Chromatin-Remodeling Factors Allow ... - Wiley Online Library

TISSUE-SPECIFIC STEM CELLS Chromatin-Remodeling Factors Allow Differentiation of Bone Marrow Cells into Insulin-Producing Cells THATAVA TAYARAMMA,a BIN MA,a MANFRED ROHDE,b HUBERT MAYERa a

Department of Gene Regulation and Differentiation and bDepartment of Microbial Pathogenesis, German Research Center for Biotechnology, Braunschweig, Germany Key Words. Bone marrow • Histone deacetylation inhibitors • Trichostatin A • Insulin-producing cells • Diabetes

ABSTRACT Type 1 diabetes is caused by the destruction of pancreatic ␤-cells by T cells of the immune system. Islet transplantation is a promising therapy for diabetes mellitus. Bone marrow stem cells (BMSC) have the capacity to differentiate into various cell lineages including endocrine cells of the pancreas. To investigate the conditions that allow BMSC to differentiate into insulin-producing cells, a novel in vitro method was developed by using the histone deacetylase inhibitor, trichostatin A (TSA). BMSC, cultured in presence of TSA, differentiated into isletlike clusters under appropriate culture conditions. These isletlike clusters were similar to the cells of the islets of the pancreas. The islet-like clusters showed endocrine gene expression typical for pancreatic ␤-cell development and function, such as

insulin (I and II), glucagon, somatostatin, GLUT-2, pancreatic duodenal homeobox-1 (PDX-1), and Pax 4. Immunocytochemistry confirmed islet-like clusters contained pancreatic hormones. The colocalization of insulin and C-peptide was also observed. Enzyme-linked immunosorbent assay analysis demonstrated that insulin secretion was regulated by glucose. Western blot analysis demonstrated the presence of stored insulin. Electron microscopy of the islet-like cells revealed an ultrastructure similar to that of pancreatic ␤-cells, which contain insulin granules within secretory vesicles. These findings suggest that histone-deacetylating agents could allow the differentiation of BMSC into insulin-producing ␤-cells. STEM CELLS 2006;24:2858 –2867

INTRODUCTION

The key goals of stem cell research are to understand the manner in which differentiation is controlled and the direction that cellular differentiation will take. This should result in a better understanding of stem cell biology and allow optimal use of these cells in cell therapy. Stem cell plasticity depends on many factors, such as chromatin structure. The actual chromatin structure and its remodeling is an important mode of the epigenetic control of gene expression. The chromatin modulation includes histone acetylation and DNA methylation. Histone acetylation or DNA methylation modifiers modulate a wide variety of cellular functions, including cell differentiation. Histone deacetylase (HDAC) inhibitors are known as agents that modulate the expression of genes by increasing histone acetylation, thereby regulating chromatin structure and transcription [17, 18]. These HDAC inhibitors induce differentiation and inhibit cell proliferation. Trichostatin A (TSA) is a hydroxamic acid and a potent HDAC inhibitor that has been identified as having potential therapeutic value against cancer in screens for agents that induce differentiation of murine erythroleukemia cells [19, 20].

Diabetes mellitus (DM) is a common metabolic disorder affecting millions of people worldwide and is characterized by abnormally high levels of glucose in blood. Type 1 diabetes (insulin-dependent) is an autoimmune disease, resulting from the body’s own immune system destroying pancreatic ␤-cells and causing absolute insulin deficiency. An effective treatment for Type 1 diabetes is islet cell transplantation [1]. However, the limited supply of suitable donors for pancreatic islets severely limits this approach. Alternatively, much effort has been made to increase ␤-cell mass by stimulating endogenous regeneration of islets or generating islet-like cells by in vitro differentiation procedures [2–5]. Multipotent stem cells within pancreatic islets and in nonendocrine compartments of the pancreas [6 –10] have been described to differentiate into pancreatic islet-like structures in vitro. Various tissue precursor cells, such as hepatic oval cells [11] and splenocytes [12], have been reported to differentiate into insulin-producing cells. In addition, embryonic stem (ES) cells and bone marrowderived stem cells (BMSC) are reported as an alternative cell source for obtaining pancreatic endocrine hormone-producing cells under appropriate in vitro and in vivo conditions [13–16].

Correspondence: Thatava Tayaramma, Ph.D., German Research Center for Biotechnology, Department of Gene Regulation and Differentiation, Mascheroder weg-1, D-38124, Braunschweig, Germany. Telephone: 49-531-6181-213; Fax: 49-531-6181-202; e-mail: [email protected] or [email protected] Received February 25, 2006; accepted for publication August 22, 2006; first published online in STEM CELLS EXPRESS September 21, 2006. ©AlphaMed Press 1066-5099/2006/$20.00/0 doi: 10.1634/stemcells.2006-0109

STEM CELLS 2006;24:2858 –2867 www.StemCells.com

Tayaramma, Ma, Rohde et al. We have evaluated the effect of TSA on the differentiation of BMSC into insulin-producing cells. We have found that treatment with TSA and subsequent cell culture conditions can induce the differentiation of BMSC into insulin-producing cells and other cell types typical of endocrine pancreas. Here we show that differentiated bone marrow cells express transcription factors specific for development of endocrine pancreas, such as PDX-1, hepatocyte nuclear factor (HNF)-3␤, and Pax 4. The bone marrow-derived islet-like cells secrete insulin in response to glucose stimulus. This is a novel and efficient method for the differentiation of bone marrow cells into insulin-producing cells within a short time period of 10 days. These cells could be a potentially unlimited source for the transplantation of insulinproducing cells into Type 1 diabetes patients.

MATERIALS

AND

METHODS

2859

HBSS gradient. Islets were aspirated from the Ficoll/HBSS interface and washed with cold HBSS. Islet cell viability was confirmed by trypan blue exclusion.

Dithizone Staining Dithizone (DTZ) (Merck & Co., Whitehouse Station, NY, http:// www.merck.com) stock solution was prepared as previously reported [22] by dissolving 50 mg of DTZ in 5 ml of dimethyl sulfoxide, was sterile-filtered through a 0.2 ␮m nylon filter, and was stored at ⫺20°C. In vitro DTZ staining was performed by adding 10 ␮l of stock solution to 1 ml of culture medium. The culture dishes were then incubated at 37°C for 30 min in DTZcontaining medium. After the dishes were rinsed three times with HBSS, crimson-red-stained clusters were examined with a stereomicroscope, and the number of DTZ-stained cells in cultures were determined.

Mice

Fluorescence Labeling with Newport Green DCF

C57BL/6 female mice of 6 –10-weeks old were purchased from the Harlan Winklemann (GmbH, Borchen, Deutschland, http:// www.harlan-winkelmann.de) and maintained in a specific pathogen-free facility until used. Mice were treated in accordance with institute guidelines.

The ester form of Newport Green (Molecular Probes Europe; Leiden, The Netherlands, http://probes.invitrogen.com) (diacetate form, NG-Ac) is cell-permeable. Inside the cell, this ester is cleaved by esterases to yield a cell-impermeant fluorescent indicator able to bind zinc ions. Cells in HBSS were exposed for 30 min at 37°C to 25 mM NG-Ac and 1.5 ␮l/ml Pluronic F127 (Molecular Probes) to enhance the penetration of the probe. After being washed in HBSS, the cells were analyzed by confocal microscopy.

BMSC Isolation and In Vitro Differentiation Conditions Bone marrow (BM) was flushed out with Dulbecco’s modified Eagle’s medium (DMEM) (MP Biomedicals, Germany, http:// www.mpbio.com) in the presence of 10% fetal bovine serum from the tibias and femurs of C57BL/6 mice. The BM cells were plated at a density of 1 ⫻ 106 cells/cm2 in 6-well plate in normal DMEM containing 10% fetal bovine serum, 100 U/ml penicillin, 100 ␮g/ml streptomycin, and 100 U/ml glutamine and allowed to adhere to the cell culture plate. After 12 hours of incubation, nonadherent cells were washed out with fresh serum-free DMEM medium. The cells were treated with TSA (kindly donated by Prof. Dr. Ju¨rgen Bode, GBF, Braunschweig) at a concentration of 55 nM in serum-free DMEM for 3 days. For the induction of differentiation, TSA was withdrawn from the medium, and the cells were then cultured for an additional 7 days in specific induction medium containing a 1:1 ratio of DMEM:DMEM/F12 (Gibco, Grand Island, NY, http://www. invitrogen.com) supplemented with 10% fetal bovine serum and 10 nM glucagon-like peptide-1 (GLP-1), in presence of high (25 mM) glucose.

Islet Isolation from Pancreas Mouse pancreatic islets were isolated as described previously [21]. Briefly, 6 –10-week-old female C57BL/6 mice were killed by cervical dislocation. The pancreata were removed, without ductal injection or distention, cut into small pieces, and washed three times with ice-cold Hanks’ balanced salt solution (HBSS) to remove released trypsin. The resulting pieces were suspended in 5 ml of HBSS, containing 2,500 U of collagenase Type 2 (Worthington Biochemical Corporation, Lakewood, NJ, http:// www.worthington-biochem.com) and digested with shaking at 37°C for 30 min. The digested sample was washed three times with cold HBSS. Islet purification was performed by centrifugation at 800g for 15 min at 4°C on a discontinuous Ficoll (Sigma, St. Louis, http://www.sigma-aldrich.com) (1.08 mg/ml)/ www.StemCells.com

Immunocytochemistry Immunocytochemistry was performed to detect the expression of insulin and other pancreas-specific hormones in differentiated cells. Cells cultured in Lab-Tek chambered slides (Nunc International, Rochester, NY, http://www.nuncbrand.com) were fixed with 4% paraformaldehyde in phosphatebuffered saline processed for immunofluorescence microscopy. The primary antibodies used for analysis were rabbit anti-insulin (sc-9168, Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com), goat anti-glucagon (C-18) (sc7779 Santa Cruz Biotechnology), rabbit anti-somatostatin (FL-116) (catalog number sc-20999 Santa Cruz Biotechnology), and goat anti-rat-C-peptide (Linco Research, catalog number 4023-01), goat monoclonal anti-insulin A (C-12), (catalog number sc-7839 Santa Cruz Biotechnology), rabbit anti-Ngn3 (kind gift from Prof. Michael S. German, UCSF Diabetes Center). The following secondary antibodies were used according to the manufacturer’s recommendations: antigoat IgG (whole molecule-Cy3 conjugate (catalog number C 2821, Sigma), anti-rabbit IgG (whole molecule)-Cy3 conjugated (catalog number C2306, Sigma), donkey anti-rabbit IgG (H⫹L) labeled with Alexa Fluor 488 (catalog number A21206, Molecular Probes Inc., Eugene, OR, http://probes. invitrogen.com), and donkey anti-goat IgG-rhodamine (sc2094, Santa Cruz Biotechnology). Goat anti-rabbit IgG, Fc fragment-specific (Jackson ImmunoResearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com, code number 111-001-008), and rabbit anti-goat IgG (Nordic Immunological Laboratories, The Netherlands, http://www. nordiclabs.nl/index.htm) were used as isotype controls. Paraffin sections of adult mouse pancreas were taken as positive controls for the pancreatic hormones insulin, glucagon, and somatostatin. Paraffin sections from 16.5-day mouse em-

Murine Bone Marrow-Derived Insulin-Producing Cells

2860

Table 1. Primers for polymerase chain reaction Sr. no

Gene

1

Glut-2

2

IAPP

3

Isl-1

4

Ngn3

5

Pax 4

6

Pdx-1

7

HNF-3␤

8

Proinsulin I

9

Proinsulin II

10

Glucagon

11

Pancreatic polypeptide

12

Somatostatin

13

M-Sur1

14

␤-Actin

Primer sequence 5⬘-ACAGAGCTACAATGCAACGTGG 5⬘-CAACCAGAATGCCAATGACGAT 5⬘-TGATATTGCTGCCTCGGACC 5⬘-GGAGGACTGGACCAAGGTTG 5⬘-GTTTGTACGGGATCAAATGC 5⬘-ATGCTGCGTTTCTTGTCCTT 5⬘-TGGCGCCTCATCCCTTGGATG 5⬘-AGTCACCCACTTCTGCTTCG 5⬘-ACCAGAGCTTGCACTGGACT 5⬘-CCCATTTCAGCTTCTCTTGC 5⬘-TGGATAAGGGAATTGCTTAACCT 5⬘-TTGGAACGCTCAAGTTTGTA 5⬘-AGAAGCAACTGGCACTGAAGGA 5⬘-GTAGTGCATGACCTGTTCGTAG 5⬘-GTTGGTGCACTTCCTACCCCTG 5⬘-GTAGAGGGAGCAGATGCTGGTG 5⬘-GTGGATGCGCTTCCTGCCCCTG 5⬘-GTAGAGGGAGCAGATGCTGGTG 5⬘-CACTCACAGGGCACATTCACC 5⬘-ACCAGCCACGCAATGAATTCCTT 5⬘-CTGCCTCTCCCTGTTTCTC 5⬘-GGCTGAAGACAAGAGAGGC 5⬘-GACCTGCGAACTAGACTGAC 5⬘-TTTGGGGGAGAGGGATCAG 5⬘-CCAGACCAAGGGAAGATTCA 5⬘-GTCCTGTAGGATGATGGACA 5⬘-GTCGTCGACAACGGCTCCGGCATGTG TG 5⬘-CATTGTAGAAGGTGTGGTGCCAGATC

Length of the gene

Annealing temperature

No. of cycles

221 bp

53

35

233 bp

65

35

514 bp

50.2

35

157 bp

60

35

236 bp

62

35

582 bp

53

35

464 bp

60

35

300 bp

60

35

300 bp

60

35

221 bp

52

35

337 bp

52

35

294 bp

52.2

35

236 bp

53

35

220 bp

55

35

Abbreviations: bp, base pair; HNF, hepatocyte nuclear factor; Ngn3, neurogenin 3.

bryos were taken as positive controls for Ngn3. Stained cells were analyzed by using confocal laser-scanning microscope (LSM META510 confocal scanning laser system consisting of an Axiovert 200 M microscope).

Glucose-Regulated Insulin Secretion To estimate secreted insulin levels, differentiated bone marrow cells were preincubated for 3 hours at 37°C in Krebs Ringerbicarbonate Hepes (KRBH) buffer containing 118 mM sodium chloride, 4.7 mM potassium chloride, 1.1 mM potassium dihydrogen phosphate, 25 mM sodium hydrogen carbonate, 3.4 mM calcium chloride, 2.5 mM magnesium sulfate, 10 mM Hepes, and 2 mg/ml bovine serum albumin (BSA) supplemented with 3.8 mM glucose. For high-glucose-induced insulin release, cells were further incubated in KRBH buffer supplemented with 7, 12, and 27.7 mM glucose, and 10 ␮M tolbutamide (Sigma) for 2 hours at 37°C. The control was incubated with 3.8 mM glucose. Determination of secreted insulin was performed by using an ultrasensitive mouse insulin enzyme-linked immunosorbent assay (ELISA) kit (Mercodia, Uppsala, Sweden, http://www.mercodia.com), according to the manufacturer’s instructions. Statistical analysis was performed by Student’s t-test.

Immunoprecipitation and Western Blotting Differentiated cells were assayed for the presence of intracellular insulin. The intracellular insulin was detected by cell extraction with lysis buffer and a combination of immunoprecipitation and Western blotting as described by Yang et al. [11]. Specif-

ically, the presence of insulin in the differentiated cell lysate was determined by immunoprecipitation with the rabbit polyclonal anti-insulin antibody (sc-9168), followed by separation on 18% SDS-polyacrylamide gel, transfer to nylon membranes, and blotting with anti-insulin antibody. Cell lysate (200 ␮g) was subjected to immunoprecipitation. Insulin protein was visualized by chemiluminescence.

RNA Extraction and Reverse TranscriptionPolymerase Chain Reaction Analysis Total cellular RNA was extracted from whole bone marrow and differentiated cells by using TRIzol (Gibco/BRL), and reverse transcription was performed with 12–18 primer oligo(dT) (Invitrogen, Carlsbad, CA, http://www.invitrogen. com). cDNA samples were subjected to polymerase chain reaction (PCR) amplification with FastStart Taq DNA Polymerase (Roche Diagnostics, Basel, Switzerland, http://www. roche-applied-science.com) in 1.5 mmol/l magnesium chloride and 0.2 mmol/l dNTPs. The following transcription factors involved in endoderm development: PDX-1, Pax-4, and Ngn-3. Pancreas-specific genes, such as islet amyloid polypeptide, isl-1, insulin gene I and II, and glucose transporter two (GLUT-2) were analyzed by PCR. The cycling parameters were as follows: denaturation at 94°C for 4 min, annealing at 50ºC– 64°C for 1 min (depending on the primer), and elongation at 72°C for 1 min (35 cycles). (Primers are listed in Table 1). PCR products of insulin gene I, insulin gene II, and PDX-1 were sequenced.

Tayaramma, Ma, Rohde et al.

2861

Electron Microscopy For the postembedding detection of insulin in differentiated cells, samples were fixed with 0.2% glutaraldehyde and 0.5% formaldehyde for 1 hour on ice, washed with PBS containing 10 mM glycine, and dehydrated with a graded series of ethanol following the progressive lowering of temperature protocol (PLT method). Samples were then embedded in Lowicryl K4M (catalog number 15,923, Polysciences, Inc., Warrington, PA, http://www.polysciences.com) at ⫺35°C, polymerized with UV light (366 nm, 2 days at ⫺35°C, 2 days at room temperature), and ultrathin sections were cut with glass knives and collected onto formvar-coated copper grids (300 mesh). Sections were incubated with rabbit polyclonal anti-insulin antibody (sc-9168), (100 ␮g IgG protein/ml) overnight at 4°C, washed with PBS, and incubated with gold markers (protein A/G gold, 15 nm in diameter from Biocell Cardiff, U.K., http://www. biocell.com) for 30 min at room temperature. After being washed with PBS containing 0.01% Triton X-100 and then with distilled water, samples were air-dried. Counterstaining was done with 4% uranyl acetate for 1 min. Samples were then examined in a Zeiss transmission electron microscope EM910 (Zeiss, Oberkochen, Germany, http://www.zeiss.com) at an acceleration voltage of 80 kV.

RESULTS Chromatin-Remodeling Factors Induce Bone Marrow Cells to Differentiate into Islet-Like Cell Clusters To evaluate the effect of chromatin-remodeling factor TSA in the differentiation of BMSC into insulin-producing cells, murine bone marrow cells were cultured in the presence of TSA for 3 days and then maintained for an additional 7 days in differentiation induction medium containing 10 nM GLP-1 and high (25 mM) glucose. After removal of TSA, cells cultured under highglucose medium started forming spheroid-like cell clusters from day 5 and attained their maximum size and maximal number at day 10. Bone marrow cells treated with TSA in the presence of high glucose formed islet-like clusters; this was not seen in control bone marrow cells cultured in the absence of TSA and high glucose (Fig. 1A). Cells cultured in the presence of TSA for 3 days had normal cell morphology (Fig. 1B). The threedimensional cellular morphology of the cell clusters after 10 days culturing resembled pancreatic islet-like clusters (Fig. 1C, 1D), as described [6, 11]. The number of clusters formed was 300 –350 per well in 6-well culture plates. The observed cellular morphology showed that bone marrow cells could differentiate into islet-like cells in the presence of TSA followed by culture in high-glucose medium.

Detection of Insulin-Containing Cells in Islet-Like Clusters by Using Zinc-Chelating Dyes To evaluate the insulin-producing cells in cultures, we stained the differentiated cells with the zinc-chelating agent DTZ and zinc-dependent fluorescence indicator Newport green (NG-Ac). Zinc is an integral part of insulin crystals forming the 2-Zninsulin hexamer and is required by pancreatic ␤-cells for packaging insulin. Free ionized zinc is present in the extragranular space as a reservoir for granular zinc. We took advantage of these zinc pools to identify cells harboring insulin synthesis in www.StemCells.com

Figure 1. Chromatin-remodeling factors induce bone marrow cells to differentiate into islet-like cell clusters. Bone marrow (BM) cells were treated in serum-free medium for 3 days with histone deacetylase inhibitor trichostatin A (TSA) and then cultured in 10% fetal bovine serum, glucagon-like peptide-1, and high glucose (25 mM) for an additional 7 days. BM cells were cultured for 10 days without any specific treatment (A). BM cells were cultured for 3 days in the presence of TSA (B). BM cells cultured for 10 days (3 ⫹ 7), 3 days in presence of TSA and subsequently cultured for an additional 7 days in highglucose medium formed islet like clusters (⫻10) (C). At higher magnification, a single islet-like cluster appears to have defined edges and structure (⫻40) (D). Scale bar for (A–C): 100 ␮m and (D): 50 ␮m.

our cultures. DTZ is a zinc-binding substance, and pancreatic islets are known to stain crimson red after DTZ treatment. We first determined whether the isolated pancreatic islets from mouse were stained with DTZ and found that most islet cells were stained crimson red (data not shown). Differentiated individual cells and, in particular, cells in the islet-like clusters were distinctly stained crimson red by DTZ (Fig. 2A). Undifferentiated BMSC were not stained by DTZ (Fig. 2B). Newport Green (NG-Ac) ester stains living cells at enzymatic cleavage by cellular esterases and subsequent binding to Zn2⫹. Confocal images of differentiated cells stained with NG-Ac exhibited green fluorescence dots (Fig. 2C). Staining of individual cells revealed a heterogeneous intensity of fluorescence with cytoplasmic dots. The control culture did not show significant staining with NG-Ac (Fig. 2D). No nuclear staining was noted. The presence of positive cells for zinc-specific dyes such as DTZ and NG-Ac suggest that insulin-producing cells can be derived from BMSCs.

Immunofluorescence Analysis for the Detection of Pancreas-Specific Hormones To investigate the expression of pancreatic hormones, immunofluorescence analysis was performed for insulin, C-peptide, glucagon, and somatostatin in differentiated cells. Immunostaining of individual islet-like clusters, which were formed in BMSC cultures treated with TSA and subsequently cultured under high-glucose conditions, revealed large numbers of insulin-positive cells. Double-immunofluorescence analysis showed insulin (Fig. 3A, left), C-peptide (Fig. 3A, middle), and colabeling (Fig. 3A, right) in the same cell of islet-like clusters. Cells were counterstained by 4⬘,6-diamidino-2-phenylindole to

2862

Figure 2. Detection of insulin-containing cells in islet-like clusters. Insulin-containing cells in islet-like clusters were detected by using zinc-chelating agents, dithizone (DTZ), and Newport green (NG-Ac). On day 10, cells distinctly stained crimson-red by DTZ are apparent in the differentiated islet-like clusters (A). Undifferentiated BM cells are not stained (B). Individual differentiated cells stained with NG-Ac, showing green fluorescence within cytoplasmic dots (C). Control differentiated cells for NG-Ac stain (D). Scale bar for (A, B): 20 ␮m and for (C, D): 10 ␮m.

reveal the nucleus (Fig. 3A). The colocalization of C-peptide demonstrated de novo insulin synthesis and excluded the possibility that cells only absorbed and concentrated insulin from the medium. Glucagon- and somatostatin-expressing cells were also present in the culture. The staining for glucagon was found in the periphery of the cell (Fig. 3B), whereas somatostatin was dispersed in the cytoplasm (Fig. 3C). Costaining for insulin (Fig. 3D, left) and somatostatin (Fig. 3D, middle) showed that somatostatin-expressing cells were fewer in number than insulinexpressing cells in culture (Fig. 3D, right). Transcription factor Ngn3 was distinctly present in the nucleus of differentiated islet-like cells (Fig. 3E, middle and right). Matched isotype controls for rabbit IgG (Fig. 3F) and goat IgG (Fig. 3G) antibodies and without primary antibody (Fig. 3H) served as negative controls. These results indicate that (a) differentiated BMSC synthesize insulin de novo as indicated by the presence of C-peptide, (b) other pancreatic hormone-producing cells are also present in the culture after 10 days, and (c) the transcription factor Ngn3 is expressed by the islet-like cells. Adult mouse pancreas was used as a positive control for the hormones insulin, C-peptide, glucagon, and somatostatin (see supplemental materials). The transcription factor Ngn-3 was only detected at 16.5 days in mouse embryos (see supplemental materials), also as a positive control.

Protein Analysis of Islet-Like Clusters To determine whether the differentiated BM cells were responsive to glucose challenge, insulin release upon exposure to high glucose was measured by using an ultrasensitive mouse insulin

Murine Bone Marrow-Derived Insulin-Producing Cells ELISA. To enhance the sensitivity of these cells to the highglucose challenge, the differentiated cells were switched to KRBH buffer containing 0.5% BSA and incubated in the presence of 3.8 mM glucose for 3 hours, then stimulated by the addition of 7, 12, or 27.7 mM glucose to the medium for 2 hours in individual experiments. Bone marrow-derived islet-like cells, after exposure to high glucose, secreted insulin in a glucose concentration-dependent manner (Fig. 4A). These data demonstrate that islet-like cell clusters derived from BMSC can secrete insulin in a glucose-regulated manner under appropriate conditions. To analyze the synthesized and stored insulin in differentiated clusters, cell lysates of day 3 culture and day 10 cultures after TSA treatment were subjected to immunoprecipitation and Western blot analysis. Cell lysates of differentiated bone marrow cells contained stored insulin. The cell lysate of day 3 cultures contained an immunopositive band, and the band intensity increased in cell lysate at day 10 (Fig. 4B). Pancreatic tissue served as a positive control. In contrast, no corresponding band was detectable in untreated BMSC (Fig. 4B). The data of insulin protein analysis suggest that the differentiated BMSC in islet-like clusters synthesize and store insulin.

Endocrine-Specific Gene Expression in Differentiated Islet-Like Clusters Pancreatic development and gene expression are regulated by specialized transcription factors. To determine whether endocrine-specific transcription factors and pancreas-specific genes were expressed during the differentiation into islet-like clusters, reverse transcription-polymerase chain reaction (RT-PCR) analysis was performed at different time points. Transcripts for PDX-1, Pax 4, HNF-3␤, and Isl-1 were not detectable in undifferentiated BMSC; in contrast, they were upregulated in differentiated cells (Fig. 5). PDX-1 and HNF-3␤ transcripts were expressed on day 3 after TSA treatment and increased by day 10. The PCR product of neurogenin 3 (Ngn-3) was visible in both undifferentiated and differentiated BMSCs. PCR products for insulin gene I and II were visible at day 3 and day 10. In contrast, other major islet-specific hormones, such as IAPP, glucagon, and somatostatin, were not detected on day 3 of differentiation and were expressed only on day 10. PCR products for the ATP-sensitive potassium channel-specific sulfonylurea receptor (SUR1) and glucose transporter gene (GLUT-2) were detectable only at day 10. Gene expression analysis in BMSC derived islet-like clusters was similar to that in the pancreas tissue and confirmed the differentiation of BMSC into islet-like cells in vitro upon treatment with TSA and subsequent culturing in high glucose.

Ultrastructural Analysis of Insulin-Producing Cell Clusters Ultrastructural analysis of bone marrow derived insulin-producing clusters was also performed. Differentiated cells at low magnification revealed a structure typical of a secretory cell, with secretory vesicles containing dense granules (Fig. 6A). Immunogold electron microscopy showed insulin within the small secretary vesicles of the insulin-producing clusters (Fig. 6B). Gold-labeling detected faint globular structures of differing size filled with low-density material. Positive and negative controls for immunogold labeling are presented in Figure 6C


2863

Figure 3. Immunofluorescence analysis for detection of pancreas-specific hormones. Confocal microscopy of immunoassaying for insulin, C-peptide, glucagon, and somatostatin on day 10 islet-like clusters. Double-immunofluorescence analysis revealed insulin (green) (A), C-peptide (red), nuclear staining with DAPI (cyan); merging of both channels green (insulin) and red (C-peptide) was seen as yellow. Glucagon-positive cells (red) (B) and somatostatin-positive cells (red) (C) are present in the cultures. Costaining of insulin in green and somatostatin (D) in red was observed. Transcription factor Ngn3 was present in the nucleus of differentiated cells (E). In the isotype control for rabbit IgG (Isotype 1 [F]) and goat IgG (Isotype 2 [G]) and in the negative control (no primary antibody [H]), no immunostaining is observed; all were counterstained by using DAPI (cyan) for nuclear staining. Scale bar 10 ␮m. Abbreviations: DAPI, 4⬘,6-diamidino-2-phenylindole; Ngn3, neurogenin 3.

www.StemCells.com

2864

Murine Bone Marrow-Derived Insulin-Producing Cells

Figure 4. Protein analysis of bone marrow derived islet-like clusters. Analysis of secretion of insulin following high glucose challenge of differentiated BM-derived islet-like cultures. (A): Enzyme-linked immunosorbent assay for insulin in Krebs Ringer-bicarbonate Hepes (KRBH) buffer from BM-derived islet-like clusters exposed to low- (3.8 mM) and high-glucose conditions 7, 12, and 27 mM for 2 hours. After exposure to high glucose, BM-derived islet-like clusters secreted insulin. Statistical significance tested by Student’s t-test: *, p ⬍ .05; **, p ⬍ .01; ***, p ⬍ .001. (B): Western blot analysis for insulin was performed for differentiated islet-like clusters following collection of cell lysates. Note the stronger band for insulin at day 10 after culturing in the presence of glucagon-like peptide-1. Abbreviations: BM, bone marrow; D, day.

and 6D, respectively. Adult pancreatic ␤-cells showed several positive immunogold particles per secretory granule (Fig. 6C). The pancreatic ␤-cells were examined only to validate the assay and not to correlate insulin-positive signals between the ␤-cells and the BM-derived insulin-producing clusters. Control bone marrow cells showed no specific labeling, but a few nonspecific particles were seen in the cytoplasm (Fig. 6D). The ultrastructural study of differentiated bone marrow cells showed features typical of an adult ␤-cell, and insulin granules were observed within the secretory vesicles suggesting that these cells had differentiated into pancreatic ␤-like cells capable of synthesizing insulin.

DISCUSSION

Stem cells are a potential source for ␤-cell replacement therapy. Fundamental processes in the determination of the differentiation pathways of stem cells remain to be elucidated and robust and reliable differentiation protocols need to be established. ES cells can be differentiated into insulin-producing cells by manipulating culture conditions [23, 24]. Rajagopal et al. have claimed that ES cells merely absorb insulin from the medium [25, 26]. However, the insulin detected by immunocytochemistry in the differentiated cells must be actively synthesized and secreted, rather than merely being taken up passively from the medium (see below). Nevertheless, many problems in the control of differentiation and teratoma formation in insulin-producing cells derived from ES cells remain to be overcome. Bone marrow trans-

Figure 5. Endocrine-specific gene expression in differentiated isletlike clusters. Reverse transcription-polymerase chain reaction (RTPCR) analysis of pancreas-specific gene expression at several differentiation stages. Total RNA isolated from both undifferentiated and differentiated bone marrow cells was subjected to RT-PCR analysis with primers for the indicated genes. Lane BM: undifferentiated whole bone marrow. Lane D3: day 3 culture with trichostatin A only. Lane D10: day 10 culture with high glucose and glucagon-like peptide-1. Lane Pancreas: adult mouse pancreas (positive control). Lane -ve RT: No template (negative control). Abbreviations: BM, bone marrow; D, day.

plantation generates islet cells in recipient mice. These cells express insulin and genetic markers of ␤-cells [27, 28]. Few reports have shown that bone marrow cells can differentiate in vitro under controlled conditions into insulin-expressing cells. Such cells transplanted under the kidney capsule of diabetic animals have been demonstrated to regulate glucose levels in the blood [29]. ES cells contain potent chromatin-remodeling activities; observations suggest that chromatin dynamics may be especially important for early lineage decisions. Chromatin dynamics are also involved in the differentiation of adult stem cells. For a better understanding of stem cell differentiation, chromatin dynamics should be considered. The reprogramming of bone mar-


2865

Figure 6. Ultrastructural analysis of insulin-producing cell clusters. Postembedding immunogold staining for insulin in day 10 islet-like clusters. Secretory granules (arrow) are densely packed within the cytoplasm of the differentiated cell (A). At higher magnification, insulin granules (arrows) are seen in secretory vesicles (B). (C): Positive control for immunogold labeling: positive granules (arrows) in adult mouse pancreas. (D): Negative control. Abbreviation: N, nucleus.

row cells to insulin-producing cells may depend on chromatin modulation. To evaluate whether chromatin reprogramming can contribute to the in vitro differentiation of BM cells into insulinproducing cells, we have added the histone deacetylase inhibitor (HDACi), TSA, to the culture medium. TSA is the best-studied HDACi [30]. TSA has been reported to suppress the growth of pancreatic adenocarcinoma cells [31]. Human hematopoietic stem cells (HSCs) and progenitor cells (HPCs), when exposed to 5-aza-2⬘-deoxycytidine and TSA in vitro, show significant expansion of a subset of CD34 cells [32]. It was reported at the site of insulin one promoter in pancreatic ␤-cells, lysine acetylation of histone H3 is required for establishment of an open chromatin structure [33]. Lumelsky et al. also presented in a recent publication, the level of histone H3 acetylation at the site of insulin one gene promoter was significantly higher in long-term proliferating islet progenitor-like cells [34]. We have explored the possibility of using mouse BM cells as a source for insulin-producing cells following treatment with TSA and subsequently culturing in the presence of high glucose and GLP-1. Glucose is well known as a growth factor for ␤-cells [35] and promotes ␤-cell replication in vitro and in vivo at concentrations of 20 –30 mmol/l [36]. GLP-1 is an incretin hormone capable of converting intestinal epithelial cells into functional insulin-producing cells [37]. We have generated functional insulin-producing cells from bone marrow and confirmed the presence of insulin production by RT-PCR, immunofluorescence, Western blot, and electron microscopy combined with immunogold anti-insulin labeling. Furthermore, we have tested the functionality of the in vitro generated insulinproducing cells from BM by measuring insulin release in response to high glucose concentrations. Taken collectively, these studies provide evidence that the BM contains pluripotent cells capable of being reprogrammed in vitro by TSA to become functional insulin-producing cells. In addition, we have found transcripts for proinsulin I and II by RT-PCR. Transcripts for the component of the ATPsensitive K⫹ channel SUR1 and for GLUT-2, which participates in the signal-mediated secretion of insulin in pancrewww.StemCells.com

atic ␤-cells, have also been detected. Moreover, transcription factors Isl 1, Pax 4, Ngn-3, and IAPP are present at later stages of differentiation. Pax4 is required for the development of cells restricted to the ␤- and ␦-cell lineages. Mice lacking Pax4 fail to develop any ␤-cells and become diabetic [38]. We have also detected the expression of glucagon and somatostatin by PCR, and the respective proteins have been detected by immunocytochemistry. We have found fewer somatostatin-positive and glucagon-positive cells in culture by single-staining immunocytochemistry in contrast to the large number of insulin-positive cells in culture. Double-staining for insulin and somatostatin has confirmed the low numbers of somatostatin-containing cells. PDX-1 gene expression is low at earlier stages of differentiation; PDX-1 expression at these stages may initiate a cascade of events leading to insulin transcription. Although at later stages, high levels of PDX-1 transcript are found, we have failed to detect PDX-1 by immunocytochemistry. This is not surprising, because it is known that when the mRNA for PDX-1 is highly abundant, immunoreactivity can hardly be detected [39]. During embryogenesis, only cells expressing Ngn-3 are islet progenitors [40]. Indeed, we have revealed the expression of the Ngn-3 transcription factor exclusively in the nucleus of the islet-like cells. Other reports [41] have shown that cells coexpressing glucagon, insulin, and Ngn-3 eventually become mature ␤-cells. The islet-like clusters in our culture system express insulin, glucagon, and Ngn-3 and thus strongly resemble pancreatic precursor cells. This has been confirmed by our immunocytochemical results. The C-peptide expression in islet-like cells and glucosedependent insulin release provide evidence that these cells synthesize and release insulin. C-peptide is a by-product of insulin synthesis, and therefore the demonstration of C-peptide and the colocalization of insulin are reliable methods to investigate pancreatic differentiation in vitro. In our study, BM cell differentiate into cells coexpressing insulin and C-peptide; this rules out the uptake of insulin from the culture medium, as mentioned

2866

above. The data confirm that BM cells are capable of differentiating into pancreatic cells producing insulin de novo. This secretion of insulin seems to be dependent upon the concentration of glucose in the medium; however, osmotic effects cannot be ruled out, although Lumelsky et al. [14] have found no such effects with sucrose. In summary, we have shown here that murine BM cells are reprogrammed into insulin-producing cells by treatment with the histone deacetylase inhibition factor, TSA, and following culture in a high-glucose medium. However, further research has to be carried out on chromatin-remodeling factors to understand the mechanisms that are involved in cell-fate determination. The insulin-producing cells secrete a substantial amount of insulin, and immunostaining and RT-PCR analysis have re-

REFERENCES 1

Shapiro AM, Lakey JR, Ryan EA et al. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid free immunosuppressive regimen. N Engl J Med 2000;343:230 –238.

2

Bonner-Weir S, Taneja M, Weir GC et al. In vitro cultivation of human islets from expanded ductal tissue. Proc Natl Acad Sci U S A 2000;97: 7999 – 8004.

3

Abraham EJ, Leech CA, Lin JC et al. Insulinotropic hormone glucagon-like peptide-1 differentiation of human pancreatic islet-derived progenitor cells into insulin producing cells. Endocrinology 2002;143:3152–3161.

4

Schmied BM, Ulrich A, Matsuzaki H et al. Transdifferentiation of human islet cells in a long-term culture. Pancreas 2001;23:157–171.

5

Lipsett M, Finegood DT. Beta-cell neogenesis during prolonged hyperglycemia in rats. Diabetes 2002;51:1834 –1841.

6

Ramiya VK, Maraist M, Arfors KE et al. Reversal of insulin-dependent diabetes using islets generated in vitro from pancreatic stem cells. Nat Med 2000;6:278 –282.

7

Schwitzgebel VM, Scheel DW, Conners JR et al. Expression of neurogenin3 reveals an islet cell precursor population in the pancreas. Development 2000;127:3533–3542.

8

Zulewski H, Abraham EJ, Gerlach MJ et al. Multipotential nestinpositive stem cells isolated from adult pancreatic islets differentiate ex vivo into pancreatic endocrine, exocrine, and hepatic phenotypes. Diabetes 2001;50:521–533.

9

Gao R, Ustinov J, Pulkkinen MA et al. Characterization of endocrine progenitor cells and critical factors for their differentiation in human adult pancreatic cell culture. Diabetes 2003;52:2007–2015.

10 Hardikar AA, Marcus-Samuels B, Geras-Raaka E et al. Human pancreatic precursor cells secrete FGF2 to stimulate clustering into hormoneexpressing islet-like cell aggregates. Proc Natl Acad Sci U S A 2003; 100:7117–7122. 11 Yang L, Li S, Hatch H et al. In vitro trans-differentiation of adult hepatic stem cells into pancreatic endocrine hormone producing cells. Proc Natl Acad Sci U S A 2002;99:8078 – 8083. 12 Kodama S, Kuhtreiber W, Fujimura S et al. Islet regeneration during the reversal of autoimmune diabetes in NOD mice. Science 2003;302: 1223–1227. 13 Soria B, Roche E, Berna G et al. Insulin-secreting cells derived from embryonic stem cells normalize glycemia in streptozotocin-induced diabetic mice. Diabetes 2000;49:157–162. 14 Lumelsky N, Blondel O, Laeng P et al. Differentiation of embryonic stem cells to insulin-secreting structures similar to pancreatic islets. Science 2001;292:1389 –1394.

Murine Bone Marrow-Derived Insulin-Producing Cells vealed that islet-like cells derived from bone marrow are similar to pancreatic cells. The development of new protocols to differentiate BM to pancreatic ␤-like cells may enable the use of these cells for future cell therapy of type 1 diabetes.

ACKNOWLEDGMENTS The work was supported by a Grant MA 852/7-3 from Deutsche Forschungsgemeinschaft (DFG). We thank P. Paul Mu¨ller and Dr. Theresa Jones for critically reading the manuscript and Prof. Michael S. German UCSF Diabetes Center, CA for providing aliquots of Ngn3 antibody.

DISCLOSURES The authors indicate no potential conflicts of interest.

17 Richon VM, O’Brien JP. Histone deacetylase inhibitors: a new class of potential therapeutic agents for cancer treatment. Clin Cancer Res 2002; 8:662– 664. 18 Marks P, Rifkind RA, Richon VM et al. Histone deacetylases and cancer: causes and therapies. Nat Rev Cancer 2001;1:194 –202. 19 Yoshida M, Nomura S, Beppu T. Effects of trichostatins on differentiation of murine erythroleukemia cells. Cancer Res 1987;47:3688 –3691. 20 Yoshida M, Kijima M, Akita M et al. Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A. J Biol Chem 1990;265:17174 –17179. 21 Gotoh M, Maki T, Kiyoizumi T et al. An improved method for isolation of mouse pancreatic islets. Transplantation 1985;40:437– 438. 22 Shiroi A, Yoshikawa M, Yokota H et al. Identification of insulinproducing cells derived from embryonic stem cells by zinc-chelating dithizone. STEM CELLS 2002;20:284 –292. 23 Leo´n-Quinto T, Jones J, Skoudy A et al. In vitro directed differentiation of mouse embryonic stem cells into insulin-producing cells. Diabetologia 2004;47:1442–1451. 24 Blyszczuk P, Czyz J, Kania G et al. Expression of Pax4 in embryonic stem cells promotes differentiation of nestin-positive progenitor and insulin-producing cells. Proc Natl Acad Sci U S A 2003;100:998 –1003. 25 Rajagopal J, Anderson WJ, Kume S et al. Insulin staining of ES cell progeny from insulin uptake. Science 2003;299: 363. 26 Hansson M, Tonning A, Frandsen U et al. Artifactual insulin release from differentiated embryonic stem cells. Diabetes 2004;53:2603–2609. 27 Ianus A, Holz GG, Theise ND et al. In vivo derivation of glucosecompetent pancreatic endocrine cells from bone marrow without evidence of cell fusion. J. Clin. Invest. 2003;111:843– 850. 28 Hess D, Li L, Martin M et al. Bone marrow-derived stem cells initiate pancreatic regeneration. Nat Biotechnol. 2003;21:763–770. 29 Tang DQ, Cao LZ, Brant R et al. In vivo and in vitro characterization of insulin-producing cells obtained from murine bone marrow. Diabetes 2004;53:1721–1732. 30 Xiao H, Hasegawa T, Isobe K. Both sp1 and sp3 are responsible for p21w af1 promoter activity induced by histone deacetylases inhibitor in NIH3T3 cells. J Cell Biochem 1999;73:291–302. 31 Massimo D, Chiara C, Laura F et al. Trichostatin A, an inhibitor of histone deacetylases, strongly suppresses growth of pancreatic adenocarcinoma cells. Mol Carcinog 2003;38:59 – 69.

15 Assady S, Maor G, Amit M et al. Insulin production by human embryonic stem cells. Diabetes 2001;50:1691–1697.

32 Mohammed M, Nadim M, Donald L et al. Modification of hematopoietic stem cell fate by 5aza 2-deoxycytidine and trichostatin A. Blood 2004; 103:4102– 4110.

16 Oh SH, Muzzonigro TM, Bae SH et al. Adult bone marrow-derived cells trans-differentiating into insulin-producing cells for the treatment of type I diabetes. Lab Invest 2004;84:607– 617.

33 Chakrabarti SK, Francis J, Ziesmann SM et al. Covalent histone modifications underlie the developmental regulation of insulin gene transcription in pancreatic beta cells. J Biol Chem 2003;278:23617–23623.


2867

34 Ta M, Choi Y, Atouf F et al. The defined combination of growth factors controls generation of long-term-replicating islet progenitor-like cells from cultures of adult mouse pancreas. STEM CELLS 2006;24:1738–1749.

38 Sosa-Pineda B, Chowdhury K, Torres M et al. The Pax4 gene is essential for differentiation of insulin producing beta cells in the mammalian pancreas. Nature 1997;386:399 – 402.

35 Soria B. In-vitro differentiation of pancreatic beta cells. Differentiation 2001;68:205–219.

39 Stoffers DA, Heller RS, Miller CP et al. Developmental expression of the homeodomain protein IDX-1 in mice transgenic for an IDX-1 promoter/ lacZ transcriptional reporter. Endocrinology 1999;140:5374 –5381.

36 Bonner-Weir S, Deery D, Leahy JL et al. Compensatory growth of pancreatic ␤-cells in adult rats after short-term glucose infusion. Diabetes 1989;38:49 –53. 37 Suzuki A, Nakauchi H, Taniguchi H. Glucagon-like peptide 1 (1–37) converts intestinal epithelial cells into insulin-producing cells. Proc Natl Acad Sci U S A 2003;100:5034 –5039.

40 Gu G, Dubauskaite J, Melton DA. Direct for the evidence for the pancreatic lineage: NGN 3⫹ cells are islet progenitors and are distinct from duct progenitors. Development 2002;129:2447–2457. 41 Chiang MK, Melton DA. Single-cell transcript analysis of pancreas development. Dev Cell 2003;4:383–393.

See www.StemCells.com for supplemental material available online.

www.StemCells.com