- Email: [email protected]
SOX9 regulates endocrine cell differentiation during human fetal pancreas development
The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Contents lists available at SciVerse ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
SOX9 regulates endocrine cell differentiation during human fetal pancreas development Erin McDonald a , Jinming Li a,b , Mansa Krishnamurthy a , George F. Fellows d , Cynthia G. Goodyer e , Rennian Wang a,b,c,∗ a
Children’s Health Research Institute, University of Western Ontario, London, ON N6C 2V5, Canada Physiology & Pharmacology, University of Western Ontario, London, ON N6C 2V5, Canada c Medicine, University of Western Ontario, London, ON N6C 2V5, Canada d Obstetrics and Gynecology, University of Western Ontario, London, ON N6C 2V5, Canada e Department of Pediatrics, McGill University, Montreal, QC H3Z 2Z3, Canada b
a r t i c l e
i n f o
Article history: Received 29 April 2011 Received in revised form 7 September 2011 Accepted 22 September 2011 Available online 29 September 2011 Keywords: Human fetal pancreas Human islet-epithelial cells Islet transcription factors Akt/GSK3 pathway Beta- and alpha-cell differentiation
a b s t r a c t The transition of pancreatic progenitor cells to mature endocrine cells is regulated by the sequential activation and interaction of several transcription factors. In mice, the transcription factor Sox9 has been shown to support endocrine cell differentiation. However, the functional role of SOX9 during pancreas development in the human has yet to be determined. The present study was to characterize SOX9 expression during human fetal pancreas development and examine its functional role by transfection with SOX9 siRNA or SOX9 expression vectors. Here we report that SOX9 was most frequently expressed in PDX1+ cells (60–83%) and least in mature endocrine cells (<1–14%). The proliferation of SOX9+ cells was significantly higher at 8-10weeks than at 14–21weeks (p < 0.05) or 20–21weeks (p < 0.01). SOX9 frequently co-localized with FOXA2, NGN3 and transcription factors linked to NGN3 (NKX2.2, NKX6.1, PAX6). siRNA knockdown of SOX9 significantly decreased islet-epithelial cell proliferation, NGN3, NKX6.1, PAX6 and INS mRNA levels and the number of NGN3+ and insulin+ cells (p < 0.05) while increasing GCG mRNA and glucagon+ cells (p < 0.05). Examination of SOX9 associated signaling pathways revealed a decrease in phospho-Akt (p < 0.01), phospho-GSK3 (p < 0.01) and cyclin D1 (p < 0.01) with a decrease in nuclear -catenin+ (p < 0.05) cells following SOX9 siRNA knockdown. In contrast, over-expression of SOX9 significantly increased the number of islet cells proliferating, NGN3, NKX6.1, PAX6 and INS mRNA levels, the phospho-Akt/GSK3 cascade and the number of insulin+ cells. Our results demonstrated that SOX9 is important for the expression of NGN3 and molecular markers of endocrine cell differentiation in the human fetal pancreas. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction One major cause of the progression of diabetes is a decline in pancreatic beta-cell mass and function. Therapeutic strategies aimed at repopulating insulin-producing cells show great potential for restoring normal blood glucose levels and, thus, extensive research efforts are focused on developing methods to differentiate progenitor cells into beta-cells (Hao et al., 2006; Jiang et al., 2007; Meier et al., 2005) and to maintain their viability and function (Bonner-Weir and Sharma, 2002; Heit and Kim, 2004). However, such an undertaking requires a thorough analysis of transcription factors that control pancreatic organogenesis and tissue
∗ Corresponding author at: Victoria Research Laboratories, Room A5-140, 800 Commissioners Road East, London, Ontario N6C 2V5, Canada. Tel.: +1 519 685 8500x55098; fax: +1 519 685 8186. E-mail address: [email protected] (R. Wang). 1357-2725/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2011.09.008
maintenance. One important player in this complex process is the SOX, or sex-determining region on Y box, family of transcription factors which is involved in cell fate determination and the development of several tissues during embryogenesis (Akiyama et al., 2002; De Santa Barbara et al., 1998; van de Wetering et al., 1993; Wegner, 1999). These transcription factors share a high mobility group (HMG) DNA binding domain which alone fulfills the functions of DNA binding and bending, protein interactions and nuclear import or export (Wilson and Koopman, 2002). Several members of this family are under investigation for their role in murine pancreas development (Lioubinski et al., 2003; McDonald et al., 2009), including Sox4, Sox6, Sox9 and Sox17 with regards to endoderm specification (Kanai-Azuma et al., 2002), maintaining pancreatic progenitor cell status and supporting endocrine cell differentiation (Lynn et al., 2007; Seymour et al., 2007; Wilson et al., 2005), beta-cell proliferation and insulin secretion (Iguchi et al., 2005, 2007) and endocrine cell differentiation (Mavropoulos et al., 2005; Wilson et al., 2005). Only SOX9 expression has been
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
confirmed in the developing and adult human pancreas (Piper et al., 2002). During murine pancreas development, Sox9 is expressed in Pdx1+ progenitor cells, but is absent from committed endocrine precursors or differentiated cells (Lioubinski et al., 2003; Lynn et al., 2007; Seymour et al., 2007; Lee and Saint-Jeannet, 2003). The inactivation of SOX9 in the mouse pancreas leads to significant depletion of pancreatic progenitor cells, resulting in pancreatic hypoplasia (Wilson et al., 2005). SOX9 has also been shown to contribute to the proliferation, survival and maintenance of pancreatic progenitor cells in mouse model systems by binding to the NGN3 promoter and activating NGN3 gene expression, resulting in stable expression of a network of transcription factors that preserve pancreatic progenitor cell populations (Lynn et al., 2007; Seymour et al., 2007). Studies using lineage-tracing in mice have demonstrated that SOX9+ cells, found in the epithelial cords, give rise to both endocrine and exocrine cells of the pancreas (Seymour et al., 2008). However, a 50% reduction in SOX9 gene expression in pancreatic progenitors revealed significant decreases in endocrine cell mass while exocrine compartments remained unchanged. These findings suggest that, in the rodent, SOX9 is an essential transcription factor that governs the adoption of an endocrine phenotype and is less important for the development of the exocrine pancreas (Seymour et al., 2008). Although the role of Sox9 in the murine pancreas is well studied, knowledge of the function of this transcription factor during human pancreas development is limited. Campomelic dysplasia, a disorder resulting in severe skeletal defects and partial sex reversal, is caused by SOX9 haploinsufficiency (Piper et al., 2002; Seymour et al., 2008). Pancreatic analyses performed at term revealed significant abnormalities, including reductions in islet diameter, decreased endocrine cell clustering resulting in poorly formed islets, and variable expression of beta-cell maturity markers, including insulin (Piper et al., 2002). These findings suggest that normal SOX9 expression is important for islet formation in the human as well as the rodent. Although, signaling molecules which govern Sox9 expression and regulation have been explored in detail in different cell types including bone, heart, gonads, lymphocytes and glial cells (McDonald et al., 2009), signaling cascades which regulate SOX9 expression in the pancreas are yet to be elucidated. It has been reported that SOX9 plays a dual role during intestinal epithelium development and that this is related to the Wnt signaling pathway (Bastide et al., 2007; Mori-Akiyama et al., 2007): SOX9 mediates Wnt-dependent paneth cell differentiation and prevents over-activity of the Wnt-dependent transcriptional program in mouse models (Bastide et al., 2007). In cartilaginous tissue, the PI3K/Akt pathway has been shown to maintain SOX9 expression and perturbation of this cascade led to reduced SOX9 expression and transcriptional activity (Cheng et al., 2009). Recent studies have shown that SOX9 knockdown in vitro and deletion in vivo reduced PI3 K expression and phosphorylation, leading to chondrocyte death – a response which was reversed with the overexpression of SOX9 (Ikegami et al., 2011). Moreover, functional insulin signaling is required for SOX9 expression and, in turn, for proper testicular development and maturation (Nef et al., 2003; Ramocki et al., 2008). Overexpression of insulin receptor substrate-1 (IRS-1) increased SOX9 expression in intestinal epithelial cells and protected against apoptosis of crypt stem or progenitor cells (Ramocki et al., 2008). Glycogen synthase 3 (GSK3) is a downstream molecule of both the PI3K/Akt and Wnt pathways and has been shown to have an inhibitory effect on downstream molecules (Mendez and Garcia-Segura, 2006; Zeng et al., 2005). Thus alterations in PI3K/Akt, Wnt and GSK3 signaling cascades were examined following SOX9 siRNA and overexpression in the present study.
73
The present study has characterized the spatial and temporal expression of SOX9 in the developing human fetal pancreas from 8 to 21 weeks of fetal age and examined its functional role during islet cell differentiation using in vitro loss and gain of function experiments. Our results show that SOX9 is important for endocrine cell differentiation. 2. Tissues, materials and methods 2.1. Pancreatic tissue collection and cell culture Human fetal pancreatic tissues (8–21 weeks fetal age) were collected according to protocols approved by the Health Sciences Research Ethics Board at the University of Western Ontario and the Research Ethics Board of the Royal Victoria Hospital at the McGill University Health Centre, in accordance with the Canadian Council on Health Sciences Research Involving Human Subjects guidelines. Tissues were immediately processed for immunohistology, RNA or protein extraction, with a minimum of five pancreases per age group (Al-Masri et al., 2010; Lyttle et al., 2008; Saleem et al., 2009). Fetal pancreatic tissues from 14 to 16 weeks fetal age, a stage at which there is an enriched SOX9+ /PDX1+ cell population (Lyttle et al., 2008), were dissected and immediately digested using dissociation buffer containing collagenase V (1 mg/ml, Sigma, St. Louis, MO, USA), as described previously (Li et al., 2006). The isolated islet-epithelial cell clusters were cultured in CMRL1066 media containing 10% fetal bovine serum (FBS) (Invitrogen, Burlington, ON, Canada). 2.1.1. Transient transfection of SOX9 siRNA Freshly isolated islet-epithelial cell clusters from 14 to 16 week fetal pancreases were recovered overnight and transiently transfected for 30 h in antibiotic-free medium with 60 nM SOX9(h) siRNA (sc-36533) or control siRNA (sc-37007, proprietary sequence), commercially obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA), using an siRNA transfection kit (Li et al., 2007). A pool of three sequences for human SOX9 siRNA (Santa Cruz) were used as listed in Supplementary Table 1. Islet epithelial-cell clusters were harvested at 48 h following transfection. Transfection efficiency was monitored using fluorescein-conjugated control siRNA (Santa Cruz) and qRT-PCR analysis for SOX9 mRNA, with approximately 50–60% of the islet-epithelial cluster cells being transfected, as described previously (Al-Masri et al., 2010; Li et al., 2007; Saleem et al., 2009). 2.1.2. Transient over-expression of SOX9 Full-length human SOX9 cDNA (a kind gift from Dr. Philippe Jay, Institut de Génomique Fonctionnelle, UMR5203, France) was inserted into the KpnI and XbaI sites of pcDNA3.1 (Clontech, Mountain View, CA, USA), as described previously (Bastide et al., 2007). Isolated human fetal islet-epithelial cell clusters were transfected with 1 g of either pcDNA3.1-SOX9 or the empty (control) vector using Lipofectamine Plus reagent (InVitrogen) in a serum- and antibiotic-free medium for 48 h (Saleem et al., 2009). At the end of the culture period, islets were harvested and processed for protein or RNA extraction or fixed for immunocytochemistry studies. Three to six different preparations from a pool of 2–3 fetal pancreatic isolations per experimental group were used for analyses, with an average of 500–1000 cells counted per experimental group (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). 2.2. Immunofluorescent and morphometric analyses Pancreatic tissues and islets were fixed in 4% paraformaldehyde. Cell pellets were embedded in 2% agarose, followed by
74
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 1. Expression pattern of SOX9 during human fetal pancreatic development. (A) Real-time RT-PCR analysis of SOX9 mRNA expression. Data are normalized to 18S rRNA subunit and expressed as means ± SEM (n = 5–9 pancreases/age group). (B) Western blot analyses of SOX9 protein in human fetal pancreatic tissue. Data are normalized to calnexin and expressed as means ± SEM (n = 3 pancreases/age group). *p < 0.05 vs. 10–11 weeks. (C–E) Double immunofluorescence of SOX9 and morphometric analyses of SOX9+ cells in the developing human pancreas. Representative images for SOX9 (green) with CK19, insulin, glucagon, PDX1 or Ki67 (red) in a 14 weeks human fetal pancreas. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrows indicate double positive cells. Bar graph data are expressed as means ± SEM percentage of SOX9+ cells in the duct and endocrine region (C) and the number of PDX1+ (D) and Ki67+ (E) cells in the SOX9-expressing cell population (n = 5–7 pancreases/age group). *p < 0.05, **p < 0.01, ***p < 0.001 vs. 8–11 weeks.
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
75
Fig. 2. Co-localization of SOX9 with NGN3 during human fetal pancreatic development. (A) Representative images of co-staining for SOX9 (green) and NGN3 (red) in a 12 week human fetal pancreas. Nuclei were stained with DAPI in blue. Scale bar: 25 m. Magnified images for each corresponding figure are shown in the bottom panel. Scale bar: 10 m. Arrows indicate double positive cells and arrow-heads show single stained cells. (B) Quantitative analysis of NGN3+ /SOX9+ cells relative to the total number of SOX9+ cells counted. Data are expressed as means ± SEM (n = 4–7 pancreases/age group). *p < 0.05 vs. 8–9 weeks.
paraffin embedding (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). Sections (5 m) were cut throughout the entire length of the pancreas and were stained with appropriate dilutions of primary antibodies (Supplementary Table 2). To identify the expression of SOX9 in various pancreatic compartments and the co-localization with transcription factors or other SOXs in the developing human fetal pancreas, double immunofluorescence staining was performed. The cells expressing these factors in the ductal and endocrine cell region and in cultured isletepithelial cells were determined by cell counting in which the counter was blind to the age and experimental groups of the samples to minimize bias (Al-Masri et al., 2010). Data are expressed as the mean percentage of SOX9+ cells in CK19+ , insulin+ or glucagon+ cell populations. The number of proliferating (Ki67), PDX1+ and NGN3+ cells co-expressing SOX9 was also examined and expressed as the mean percentage of proliferating, PDX1+ or NGN3+ cells within the SOX9+ cell population. The number of nuclear -catenin expressing cells were also examined (Al-Masri et al., 2010). 2.3. Protein extraction and western blot analysis. Total protein from human fetal pancreatic tissues and isolated fetal islets were extracted (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). Equal amounts (30 g) of extracted proteins from each experimental group were separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane (Bio-Rad, Mississauga, ON, Canada).
Membranes were incubated with appropriate dilutions of primary antibodies (Supplementary Table 2), followed by the application of appropriate horse radish peroxidase-conjugated secondary antibodies (Santa Cruz). Proteins were detected using ECLTM -Plus Western blot detection reagents (Perkin Elmer, Wellesley, MA, USA). Densitometric quantification of bands at subsaturation levels was performed using Syngenetool gel analysis software (Syngene, Cambridge, UK) and normalized to appropriate loading controls, which included the total signaling protein content or a housekeeping protein (calnexin). Data are provided as the relative expression level of phosphorylated proteins to total protein levels or protein levels to the loading control (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). 2.4. RNA extraction and real-time RT-PCR RNA was extracted from human fetal pancreatic tissues using TRIZOL reagent (Invitrogen). Islet RNA was extracted using the RNAqueous-4PCR kit (Invitrogen) (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). For each RT reaction, 2 g of DNA-free RNA were used with random hexamers/oligo(dT) primers and Superscript reverse transcriptase. Real-time PCR analyses were performed using the iQ SYBR Green Supermix kit in Chromo4 Real time PCR (Bio-Rad). Primers used are listed in Supplementary Table 3. Relative gene expression was calculated and normalized to the internal standard gene, 18S rRNA, with at least five repeats per age or experimental group (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009).
76
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 3. Co-localization of SOX9 with NGN3-linked transcription factors during human fetal pancreatic development. Representative images of SOX9 (green) and FOXA2, NKX6.1, NKX2.2 or PAX6 (red) co-localization at 12 weeks of human fetal pancreatic development. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrow indicates double-stained cells, arrow-heads show single stained cells and asterisks indicate non-specific stained dots.
2.5. Statistical analysis Data were expressed as means ± SEM. Statistical significance was determined using either a paired student’s t-test or one-way ANOVA followed by the post hoc least significant differences (LSD) and Tukey’s multiple comparison tests. Differences were considered to be statistically significant when p < 0.05. 3. Results 3.1. Temporal and spatial expression of SOX9 during human fetal pancreas development Using qRT-PCR, western blotting and immunofluorescent staining, we characterized the temporal expression pattern of SOX9 in the developing human fetal pancreas from 8 to 21 weeks of fetal age. Pancreatic SOX9 mRNA levels remained relatively
stable during this period (Fig. 1A); however, SOX9 protein levels decreased gradually with a significant reduction at 19–20 weeks when compared to the expression level at 10–11 weeks (p < 0.05, Fig. 1B). Double immunofluorescence staining of SOX9 with pancreatic cell markers revealed that the majority of pancreatic CK19+ cells contained SOX9 at 8–11 weeks and this co-expression declined significantly at 14–16 weeks (p < 0.01) and 20–21 weeks (p < 0.001) (Fig. 1C), by which time SOX9 expression was restricted to a subpopulation of CK19+ cells. The percentage of insulin+ and glucagon+ cells co-expressing SOX9 is considerably lower than that of CK19+ cells at all stages of development and this decreased significantly, reaching almost undetectable levels at midgestation (Fig. 1C). Co-localization studies of PDX1 with SOX9 showed PDX1 to be highly expressed in SOX9+ cell population (∼83% colocalization) at 8–11 weeks of development but that this had decreased significantly by 14–16 weeks (p < 0.05) and even more by 20–21 weeks (p < 0.01) (Fig. 1D). Furthermore, the proliferation
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
77
Fig. 4. Co-localization of SOX9 with other SOX transcription factors during human fetal pancreatic development. Representative images of SOX9 (green) and SOX4, SOX6, SOX11 or SOX17 (red) co-localization at 14 weeks of human fetal pancreatic development. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrow indicates doublestained cells, arrow-heads show single stained cells and asterisks indicate non-specific stained dots.
of SOX9+ cells was examined using the proliferation marker Ki67 (Fig. 1E). SOX9+ cells showed the highest proliferative capacity at 8–11 weeks (10 ± 1%); this declined significantly at 14–16 weeks (4 ± 1%, p < 0.05) and further at 20–21 weeks (2 ± 0.4%, p < 0.01) (Fig. 1E). 3.2. Co-localization of SOX9 with NGN3 and NGN3-linked as well as other SOX transcription factors in the human fetal pancreas Using immunofluorescence staining to examine the colocalization of SOX9 with NGN3 (Fig. 2), FOXA2, NKX2.2, NKX6.1 and PAX6 (Fig. 3), we found that SOX9 was frequently co-expressed with transcription factors involved in endocrine cell development from 8 to 21 weeks of fetal age. Quantitative analysis showed approximately 5% of the SOX9+ cells stained for NGN3 in 8–9 week old human fetal pancreas with a significant reduction at 14–21 weeks (p < 0.05, Fig. 2B), suggesting that SOX9 interactions with
endocrine cell specific transcription factors may be important in regulating islet cell differentiation during early stages of human fetal pancreatic development (Wegner, 1999). A large number of SOX transcription factors are expressed in the murine pancreas, but whether these are also expressed in the human fetal pancreas and co-express SOX9 has not been explored. Double immunofluorescent staining for SOX9 with SOX4, SOX6, SOX11 or SOX17 at 14 weeks fetal age showed frequent coexpression of SOX9 with SOX4 and SOX17 (Fig. 4). Most of the SOX11+ cells in the ductal regions co-expressed SOX9 while only a small number of SOX9+ cells co-expressed SOX6 (Fig. 4). 3.3. Knockdown of SOX9 mRNA levels results in the reduction of beta-cell differentiation in human fetal islets In order to explore the functional role of SOX9 in human fetal endocrine cell differentiation, we performed an siRNA knockdown
78
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 5. Effect of SOX9 siRNA on beta-cell differentiation in isolated human fetal islet-epithelial cell clusters. (A) Double immunofluorescence images of SOX9 (green) and CK19 (red), SOX9+ cell counts, western blot and qRT-PCR analyses of SOX9 expression in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Data were normalized to control siRNA and are expressed as means ± SEM (n = 3–5 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group). Representative blots are displayed. Scale bar: 25 m. (B) qRT-PCR analysis of transcription factors associated with pancreatic endocrine cell differentiation in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Data are expressed as means ± SEM (n = 4–6 experiments/treatment group; *p < 0.05 vs. control siRNA group).
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
79
Fig. 5. (Continued ) (C) The percentage of NGN3+ , insulin+ , glucagon+ and Ki67+ cells in human fetal islet-epithelial cell clusters treated with either control or SOX9 siRNA for 48 h. Data are normalized to control siRNA and expressed as means ± SEM (n = 5 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group).
of SOX9 in 14–16 week human fetal islet-epithelial cells. The efficiency of the knockdown was determined by qRT-PCR, western blot and immunostaining and showed at least a 50% reduction in SOX9 expression at both mRNA and protein levels (p < 0.05–0.01, Fig. 5A). Following knockdown of SOX9 expression, several transcription factors and hormones important to the development of the endocrine pancreas were examined by qRT-PCR (Fig. 5B). Cells treated with SOX9 siRNA showed no significant changes in PDX1, FOXA2, NEUROD1, PAX4 or NKX2.2 mRNA expression when compared to control siRNA groups (Fig. 5B). However, a significant decrease in NGN3 and INS mRNA levels (p < 0.05, Fig. 5B) and in the number of NGN3+ (p < 0.05) and insulin+ cells (p < 0.01) (Fig. 5C) in comparison with control siRNA groups was observed, along with decreased expression of NKX6.1 and PAX6 mRNA (p < 0.05, Fig. 5B). Interestingly, knockdown of SOX9 expression induced a significant increase in TCF1, HES1 and GCG mRNA (p < 0.05, Fig. 5B) along with a significant increase in the number of glucagon+ cells (p < 0.01, Fig. 5C). The proliferation capacity of the islet-epithelial cells following SOX9 siRNA transfection was ∼50% reduced (p < 0.05, Fig. 5C). 3.4. Increasing SOX9 expression affects beta-cell differentiation in human fetal islets To understand whether increasing SOX9 expression could stimulate beta-cell differentiation, SOX9 was overexpressed in isolated human fetal islet-epithelial cells using transient transfection of a full-length human SOX9 (SOX9) construct or empty vector (Ctrl). qRT-PCR and western blot analyses showed an increase in SOX9 mRNA (2.8-fold, p < 0.05) (Fig. 6A) and protein (2.2-fold, p < 0.01) (Fig. 6B) levels, and the number of SOX9+ cells (1.8-fold, p < 0.05) (Fig. 6C) in the over-expression group as compared to controls after 48 h. qRT-PCR analyses revealed a significant increase in NGN3, PAX6, NKX6.1 and INS mRNA levels in the islet-epithelial cells following SOX9 overexpression (p < 0.05; Fig. 6D), while genes for TCF1, HES1 and GCG remained unchanged (Fig. 6D).
3.5. SOX9 affects Akt/GSK3ˇ signaling during human fetal pancreas development To explore if signaling pathways may be altered following either knockdown or over-expression of SOX9, members of the Akt and Wnt signaling pathways, including phospho-Akt, phosph-GSK3, nuclear -catenin (active) and cyclin D1 were examined. Partial reduction of SOX9 expression by siRNA transfection was associated with a decrease in phospho-Akt (p < 0.05), phospho-GSK3B (p < 0.05), cyclin D1 (p < 0.01) (Fig. 7A) and the number of nuclear localized -catenin (p < 0.01; Fig. 7B) when compared to control siRNA groups. In contrast, in the SOX9 over-expression experimental group there was a significant increase in phosphorylated Akt and GSK3 (p < 0.05) along with cyclin D1 protein levels (p < 0.01) (Fig. 8A), in parallel with a significant increase in the percentage of cells expressing the proliferation marker Ki67 (p < 0.01, Fig. 8B), but no change in the number of n-catenin+ cells (data not shown). To further examine whether changes in phospho-Akt expression were a result of insulin signaling, insulin treated islet-epithelial cells were analyzed for SOX9 expression. The expression of SOX9 mRNA and the number of SOX9+ cells in insulin-treated human fetal isletepithelial cells significantly increased when compared to untreated groups (Supplementary Fig. 1) (Al-Masri et al., 2010). These results were further confirmed by blocking the PI3K/Akt cascade using the PI3K inhibitor, wortmannin, which abrogated the increase in the number of SOX9+ cells in human fetal islet-epithelial cell clusters following insulin treatment (Supplementary Fig. 1) (Al-Masri et al., 2010; Li et al., 2006). In addition, we examined if the Notch signaling pathway is involved in SOX9 regulation of the human fetal pancreas. qRTPCR analyses of NOTCH1, JAG1 and DLL1 were performed but no statistical differences were observed in their mRNA expression following knockdown or overexpression of SOX9 versus control groups (Supplementary Fig. 2A and C). There was also no change in notch intracellular domain (NICD) expression after SOX9 siRNA treatment, as determined by immunofluorescence staining (Supplementary Fig. 2B).
80
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 6. Over-expression of SOX9 increases the expression of NGN3 and its linked transcription factors in isolated human fetal islet-epithelial cell clusters. (A) qRT-PCR and (B) western blot analysis of SOX9 expression, as well as (C) the percentage of SOX9+ cells in islet-epithelial cell clusters transfected with either full-length human SOX9 cDNA or control vector for 48 h. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group); *p < 0.05; **p < 0.01 vs. control group. (D) qRT-PCR analysis of NGN3 and its linked transcription factor genes. Data are normalized to 18S rRNA subunit and expressed as means ± SEM (n = 3–4 experiments/treatment group); *p < 0.05 vs. control group.
4. Discussion SOX9 is expressed in a wide variety of human tissues but, to date, studies in the human pancreas have been limited (Piper et al., 2002). The present study demonstrates that SOX9 is detectable in the human pancreas from as early as the 8th week of fetal life, primarily in PDX1+ ductal cells; it is frequently found co-localizing with NGN3 and other key islet beta-cell progenitor markers. Interestingly, knockdown studies of SOX9 revealed significant decreases in cells expressing beta-cell markers paralleled by increases in alpha cell phenotypes. SOX9 over-expression experiments confirmed that expression of this transcription factor is critical for beta cell differentiation in the human fetal pancreas. The highest expression of SOX9 mRNA and protein was observed in the 8–11 week fetal pancreas; while the mRNA levels remained relatively stable, the protein levels steadily declined by midgestation. The mechanism by which SOX9 protein levels decrease during development has been reported to be pleiotrophic and includes processes of sumoylation, ubiquitination and proteasome degradation (Hattori et al., 2006; Akiyama et al., 2005a,b). In addition, the proliferative capacity of SOX9+ cells was highest during 8–11 weeks, with a rapid decline by 20–21 weeks. These results concur with a report by Piper et al. (2002) and studies conducted in mice (Lioubinski et al., 2003; Lynn et al., 2007;
Seymour et al., 2007) revealing that SOX9 expression is highest during embryogenesis and declines during maturation of the pancreas. The frequent co-localization of SOX9 with PDX1+ and CK19+ cells observed during early human fetal pancreas development is also consistent with previous studies in the human (Piper et al., 2002) and murine pancreas (Lioubinski et al., 2003; Lynn et al., 2007; Seymour et al., 2007) showing that SOX9 is expressed in endocrine progenitor and undifferentiated CK19+ cell populations. In addition, it has been shown in mice that all cells of the pancreas are derived from SOX9 expressing precursors (Lynn et al., 2007; Jørgensen et al., 2007). Sox9 has also been shown to be critical for expression of a transcription factor gene network in pancreatic progenitor cells in mice (Lynn et al., 2007; Seymour et al., 2007) that initiates differentiation of progenitors into endocrine cell types (Lynn et al., 2007). Studies by Lynn et al. revealed that direct Sox9 binding to the Ngn3 promoter activates Ngn3, resulting in expression of several downstream transcription factors, including Foxa2, Onecut1 and Tcf2 (Lynn et al., 2007). In the human fetal pancreas, co-expression of SOX9 with NGN3 and NGN3-linked transcription factors such as FOXA2, PAX6 and NKX2.2 was observed frequently. In contrast, SOX9 had limited co-expression with insulin and glucagon in the islet cells, suggesting that SOX9 is more involved
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
81
Fig. 7. Knockdown of SOX9 expression decreases the phosphorylation of Akt/GSK3 signals in isolated human fetal islet-epithelial cell clusters. (A) Western blot analysis of phosphorylated [P] and total [T] Akt; phosphorylated [P] and total [T] GSK3, and cyclin D1; (B) immunofluorescence staining for PY489--catenin (n-catenin) and the percentage of n-catenin+ cells in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Representative blots and images are shown. Scale bar: 25 m. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
in progenitor cell populations and is not required for maintenance of mature endocrine cells in the human fetal pancreas (Piper et al., 2002). To understand the functional role for SOX9 in human fetal islets, the effects of SOX9 knockdown and over-expression were examined. Knockdown or over-expression of SOX9 did not change PDX-1 expression, indicating that SOX9 and PDX1 do not regulate each other during human fetal pancreas development. These findings are similar to those reported in mouse model systems: no differences in Sox9 expression were observed in PDX1 null embryos, despite their co-expression in pluripotent pancreatic progenitors (Seymour et al., 2007). Following SOX9 siRNA knockdown of the human fetal islet cells, there were significant decreases in NGN3, NKX6.1 and PAX6 mRNA as well as the number of cells expressing NGN3; these data parallel previous findings from mouse studies (Lynn et al., 2007; Seymour et al., 2007). In contrast, no significant changes in FOXA2, NEUROD1, PAX4 or NKX2.2 mRNA levels were observed, suggesting that the network of transcription factors that SOX9 regulates differs significantly from that reported in the mouse (Lynn et al., 2007; Seymour et al., 2007).
Interestingly, siRNA knockdown in human fetal islet cells also resulted in an increase in glucagon mRNA and protein expression. These data are inconsistent with a report in the literature that partial Sox9 gene deletion in the murine pancreas leads to defects in both alpha- and beta-cell mass (Seymour et al., 2008). However, we previously observed a dramatic increase in alpha-cell proliferation and cell mass from 14 to 21 weeks of fetal life, suggesting that the differentiation of alpha-cells occurs earlier than that of beta-cells in the human pancreas (Lyttle et al., 2008). Thus, it is possible that the effects of SOX9 on human alpha- and beta-cell differentiation in the present study may be due to the timing of SOX9 knockdown (14–16 weeks). Moreover, the time point at which SOX9 was manipulated has influenced changes in NGN3 expression and that may impact the type of endocrine cell that is induced to differentiate. In mice, Sox9+ cells express Ngn3 early in development, thus increasing the differentiation of alpha cells. However, inducing Ngn3 expression in progenitor cells later in murine development also produced beta- and PP-cells (Jørgensen et al., 2007). Furthermore, increases in NGN3, NKX6.1, PAX6 and insulin mRNA were observed in SOX9 over-expressed islet cells. This increase in SOX9 did not affect glucagon gene expression, suggesting that SOX9 may have less or no
82
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 8. Over-expression of SOX9 increases the phosphorylation of Akt/GSK3 and cell proliferation. (A) Western blot analysis of phosphorylated [P] and total [T] Akt; phosphorylated [P] and total [T] GSK3, and cyclin D1 expression in islet-epithelial cell clusters transfected with either full-length human SOX9 cDNA or control vector for 48 hr. (B) The percentage of Ki67+ cells. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control group).
influence over alpha-cell differentiation after 14–16 weeks of pancreatic development. Signaling pathways associated with SOX9 expression in the human pancreas have not been previously explored. Upon knockdown of SOX9 mRNA, a significant decrease in GSK3 phosphorylation was observed, along with decreases in n-catenin and cyclin D expression and a reduction in cell proliferation labeled by Ki67. Conversely, increasing SOX9 products in isolated human fetal islet-epithelial cells resulted in the up-regulation of phosphorylated GSK3/cyclin D1 and Ki67+ cells, suggesting that SOX9 promotes pancreatic progenitor cell proliferation via inhibition of GSK3 activity. However, it remains to be determined whether the effects observed in GSK3 phosphorylation are through the PI3 K/Akt or Wnt signaling cascades. Our results also showed significant changes in the phosphorylation of Akt in both knockdown and overexpression studies of SOX9, indicating that the PI3K/Akt signaling pathway is involved in the expression and regulation of SOX9 expression. Furthermore, insulin treatment of human fetal islet-epithelial cells resulted in an increase in SOX9 mRNA, along with an increase in the number of SOX9+ cells when compared to untreated groups. This increase in SOX9 expression was blocked by wortmannin, a PI3K inhibitor, demonstrating the involvement of PI3K/Akt signaling pathway in SOX9 expression. The increase in SOX9 mRNA and protein expression following insulin treatment correlate with the present observations that over-expression or knockdown of SOX9 gene affects the insulin gene and protein expression, suggesting that changes in Akt phosphorylation by SOX9 may be associated with beta-cell differentiation and insulin signaling. Several studies have demonstrated that Sox9 is able to physically interact with -catenin, modulating c-Myc and cyclin D1 genes (Akiyama et al., 2005a,b; Bastide et al., 2007), while insulin signaling via the PI3K/Akt pathway has the ability to repress GSK3 function allowing -catenin accumulation (Schepers et al., 2002). Recent studies have also identified several binding sites for SOX9 on the human PI3K gene specifically around the transcription start site (Ikegami
et al., 2011). Taken together, these findings further support that SOX9 signals via the Akt/GSK3  pathway to maintain progenitor cell populations in the human fetal pancreas. In summary, we have investigated the expression and potential function of SOX9 in the early to mid-gestation human fetal pancreas. Our data suggest that SOX9 is important for endocrine cell differentiation in the developing human pancreas. Acknowledgements We greatly appreciate the gifts of a full-length human SOX9 cDNA from Dr. Philippe Jay (Institut de Génomique Fonctionnelle, Centre National de la Recherche Scientifi que UMR5203, Université de Montpellier I and Université de Montpellier II, Montpellier, France), and anti-PDX1 antibody from Dr. Christopher Wright (University of Vanderbilt, Nashville, USA). This work was supported by grants from the Canadian Diabetes Association (CDA). Dr. Wang is supported by a New Investigator Award from CIHR. We thank the Department of Pathology at London Health Science Centre for providing human fetal pancreatic tissue sections. The authors declare that there is no duality of interest associated with this manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biocel.2011.09.008. References Al-Masri M, Krishnamurthy M, Li J, Fellows GF, Dong HH, Goodyer CG, et al. Effect of forkhead box O1 (FOXO1) on beta cell development in the human fetal pancreas. Diabetologia 2010;53:699–711. Akiyama H, Chaboissier MC, Martin JF, Schedl A, de Crombrugghe B. The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 2002;16:2813–28.
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83 Akiyama H, Kim JE, Nakashima K, Balmes G, Iwai N, Deng JM, et al. Osteochondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci USA 2005a;102:14665–70. Akiyama H, Kamitani T, Yang X, Kandyil R, Bridgewater LC, Fellous M, et al. The transcription factor Sox9 is degraded by the ubiquitin-proteosome system and is stabilized by a mutation in a ubiquitin target site. Matrix Biol 2005b;23:499–505. Bastide P, Darido C, Pannequin J, Kist R, Robine S, Marty-Double C, et al. Sox9 regulates cell proliferation and is required for Paneth cell differentiation in the intestinal epithelium. J Cell Biol 2007;178(4):635–48. Bonner-Weir S, Sharma A. Pancreatic stem cells. J Pathol 2002;197:519–26. Cheng CC, Uchiyama Y, Hiyama A, Gajghate S, Shapiro IM, Risbud MV. PI3K regulates aggrecan gene expression by modulating Sox9 expression and activity in nucleus pulposus cells of the intervertebral disc. J Cell Physiol 2009;221:668–76. De Santa Barbara P, Bonneaud N, Boizet B, Desclozeaux M, Moniot B, Sudbeck P, et al. Direct interaction of SRY-related protein SOX9 and steroidogenic factor 1 regulates transcription of the human anti-Mullerian hormone gene. Mol Cell Biol 1998;18:6653–65. Hattori T, Eberspaecher H, Lu J, Zhang R, Nishida T, Kahyo T, et al. Interactions between PIAS Proteins and SOX9 Result in an Increase in the Cellular Concentrations of SOX9. J Biol Chem 2006;20:14417–28. Hao E, Tyrberg B, Itkin-Ansari P, Lakey JR, Geron I, Monosov EZ, et al. Beta-cell differentiation from nonendocrine epithelial cells of the adult human pancreas. Nat Med 2006;12:310–4. Heit JJ, Kim SK. Embryonic stem cells and islet replacement in diabetes mellitus. Pediatr Diabetes 2004;5:5–15. Iguchi H, Ikeda Y, Okamura M, Tanaka T, Urashima Y, Ohguchi H, et al. SOX6 attenuates glucose-stimulated insulin secretion by repressing PDX1 transcriptional activity and is down-regulated in hyperinsulinemic obese mice. J Biol Chem 2005;280:37669–80. Iguchi H, Urashima Y, Inagaki Y, Ikeda Y, Okamura M, Tanaka T, et al. SOX6 suppresses cyclin D1 promoter activity by interacting with beta-catenin and histone deacetylase 1, and its down-regulation induces pancreatic beta-cell proliferation. J Biol Chem 2007;282:19052–61. Ikegami D, Akiyama H, Suzuki A, Nakamura T, Nakano T, Yoshikawa H, et al. Sox9 sustains chondrocyte survival and hypertrophy in part through Pik3ca-Akt pathways. Development 2011;138:1507–19. Jiang J, Au M, Eshpeter A, Korbutt G, Fisk G, Majumdar AS. Generation of Insulinproducing islet-like clusters from human embryonic stem cells. Stem Cells 2007;25:1940–53. Jørgensen MC, Ahnfelt-Rønne J, Hald J, Madsen OD, Serup P, Hecksher-Sørensen J. An illustrated review of early pancreas development in the mouse. Endocr Rev 2007;28:685–705. Kanai-Azuma M, Kanai Y, Gad JM, Tajima Y, Taya C, Kurohmaru M, et al. Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 2002;129:2367–79. Lee YH, Saint-Jeannet JP. Sox9 a novel pancreatic marker in xenopus. Int J Dev Biol 2003;47:459–62. Li J, Goodyer CG, Fellows F, Wang R. Stem cell factor/c-kit interactions regulate human islet-epithelial clusters proliferation and differentiation. Int J Biochem Cell Biol 2006;38:961–72. Li J, Quirt J, Do HQ, Lyte K, Fellows F, Goodyer CG, et al. Expression of c-Kit receptor tyrosine kinase and effect on -cell development in the human fetal pancreas. Am J Physiol Endocrinol Metab 2007;293:E475–83. Lioubinski O, Muller M, Wegner M, Sander M. Expression of SOX transcription factors in the developing mouse pancreas. Dev Dyn 2003;227:402–8.
83
Lynn FC, Smith SB, Wilson ME, Yang KY, Nekrep N, German MS. Sox9 coordinates a transcriptional network in pancreatic progenitor cells. Proc Natl Acad Sci USA 2007;104:10500–5. Lyttle BM, Li J, Krishnamurthy M, Fellows F, Wheeler MB, Goodyer CG, et al. Transcription factor expression in the developing human fetal endocrine pancreas. Diabetologia 2008;51:1169–80. Mavropoulos A, Devos N, Biemar F, Zecchin E, Argenton F, Edlund H, et al. Sox4b is a key player of pancreatic alpha cell differentiation in zebrafish. Dev Biol 2005;285:211–23. McDonald E, Krishnamurthy M, Goodyer CG, Wang R. The emerging role of SOX transcription factors in pancreatic endocrine cell development and function. Stem Cells Dev 2009;18:1379–88. Meier JJ, Bhushan A, Butler PC. The potential for stem cell therapy in diabetes. Pediatr Res 2005;59:65–73. Mendez P, Garcia-Segura LMO. Phosphatidylinositol 3-kinase and glycogen synthase kinase 3 regulate estrogen receptor-mediated transcription in neuronal cells. Endocrinology 2006;147:3027–39. Mori-Akiyama Y, van den Born M, van Es JH, Hamilton SR, Adams HP, Zhang J, et al. SOX9 is required for the differentiation of paneth cells in the intestinal epithelium. Gastroenterology 2007;133:539–46. Nef S, Verma-Kurvari S, Merenmies J, Vassalli J, Efstratiadis A, Accili D, et al. Testis determination requires insulin receptor family function in mice. Nature 2003;426:291–5. Piper K, Ball SG, Keeling JW, Mansoor S, Wilson DI, Hanley NA. Novel SOX9 expression during human pancreas development correlates to abnormalities in campomelic dysplasia. Mech Dev 2002;116:223–6. Ramocki NM, Wilkins HR, Magness ST, Simmons JG, Scull BP, Lee GH, et al. Insulin receptor substrate-1 deficiency promotes apoptosis in the putative intestinal crypt stem cell region, limits Apcmin/+ tumors and regulates Sox9. Endocrinology 2008;149:261–7. Saleem S, Li J, Yee SP, Fellows GF, Goodyer CG, Wang R. beta1 integrin/FAK/ERK signalling pathway is essential for human fetal islet cell differentiation and survival. J Pathol 2009;219:182–92. Schepers GE, Teasdale RD, Koopman P. Twenty pairs of sox: extent, homology, and nomenclature of the mouse and human sox transcription factor gene families. Dev Cell 2002;3:167–70. Seymour PA, Freude KK, Tran MN, Mayes EE, Jensen J, Kist R, et al. SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proc Natl Acad Sci USA 2007;104:1865–70. Seymour PA, Freude KK, Dubois CL, Shih HP, Patel NA, Sander M. A dosagedependent requirement for Sox9 in pancreatic endocrine cell formation. Dev Biol 2008;323:19–30. van de Wetering M, Oosterwegel M, van Norren K, Clevers H. Sox-4 an Srylike HMG box protein, is a transcriptional activator in lymphocytes. EMBO J 1993;12:3847–54. Wegner M. From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res 1999;27:1409–20. Wilson M, Koopman P. Matching S.O.X. Partner proteins and co-factors of the SOX family of transcriptional regulators. Curr Opin Genet Dev 2002;12: 441–6. Wilson ME, Yang KY, Kalousova A, Lau J, Kosaka Y, Lynn FC, et al. The HMG box transcription factor Sox4 contributes to the development of the endocrine pancreas. Diabetes 2005;54:3402–9. Zeng X, Tamai K, Doble B, Li S, Huang H, Habas R, et al. A dual-kinase mechanism for Wnt co-receptor phosphorylation and activation. Nature 2005;438:873–7.
Contents lists available at SciVerse ScienceDirect
The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel
SOX9 regulates endocrine cell differentiation during human fetal pancreas development Erin McDonald a , Jinming Li a,b , Mansa Krishnamurthy a , George F. Fellows d , Cynthia G. Goodyer e , Rennian Wang a,b,c,∗ a
Children’s Health Research Institute, University of Western Ontario, London, ON N6C 2V5, Canada Physiology & Pharmacology, University of Western Ontario, London, ON N6C 2V5, Canada c Medicine, University of Western Ontario, London, ON N6C 2V5, Canada d Obstetrics and Gynecology, University of Western Ontario, London, ON N6C 2V5, Canada e Department of Pediatrics, McGill University, Montreal, QC H3Z 2Z3, Canada b
a r t i c l e
i n f o
Article history: Received 29 April 2011 Received in revised form 7 September 2011 Accepted 22 September 2011 Available online 29 September 2011 Keywords: Human fetal pancreas Human islet-epithelial cells Islet transcription factors Akt/GSK3 pathway Beta- and alpha-cell differentiation
a b s t r a c t The transition of pancreatic progenitor cells to mature endocrine cells is regulated by the sequential activation and interaction of several transcription factors. In mice, the transcription factor Sox9 has been shown to support endocrine cell differentiation. However, the functional role of SOX9 during pancreas development in the human has yet to be determined. The present study was to characterize SOX9 expression during human fetal pancreas development and examine its functional role by transfection with SOX9 siRNA or SOX9 expression vectors. Here we report that SOX9 was most frequently expressed in PDX1+ cells (60–83%) and least in mature endocrine cells (<1–14%). The proliferation of SOX9+ cells was significantly higher at 8-10weeks than at 14–21weeks (p < 0.05) or 20–21weeks (p < 0.01). SOX9 frequently co-localized with FOXA2, NGN3 and transcription factors linked to NGN3 (NKX2.2, NKX6.1, PAX6). siRNA knockdown of SOX9 significantly decreased islet-epithelial cell proliferation, NGN3, NKX6.1, PAX6 and INS mRNA levels and the number of NGN3+ and insulin+ cells (p < 0.05) while increasing GCG mRNA and glucagon+ cells (p < 0.05). Examination of SOX9 associated signaling pathways revealed a decrease in phospho-Akt (p < 0.01), phospho-GSK3 (p < 0.01) and cyclin D1 (p < 0.01) with a decrease in nuclear -catenin+ (p < 0.05) cells following SOX9 siRNA knockdown. In contrast, over-expression of SOX9 significantly increased the number of islet cells proliferating, NGN3, NKX6.1, PAX6 and INS mRNA levels, the phospho-Akt/GSK3 cascade and the number of insulin+ cells. Our results demonstrated that SOX9 is important for the expression of NGN3 and molecular markers of endocrine cell differentiation in the human fetal pancreas. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction One major cause of the progression of diabetes is a decline in pancreatic beta-cell mass and function. Therapeutic strategies aimed at repopulating insulin-producing cells show great potential for restoring normal blood glucose levels and, thus, extensive research efforts are focused on developing methods to differentiate progenitor cells into beta-cells (Hao et al., 2006; Jiang et al., 2007; Meier et al., 2005) and to maintain their viability and function (Bonner-Weir and Sharma, 2002; Heit and Kim, 2004). However, such an undertaking requires a thorough analysis of transcription factors that control pancreatic organogenesis and tissue
∗ Corresponding author at: Victoria Research Laboratories, Room A5-140, 800 Commissioners Road East, London, Ontario N6C 2V5, Canada. Tel.: +1 519 685 8500x55098; fax: +1 519 685 8186. E-mail address: [email protected] (R. Wang). 1357-2725/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2011.09.008
maintenance. One important player in this complex process is the SOX, or sex-determining region on Y box, family of transcription factors which is involved in cell fate determination and the development of several tissues during embryogenesis (Akiyama et al., 2002; De Santa Barbara et al., 1998; van de Wetering et al., 1993; Wegner, 1999). These transcription factors share a high mobility group (HMG) DNA binding domain which alone fulfills the functions of DNA binding and bending, protein interactions and nuclear import or export (Wilson and Koopman, 2002). Several members of this family are under investigation for their role in murine pancreas development (Lioubinski et al., 2003; McDonald et al., 2009), including Sox4, Sox6, Sox9 and Sox17 with regards to endoderm specification (Kanai-Azuma et al., 2002), maintaining pancreatic progenitor cell status and supporting endocrine cell differentiation (Lynn et al., 2007; Seymour et al., 2007; Wilson et al., 2005), beta-cell proliferation and insulin secretion (Iguchi et al., 2005, 2007) and endocrine cell differentiation (Mavropoulos et al., 2005; Wilson et al., 2005). Only SOX9 expression has been
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
confirmed in the developing and adult human pancreas (Piper et al., 2002). During murine pancreas development, Sox9 is expressed in Pdx1+ progenitor cells, but is absent from committed endocrine precursors or differentiated cells (Lioubinski et al., 2003; Lynn et al., 2007; Seymour et al., 2007; Lee and Saint-Jeannet, 2003). The inactivation of SOX9 in the mouse pancreas leads to significant depletion of pancreatic progenitor cells, resulting in pancreatic hypoplasia (Wilson et al., 2005). SOX9 has also been shown to contribute to the proliferation, survival and maintenance of pancreatic progenitor cells in mouse model systems by binding to the NGN3 promoter and activating NGN3 gene expression, resulting in stable expression of a network of transcription factors that preserve pancreatic progenitor cell populations (Lynn et al., 2007; Seymour et al., 2007). Studies using lineage-tracing in mice have demonstrated that SOX9+ cells, found in the epithelial cords, give rise to both endocrine and exocrine cells of the pancreas (Seymour et al., 2008). However, a 50% reduction in SOX9 gene expression in pancreatic progenitors revealed significant decreases in endocrine cell mass while exocrine compartments remained unchanged. These findings suggest that, in the rodent, SOX9 is an essential transcription factor that governs the adoption of an endocrine phenotype and is less important for the development of the exocrine pancreas (Seymour et al., 2008). Although the role of Sox9 in the murine pancreas is well studied, knowledge of the function of this transcription factor during human pancreas development is limited. Campomelic dysplasia, a disorder resulting in severe skeletal defects and partial sex reversal, is caused by SOX9 haploinsufficiency (Piper et al., 2002; Seymour et al., 2008). Pancreatic analyses performed at term revealed significant abnormalities, including reductions in islet diameter, decreased endocrine cell clustering resulting in poorly formed islets, and variable expression of beta-cell maturity markers, including insulin (Piper et al., 2002). These findings suggest that normal SOX9 expression is important for islet formation in the human as well as the rodent. Although, signaling molecules which govern Sox9 expression and regulation have been explored in detail in different cell types including bone, heart, gonads, lymphocytes and glial cells (McDonald et al., 2009), signaling cascades which regulate SOX9 expression in the pancreas are yet to be elucidated. It has been reported that SOX9 plays a dual role during intestinal epithelium development and that this is related to the Wnt signaling pathway (Bastide et al., 2007; Mori-Akiyama et al., 2007): SOX9 mediates Wnt-dependent paneth cell differentiation and prevents over-activity of the Wnt-dependent transcriptional program in mouse models (Bastide et al., 2007). In cartilaginous tissue, the PI3K/Akt pathway has been shown to maintain SOX9 expression and perturbation of this cascade led to reduced SOX9 expression and transcriptional activity (Cheng et al., 2009). Recent studies have shown that SOX9 knockdown in vitro and deletion in vivo reduced PI3 K expression and phosphorylation, leading to chondrocyte death – a response which was reversed with the overexpression of SOX9 (Ikegami et al., 2011). Moreover, functional insulin signaling is required for SOX9 expression and, in turn, for proper testicular development and maturation (Nef et al., 2003; Ramocki et al., 2008). Overexpression of insulin receptor substrate-1 (IRS-1) increased SOX9 expression in intestinal epithelial cells and protected against apoptosis of crypt stem or progenitor cells (Ramocki et al., 2008). Glycogen synthase 3 (GSK3) is a downstream molecule of both the PI3K/Akt and Wnt pathways and has been shown to have an inhibitory effect on downstream molecules (Mendez and Garcia-Segura, 2006; Zeng et al., 2005). Thus alterations in PI3K/Akt, Wnt and GSK3 signaling cascades were examined following SOX9 siRNA and overexpression in the present study.
73
The present study has characterized the spatial and temporal expression of SOX9 in the developing human fetal pancreas from 8 to 21 weeks of fetal age and examined its functional role during islet cell differentiation using in vitro loss and gain of function experiments. Our results show that SOX9 is important for endocrine cell differentiation. 2. Tissues, materials and methods 2.1. Pancreatic tissue collection and cell culture Human fetal pancreatic tissues (8–21 weeks fetal age) were collected according to protocols approved by the Health Sciences Research Ethics Board at the University of Western Ontario and the Research Ethics Board of the Royal Victoria Hospital at the McGill University Health Centre, in accordance with the Canadian Council on Health Sciences Research Involving Human Subjects guidelines. Tissues were immediately processed for immunohistology, RNA or protein extraction, with a minimum of five pancreases per age group (Al-Masri et al., 2010; Lyttle et al., 2008; Saleem et al., 2009). Fetal pancreatic tissues from 14 to 16 weeks fetal age, a stage at which there is an enriched SOX9+ /PDX1+ cell population (Lyttle et al., 2008), were dissected and immediately digested using dissociation buffer containing collagenase V (1 mg/ml, Sigma, St. Louis, MO, USA), as described previously (Li et al., 2006). The isolated islet-epithelial cell clusters were cultured in CMRL1066 media containing 10% fetal bovine serum (FBS) (Invitrogen, Burlington, ON, Canada). 2.1.1. Transient transfection of SOX9 siRNA Freshly isolated islet-epithelial cell clusters from 14 to 16 week fetal pancreases were recovered overnight and transiently transfected for 30 h in antibiotic-free medium with 60 nM SOX9(h) siRNA (sc-36533) or control siRNA (sc-37007, proprietary sequence), commercially obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA), using an siRNA transfection kit (Li et al., 2007). A pool of three sequences for human SOX9 siRNA (Santa Cruz) were used as listed in Supplementary Table 1. Islet epithelial-cell clusters were harvested at 48 h following transfection. Transfection efficiency was monitored using fluorescein-conjugated control siRNA (Santa Cruz) and qRT-PCR analysis for SOX9 mRNA, with approximately 50–60% of the islet-epithelial cluster cells being transfected, as described previously (Al-Masri et al., 2010; Li et al., 2007; Saleem et al., 2009). 2.1.2. Transient over-expression of SOX9 Full-length human SOX9 cDNA (a kind gift from Dr. Philippe Jay, Institut de Génomique Fonctionnelle, UMR5203, France) was inserted into the KpnI and XbaI sites of pcDNA3.1 (Clontech, Mountain View, CA, USA), as described previously (Bastide et al., 2007). Isolated human fetal islet-epithelial cell clusters were transfected with 1 g of either pcDNA3.1-SOX9 or the empty (control) vector using Lipofectamine Plus reagent (InVitrogen) in a serum- and antibiotic-free medium for 48 h (Saleem et al., 2009). At the end of the culture period, islets were harvested and processed for protein or RNA extraction or fixed for immunocytochemistry studies. Three to six different preparations from a pool of 2–3 fetal pancreatic isolations per experimental group were used for analyses, with an average of 500–1000 cells counted per experimental group (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). 2.2. Immunofluorescent and morphometric analyses Pancreatic tissues and islets were fixed in 4% paraformaldehyde. Cell pellets were embedded in 2% agarose, followed by
74
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 1. Expression pattern of SOX9 during human fetal pancreatic development. (A) Real-time RT-PCR analysis of SOX9 mRNA expression. Data are normalized to 18S rRNA subunit and expressed as means ± SEM (n = 5–9 pancreases/age group). (B) Western blot analyses of SOX9 protein in human fetal pancreatic tissue. Data are normalized to calnexin and expressed as means ± SEM (n = 3 pancreases/age group). *p < 0.05 vs. 10–11 weeks. (C–E) Double immunofluorescence of SOX9 and morphometric analyses of SOX9+ cells in the developing human pancreas. Representative images for SOX9 (green) with CK19, insulin, glucagon, PDX1 or Ki67 (red) in a 14 weeks human fetal pancreas. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrows indicate double positive cells. Bar graph data are expressed as means ± SEM percentage of SOX9+ cells in the duct and endocrine region (C) and the number of PDX1+ (D) and Ki67+ (E) cells in the SOX9-expressing cell population (n = 5–7 pancreases/age group). *p < 0.05, **p < 0.01, ***p < 0.001 vs. 8–11 weeks.
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
75
Fig. 2. Co-localization of SOX9 with NGN3 during human fetal pancreatic development. (A) Representative images of co-staining for SOX9 (green) and NGN3 (red) in a 12 week human fetal pancreas. Nuclei were stained with DAPI in blue. Scale bar: 25 m. Magnified images for each corresponding figure are shown in the bottom panel. Scale bar: 10 m. Arrows indicate double positive cells and arrow-heads show single stained cells. (B) Quantitative analysis of NGN3+ /SOX9+ cells relative to the total number of SOX9+ cells counted. Data are expressed as means ± SEM (n = 4–7 pancreases/age group). *p < 0.05 vs. 8–9 weeks.
paraffin embedding (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). Sections (5 m) were cut throughout the entire length of the pancreas and were stained with appropriate dilutions of primary antibodies (Supplementary Table 2). To identify the expression of SOX9 in various pancreatic compartments and the co-localization with transcription factors or other SOXs in the developing human fetal pancreas, double immunofluorescence staining was performed. The cells expressing these factors in the ductal and endocrine cell region and in cultured isletepithelial cells were determined by cell counting in which the counter was blind to the age and experimental groups of the samples to minimize bias (Al-Masri et al., 2010). Data are expressed as the mean percentage of SOX9+ cells in CK19+ , insulin+ or glucagon+ cell populations. The number of proliferating (Ki67), PDX1+ and NGN3+ cells co-expressing SOX9 was also examined and expressed as the mean percentage of proliferating, PDX1+ or NGN3+ cells within the SOX9+ cell population. The number of nuclear -catenin expressing cells were also examined (Al-Masri et al., 2010). 2.3. Protein extraction and western blot analysis. Total protein from human fetal pancreatic tissues and isolated fetal islets were extracted (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). Equal amounts (30 g) of extracted proteins from each experimental group were separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane (Bio-Rad, Mississauga, ON, Canada).
Membranes were incubated with appropriate dilutions of primary antibodies (Supplementary Table 2), followed by the application of appropriate horse radish peroxidase-conjugated secondary antibodies (Santa Cruz). Proteins were detected using ECLTM -Plus Western blot detection reagents (Perkin Elmer, Wellesley, MA, USA). Densitometric quantification of bands at subsaturation levels was performed using Syngenetool gel analysis software (Syngene, Cambridge, UK) and normalized to appropriate loading controls, which included the total signaling protein content or a housekeeping protein (calnexin). Data are provided as the relative expression level of phosphorylated proteins to total protein levels or protein levels to the loading control (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). 2.4. RNA extraction and real-time RT-PCR RNA was extracted from human fetal pancreatic tissues using TRIZOL reagent (Invitrogen). Islet RNA was extracted using the RNAqueous-4PCR kit (Invitrogen) (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009). For each RT reaction, 2 g of DNA-free RNA were used with random hexamers/oligo(dT) primers and Superscript reverse transcriptase. Real-time PCR analyses were performed using the iQ SYBR Green Supermix kit in Chromo4 Real time PCR (Bio-Rad). Primers used are listed in Supplementary Table 3. Relative gene expression was calculated and normalized to the internal standard gene, 18S rRNA, with at least five repeats per age or experimental group (Al-Masri et al., 2010; Li et al., 2006, 2007; Saleem et al., 2009).
76
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 3. Co-localization of SOX9 with NGN3-linked transcription factors during human fetal pancreatic development. Representative images of SOX9 (green) and FOXA2, NKX6.1, NKX2.2 or PAX6 (red) co-localization at 12 weeks of human fetal pancreatic development. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrow indicates double-stained cells, arrow-heads show single stained cells and asterisks indicate non-specific stained dots.
2.5. Statistical analysis Data were expressed as means ± SEM. Statistical significance was determined using either a paired student’s t-test or one-way ANOVA followed by the post hoc least significant differences (LSD) and Tukey’s multiple comparison tests. Differences were considered to be statistically significant when p < 0.05. 3. Results 3.1. Temporal and spatial expression of SOX9 during human fetal pancreas development Using qRT-PCR, western blotting and immunofluorescent staining, we characterized the temporal expression pattern of SOX9 in the developing human fetal pancreas from 8 to 21 weeks of fetal age. Pancreatic SOX9 mRNA levels remained relatively
stable during this period (Fig. 1A); however, SOX9 protein levels decreased gradually with a significant reduction at 19–20 weeks when compared to the expression level at 10–11 weeks (p < 0.05, Fig. 1B). Double immunofluorescence staining of SOX9 with pancreatic cell markers revealed that the majority of pancreatic CK19+ cells contained SOX9 at 8–11 weeks and this co-expression declined significantly at 14–16 weeks (p < 0.01) and 20–21 weeks (p < 0.001) (Fig. 1C), by which time SOX9 expression was restricted to a subpopulation of CK19+ cells. The percentage of insulin+ and glucagon+ cells co-expressing SOX9 is considerably lower than that of CK19+ cells at all stages of development and this decreased significantly, reaching almost undetectable levels at midgestation (Fig. 1C). Co-localization studies of PDX1 with SOX9 showed PDX1 to be highly expressed in SOX9+ cell population (∼83% colocalization) at 8–11 weeks of development but that this had decreased significantly by 14–16 weeks (p < 0.05) and even more by 20–21 weeks (p < 0.01) (Fig. 1D). Furthermore, the proliferation
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
77
Fig. 4. Co-localization of SOX9 with other SOX transcription factors during human fetal pancreatic development. Representative images of SOX9 (green) and SOX4, SOX6, SOX11 or SOX17 (red) co-localization at 14 weeks of human fetal pancreatic development. Nuclei were stained with DAPI (blue). Scale bar: 25 m. Arrow indicates doublestained cells, arrow-heads show single stained cells and asterisks indicate non-specific stained dots.
of SOX9+ cells was examined using the proliferation marker Ki67 (Fig. 1E). SOX9+ cells showed the highest proliferative capacity at 8–11 weeks (10 ± 1%); this declined significantly at 14–16 weeks (4 ± 1%, p < 0.05) and further at 20–21 weeks (2 ± 0.4%, p < 0.01) (Fig. 1E). 3.2. Co-localization of SOX9 with NGN3 and NGN3-linked as well as other SOX transcription factors in the human fetal pancreas Using immunofluorescence staining to examine the colocalization of SOX9 with NGN3 (Fig. 2), FOXA2, NKX2.2, NKX6.1 and PAX6 (Fig. 3), we found that SOX9 was frequently co-expressed with transcription factors involved in endocrine cell development from 8 to 21 weeks of fetal age. Quantitative analysis showed approximately 5% of the SOX9+ cells stained for NGN3 in 8–9 week old human fetal pancreas with a significant reduction at 14–21 weeks (p < 0.05, Fig. 2B), suggesting that SOX9 interactions with
endocrine cell specific transcription factors may be important in regulating islet cell differentiation during early stages of human fetal pancreatic development (Wegner, 1999). A large number of SOX transcription factors are expressed in the murine pancreas, but whether these are also expressed in the human fetal pancreas and co-express SOX9 has not been explored. Double immunofluorescent staining for SOX9 with SOX4, SOX6, SOX11 or SOX17 at 14 weeks fetal age showed frequent coexpression of SOX9 with SOX4 and SOX17 (Fig. 4). Most of the SOX11+ cells in the ductal regions co-expressed SOX9 while only a small number of SOX9+ cells co-expressed SOX6 (Fig. 4). 3.3. Knockdown of SOX9 mRNA levels results in the reduction of beta-cell differentiation in human fetal islets In order to explore the functional role of SOX9 in human fetal endocrine cell differentiation, we performed an siRNA knockdown
78
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 5. Effect of SOX9 siRNA on beta-cell differentiation in isolated human fetal islet-epithelial cell clusters. (A) Double immunofluorescence images of SOX9 (green) and CK19 (red), SOX9+ cell counts, western blot and qRT-PCR analyses of SOX9 expression in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Data were normalized to control siRNA and are expressed as means ± SEM (n = 3–5 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group). Representative blots are displayed. Scale bar: 25 m. (B) qRT-PCR analysis of transcription factors associated with pancreatic endocrine cell differentiation in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Data are expressed as means ± SEM (n = 4–6 experiments/treatment group; *p < 0.05 vs. control siRNA group).
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
79
Fig. 5. (Continued ) (C) The percentage of NGN3+ , insulin+ , glucagon+ and Ki67+ cells in human fetal islet-epithelial cell clusters treated with either control or SOX9 siRNA for 48 h. Data are normalized to control siRNA and expressed as means ± SEM (n = 5 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group).
of SOX9 in 14–16 week human fetal islet-epithelial cells. The efficiency of the knockdown was determined by qRT-PCR, western blot and immunostaining and showed at least a 50% reduction in SOX9 expression at both mRNA and protein levels (p < 0.05–0.01, Fig. 5A). Following knockdown of SOX9 expression, several transcription factors and hormones important to the development of the endocrine pancreas were examined by qRT-PCR (Fig. 5B). Cells treated with SOX9 siRNA showed no significant changes in PDX1, FOXA2, NEUROD1, PAX4 or NKX2.2 mRNA expression when compared to control siRNA groups (Fig. 5B). However, a significant decrease in NGN3 and INS mRNA levels (p < 0.05, Fig. 5B) and in the number of NGN3+ (p < 0.05) and insulin+ cells (p < 0.01) (Fig. 5C) in comparison with control siRNA groups was observed, along with decreased expression of NKX6.1 and PAX6 mRNA (p < 0.05, Fig. 5B). Interestingly, knockdown of SOX9 expression induced a significant increase in TCF1, HES1 and GCG mRNA (p < 0.05, Fig. 5B) along with a significant increase in the number of glucagon+ cells (p < 0.01, Fig. 5C). The proliferation capacity of the islet-epithelial cells following SOX9 siRNA transfection was ∼50% reduced (p < 0.05, Fig. 5C). 3.4. Increasing SOX9 expression affects beta-cell differentiation in human fetal islets To understand whether increasing SOX9 expression could stimulate beta-cell differentiation, SOX9 was overexpressed in isolated human fetal islet-epithelial cells using transient transfection of a full-length human SOX9 (SOX9) construct or empty vector (Ctrl). qRT-PCR and western blot analyses showed an increase in SOX9 mRNA (2.8-fold, p < 0.05) (Fig. 6A) and protein (2.2-fold, p < 0.01) (Fig. 6B) levels, and the number of SOX9+ cells (1.8-fold, p < 0.05) (Fig. 6C) in the over-expression group as compared to controls after 48 h. qRT-PCR analyses revealed a significant increase in NGN3, PAX6, NKX6.1 and INS mRNA levels in the islet-epithelial cells following SOX9 overexpression (p < 0.05; Fig. 6D), while genes for TCF1, HES1 and GCG remained unchanged (Fig. 6D).
3.5. SOX9 affects Akt/GSK3ˇ signaling during human fetal pancreas development To explore if signaling pathways may be altered following either knockdown or over-expression of SOX9, members of the Akt and Wnt signaling pathways, including phospho-Akt, phosph-GSK3, nuclear -catenin (active) and cyclin D1 were examined. Partial reduction of SOX9 expression by siRNA transfection was associated with a decrease in phospho-Akt (p < 0.05), phospho-GSK3B (p < 0.05), cyclin D1 (p < 0.01) (Fig. 7A) and the number of nuclear localized -catenin (p < 0.01; Fig. 7B) when compared to control siRNA groups. In contrast, in the SOX9 over-expression experimental group there was a significant increase in phosphorylated Akt and GSK3 (p < 0.05) along with cyclin D1 protein levels (p < 0.01) (Fig. 8A), in parallel with a significant increase in the percentage of cells expressing the proliferation marker Ki67 (p < 0.01, Fig. 8B), but no change in the number of n-catenin+ cells (data not shown). To further examine whether changes in phospho-Akt expression were a result of insulin signaling, insulin treated islet-epithelial cells were analyzed for SOX9 expression. The expression of SOX9 mRNA and the number of SOX9+ cells in insulin-treated human fetal isletepithelial cells significantly increased when compared to untreated groups (Supplementary Fig. 1) (Al-Masri et al., 2010). These results were further confirmed by blocking the PI3K/Akt cascade using the PI3K inhibitor, wortmannin, which abrogated the increase in the number of SOX9+ cells in human fetal islet-epithelial cell clusters following insulin treatment (Supplementary Fig. 1) (Al-Masri et al., 2010; Li et al., 2006). In addition, we examined if the Notch signaling pathway is involved in SOX9 regulation of the human fetal pancreas. qRTPCR analyses of NOTCH1, JAG1 and DLL1 were performed but no statistical differences were observed in their mRNA expression following knockdown or overexpression of SOX9 versus control groups (Supplementary Fig. 2A and C). There was also no change in notch intracellular domain (NICD) expression after SOX9 siRNA treatment, as determined by immunofluorescence staining (Supplementary Fig. 2B).
80
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 6. Over-expression of SOX9 increases the expression of NGN3 and its linked transcription factors in isolated human fetal islet-epithelial cell clusters. (A) qRT-PCR and (B) western blot analysis of SOX9 expression, as well as (C) the percentage of SOX9+ cells in islet-epithelial cell clusters transfected with either full-length human SOX9 cDNA or control vector for 48 h. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group); *p < 0.05; **p < 0.01 vs. control group. (D) qRT-PCR analysis of NGN3 and its linked transcription factor genes. Data are normalized to 18S rRNA subunit and expressed as means ± SEM (n = 3–4 experiments/treatment group); *p < 0.05 vs. control group.
4. Discussion SOX9 is expressed in a wide variety of human tissues but, to date, studies in the human pancreas have been limited (Piper et al., 2002). The present study demonstrates that SOX9 is detectable in the human pancreas from as early as the 8th week of fetal life, primarily in PDX1+ ductal cells; it is frequently found co-localizing with NGN3 and other key islet beta-cell progenitor markers. Interestingly, knockdown studies of SOX9 revealed significant decreases in cells expressing beta-cell markers paralleled by increases in alpha cell phenotypes. SOX9 over-expression experiments confirmed that expression of this transcription factor is critical for beta cell differentiation in the human fetal pancreas. The highest expression of SOX9 mRNA and protein was observed in the 8–11 week fetal pancreas; while the mRNA levels remained relatively stable, the protein levels steadily declined by midgestation. The mechanism by which SOX9 protein levels decrease during development has been reported to be pleiotrophic and includes processes of sumoylation, ubiquitination and proteasome degradation (Hattori et al., 2006; Akiyama et al., 2005a,b). In addition, the proliferative capacity of SOX9+ cells was highest during 8–11 weeks, with a rapid decline by 20–21 weeks. These results concur with a report by Piper et al. (2002) and studies conducted in mice (Lioubinski et al., 2003; Lynn et al., 2007;
Seymour et al., 2007) revealing that SOX9 expression is highest during embryogenesis and declines during maturation of the pancreas. The frequent co-localization of SOX9 with PDX1+ and CK19+ cells observed during early human fetal pancreas development is also consistent with previous studies in the human (Piper et al., 2002) and murine pancreas (Lioubinski et al., 2003; Lynn et al., 2007; Seymour et al., 2007) showing that SOX9 is expressed in endocrine progenitor and undifferentiated CK19+ cell populations. In addition, it has been shown in mice that all cells of the pancreas are derived from SOX9 expressing precursors (Lynn et al., 2007; Jørgensen et al., 2007). Sox9 has also been shown to be critical for expression of a transcription factor gene network in pancreatic progenitor cells in mice (Lynn et al., 2007; Seymour et al., 2007) that initiates differentiation of progenitors into endocrine cell types (Lynn et al., 2007). Studies by Lynn et al. revealed that direct Sox9 binding to the Ngn3 promoter activates Ngn3, resulting in expression of several downstream transcription factors, including Foxa2, Onecut1 and Tcf2 (Lynn et al., 2007). In the human fetal pancreas, co-expression of SOX9 with NGN3 and NGN3-linked transcription factors such as FOXA2, PAX6 and NKX2.2 was observed frequently. In contrast, SOX9 had limited co-expression with insulin and glucagon in the islet cells, suggesting that SOX9 is more involved
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
81
Fig. 7. Knockdown of SOX9 expression decreases the phosphorylation of Akt/GSK3 signals in isolated human fetal islet-epithelial cell clusters. (A) Western blot analysis of phosphorylated [P] and total [T] Akt; phosphorylated [P] and total [T] GSK3, and cyclin D1; (B) immunofluorescence staining for PY489--catenin (n-catenin) and the percentage of n-catenin+ cells in islet-epithelial cell clusters transfected with either control or SOX9 siRNA for 48 h. Representative blots and images are shown. Scale bar: 25 m. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control siRNA group). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
in progenitor cell populations and is not required for maintenance of mature endocrine cells in the human fetal pancreas (Piper et al., 2002). To understand the functional role for SOX9 in human fetal islets, the effects of SOX9 knockdown and over-expression were examined. Knockdown or over-expression of SOX9 did not change PDX-1 expression, indicating that SOX9 and PDX1 do not regulate each other during human fetal pancreas development. These findings are similar to those reported in mouse model systems: no differences in Sox9 expression were observed in PDX1 null embryos, despite their co-expression in pluripotent pancreatic progenitors (Seymour et al., 2007). Following SOX9 siRNA knockdown of the human fetal islet cells, there were significant decreases in NGN3, NKX6.1 and PAX6 mRNA as well as the number of cells expressing NGN3; these data parallel previous findings from mouse studies (Lynn et al., 2007; Seymour et al., 2007). In contrast, no significant changes in FOXA2, NEUROD1, PAX4 or NKX2.2 mRNA levels were observed, suggesting that the network of transcription factors that SOX9 regulates differs significantly from that reported in the mouse (Lynn et al., 2007; Seymour et al., 2007).
Interestingly, siRNA knockdown in human fetal islet cells also resulted in an increase in glucagon mRNA and protein expression. These data are inconsistent with a report in the literature that partial Sox9 gene deletion in the murine pancreas leads to defects in both alpha- and beta-cell mass (Seymour et al., 2008). However, we previously observed a dramatic increase in alpha-cell proliferation and cell mass from 14 to 21 weeks of fetal life, suggesting that the differentiation of alpha-cells occurs earlier than that of beta-cells in the human pancreas (Lyttle et al., 2008). Thus, it is possible that the effects of SOX9 on human alpha- and beta-cell differentiation in the present study may be due to the timing of SOX9 knockdown (14–16 weeks). Moreover, the time point at which SOX9 was manipulated has influenced changes in NGN3 expression and that may impact the type of endocrine cell that is induced to differentiate. In mice, Sox9+ cells express Ngn3 early in development, thus increasing the differentiation of alpha cells. However, inducing Ngn3 expression in progenitor cells later in murine development also produced beta- and PP-cells (Jørgensen et al., 2007). Furthermore, increases in NGN3, NKX6.1, PAX6 and insulin mRNA were observed in SOX9 over-expressed islet cells. This increase in SOX9 did not affect glucagon gene expression, suggesting that SOX9 may have less or no
82
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83
Fig. 8. Over-expression of SOX9 increases the phosphorylation of Akt/GSK3 and cell proliferation. (A) Western blot analysis of phosphorylated [P] and total [T] Akt; phosphorylated [P] and total [T] GSK3, and cyclin D1 expression in islet-epithelial cell clusters transfected with either full-length human SOX9 cDNA or control vector for 48 hr. (B) The percentage of Ki67+ cells. Data are expressed as means ± SEM (n = 3–6 experiments/treatment group; *p < 0.05; **p < 0.01 vs. control group).
influence over alpha-cell differentiation after 14–16 weeks of pancreatic development. Signaling pathways associated with SOX9 expression in the human pancreas have not been previously explored. Upon knockdown of SOX9 mRNA, a significant decrease in GSK3 phosphorylation was observed, along with decreases in n-catenin and cyclin D expression and a reduction in cell proliferation labeled by Ki67. Conversely, increasing SOX9 products in isolated human fetal islet-epithelial cells resulted in the up-regulation of phosphorylated GSK3/cyclin D1 and Ki67+ cells, suggesting that SOX9 promotes pancreatic progenitor cell proliferation via inhibition of GSK3 activity. However, it remains to be determined whether the effects observed in GSK3 phosphorylation are through the PI3 K/Akt or Wnt signaling cascades. Our results also showed significant changes in the phosphorylation of Akt in both knockdown and overexpression studies of SOX9, indicating that the PI3K/Akt signaling pathway is involved in the expression and regulation of SOX9 expression. Furthermore, insulin treatment of human fetal islet-epithelial cells resulted in an increase in SOX9 mRNA, along with an increase in the number of SOX9+ cells when compared to untreated groups. This increase in SOX9 expression was blocked by wortmannin, a PI3K inhibitor, demonstrating the involvement of PI3K/Akt signaling pathway in SOX9 expression. The increase in SOX9 mRNA and protein expression following insulin treatment correlate with the present observations that over-expression or knockdown of SOX9 gene affects the insulin gene and protein expression, suggesting that changes in Akt phosphorylation by SOX9 may be associated with beta-cell differentiation and insulin signaling. Several studies have demonstrated that Sox9 is able to physically interact with -catenin, modulating c-Myc and cyclin D1 genes (Akiyama et al., 2005a,b; Bastide et al., 2007), while insulin signaling via the PI3K/Akt pathway has the ability to repress GSK3 function allowing -catenin accumulation (Schepers et al., 2002). Recent studies have also identified several binding sites for SOX9 on the human PI3K gene specifically around the transcription start site (Ikegami
et al., 2011). Taken together, these findings further support that SOX9 signals via the Akt/GSK3  pathway to maintain progenitor cell populations in the human fetal pancreas. In summary, we have investigated the expression and potential function of SOX9 in the early to mid-gestation human fetal pancreas. Our data suggest that SOX9 is important for endocrine cell differentiation in the developing human pancreas. Acknowledgements We greatly appreciate the gifts of a full-length human SOX9 cDNA from Dr. Philippe Jay (Institut de Génomique Fonctionnelle, Centre National de la Recherche Scientifi que UMR5203, Université de Montpellier I and Université de Montpellier II, Montpellier, France), and anti-PDX1 antibody from Dr. Christopher Wright (University of Vanderbilt, Nashville, USA). This work was supported by grants from the Canadian Diabetes Association (CDA). Dr. Wang is supported by a New Investigator Award from CIHR. We thank the Department of Pathology at London Health Science Centre for providing human fetal pancreatic tissue sections. The authors declare that there is no duality of interest associated with this manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biocel.2011.09.008. References Al-Masri M, Krishnamurthy M, Li J, Fellows GF, Dong HH, Goodyer CG, et al. Effect of forkhead box O1 (FOXO1) on beta cell development in the human fetal pancreas. Diabetologia 2010;53:699–711. Akiyama H, Chaboissier MC, Martin JF, Schedl A, de Crombrugghe B. The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev 2002;16:2813–28.
E. McDonald et al. / The International Journal of Biochemistry & Cell Biology 44 (2012) 72–83 Akiyama H, Kim JE, Nakashima K, Balmes G, Iwai N, Deng JM, et al. Osteochondroprogenitor cells are derived from Sox9 expressing precursors. Proc Natl Acad Sci USA 2005a;102:14665–70. Akiyama H, Kamitani T, Yang X, Kandyil R, Bridgewater LC, Fellous M, et al. The transcription factor Sox9 is degraded by the ubiquitin-proteosome system and is stabilized by a mutation in a ubiquitin target site. Matrix Biol 2005b;23:499–505. Bastide P, Darido C, Pannequin J, Kist R, Robine S, Marty-Double C, et al. Sox9 regulates cell proliferation and is required for Paneth cell differentiation in the intestinal epithelium. J Cell Biol 2007;178(4):635–48. Bonner-Weir S, Sharma A. Pancreatic stem cells. J Pathol 2002;197:519–26. Cheng CC, Uchiyama Y, Hiyama A, Gajghate S, Shapiro IM, Risbud MV. PI3K regulates aggrecan gene expression by modulating Sox9 expression and activity in nucleus pulposus cells of the intervertebral disc. J Cell Physiol 2009;221:668–76. De Santa Barbara P, Bonneaud N, Boizet B, Desclozeaux M, Moniot B, Sudbeck P, et al. Direct interaction of SRY-related protein SOX9 and steroidogenic factor 1 regulates transcription of the human anti-Mullerian hormone gene. Mol Cell Biol 1998;18:6653–65. Hattori T, Eberspaecher H, Lu J, Zhang R, Nishida T, Kahyo T, et al. Interactions between PIAS Proteins and SOX9 Result in an Increase in the Cellular Concentrations of SOX9. J Biol Chem 2006;20:14417–28. Hao E, Tyrberg B, Itkin-Ansari P, Lakey JR, Geron I, Monosov EZ, et al. Beta-cell differentiation from nonendocrine epithelial cells of the adult human pancreas. Nat Med 2006;12:310–4. Heit JJ, Kim SK. Embryonic stem cells and islet replacement in diabetes mellitus. Pediatr Diabetes 2004;5:5–15. Iguchi H, Ikeda Y, Okamura M, Tanaka T, Urashima Y, Ohguchi H, et al. SOX6 attenuates glucose-stimulated insulin secretion by repressing PDX1 transcriptional activity and is down-regulated in hyperinsulinemic obese mice. J Biol Chem 2005;280:37669–80. Iguchi H, Urashima Y, Inagaki Y, Ikeda Y, Okamura M, Tanaka T, et al. SOX6 suppresses cyclin D1 promoter activity by interacting with beta-catenin and histone deacetylase 1, and its down-regulation induces pancreatic beta-cell proliferation. J Biol Chem 2007;282:19052–61. Ikegami D, Akiyama H, Suzuki A, Nakamura T, Nakano T, Yoshikawa H, et al. Sox9 sustains chondrocyte survival and hypertrophy in part through Pik3ca-Akt pathways. Development 2011;138:1507–19. Jiang J, Au M, Eshpeter A, Korbutt G, Fisk G, Majumdar AS. Generation of Insulinproducing islet-like clusters from human embryonic stem cells. Stem Cells 2007;25:1940–53. Jørgensen MC, Ahnfelt-Rønne J, Hald J, Madsen OD, Serup P, Hecksher-Sørensen J. An illustrated review of early pancreas development in the mouse. Endocr Rev 2007;28:685–705. Kanai-Azuma M, Kanai Y, Gad JM, Tajima Y, Taya C, Kurohmaru M, et al. Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 2002;129:2367–79. Lee YH, Saint-Jeannet JP. Sox9 a novel pancreatic marker in xenopus. Int J Dev Biol 2003;47:459–62. Li J, Goodyer CG, Fellows F, Wang R. Stem cell factor/c-kit interactions regulate human islet-epithelial clusters proliferation and differentiation. Int J Biochem Cell Biol 2006;38:961–72. Li J, Quirt J, Do HQ, Lyte K, Fellows F, Goodyer CG, et al. Expression of c-Kit receptor tyrosine kinase and effect on -cell development in the human fetal pancreas. Am J Physiol Endocrinol Metab 2007;293:E475–83. Lioubinski O, Muller M, Wegner M, Sander M. Expression of SOX transcription factors in the developing mouse pancreas. Dev Dyn 2003;227:402–8.
83
Lynn FC, Smith SB, Wilson ME, Yang KY, Nekrep N, German MS. Sox9 coordinates a transcriptional network in pancreatic progenitor cells. Proc Natl Acad Sci USA 2007;104:10500–5. Lyttle BM, Li J, Krishnamurthy M, Fellows F, Wheeler MB, Goodyer CG, et al. Transcription factor expression in the developing human fetal endocrine pancreas. Diabetologia 2008;51:1169–80. Mavropoulos A, Devos N, Biemar F, Zecchin E, Argenton F, Edlund H, et al. Sox4b is a key player of pancreatic alpha cell differentiation in zebrafish. Dev Biol 2005;285:211–23. McDonald E, Krishnamurthy M, Goodyer CG, Wang R. The emerging role of SOX transcription factors in pancreatic endocrine cell development and function. Stem Cells Dev 2009;18:1379–88. Meier JJ, Bhushan A, Butler PC. The potential for stem cell therapy in diabetes. Pediatr Res 2005;59:65–73. Mendez P, Garcia-Segura LMO. Phosphatidylinositol 3-kinase and glycogen synthase kinase 3 regulate estrogen receptor-mediated transcription in neuronal cells. Endocrinology 2006;147:3027–39. Mori-Akiyama Y, van den Born M, van Es JH, Hamilton SR, Adams HP, Zhang J, et al. SOX9 is required for the differentiation of paneth cells in the intestinal epithelium. Gastroenterology 2007;133:539–46. Nef S, Verma-Kurvari S, Merenmies J, Vassalli J, Efstratiadis A, Accili D, et al. Testis determination requires insulin receptor family function in mice. Nature 2003;426:291–5. Piper K, Ball SG, Keeling JW, Mansoor S, Wilson DI, Hanley NA. Novel SOX9 expression during human pancreas development correlates to abnormalities in campomelic dysplasia. Mech Dev 2002;116:223–6. Ramocki NM, Wilkins HR, Magness ST, Simmons JG, Scull BP, Lee GH, et al. Insulin receptor substrate-1 deficiency promotes apoptosis in the putative intestinal crypt stem cell region, limits Apcmin/+ tumors and regulates Sox9. Endocrinology 2008;149:261–7. Saleem S, Li J, Yee SP, Fellows GF, Goodyer CG, Wang R. beta1 integrin/FAK/ERK signalling pathway is essential for human fetal islet cell differentiation and survival. J Pathol 2009;219:182–92. Schepers GE, Teasdale RD, Koopman P. Twenty pairs of sox: extent, homology, and nomenclature of the mouse and human sox transcription factor gene families. Dev Cell 2002;3:167–70. Seymour PA, Freude KK, Tran MN, Mayes EE, Jensen J, Kist R, et al. SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proc Natl Acad Sci USA 2007;104:1865–70. Seymour PA, Freude KK, Dubois CL, Shih HP, Patel NA, Sander M. A dosagedependent requirement for Sox9 in pancreatic endocrine cell formation. Dev Biol 2008;323:19–30. van de Wetering M, Oosterwegel M, van Norren K, Clevers H. Sox-4 an Srylike HMG box protein, is a transcriptional activator in lymphocytes. EMBO J 1993;12:3847–54. Wegner M. From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res 1999;27:1409–20. Wilson M, Koopman P. Matching S.O.X. Partner proteins and co-factors of the SOX family of transcriptional regulators. Curr Opin Genet Dev 2002;12: 441–6. Wilson ME, Yang KY, Kalousova A, Lau J, Kosaka Y, Lynn FC, et al. The HMG box transcription factor Sox4 contributes to the development of the endocrine pancreas. Diabetes 2005;54:3402–9. Zeng X, Tamai K, Doble B, Li S, Huang H, Habas R, et al. A dual-kinase mechanism for Wnt co-receptor phosphorylation and activation. Nature 2005;438:873–7.