Efficient differentiation of AR42J cells towards insulin-producing cells using pancreatic transcription factors in combination with growth factors

Molecular and Cellular Endocrinology 358 (2012) 69–80

Contents lists available at SciVerse ScienceDirect

Molecular and Cellular Endocrinology journal homepage: www.elsevier.com/locate/mce

Efficient differentiation of AR42J cells towards insulin-producing cells using pancreatic transcription factors in combination with growth factors Maria João Lima, Hilary M. Docherty, Yuanxiao Chen 1, Kevin Docherty ⇑ School of Medical Sciences, University of Aberdeen, Institute of Medical Sciences, Foresterhill, Aberdeen AB25 2ZD, UK

a r t i c l e

i n f o

Article history: Received 13 December 2011 Received in revised form 24 February 2012 Accepted 26 February 2012 Available online 10 March 2012 Keywords: Stem cells Diabetes Islets of langerhans Endocrine pancreas Pdx1

a b s t r a c t The AR42J-B13 rat pancreatic acinar cell line was used to identify pancreatic transcription factors and exogenous growth factors (GFs) that might facilitate the reprogramming of exocrine cells into islets. Adenoviruses were used to induce exogenous expression of the pancreatic transcription factors (TFs) Pdx1, MafA, Ngn3 and Pax4. Individually Pdx1, MafA and Pax4 had no effect on the expression of endocrine markers, whilst adeno-Ngn3 on its own increased the expression of Pax4, Ngn3 and NeuroD. In combination the four TFs had a significant effect on the expression of insulin 1 and 2 that was associated with a change in cell morphology from a rounded to a spindle-like shape. Amongst a range of growth factors, Betacellulin and Nicotinamide were shown to enhance the effects of the four TFs. The presence of adeno-Pax4 in the differentiation cocktail was important in limiting the expression of glucagon and in generating glucose sensitive insulin secretion. Further experiments asked whether the adenoviral TFs could be replaced by protein transduction domain (PTD)-containing TFs. The results showed that the PTD-TFs could mimic in part the effects of the adeno-TFs, but the resultant cells did not undergo the important morphological change associated with differentiation to endocrine lineages and levels of endogenous markers were very much lower. In summary, the results describe a cocktail of four TFs and two GFs that can be used to induce formation of glucose sensitive insulin secreting cells from ARJ42 cells, and demonstrate that it would be difficult to replace adenoviral transduction with PTD-TFS. Ó 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction The major therapeutic challenges in type 1 diabetes are, in the short term avoiding unexpected episodes of hypoglycaemia and in the long term avoiding the late complications of the disease that affect eye, kidney, and peripheral nerve function. Recent successes in islet transplantation (Shapiro et al., 2000), albeit limited (Ryan et al., 2005), have raised hope these challenges might be met. However, the dependence on the rare availability of cadaveric tissue has stimulated a drive towards generating an alternative source of islets. An understanding of the transcription factors that control the developing pancreas (Gittes, 2009) has provided protocols that reproducibly generate pancreatic progenitors from embryonic stem cells which can develop into functional b-cells when engrafted in mice (Kroon et al., 2008). At the same time the discovery that cells of one tissue type could be reprogrammed into those of other tissues (Takahashi and Yamanaka, 2006) has led to new therapeutic options. Thus acinar (Zhou et al., 2008) and liver (Yechoor et al., 2009; Yatoh et al., 2007) cells, which are developmentally ⇑ Corresponding author. Tel.: +44 1224 437359; fax: +44 1224 437344. E-mail addresses: [email protected], [email protected] (K. Docherty). Present address: Department of Cell Biology and Medical Genetics, Kunming Medical College, Kunming 650500, China. 1

0303-7207/$ - see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mce.2012.02.024

closely related to pancreatic endocrine cells, have been reprogrammed to b-cells through a specific combination of pancreatic transcription factors that include Pdx1, Ngn3, NeuroD and MafA. Pdx1 is present in early progenitors that give rise to all cells of the pancreas, Ngn3 directs cells towards an endocrine fate, while later expression of NeuroD, MafA, and Pdx1 in a second wave of expression, act as b-cell differentiation factors (Bernardo et al., 2008). Because of the risk of inflammation and resultant pancreatitis in vivo reprogramming of acinar tissue using viral vectors is unrealistic for treatment of type 1 diabetes, however, the in vitro reprogramming of exocrine tissue may be of therapeutic value. AR42J is a rat cell line that was originally derived from a chemically induced pancreatic tumour (Longnecker et al., 1979). AR42J cells can be induced to differentiate towards hepatocyte (Wallace et al., 2010) and endocrine pancreatic lineages (Mashima et al., 1996a). Here we used these cells in an attempt to optimise conditions for the efficient reprogramming of pancreatic exocrine cells towards functional pancreatic endocrine cells. Our strategy was to infect the cells with adenoviruses containing pancreatic transcription factors (TFs) and to measure effects on phenotypic markers that are characteristic of pancreatic endocrine lineages. An additional aim was to determine whether protein transduction domain (PTD) mediated uptake of TFs might provide a better alternative to the use of adenoviruses. The most commonly used

70

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

PTD is the 11 amino acid peptide derived from the TAT protein of the TAT/HIV-1 transactivator protein (Gump and Dowdy, 2007). Recombinant proteins that have been engineered to contain the TAT sequence are able to directly cross the cell membrane, and enter the nucleus of a variety of cell types (Manceur et al., 2007; Krosl et al., 2005).

A280. Peak protein fractions were diluted 40 in 20 mM Tris pH 7.6, left overnight at 4 °C, and re-concentrated using 10000 Da Millipore Amicon Centrifugal filter units (Millipore, Livingston, UK). The final protein concentration was assessed by the Biorad protein assay (Biorad, Hertfordshire, UK). 2.4. SDS–PAGE and Western blotting

2. Materials and methods 2.1. Cell culture AR42J-B13, a sub clone of ARJ42 that was selected on the basis of its enhanced capability to differentiate into insulin-secreting cells (Mashima et al., 1996a), was cultured in DMEM supplemented with 10% Foetal Calf Serum (Invitrogen, Paisley, UK). Cells were plated at a density of 5  106 cells per well of 6 well-culture tissue dishes (Greiner, Stonehouse, UK). For adenoviral infection, the cells were incubated with adenoviruses at a multiplicity of infection of 100, for 4 h at 37 °C before replacing with fresh culture medium. For incubation with recombinant PTD transcription factors the cells were incubated for 48 h with medium containing 1 lM protein. INS-1 cells (Asfari et al., 1992) were cultured in RPMI medium supplemented with 10% FCS, 10 mM Hepes buffer (all from Invitrogen), 1 mM sodium pyruvate (Thermo Scientifique, Loughborough, UK) and 50 lM b-Mercaptoethanol (Sigma Aldrich, Dorset, UK). 2.2. Preparation of adenoviruses Recombinant adenoviruses containing Pdx1, MafA, Ngn3 and Pax4 (kindly provided by Dr. H. Heimberg, ULB, Brussels) were prepared using the Ad-EasyTM system (Agilent Technologies, Edinburgh, UK). HEK293 cells (ATCC, Teddington, UK) were cultured in DMEM supplemented with 10% FCS (Gibco, Invitrogen). Cells were incubated in culture medium containing 1–5 ll of the original viral titer for 90 min and cultured for another 4 days before harvesting. Collection of viral particles was done after lysing the cells with three freeze–thaw cycles. Viral titer was determined using an End Point Dilution Assay (Clontech, Saint-Germain-enLaye, France). 2.3. Vector construction and PTD-protein purification The His6-Pdx1 construct was generated by inserting the coding sequence of human Pdx1 into a pQE-31 vector (Docherty et al., 2005). The His6-TAT-Pax4 and the His6-TAT-MafA constructs were generated by inserting the rat Pax4 and mouse MafA coding sequences into a pEQ-TAT vector. The TAT-Ngn3 construct encoding mouse Ngn3 was a kind gift from Dr. M.S. German, UCSF. These DNA constructs were used to transform E. coli strains BL21 (DE3) (Invitrogen, for MafA and Pax4) or M13pREp4 (Qiagen, West Sussex, UK, for Pdx1 and Ngn3). Then, 200 ml of overnight cultures expressing the protein of interest were used to inoculate 2 L of LB-Amp/Kan followed by growth for 1.5 h at 37 °C (25 °C for Pdx1 and Ngn3 cultures). Next, 0.5 mM isopropyl b-D-1 thiogalactopyranoside (IPTG, Sigma Aldrich) was added 4 h before harvesting. Cells were centrifuged at 4000 rpm at 4 °C for 20 min and stored at 80 °C. Cell pellets were re-suspended in lysis buffer (8 M urea, 0.1 M NaH2PO4 pH 8.0) and incubated for 1 h at RT. Lysates were centrifuged at 10000 rpm for 30 min at 4 °C. The supernatant was applied to a His-select affinity column (Sigma Aldrich) pre-equilibrated with lysis buffer. The column was washed with 50 ml washing buffer (8 M urea, 0.1 M NaH2PO4 pH 6.3) and proteins eluted with 20 ml elution buffer (8 M urea, 0.1 M NaH2PO4 pH 4.5). Collected fractions were monitored for absorbance at

Purified protein aliquots were diluted in NuPage Loading Buffer, run for 1 h on a 10%, 1 mm, bis-tris polyacrylamide gel (both from Invitrogen) and stained with Coomassie Brilliant Blue. Proteins were then transferred onto a nitrocellulose membrane for 1 h at 30 V. The membrane was blocked for 1 h in 5% skimmed milk and incubated with antibodies against Pdx1, MafA, Ngn3 or Pax4 (Table 1) overnight at 4 °C. Membranes were further incubated for 1 h with anti-rabbit or anti-goat IgG Horseradish Peroxidase conjugated antibodies (1:25000, Zymed laboratories, S. Francisco, USA) and developed using an ECL detection kit (Thermo Scientifique). 2.5. RT-qPCR RNA was extracted using TrizolÒ reagent (Invitrogen). After digestion with DNase I (Invitrogen) to remove any contaminating DNA, 1 lg of RNA was used for cDNA synthesis. Quantitative Polymerase Chain Reactions (RT-qPCRs) were then performed using the TaqMan gene expression assays (Applied Biosystems, Paisley, UK). Real-time PCR mixtures were prepared as described by the manufacturer (SensiMiX, Bioline, London, UK) for each gene, denatured at 95 °C for 15 s and then cycled at 95 °C for 30 s, 60 °C for 30 s and 72 °C for 10 s during 50 cycles, followed by final extension at 72 °C for 10 min. QPCRs were run in a Roche Lightcycler 480Ò in triplicate and normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH) in the same run. Data was analysed using the 2 DDCT method (Pfaffl, 2001). Statistical analysis was performed using the Student’s t-test or one-way Anova followed by the Dunnet’s post hoc test, as appropriate. 2.6. Immunocytochemistry Cells were fixed in 24-well plates with 4% paraformaldehyde in phosphate buffered saline (PBS) followed by permeabilisation in ice cold methanol. Wells were washed twice with PBS and blocked in 4% BSA for 2 h (Sigma Aldrich). Fixed cells were incubated with primary antibodies (Table 1) overnight at 4 °C. The fixed cells were then incubated with goat anti rabbit IgG Alexa Fluor 488, goat anti rabbit IgG Alexa Fluor 594, goat anti rabbit IgG Alexa Fluor 350, donkey anti goat IgG Alexa Fluor 488, goat anti mouse IgG Alexa Fluor 594, goat anti mouse IgG Alexa Fluor 488 or goat anti guinea pig IgG Alexa Fluor 594 (all from Invitrogen) secondary antibodies for 1 h at room temperature. Each well was washed in PBS and mounted in Vectashield mounting medium hardset (Vector Labs, Peterborough, UK) with 4,6-diamidino-2-phenylindole (DAPI) and

Table 1 List of antibodies used for immunocytochemistry. Antigen

Host species

Dilution

Supplier

Pdx1 MafA Pax4 Ngn3 Insulin C-peptide Amylase Glucagon

Rabbit Rabbit Goat Rabbit Mouse Guinea Pig Rabbit Mouse

1/1000 1/200 1/200 1/200 1/1000 1/50 1/200 1/1000

Chris Wright Santa Cruz Biosciences Santa Cruz Biosciences Abcam Sigma Aldrich Abcam Abcam Sigma Aldrich

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

covered with a 13 mm cover slip. Fluorescence was observed with an inverted fluorescent microscope (Zeiss Axiovert S100, Hertfordshire, UK). 2.7. Glucose-stimulated insulin release Cells were incubated in 0.5 ml DMEM containing 2.5 mM Glucose in the absence of serum for 3.5 h at 37 °C. The supernatants were collected and the same cells were further incubated at 37 °C with 0.5 ml DMEM containing 25 mM Glucose in the absence of serum for another 3.5 h. Supernatants were collected and stored at 20 °C until analysed. Insulin levels were measured using a Rat/ Mouse ELISA kit (Millipore, Livingston, UK) and the values were normalized to DNA content of each well. DNA extraction was done using Trizol ReagentÒ, according to the manufacturer instructions (Invitrogen). Statistical analysis was performed using a two-way Anova followed by the Bonferroni post hoc test. 3. Results Adenoviruses containing Pdx1, MafA, Ngn3 and Pax4 were used to reprogramme AR42J-B13 cells towards endocrine lineages. Efficient expression of each of these transcription factors (TFs) was detected by immunocytochemistry in 80% of the cells three days after infection (Fig. 1). Cells that were not infected with the adenoviruses contained low levels of endogenous Pdx1 (but no GFP) and no detectable levels of the other three TFs (Fig. 1B). The presence of Pdx1 in AR42J-B13 cells has been previously described (Palgi et al., 2000; Ogihara et al., 2008). These data indicate that the adenoviruses were able to efficiently infect the exocrine cell line AR42JB13, leading to the exogenous expression of b-cell specific transcription factors in these cells. Individually Pdx1, MafA and Pax4 had little effect on the expression of endocrine markers, whilst adeno-Ngn3 on its own increased the expression of Pax4 and Ngn3 (Fig. 2). In combination the four TFs had a significant effect on the expression of insulin 1, insulin 2 and MafA as demonstrated by RT/qPCR (Fig. 2A) and immunocytochemistry (ICC – Fig. 2B). Expression of endogenous Pdx1 and NeuroD, which was already high within the untreated AR42J-B13 cells (Palgi et al., 2000; Ogihara et al., 2008), was unaffected by infection with the adeno-TFs, whether added individually or in combination (Fig. 2A). Collectively, these data indicate that, while individually the pancreatic TFs are not able to induce endogenous expression of b-cell markers, when the four TFs are added in combination expression of insulin 1 and 2 genes is significantly increased. It has previously been shown that Activin A in combination with Hepatocyte Growth Factor (HGF) or Betacellulin can promote the in vitro differentiation AR42J-B13 cells towards insulin-secreting cells (Mashima et al., 1996a,b), while the combination of Leukemia Inhibitory Factor (LIF) and Epidermal Growth Factor (EGF) was able to drive cultured rodent exocrine cells towards insulin producing cells (Baeyens et al., 2005). In the present study the combination of Betacellulin and Nicotinamide had a robust effect on the expression of insulin 1, insulin 2 and Ngn3 (Fig. 3). In contrast to the previous studies addition of LIF and EGF, and combinations of Activin A and HGF or Betacellulin, had no effect on insulin expression levels (Fig. 3). Interestingly, the effect of Betacellulin and Nicotinamide was associated with a concomitant decrease in the levels of endogenous Pdx1, which is in accordance with the generation of a Ngn3+ precursor endocrine cell population (Servitja and Ferrer, 2004). Administration of the four TFs in combination with Betacellulin and Nicotinamide (4TFs + 2GFs) had a marked effect on the expression of insulin 1, insulin 2, Ngn3, MafA and Pax4 above that seen with the 4TFs alone (Fig. 4). There was no effect on

71

the expression of endogenous NeuroD, but as in the previous experiment, there was a decrease in the expression of Pdx1 in the presence of 4TFs and 2GFs. The levels of expression of all these markers following treatment of AR42J-B13 cells with 4TFs and 2GFs was comparable to those seen in the insulinoma rodent b-cell line (Fig. 4B). There was a marked change in the morphology of the cells treated with 4TFs and 2GFs. Non-treated cells were roundshaped and proliferated in cluster-like structures, which is the typical morphology of AR42J-B13. Similar cell clusters were observed in cells treated with the 4TFs, although a few rare and isolated spindle-shaped cells were also present (Fig. 5A). On the other hand, cultures that had been treated with 4TFs and 2GFs were more dispersed, with few round shaped cells and a predominance of more spindle-shaped cells (Fig. 5A). Interestingly, these spindle-shaped cells, but not the rounded cells, stained strongly for both Pdx1 and insulin by ICC (Fig. 5B). Importantly, spindle-shaped cells also stained for C-peptide, confirming that insulin presence was not the result of uptake from the media, but not for amylase (Fig. 5C). The rationale for including Pax4 in our adeno-TF cocktail was that Pax4 would favour production of b-cells at the expense of acells. This proved to be the case, since omitting Pax4 led to an increase in glucagon expression as measured by RT/qPCR and ICC (Fig. 6A and B). Interestingly, the inclusion of Pax4 generated cells that secreted insulin in response to glucose, although in the absence of Pax4 higher amounts of insulin were released to the medium (Fig. 6C). The amount of insulin released by the cells infected with 4TFs and 2GFs was 510 ng/mg DNA, which compares well with other studies (Ogihara et al., 2008). In a further series of experiments we determined whether AR42J-B13 cells could be reprogrammed using PTD-TFs. The advantage of this approach is that the TFs could theoretically be added to the culture media in a temporal and concentration dependent manner that best mimicked pancreatic development. Pdx1 contains a PTD domain in its sequence and it has been shown to be efficiently taken up by a number of cells (Noguchi et al., 2003). PTD-containing MafA, Ngn3 and Pax4 constructs were generated by fusing the sequence of the pancreatic transcription factors to the TAT domain of the transactivator protein of the HIV-1 virus (Fig. 7A). After expression of the recombinant proteins in E. coli, each transcription factor was purified on a histidine affinity column, and the purity of each protein was assessed by SDS–PAGE (Fig. 7B). The identity of the purified proteins was subsequently confirmed by Western blotting and mass spectrometry (data not shown). The purified proteins (1 lM) were efficiently taken up into the cells within 4 h where they could be visualised by ICC in the majority of the cells (Fig. 7C). A combination of the 4 PTD-TFs and 2GFs resulted in increased expression levels of insulin 1, insulin 2, Pax4, NeuroD, and Pdx1 above that seen in untreated cells, whilst the expression of endogenous Ngn3 was decreased (Fig. 8). Although insulin could be detected by ICC (Fig. 8), expression levels within the cells were below the sensitivity of the ELISA. These results suggest that PTD-TFs could be used to investigate mechanisms of reprogramming but their usefulness is limited by the low expression levels of phenotypic markers relative to those observed using adeno-TFs.

4. Discussion The major finding of this study was that AR42J-B13 exocrine cells could be reprogrammed into cells that expressed and secreted insulin in response to glucose using four transcription factors, namely Pdx1, MafA, Ngn3 and Pax4, in combination with two factors, Betacellulin and Nicotinamide. The reprogrammed cells also expressed transcription factors such as NeuroD, MafA, Pax4 and Ngn3 that are typically expressed in pancreatic endocrine cells

72

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

Fig. 1. Pancreatic transcription factors are expressed in AR42J-B13 cells following adenoviral infection. Expression of the transcription factors Pdx1, MafA, Pax4 and Ngn3 in the pancreatic exocrine cell line AR42J-B13. Treated cells (a-l, Ad-TF) were infected with adenoviruses expressing each transcription factor and harvested after three days. (A) Immunocytochemistry was performed using antibodies against Pdx1, MafA, Pax4 and Ngn3 in transfected and untransfected cells (m–x, N/A). The adenovirus expressing Pdx1 contained a GFP sequence (e) and green fluorescence was present due to the exogenous protein expression. No GFP was detected in untransfected AR42J-B13 cells (q). 4 inlets of stained cells are shown in panels i–l and u–x. Cell nuclei were counterstained with DAPI. Data are representative of triplicate experiments. (B) Percentage of cells positive for each transcription factor (Ad-TF+). The number of cells double positive for Pdx1 and GFP is represented by a green bar (GFP+). Data are representative of triplicate experiments and are represented as mean ± SEM.

and adopted a spindle-shaped morphology that was more typical of islet cells. It was previously shown that the in vivo administration to mice of three transcription factors, namely Pdx1, Ngn3 and MafA, resulted in reprogramming of exocrine cells towards b-like cells (Zhou et al., 2008). We extend these studies to show that exocrine cells can be efficiently reprogrammed in vitro and

that the inclusion of Pax4, in keeping with its role in the late maturation of b-cells (Wang et al., 2004; SosaPineda et al., 1997), is important in generating functional glucose responsive b-like cells. AR42J is a rat cell line derived from a chemically induced pancreatic carcinoma (Longnecker et al., 1979). It possesses both exocrine and neuroendocrine properties and can be induced to

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

73

Fig. 2. Pancreatic transcription factors induce insulin expression in AR42J-B13 cells. (A) AR42J-B13 cells were incubated with each of the adenoviruses expressing the pancreatic transcription factors Pdx1 (P), MafA (M), Pax4 (Px4) and Ngn3 (N) for a period of seven days. Untransfected cells (N/A) and cells treated simultaneously with the four transcription factors (4TFs) were also analysed. Cells were harvested and the mRNA was extracted, followed by analysis of expression of different b-cell markers by RT/qPCR. Data represent the average of triplicate experiments and are expressed as mean ± SEM, where p < 0.001 (⁄⁄⁄), p < 0.01 (⁄⁄) or p < 0.05 (⁄). (B) Immunocytochemistry for Insulin in AR42J-B13 cells in untransfected (a–c, N/A) or cells treated with the simultaneous addition of the four transcription factors (d–f, 4TFs) harvested seven days after adenoviral infection. Cell nuclei were conterstained with DAPI. Scale bar = 50 lm.

74

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

Fig. 3. Betacellulin and nicotinamide induce insulin and Ngn3 expression in AR42J-B13 cells. AR42J-B13 cells were incubated with combinations of Activin A (Act. A) and Hepatocyte Growth Factor (HFG), Act. A and Betacellulin, Epidermal Growth Factor (EGF) and Leukemia Inhibitory Factor (LIF) or Betacellulin and Nicotinamide. Untreated (N/A) and treated cells were harvested and the mRNA was extracted, followed by analysis of expression of different b-cell markers by RT/qPCR. Data represent the average of triplicate experiments and are expressed as mean ± SEM, where p < 0.001 (⁄⁄⁄).

undergo conversion to hepatocytes following treatment with dexamethasone (Shen et al., 2000). It can also be converted into insulin producing cells using combinations of Activin A, Betacellulin and HGF (Mashima et al., 1996a,b). This latter conversion is associated with a change in cell morphology from a rounded to spindle-like shape. Several other studies (Zhang et al., 2001; Matsuoka et al., 2007) although not all (Palgi et al., 2000), have been able to confirm these findings while others have shown effects of GLP1 and its analogue Exendin 4 (Zhou et al., 1999) as well as vinca alkaloids on insulin expression in AR42J cells (Umezawa et al., 2003). In the present study we failed to see an effect of Activin A in combination with Betacellulin or HGF on the expression of insulin. However, we were able to detect a strong effect of Betacellulin in combination with Nicotinamide. The reason for the difference between this and previous studies (Mashima et al., 1996a,b) is unclear, but possible explanation might include clonal differences in the AR42J cell line. The fact that the combination of Betacellulin and Nicotinamide increased the expression of Ngn3 with a concomitant fall in the endogenous levels of Pdx1 is in keeping with the pattern of expression of these factors in the developing embryo (Servitja and Ferrer, 2004). Thus Pdx1 is expressed in two phases in a contrary pattern to that of Ngn3 (Bernardo et al., 2008). In terms of exogenous TFs, it was previously shown that transfection with Ngn3 could induce similar morphological changes to those induced by Activin A, and although Ngn3 was not able to induce insulin expression it could induce expression of Pancreatic Polypeptide (Zhang et al., 2001). In a further study administration of adeno-Ngn3 resulted in a moderate increase in NeuroD, Pax4, Nkx2.2, Pax6 and Pdx1 but with no effect on insulin or glucagon

(Ogihara et al., 2008). However a combination of adeno-Ngn3 and adeno-MafA induced insulin expression, which was not further increased by using adeno-Nkx6.1 along with adeno-MafA. A recent study has shown that the simultaneous expression of Pdx1, MafA and Ngn3 was able to generate insulin secreting cells from AR42J-B13 cells. The cells obtained were able to ameliorate diabetes in STZ-treated mice. However, these cells were glucose insensitive suggesting that an incomplete reprogramming towards the b-cell phenotype was achieved (Akinci et al., 2012). Our present study confirms the previous effects of adeno-Ngn3, adenoMafA and adeno-Pdx1 on the expression of the insulin 1 and 2 genes, on AR42J cells and extends previous findings (Ogihara et al., 2008; Akinci et al., 2012) to show that the combination of the previous three transcription factors with adeno-Pax4, Betacellulin and Nicotinamide generates robust expression of insulin, and other islet phenotypic markers. Importantly, we show here that the inclusion of adeno-Pax4 reduces glucagon expression and generates cells that secrete insulin in response to glucose. Further studies will focus on the storage and processing of the expressed (pro)insulin (Alidibbiat et al., 2008) as well as the effects of extracellular matrix (Hamamoto et al., 2011). We anticipated that the ability to use recombinant PTD-containing TFs along with GFs to induce differentiation of AR42J cells would allow one to fine tune the amount and timing of the exogenous TFs and thus provide a better control of the differentiation process. The rationale was based on previous studies showing that administration of a TAT/Ngn3 fusion protein to cultured embryonic day (E) 9.5 and E13.5 pancreatic explants resulted in efficient uptake and nuclear localisation and an increased level of endocrine differentia-

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

75

Fig. 4. Betacellulin and Nicotinamide enhance the reprogramming capacity of pancreatic transcription factors. (A) AR42J-B13 cells were incubated with a combination of adenovirus expressing Pdx1, MafA, Ngn3 and Pax4 (4TFs) or with the 4TFs and the growth factors Betacellulin and Nicotinamide (4TFs + 2GFs). Untransfected (N/A) and treated cells were harvested for RNA extraction seven days after infection and the expression of different b-cell markers was analysed by RT/qPCR. Data represent the average of triplicate experiments and are expressed as mean ± SEM, where p < 0.001 (⁄⁄⁄) or p < 0.05 (⁄). (B) Comparison of expression levels between reprogrammed AR42J-B13 cells treated with the combination of 4Tfs + 2GFs and the rat insulinoma cell line INS1. The expression levels of different b-cell markers was analysed by RT/qPCR. Data represent the average of triplicate cultures and are expressed as mean ± SEM.

tion compared to control samples (Dominguez-Bendala et al., 2005). A further proof of principle study showed that intraperitoneal

injection of recombinant Pdx1, which contains an endogenous PTD, into streptozotocin diabetic mice restored euglycaemia

76

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

Fig. 5. Insulin can be detected in reprogrammed AR42J-B13 cells. (A) Brightfield images showing the morphology of AR42J-B13 cells. Untransfected (N/A) cells have a round shaped morphology and tend to grow in cluster structures (a). After treatment with Pdx1, MafA, Ngn3 and Pax4 (4TFs), a few spindle-shaped cells (yellow arrows) appear in the culture, isolated from the clusters (b). Cultures treated with the 4 TFs, Betacellulin and Nicotinamide (4TFs + 2GFs) show a higher number of spindle-shaped cells (yellow arrows, c). Scale bar = 50 lm. (B) Immunocytochemistry for Insulin and Pdx1 in untransfected (N/A) AR42J-B13 cells (a-c), and in cells treated with the 4TFs + 2 GFs (d-f). Cells were harvested seven days after the adenoviral infection and stained with antibodies against Pdx1 and Insulin. Scale bar = 25 lm. (C) Immunocytochemistry for C-peptide and Amylase in untransfected (N/A) AR42J-B13 cells (a–c), and in cells treated with the 4TFs + 2GFs (d–f). Cells were harvested seven days after the adenoviral infection and stained with antibodies against the acinar marker Amylase and the b-cell marker C-Peptide. Scale bar = 25 lm.

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

77

Fig. 6. Pax4 expression is required for glucose sensitivity in reprogrammed AR42J-B13 cells. (A) AR42J-B13 cells were incubated with a combination of adenovirus expressing Pdx1, MafA, Ngn3 (3TFs), Pdx1, MafA, Ngn3 and Pax4 (4TFs) or with each group of transcription factors and the growth factors Betacellulin and Nicotinamide (3TFs + 2 GFs and 4TFs + 2GFs). Untransfected (N/A) and treated cells were harvested for RNA extraction seven days after infection and the expression of the a-cell marker Glucagon was analysed by RT/qPCR. Data represent the average of triplicate experiments and are expressed as mean ± SEM, where p < 0.001 (⁄⁄⁄) or p < 0.05 (⁄). (B) Immunocytochemistry for Glucagon in untransfected (N/A) AR42J-B13 cells (a–c), and in cells treated with the Transcription Factors combinations, Betacellulin and Nicotinamide (3TFs + 2GFs, d–f, and 4TFs + 2GFs, g–i). Cells were harvested seven days after the adenoviral infection and stained with antibodies against Glucagon. Scale bar = 25 lm. (C) Insulin concentration in culture media from untransfected (N/A) AR42J-B13 or in the media from cells cultured in the presence (4TFs + 2GFs) or absence (3TFs + 2GFs) of Pax4. Cells were cultured with 2.5 or 25 mM of Glucose for 3.5 h, media was collected and Insulin release was measured by ELISA. Data are shown as the average ± SEM from triplicate experiments, where p < 0.05 (⁄). (D) Insulin concentration in culture media from untransfected (N/A) and treated AR42J-B13 normalized to the total amount of DNA in each culture well. Data are shown as the average ± SEM from triplicate experiments, where p < 0.05 (⁄).

78

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

Fig. 7. Purification and uptake of recombinant pancreatic protein transduction domain containing transcription factors. (A) Schematic representation of TAT-Ngn3, TAT-Pax4, TAT-MafA and Pdx1 sequences. The PTD box represents the antennapedia-like protein transduction domain of Pdx1. A histidine tag (His6) was added to the sequence of each protein for binding to the affinity column. (B) The purity of recombinant TAT-Ngn3, TAT-Pax4, TAT-MafA and Pdx1 was assessed on a denaturing polyacrylamide gel stained with Coomassie Brilliant Blue. (C) Immunocytochemistry of AR42J-B13 cells incubated with each of the pancreatic transcription factors for assessing the efficiency of protein uptake. AR42J-B13 cells were incubated with 1 lM of recombinant TAT-Ngn3, TAT-Pax4, TAT-MafA or Pdx1 for a period of four hours. Untransfected (N/A) and treated (PTDTF) cells were harvested and stained with antibodies against Pdx1, MafA, Ngn3 and Pax4. Nuclei were counterstained with DAPI. Scale bar = 50 lm.

through a combination of b-cell regeneration and liver reprogramming (Koya et al., 2008). Our results were encouraging in that the recombinant PTD-TFs could partially mimic the effects of their adenoviral counterparts in eliciting effects on the expression of Insulin, Pax4 and NeuroD, but there was no noticeable effect on the morphology of the cells and the levels of insulin were below the sensitivity range of the ELISA. There are several possible explanations for this. It may simply be that despite being taken up efficiently by almost all the cells and localised to the nucleus the overall amount of active TF may be insufficient to match the expression levels generated by using adenoviruses. It may be possible to

increase the amount of PTD-TF added to the media but this would require production of recombinant proteins beyond the scale possible within an academic laboratory and the overall cost (versus use of adenovirus) would be prohibitive. In summary, we have generated a robust protocol for reprogramming the AR42J-B13 pancreatic exocrine cell line towards functional b-cells. This study can have implications for the reprogramming of human exocrine tissue towards functional b-cells, and for the derivation of a protocol with potential to be used in cell therapy strategies for the treatment of Type 1 Diabetes (Docherty, 2011).

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

79

Fig. 8. Insulin is expressed in AR42J-B13 cells after administration of pancreatic PTD-transcription factors. (A) AR42J-B13 cells were simultaneously incubated with the purified recombinant Pdx1, TAT-MafA, TAT-Ngn3 and TAT-Pax4 (4 PTD-TFs) transcription factors, in the presence of Betacellulin and Nicotinamide (2GFs). Untransfected (N/ A) and treated cells were harvested seven days after the recombinant transcription factor incubation and RNA was extracted, followed by analysis of expression of different bcell markers by RT/qPCR. Data represent the average of triplicate experiments and are expressed as mean ± SEM, where p < 0.001 (⁄⁄⁄), p < 0.01 (⁄⁄) or p < 0.05 (⁄). (B) Immunocytochemistry for Insulin in untransfected (N/A) AR42J-B13 cells (a–c), and in cells treated with the four recombinant PTD-Transcription Factors, Betacellulin and Nicotinamide (4 PTD-TFs + 2GFs, d–f). Cells were harvested seven days after the adenoviral infection and stained with antibodies against Insulin. Nuclei were counterstained with DAPI. Data are representative of triplicate experiments. Scale bar = 50 lm.

80

M.J. Lima et al. / Molecular and Cellular Endocrinology 358 (2012) 69–80

References Alidibbiat, A., Marriott, C.E., Scougall, K.T., Campbell, S.C., Huang, G.C., 2008. Inability to process and store proinsulin in transdifferentiated pancreatic acinar cells lacking the regulated secretory pathway. J. Endocrinol. 196, 33–43. Akinci, E., Banga, A., Greder, V., Dutton, J., Slack, J., 2012. Reprogramming of pancreatic exocrine cells towards a beta cell character using Pdx1,Ngn3 and MafA. Biochem. J.442, 539–550. Asfari, M., Janjic, D., MedaA, P., Li, G., Halban, P., Wollheim, C.B., 1992. Establishment of 2-mercaptoethanol-dependent differentiated insulin-secreting cell lines. Endocrinology 130, 167–178. Baeyens, L., De Breuck, S., Lardon, J., Mfopou, J., Rooman, I., Bouwens, L., 2005. In vitro generation of insulin-producing beta cells from adult exocrine pancreatic cells. Diabetologia 48, 49–57. Bernardo, A., Hay, C., Docherty, K., 2008. Pancreatic transcription factors and their role in the birth, life and survival of the pancreatic beta cell. Mol. Cell. Endocrinol. 294, 1–9. Docherty, K., 2011. Reprogramming Towards Pancreatic Beta Cells. In: Ainscough, J., Yamanaka, S., Takahashi, T. (Eds.), Stem Cell Biology and Regenerative Medicine. Springer, pp. 177–191. Docherty, H., Hay, C., Ferguson, L., Barrow, J., Durward, E., Docherty, K., 2005. Relative contribution of PDX-1, MafA and E47/beta 2 to the regulation of the human insulin promoter. Biochem. J. 389, 813–820. Dominguez-Bendala, J., Klein, D., Ribeiro, M., Ricordi, C., Inverardi, L., Pastori, R., 2005. TAT-mediated neurogenin 3 protein transduction stimulates pancreatic endocrine differentiation in vitro. Diabetes 54, 720–726. Gittes, G.K., 2009. Developmental biology of the pancreas: A comprehensive review. Dev. Biol. 326, 4–35. Gump, J., Dowdy, S., 2007. TAT transduction: the molecular mechanism and therapeutic prospects. Trends Mol. Med. 13, 443–448. Hamamoto, K., Yamada, S., Hara, A., Kodera, T., Seno, M., Kojima, I., 2011. Extracellular matrix modulates insulin production during differentiation of AR42J cells: functional role of Pax6 transcription factor. J. Cell. Biochem. 112, 318–329. Koya, V., Lu, S., Sun, Y., Purich, D., Atkinson, M., Yang, L., 2008. Reversal of streptozotocin-induced diabetes in mice by cellular transduction with recombinant pancreatic transcription factor pancreatic duodenal homeobox1 – a novel protein transduction domain-based therapy. Diabetes 57, 757. Kroon, E., Martinson, L.A., Kadoya, K., Bang, A.G., Kelly, O.G., Eliazer, S., Young, H., Richardson, M., Smart, N.G., Cunningham, J., Agulnick, A.D., D’Amour, K.A., Carpenter, M.K., Baetge, E.E., 2008. Pancreatic endoderm derived from human embryonic stem cells generates glucose-responsive insulin-secreting cells in vivo. Nat. Biotechnol. 26, 443–452. Krosl, G., Giard, M., Krosl, J., Humphries, R., Sauvageau, G., Roy, D., 2005. Human hematopoietic stem cells can be expanded ex vivo using recombinant TATHOXB4 protein. Biol. Blood Marrow Transplant. 11, 19. Longnecker, D.S., Lilja, A.H., French, J., Kuhlmann, E., Noll, W., Lilja, H.S., 1979. Transplantation of azaserine-induced carcinomas of pancreas in rats. Cancer Lett. 7, 197–202. Manceur, A., Wu, A., Audet, J., 2007. Flow cytometric screening of cell-penetrating peptides for their uptake into embryonic and adult stem cells. Anal. Biochem. 364, 51–59. Mashima, H., Ohnishi, H., Wakabayashi, K., Mine, T., Miyagawa, J., Hanafusa, T., 1996a. Betacellulin and activin A coordinately convert amylase-secreting pancreatic AR42J cells into insulin-secreting cells. J. Clin. Invest. 97, 1647–1654.

Mashima, H., Shibata, H., Mine, T., Kojima, I., 1996b. Formation of insulin-producing cells from pancreatic acinar AR42J cells by hepatocyte growth factor. Endocrinology 137, 3969–3976. Matsuoka, T., Kaneto, H., Stein, R., Miyatsuka, T., Kawamori, D., Henderson, E., 2007. MafA regulates expression of genes important to islet beta-cell function. Mol. Endocrinol. 21, 2764–2774. Noguchi, H., Kaneto, H., Weir, G.C., Bonner-Weir, S., 2003. PDX-1 protein containing its own antennapedia-like protein transduction domain can transduce pancreatic duct and islet cells. Diabetes 52, 1732–1737. Ogihara, T., Fujitani, Y., Uchida, T., Kanno, R., Choi, J., Hirose, T., Kawamori, R., Watada, H., 2008. Combined expression of transcription factors induces AR42JB13 cells to differentiate into insulin-producing cells. Endocr. J. 55, 691–698. Palgi, J., Stumpf, E., Otonkoski, T., 2000. Transcription factor expression and hormone production in pancreatic AR42J cells. Mol. Cell. Endocrinol. 165, 41– 49. Pfaffl, M., 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 29, 2002–2007. Ryan, E., Paty, B., Senior, P., Bigam, D., Alfadhli, E., Kneteman, N., 2005. Five-year follow-up after clinical islet transplantation. Diabetes 54, 2060–2069. Servitja, J.M., Ferrer, J., 2004. Transcriptional networks controlling pancreatic development and beta cell function. Diabetologia 47, 597–613. Shapiro, A.M., Lakey, J., Ryan, E., Korbutt, G., Toth, E., Lakey, J.R., Ryan, E.A., Korbutt, G.S., Warnock, G.L., Kneteman, N.M., Rajotte, R.V., 2000. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N. Engl. J. Med. 343, 230–238. Shen, C., Slack, J., Tosh, D., 2000. Molecular basis of transdifferentiation of pancreas to liver. Nat. Cell Biol. 2, 879–887. SosaPineda, B., Chowdhury, K., Torres, M., Oliver, G., Gruss, P., 1997. The Pax4 gene is essential for differentiation of insulin-producing beta cells In the mammalian pancreas. Nature 386, 399–402. Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Umezawa, K., Hiroki, A., Kawakami, M., Naka, H., Takei, I., Ogata, T., 2003. Induction of insulin production in rat pancreatic acinar carcinoma cells by conophylline. Biomed. Pharmacother. 57, 341–350. Wallace, K., Fairhall, E., Charlton, K., Wright, M., 2010. AR42J-B-13 cell: an expandable progenitor to generate an unlimited supply of functional hepatocytes. Toxicology 278, 277–287. Wang, J., Elghazi, L., Parker, S., Kizilocak, H., Asano, M., Sussel, L., 2004. The concerted activities of Pax4 and Nkx2.2 are essential to initiate pancreatic betacell differentiation. Dev. Biol. 266, 178–189. Yatoh, S., Akashi, T., Chan, P., Kaneto, H., Sharma, A., Bonner Weir, S., Weir, G., 2007. NeuroD and reaggregation induce beta-cell specific gene expression in cultured hepatocytes. Diabetes. Metab. Res. 23, 239–249. Yechoor, V., Liu, V., Espiritu, C., Paul, A., Oka, K., Kojima, H., Chan, L., 2009. Neurogenin3 is sufficient for transdetermination of hepatic progenitor cells into neo-islets in vivo but not transdifferentiation of hepatocytes. Dev. Cell 16, 358–373. Zhang, Y., Mashima, H., Kojima, I., 2001. Changes in the expression of transcription factors in pancreatic AR42J cells during differentiation into insulin-producing cells. Diabetes 50, S10–4. Zhou, J., Wang, X., Pineyro, M., Egan, J., 1999. Glucagon-like peptide 1 and exendin-4 convert pancreatic AR42J cells into glucagon- and insulin-producing cells. Diabetes 48, 2358–2366. Zhou, Q., Brown, J., Kanarek, A., Rajagopal, J., Melton, D., 2008. In vivo reprogramming of adult pancreatic exocrine cells to beta-cells. Nature 455, 627–632.