The sections were washed, dehydrated and examined by confocal laser scanning microscopy (Leica).

For the determination of BrdU incorporation, mice were administered with BrdU-labeling reagent 2 h before being killed on day 7 after AdV-injection. Immunohistochemical staining for BrdU was performed using a Cell Proliferation Kit (Amersham, UK).

Evaluation of histological changes

To evaluate the histological changes, four to six mice were treated with AdV in each group. Four to six randomly selected sections for each pancreas were stained with hematoxylin/eosin and anti-insulin antibody and were examined by light microscopy. On each section, normal islets adjacent to the exocrine cells were counted (Figure 5a). Then, the newly induced insulin-positive cells, which were observed near or in the newly proliferated ductal tissues (Figure 4d–f), were counted and classified into two groups: (1) a cell cluster consisting of four or more newly induced insulin-positive cells; and (2) a single or a cluster of two to three newly induced insulin-positive cells.

Statistical analysis

Statistical analysis was performed by Student's t-test. The P-values for significance were set to 0.05.

References

1. Weir GC, Bonner-Weir S. Islet transplantation as a treatment for diabetes. J Am Optom Assoc 1998; 69: 727–732.
2. Peck AB, Chaudhari M, Cornelius JG, Ramiya VK. Pancreatic stem cells: building blocks for a better surrogate islet to treat type 1 diabetes. Ann Med 2001; 33: 186–192.
3. Ramiya VK et al. Reversal of insulin-dependent diabetes using islets generated in vitro from pancreatic stem cells. Nat Med 2000; 6: 278–282. | Article | PubMed | ISI | ChemPort |
4. Bonner-Weir S et al. In vitro cultivation of human islets from expanded ductal tissue. Proc Natl Acad Sci USA 2000; 97: 7999–8004. | Article | PubMed | ChemPort |
5. Sharma A et al. The homeodomain protein IDX-1 increases after an early burst of proliferation during pancreatic regeneration. Diabetes 1999; 48: 507–513. | PubMed | ISI | ChemPort |
6. Gu D, Arnush M, Sawyer SP, Sarvetnick N. Transgenic mice expressing IFN-gamma in pancreatic beta-cells are resistant to streptozotocin-induced diabetes. Am J Physiol 1995; 269: E1089–E1094.
7. Kritzik MR et al. PDX-1 and Msx-2 expression in the regenerating and developing pancreas. J Endocrinol 1999; 163: 523–530. | PubMed | ChemPort |
8. Tashiro F, Niwa H, Miyazaki J. Constructing adenoviral vectors by using the circular form of the adenoviral genome cloned in a cosmid and the Cre-loxP recombination system. Hum Gene Ther 1999; 10: 1845–1852. | Article | PubMed |
9. Nakayama H, Yokoi H, Fujita J. Quantification of mRNA by non-radioactive RT-PCR and CCD imaging system. Nucleic Acids Res 1992; 20: 4939. | PubMed | ISI | ChemPort |
10. Grubb BR et al. Inefficient gene transfer by adenovirus vector to cystic fibrosis airway epithelia of mice and humans. Nature 1994; 371: 802–806. | Article | PubMed | ISI | ChemPort |
11. Ohno T et al. Gene therapy for vascular smooth muscle cell proliferation after arterial injury. Science 1994; 265: 781–784. | Article | PubMed | ISI | ChemPort |
12. Raper SE, DeMatteo RP. Adenovirus-mediated in vivo gene transfer and expression in normal rat pancreas. Pancreas 1996; 12: 401–410. | PubMed | ISI | ChemPort |
13. Ohlsson H, Karlsson K, Edlund T. IPF1, a homeodomain-containing transactivator of the insulin gene. EMBO J 1993; 12: 4251–4259. | PubMed | ISI | ChemPort |
14. Waeber G, Thompson N, Bonny C. Transcriptional activation of the GLUT2 gene by the IPF-1/STF-1/IDX-1 homeobox factor. Mol Endocrinol 1996; 10: 1327–1334. | Article | PubMed | ISI | ChemPort |
15. Bonny C et al. The loss of GLUT2 expression in the pancreatic beta-cells of diabetic db/db mice is associated with an impaired DNA-binding activity of islet-specific trans-acting factors. Mol Cell Endocrinol 1997; 135: 59–65.
16. Marshak S, Totary H, Cerasi E, Melloul D. Purification of the beta-cell glucose-sensitive factor that transactivates the insulin gene differentially in normal and transformed islet cells. Proc Natl Acad Sci USA 1996; 93: 15057–15062. | Article | PubMed | ChemPort |
17. McGrath KE, Palis J. Expression of homeobox genes, including an insulin promoting factor, in the murine yolk sac at the time of hematopoietic initiation. Mol Reprod Dev 1997; 48: 145–153. | Article | PubMed | ISI | ChemPort |
18. Miller CP, McGehee Jr RE, Habener JF. IDX-1: a new homeodomain transcription factor expressed in rat pancreatic islets and duodenum that transactivates the somatostatin gene. EMBO J 1994; 13: 1145–1156. | PubMed | ISI | ChemPort |
19. Petersen HV et al. Glucose stimulates the activation domain potential of the PDX-1 homeodomain transcription factor. FEBS Lett 1998; 431: 362–366.
20. Seijffers R et al. Increase in PDX-1 levels suppresses insulin gene expression in RIN 1046-38 cells. Endocrinology 1999; 140: 3311–3317. | Article | PubMed | ISI | ChemPort |
21. Ahlgren U, Jonsson J, Edlund H. The morphogenesis of the pancreatic mesenchyme is uncoupled from that of the pancreatic epithelium in IPF1/PDX1-deficient mice. Development 1996; 122: 1409–1416. | PubMed | ISI | ChemPort |
22. Grapin-Botton A, Majithia AR, Melton DA. Key events of pancreas formation are triggered in gut endoderm by ectopic expression of pancreatic regulatory genes. Genes Dev 2001; 15: 444–454. | Article | PubMed | ISI | ChemPort |
23. Offield MF et al. PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development 1996; 122: 983–995. | PubMed | ISI | ChemPort |
24. Stoffers DA, Heller RS, Miller CP, Habener JF. Developmental expression of the homeodomain protein IDX-1 in mice transgenic for an IDX-1 promoter/lacZ transcriptional reporter. Endocrinology 1999; 140: 5374–5381. | Article | PubMed | ISI | ChemPort |
25. Stoffers DA, et al. Pancreatic agenesis attributable to a single nucleotide deletion in the human IPF1 gene coding sequence. Nat Genet 1997; 15: 106–110. | Article | PubMed | ISI | ChemPort |
26. Huang HP et al. Regulation of the pancreatic islet-specific gene BETA2 (neuroD) by neurogenin 3. Mol Cell Biol 2000; 20: 3292–3307. | Article | PubMed | ISI | ChemPort |
27. Jensen J et al. Independent development of pancreatic alpha- and beta-cells from neurogenin3-expressing precursors: a role for the notch pathway in repression of premature differentiation. Diabetes 2000; 49: 163–176. | PubMed | ISI | ChemPort |
28. Lee JC et al. Regulation of the pancreatic pro-endocrine gene neurogenin3. Diabetes 2001; 50: 928–936. | PubMed | ISI | ChemPort |
29. Apelqvist A et al. Notch signalling controls pancreatic cell differentiation. Nature 1999; 400: 877–881. | Article | PubMed | ISI | ChemPort |
30. Gradwohl G, Dierich A, LeMeur M, Guillemot F. neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc Natl Acad Sci USA 2000; 97: 1607–1611. | Article | PubMed | ChemPort |
31. Sommer L, Ma Q, Anderson DJ. Neurogenins, a novel family of atonal-related bHLH transcription factors, are putative mammalian neuronal determination genes that reveal progenitor cell heterogeneity in the developing CNS and PNS. Mol Cellular Neurosci 1996; 8: 221–241.
32. Watada H et al. PDX-1 induces insulin and glucokinase gene expressions in alphaTC1 clone 6 cells in the presence of betacellulin. Diabetes 1996; 45: 1826–1831.
33. Ferber S et al. Pancreatic and duodenal homeobox gene 1 induces expression of insulin genes in liver and ameliorates streptozotocin-induced hyperglycemia. Nat Med 2000; 6: 568–572. | Article | PubMed | ISI | ChemPort |
34. Marshak S et al. Functional conservation of regulatory elements in the pdx-1 gene: PDX-1 and hepatocyte nuclear factor 3beta transcription factors mediate beta-cell-specific expression. Mol Cell Biol 2000; 20: 7583–7590.
35. Dutta S et al. PDX:PBX complexes are required for normal proliferation of pancreatic cells during development. Proc Natl Acad Sci USA 2001; 98: 1065–1070. | Article | PubMed | ChemPort |
36. Bouwens L et al. Cytokeratins as markers of ductal cell differentiation and islet neogenesis in the neonatal rat pancreas. Diabetes 1994; 43: 1279–1283.
37. Bouwens L, Braet F, Heimberg H. Identification of rat pancreatic duct cells by their expression of cytokeratins 7, 19, and 20 in vivo and after isolation and culture. J Histochem Cytochem 1995; 43: 245–253.
38. Wang RN. Duct- to islet-cell differentiation and islet growth in the pancreas of duct-ligated adult rats. Diabetologia 1995; 38: 1405–1411. | Article | PubMed | ISI | ChemPort |
39. Bouwens L. Cytokeratins and cell differentiation in the pancreas. J Pathol 1998; 184: 234–239. | Article | PubMed | ChemPort |
40. Sawamoto K et al. Generation of dopaminergic neurons in the adult brain from mesencephalic precursor cells labeled with a nestin-GFP transgene. J Neurosci 2001; 21: 3895–3903. | PubMed | ISI | ChemPort |
41. Rietze RL et al. Purification of a pluripotent neural stem cell from the adult mouse brain. Nature 2001; 412: 736–739. | Article | PubMed | ISI | ChemPort |
42. Zulewski H, et al. Multipotential nestin-positive stem cells isolated from adult pancreatic islets differentiate ex vivo into pancreatic endocrine, exocrine, and hepatic phenotypes. Diabetes 2001; 50: 521–533. | Article | PubMed | ISI | ChemPort |
43. Schwitzgebel VM et al. Expression of neurogenin3 reveals an islet cell precursor population in the pancreas. Development 2000; 127: 3533–3542. | PubMed | ISI | ChemPort |
44. Miyazaki J et al. Establishment of a pancreatic beta cell line that retains glucose-inducible insulin secretion: special reference to expression of glucose transporter isoforms. Endocrinology 1990; 127: 126–132. | PubMed | ISI | ChemPort |
45. Beards GM. A method for the purification of rotaviruses and adenoviruses from faeces. J Virol Methods 1982; 4: 343–352.
46. Vrancken Peeters MJ, Perkins AL, Kay MA. Method for multiple portal vein infusions in mice: quantitation of adenovirus-mediated hepatic gene transfer. Biotechniques 1996; 20: 278–285. | PubMed | ChemPort |
47. Niwa H, Yamamura K, Miyazaki J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 1991; 108: 193–199. | Article | PubMed | ISI | ChemPort |
48. Kawamoto S et al. A novel reporter mouse strain that expresses enhanced green fluorescent protein upon Cre-mediated recombination. FEBS Lett 2000; 470: 263–268. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We are grateful to Mr T. Mitsu (Oriental Yeast Co., Japan) and Ms M. Yamamoto (Osaka University, Japan) for technical assistance and to Dr Y. Kajimoto (Osaka University, Japan) for rabbit anti-mouse pdx-1 serum. This work was supported by a grant from the Research for the Future Program (JSPS-RFTF97I00201) of the Japan Society for the Promotion of Science (JSPS). This work was also supported by a grant from the Japanese Ministry of Education, Science, Sports and Culture.