DNA AND CELL BIOLOGY Volume 33, Number 9, 2014 ª Mary Ann Liebert, Inc. Pp. 591–598 DOI: 10.1089/dna.2014.2385
Involvement of Gene Methylation Changes in the Differentiation of Human Amniotic Epithelial Cells into Islet-Like Cell Clusters Lin Peng,1 Jian Wang,1,2 and Guangxiu Lu1,2
Insulin-dependent diabetes results from destruction of the insulin-producing b-cells of the pancreas. Islet cell transplantation is a promising cure for diabetes. Here, we induced human amniotic epithelial cells (hAECs) to differentiate into islet-like cell clusters by nicotinamide plus betacellulin in vitro, and further investigated the DNA methylation status by a Nimble MeDIP microarray before and after cell differentiation to shed light on the molecular mechanisms of this differentiation. In addition, 5-Aza-2¢deoxycytidine was used to investigate whether the differentiation of hAECs into islet-like cells occurred through demethylation. Purified hAECs (CK18 + /E-cadherin + /CD29 + /CD90 - /CD34 - /CD45 - ) were isolated from human amnia. After induction, hAECs were found to be insulin positive and sensitive to glucose, indicating successful induction to islet-like cells. The methylation status of cell cytoskeletonrelated genes was down-regulated and that of negative regulation of cell adhesion-related genes was upregulated. The methylation status of pancreas development-related genes such as HNF1a and DGAT1 was decreased in hAECs after induction. After brief demethylation, INS gene expression was up-regulated in islet-like cell clusters, suggesting that DNA methylation changes were associated with the differentiation of hAECs into islet-like cell clusters.
Introduction
D
iabetes currently affects more than 336 million people worldwide (Ashcroft and Rorsman, 2012). Type 1 diabetes is characterized by insulin deficiency due to the destruction of b-cells within pancreatic islets. Increasing the number of insulin-producing b-cells is a promising therapeutic avenue (Rastellini et al, 1997). Embryonic stem (ES) and induced pluripotent stem cells are considered the most promising cell types for clinical applications because of their pluripotent characteristics. However, the use of these cell types is limited by ethical and safety issues. Therefore, more suitable cell sources are desired to satisfy the needs of a clinical setting. Human amniotic epithelial cells (hAECs) are derived from the extra-embryonic tissues delivered at birth, which are typically disposed of as waste. Therefore, hAECs are readily accessible, their use is generally considered ethically sound, and, since hAECs are nontumorigenic when being transplanted into mice, they are considered safe (Moodley et al., 2010; Fang et al, 2012; Manuelpillai et al., 2012). In addition, hAECs have successfully been transdifferentiated into many other kinds of cells (Miki et al., 2005; Hou et al., 2008; Evron et al.,
1 2
2012); thus, hAECs may be considered a suitable cell source for clinical applications. Safety, therapeutic efficiency, and availability/sufficient quantity are the basic requirements of clinically applicable stem cells. However, it is still difficult to harvest abundant amounts of insulin secreting-cells by inducing hAECs to differentiate into pancreatic-related cells. A more thorough understanding of the mechanisms controlling this transdifferentiation is needed before hAECs can be considered for clinical applications. Recently, Gifford et al. (2013) and Xie et al. (2013) suggested that dynamic alterations in DNA methylation play an important role in ES cell differentiation. In general, DNA methylation is negatively associated with the activation status of genes (Albalat et al., 2012). Differentiation of stem cells into specialized cells requires upregulation of genes involved in creation of a specific cell phenotype, and suppression of genes responsible for pluripotency (Kouzarides, 2007). In the present study, we established an induction method using nicotinamide plus betacellulin to obtain three-dimensional (3D) islet-like cell clusters, and further investigated gene methylation changes of cells before and after induction by Roche-NimbleGen MeDIP analysis. These gene methylation changes may be
Institute of Human Reproduction and Stem Cell Engineering, Central South University, Changsha, People’s Republic of China. National Center of Human Stem Cell Research and Engineering, Changsha, People’s Republic of China.
591
592
PENG ET AL.
responsible, at least in part, for the transdifferentiation of hAECs to islet-like cell clusters.
cytometer (BD Biosciences) and analyzed on FACS Diva software (BD Biosciences).
Materials and Methods
Immunofluorescence analysis
hAEC isolation, identification, and induced differentiation to islet-like cell clusters
This study was approved by the local ethics committee of the Maternal and Child Health Hospital of Hunan Province, China. Amnia (n = 12, gestational age = 273 – 4 days) were retrieved from healthy women (excluding HIV and HBV virus infection) aged 28 – 3 years by elective cesarean delivery. Informed consent was obtained from each patient before amnion collection. To isolate the hAECs, amniotic membranes were manually stripped from the chorionic membranes of the placenta immediately after delivery, and amnia were rinsed in Dulbecco phosphate-buffered saline (DPBS; GIBCO) containing 100 U/mL penicillin/streptomycin solution (SigmaAldrich). Tissues were digested twice in 0.05% trypsin/ EDTA (T/E; GIBCO) for 15 min at 37C, then centrifuged at 230 g, and washed twice in Dulbecco modified Eagle medium (DMEM; Hyclone). The isolated cells were then plated in 100-mm cell culture dishes at a density of 0.5 · 105 cells/mL, and cultured in medium containing DMEM supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich) and 10 ng/mL epidermal growth factor (R&D Systems). In addition to hAECs, we also isolated human amniotic mesenchymal stem cells (hAMSCs). This step enabled the comparison of hAECs to hAMSCs in order to eliminate the possibility of mesenchymal cell contamination in our isolated hAECs. To isolate hAMSCs, the remaining amnia were digested overnight in dispase (2 mg/mL), collagenase I (2 mg/mL), and hyaluronic acid (2 mg/mL) (all from SigmaAldrich) at 37C to collect the hAMSCs. The isolated cells were then plated in 100-mm cell culture dishes at a density of 0.5 · 105 cells/mL in culture medium containing DMEM/ F12 (GIBCO) supplemented with 10% FBS and 10 ng/mL fibroblast growth factor-basic (Invitrogen). To induce differentiation, hAECs were passaged once and then plated at a density of 1.5 · 106 cells/petri dish (60-mm) in DMEM supplemented with 1 · N-2 Supplement (GIBCO), 10 mM nicotinamide (Sigma-Aldrich), and 10 ng/mL betacellulin (R&D Systems) for 14 days. The medium was changed every 3 days. Flow cytometry analysis
For analysis of membrane antigens, monoclonal antibodies against PE-conjugated SSEA4, PE-conjugated CD34, FITC-conjugated CD45, PE-conjugated CD29, and PEconjugated CD90 (all from Biolegend) were used, and a monoclonal insulin antibody (Sigma-Aldrich) and speciesappropriate FITC-conjugated secondary antibodies were used. Briefly, cells were digested into a single-cell suspension and then incubated with the monoclonal antibodies mentioned earlier for 45 min, respectively (Cells were incubated overnight with insulin monoclonal antibodies at 4C and then incubated with species-appropriate FITC-conjugated secondary antibodies at room temperature for 40 min protected from light.). Then, cells were washed with PBS twice before analysis. Data were collected using an FACS Aria flow
The isolated cells were immunostained to identify expression of cytokeratin 18 (CK18) and insulin. The process was as follows: First, cells were fixed by immersion in 4% paraformaldehyde (Sigma-Aldrich) at room temperature for 20 min, then permeabilized with 0.1% Triton X-100 (SigmaAldrich) for 25 min, and blocked at room temperature for 45 min (Blocking reagent; East Coast Bio). Second, cells were incubated overnight with a monoclonal CK18 antibody (Santa Cruz Biotechnology) or insulin antibody (Sigma-Aldrich) at 4C, washed twice with DPBS, and then incubated with species-appropriate FITC-, PE-, or CY3conjugated secondary antibodies at room temperature for 40 min protected from light. Cell nuclei were stained with DAPI. Cells were observed under a laser scanning microscope (Nikon TE-2000u) at the corresponding wavelength of FITC, PE, Cy3 (n = 3). RT-PCR and quantitative real-time PCR analysis
Reverse transcription-polymerase chain reaction (RT-PCR) and quantitative real-time PCR (RT-qPCR) were performed to compare transcript levels between freshly isolated hAECs (P0); hAECs before and after induction (n = 3); and hAECs after induction in the presence or absence of 5-Aza-2¢-deoxycytidine (Aza; Sigma-Aldrich; n = 3). Briefly, total RNA was isolated by Trizol reagent (Invitrogen) according to the manufacturer’s protocol, and 1 mg total RNA was reverse transcribed into cDNA using random primers (RevertAid First-Strand cDNA Synthesis Kit #K1622; Formentas). The oligonucleotide primers used for RT-PCR are shown in Table 1. Primers were generated by software Primer V; all primers were designed from two different exons and encompassed at least one intron sequence. RT-qPCR was performed using the Bio-RAD iCycler FAM-490 system using the PCR SuperMix UDG (Invitrogen). The INS primers used were A 5¢-AAGGGCTTTATT CCATCTC-3¢, S 5¢-GAAGCGTGGCATTGTGG-3¢. Levels of mRNA were normalized to ACTB. Glucose stimulated insulin release experiment
After differentiation of hAECs for 14 days, cells were incubated in DPBS for 30 min, then exposed to glucose of either 5.5 or 16.5 mM for 2 h. Insulin concentrations were measured in the collected media by chemiluminescence analysis (Roche). Cell phenotypic shift analysis
To assess cellular morphology, images of cells were taken using a NIKON microscope (TS100; ECLIPSE) every other day after the beginning of the induction. Moreover, a transmission electron microscope (TEM; JEOL-1230) was used to search for secretion granules in the cytoplasm. Treatment of hAECs with 5-Aza-2 ¢-deoxycytidine
Cells were exposed to 10 mM Aza for 72 h. Concentration and duration of exposure were selected according to
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
593
Table 1. Primer Sequences and PCR Conditions for RT-PCR Experiments Genes
Base pair
Primers
Temperature (C)
Cycles
CDH1
169
60C
35
CLDN3
107
60C
35
CLDN4
137
60C
35
SNAI1
174
60C
35
Pou5f1
168bp
64C
30
Nanog
387bp
64C
30
TRIMI7
159bp
5¢-GTTCCGCTCTGTCTTTGG-3¢ 5¢-TGGCTTCCCTCTTTCATC-3¢ 5¢-GCAGCGAGTCGTACACCTT-3¢ 5¢-CATCGGCAGCAACATCAT-3¢ 5¢-CCACCACTGCCCAAACCT-3¢ 5¢-GTGCCTTGCTCACCGAAAC-3¢ 5¢-AGGGACATTCGGGAGA-3¢ 5¢-CTGGCTGCTACAAGGC-3¢ 5¢-CTTGCTGCAGAAGTGGGTGGAGGAA-3¢ 5¢-CTGCAGTGTGGGTTTCGGGCA-3¢ 5¢-ACTGTCTCTCCTCTTCCCTCCTCC-3¢ 5¢-GTAGAGGCTGGGGTAGGTAGGTG-3¢ 5¢-GCAACAGCGCAGAGGCTATTATT-3¢ 5¢-AGGGAAGTTGGGCTCAGGACTGG- 3¢ 5¢-AGCCTTTGTGAACCAACACC-3¢ 5¢-GCTGGTAGAGGGAGCAGATG-3¢ 5¢-GGATGAAGTCTACCAAAGCTCACGC-3¢ 5¢-CCAGATCTTGATGTGTCTCTCGGTC-3¢ 5¢-CAATCGAATGCACAACCTCA-3¢ 5¢-GGGAGACTGGGGAGTAGAGG-3¢ 5¢-AGGCAGACCCACTCAGTGA-3¢ 5¢-AACAATGGCGACCTCTTCTG-3 5¢-GTTCCTCCTCCTCCTCTTCCTC-3¢ 5¢-AAGATCTGCTGTCCGGAAAAAG-3¢ 5¢-GCCTTTTATTGTGAGAGTGG-3¢ 5¢-CTCACACATCCGTTGGACAC-3¢ 5¢-CGCACCACTGGCATTGTCAT-3¢ 5¢-TTCTCCTTGATGTCACGCAC-3¢
64C
35
60C
35
60C
35
62C
35
62C
35
60C
35
60C
35
62C
30
INS
244
PDX1
217
NEUROG3
253
GCG
307
NKX6-1
381
PAX6
256
ACTB
200
PCR, polymerase chain reaction; RT-PCR, reverse transcription-polymerase chain reaction.
previous studies (Pennarossa et al., 2013) and based on preliminary experiments in which different doses and different incubation times were tested. At the end of the 72-h exposure, cells were harvested, rinsed thrice, and then plated in 60-mm petri dishes for 14 days in induction medium. NimbGen MeDIP analysis
Samples were hybridized to promoter tiling microarrays (MM9_min_promoter_array, NimbleGen Systems, Inc.) representing 20,404 putative promoters and 15,980 CpG islands. Sample labeling, hybridization, and array scanning were performed by NimbleGen Systems, Inc. according to standard procedures. For DNA methylation experiments, we considered weak and strong CpG island promoters with a log2 > 0.2 to be DNA hypermethylated. This cutoff was verified by PCR. We compared differentially methylated promoters between hAECs after induction and hAECs before induction. Here, a ratio less than or equal to 0.5 was identified as significantly hypomethylated, whereas ratios equal to or greater than 2 were identified as significantly hypermethylated. Gene ontology analysis and statistical analysis
Gene ontology (GO) analysis was used to analyze the main functions of the differentially expressed genes according to the functional classifications of NCBI. Enrichment values provide a measure of the significance of the function: as the enrichment value increases, the corresponding function is more specific, which helps us find those
GOs with more concrete function descriptions in the experiment. Within the significant category, the enrichment Re was given by Re = (nf /n)/(Nf /N), where nf is the number of differential genes within the particular category, n is the total number of genes within the same category, Nf is the number of differential genes in the entire microarray, and N is the total number of genes in the microarray. Enrichment values equal to or greater than 2 are considered significant. Statistical analysis
Data are expressed as the mean – SEM. Values between two groups were compared using two-tailed Student’s ttests. A probability value of < 0.05 was considered statistically significant. Results The isolated hAECs possessed epithelial characteristics
Cell isolation was performed on 12-term amnia that were dissected into about 10 square centimeters. The isolations of all of the amnia were successful. By using the two-step T/E enzymatic method, an average of 1.13 · 108 – 0.23 · 108 epithelial cells/dm2 could be isolated from each amnion. hAECs proliferated robustly and formed a confluent monolayer of cobblestone-shaped epithelial cells (Fig. 1A) in culture. The isolated cells were confirmed to be epithelial cells by immunofluorescence analysis for antibodies against CK18 and E-cadherin. The positive stain rate of cells to the two antigens was 100% (Fig. 1B, C), indicating a homogenous
594
PENG ET AL.
FIG. 1. Gene expression and immunocytofluorescence of isolated amniotic cells. (A) The morphology of freshly isolated hAECs (200 · ). (B, C) Immunofluorescence analysis of the epithelial cell markers CK18 and E-cadherin in isolated cells (B: fluorescence microscopy 300 · ; C: confocal microscopy 400 · ). (D) mRNA expression of the epithelial-related genes CDH1, CLDN3, and CLDN4 in isolated cells. (E, F) The flow cytometric analysis showed the SSEA4 + cells in the isolated cells. (G) mRNA expression of the stem cell totipotent-related genes Pou5f1, Nanog, and TRIMI7 in isolated cells (The freshly isolated cells were defined as P0; cells passaged once in vitro were defined as P1; and those that passaged twice in vitro were defined as P2). hAECs, human amniotic epithelial cells. Color images available online at www.liebertpub.com/dna population of the isolated cells. The isolated cells were also positive for the expression of epithelial cell-related genes, such as CDH1, CLDN3, and CLDN4 (Fig. 1D), and stem cell totipotent-related genes, such as Pou5f1, Nanog, and TRIMI7 (Fig. 1G). Flow cytometry results displayed that almost 46.03% – 2.3% isolated cells were positive for SSEA4 (Fig. 1E, F), suggesting that cells isolated were hAECs and these cells retained stem cell properties. Approximately all of
the cells isolated from the amnia were negative for CD34 (0.27% – 0.15%) and CD45 (0.27% – 0.12%), which were used as hematopoietic markers. hAECs showed high expression of CD29 (97% – 2.19%) and very low expression of CD90 (3.8% – 0.15%) (Fig. 2A–E), while the hAMSCs showed high expression of both CD29 and CD90, about 97.9% – 2.29% and 97.3% – 0.71%, respectively (Fig. 2F–J). The absence of CD34, CD45, and CD90-positive cells in
FIG. 2. The expression of CD34, CD45, CD29, and CD90 in isolated cells from amnia. (A–E) The expression of CD34, CD45, CD29, and CD90 in hAECs. (F–J) The expression of CD34, CD45, CD29, and CD90 markers in hAMSCs.
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
hAECs indicated that these isolated hAECs were not contaminated with other stem cells such as umbilical cord blood cells or embryonic fibroblasts. The phenotypic shift of hAECs during differentiation in vitro
During induction, hAEC cultures lost the typical epithelial characteristics, including cobblestone cellular morphology, but gained a spindle-shaped morphology and finally migrated into cell aggregations. These results were observed every other day using an optical microscope. Representative pictures are shown in Fig. 3A–C. The cells were seeded in Petri dishes and grew into monolayers; then, by day 8 of induction, cells had morphed into spindle-shaped cells. As induction progressed, cells migrated closely together and aggregated on day 12. Finally, the cells transitioned into spherical-like cell aggregations on day 14. Using the optical microscope ruler, most of the spheroid-like cell aggregates were 119.2 – 56.8 mm in diameter on average.
595
The hAECs retained the ability to differentiate into islet-like cells
To assess the potential of hAECs to differentiate into islet-like cells, we detected the expression of pancreatic developmentrelated genes, such as pancreatic duodenum homeobox-1 (PDX1), nk homeobox factor 6.1 (NKX6-1), neurogenin 3 (NEUROG3), paired-box gene 6 (PAX6), and the mature hormones INS and GCG. The RT-PCR results showed negative expression of these genes before induction, but positive expression after hAECs was induced for 14 days (Fig. 3D). Through immunofluorescence, we proved that the spherical cell aggregations formed by hAECs after induction stained positive for insulin (Fig. 3E–I). Meanwhile, TEM analysis of the islet-like cell clusters revealed many granules in the cytoplasm, which are likely endocrine granules (Fig. 3J–K). Glucose-stimulated insulin secreting analysis was performed to further evaluate the capacity of the induced cells to secrete insulin in response to changes in glucose concentrations. The results demonstrate that insulin secretion increases
FIG. 3. The cell phenotype and gene expression changes of hAECs during the islet-like cell cluster differentiation. (A–C) Phenotype of cells during differentiation at day 8, 10, and 14, respectively. (D) The expression of pancreatic developmentrelated genes of cells induced for 14 days. (E) The immunocytofluorescence images taken by confocal microscopy showed that, after induction, cell clusters contained insulin-positive cells (Green; 600 · ). (F) The immunocytofluorescence images of insulin + cells in induced cell clusters after digestion (Green; 200 · ). (G–I) The flow cytometric analysis showed the insulin + cells in cell clusters after induction. ( J, K) Electronic microscope images showed that the granules in the cytoplasm were close to the cell membrane of hAECs after induction (10,000 · ; black arrows). (L) Glucose-stimulated insulinsecreting tests were performed to investigate the ability of the induced cells to respond to glucose fluctuations. Induced cells showed the ability to secrete insulin when exposed to medium containing 16.5 mM glucose (about 59.5 – 3.99 uU/mL; n ‡ 3). When the difference between the test group and control group was statistically significant (p < 0.05), it was marked as *. Color images available online at www.liebertpub.com/dna
596
PENG ET AL.
to 59.5 – 3.99 uU/mL when the cells are exposed to 16.5 mM glucose for 2 h (Fig. 3L). These results indicate that induced cells can correctly transcribe and dock insulin.
sults showed that the expression of INS was 4 times higher in hAECs treated with Aza compared with nontreated cells (Fig. 4H, I).
GO analysis and gene methylation changes
Discussion
Roche-NimbleGen MeDIP results showed the ratio of HNF1a was down to 0.24 and DGAT1 was down to 0.19, while the ratio of PDE3B was approximately 3.34 in hAECs after induction compared with original hAECs (Fig. 4A–C). In addition, the ratios of MAST1, FGF3, and NOTCH1 genes were decreased remarkably in induced cells compared with noninduced cells (Fig. 4D–F). The ratio was the methylation status ratio of hAECs before and after induction. GO analysis showed that endoderm development and insulin secreting-related GOs were down-regulated, and establishment of cell polarity and negative regulation of cell adhesion-related GOs were up-regulated in hAECs after induction (Fig. 4G). Genes involved in each pathway were listed in Table 2.
Currently, islet transplantation is an effective method for treating insulin-dependent diabetes mellitus, which is apt to supersede pancreas transplantation. The diameter of human islets is *75–500 mm on average, while the clusters we obtained after hAECs induction were 119.2 – 56.8 mm in diameter on average. In addition, the clusters we obtained were a 3D structure that may be more similar to in vivo islets in structure and function, not like single-layer adherent cells obtained by previous methods (Wei et al., 2003; Miki et al., 2005; Hou et al., 2008). Meanwhile, our microarray results revealed that FGF3, MAST1, and NOTCH1 genes were hypomethylated after induction. These genes are associated with cell migration and cytoskeleton organization; thus, we speculate that they may facilitate the cellular morphology changes observed during hAEC differentiation. There are a few studies that have investigated the molecular mechanisms of hAEC differentiation into other cell types. Epigenetic modifications not only play important roles in the differentiation and functional maintenance of cell lineages, but also contribute to the development of complex diseases (Petronis, 2010). Szukiewicz et al. (2010a) suggested that histamine H2 may be involved in the
Brief demethylation step elevated the INS gene expression in hAEC-derived islet-like cell clusters
Furthermore, to assess whether demethylation would promote the differentiation of hAECs into islet-like cells or not, we measured INS gene expression in noninduced hAECs with or without treatment with Aza. The qPCR re-
FIG. 4. Gene ontology (GO) analysis and methylation profiles of genes in hAECs before and after induction. (A–C) The methylation status of pancreatic endocrine development and insulin secretion-related genes. (D–F) The methylation status of cell cytoskeleton organization-related genes. The methylation array profiles are shown in these pictures; blue areas denote the location of transcriptional start sites of the genes analyzed. (G) The GO analysis of samples after induction compared with those before induction. The X-axis is enrichment. Enrichment provides a measure of the significance of the function: as the enrichment increases, the corresponding function is more specific. Enrichment values equal to or greater than 2 are defined as meaningful, down GO enrichment was shown as negative number, while up GO was shown as a positive number. (H) The expression of insulin in hAECs before and after brief demethylation by Aza treatment. (I) The fold changes of INS gene (n ‡ 3). When the difference between the test group and control group was statistically significant (p < 0.05), it was marked as *.
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
Table 2. Up- and Down-GOs and Genes in hAECs After Induction Compared with hAECs Before Induction GO term Up-GOs and genes Negative regulation of cell adhesion
Establishment of cell polarity
Down-GOs and genes Endoderm development cAMP catabolic process Calcium iondependent exocytosis
Query ID
Enrichment
PDE3B RASA1 TGFBI LAMB1 DSCAM DSCAML1 CDH13 RAB11FIP2 KIF26B ARHGEF11 IGF1R EPHB1 PTK2
3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779
WNT9B SMAD3 ARC HNF1B PDE3B PDE4D ARHGAP17 RIMS1
2.96419616888978 2.96419616888978 2.96419616888978 2.96419616888978 4.28161668839635 4.28161668839635 3.50314092686974 3.50314092686974
hAECs, human amniotic epithelial cells; GO, gene ontology.
early stages of nicotinamide-induced differentiation of hAEC into pancreatic b-like cells (Szukiewicz et al., 2010a, 2010b). The studies of Bramswig et al. (2013) showed that a- to b-cell reprogramming could be promoted by manipulating the histone methylation signature of human pancreatic islets (Bramswig et al., 2013). Here, we found that the state of DNA methylation was different in hAECs before and after induction, suggesting a role for DNA methylation in the mechanism of hAEC differentiation into islet-like cell clusters. In humans and mice, impaired acute secretion of insulin by pancreatic b-cells is a characteristic of hepatocyte nuclear factor HNF1a deficiency (Gupta et al., 2005). Studies in mice have shown that a heterozygous mutation of the HNF1a gene causes maturity onset diabetes of the young (MODY1). The primary cause of MODY1 in these genetically altered mice is impaired acute insulin secretion by pancreatic b-cells in response to glucose loading, indicating that the loss of HNF1a leads to abnormal insulin secretion by these cells (Wang et al., 2000; Miura et al., 2006). Diacylglycerol acyltransferase 1 (DGAT1) can catalyze the final reaction of triglyceride synthesis. Triglyceride accumulation is associated with obesity and type 2 diabetes (Krapivner et al., 2010). Thus, DGAT1 is considered a potential therapeutic target for treating obesity and related metabolic disorders. Our NimbleGen MeDIP results revealed that the methylation status of HNF1a and DGAT1 genes in hAECs was twice higher than that in induced hAECs. On the other hand, the methylation status of phosphodiesterase 3B (PDE3B) was higher in hAECs after induction. PDE3B is expressed in cells that are important in the regulation of energy homeostasis, including adipocytes,
597
hepatocytes, hypothalamic cells, and b-cells. This protein is one of the several enzymes that hydrolyzes cAMP—a key mediator in the regulation of lipolysis, glycogenolysis, gluconeogenesis, and pancreatic b-cell insulin secretion (Degerman et al., 2011). Increased cAMP levels potentiate insulin secretion via the combined action of PKA and Epac2 (Kai et al., 2013). Though the NimbleGen MeDIP results did not show methylation status differences of insulin and PDX1 genes in hAECs before and after induction, the pancreatic b-cell development-related gene MAFA was partly hypomethylated in hAECs after induction. MAFA regulates many genes that are essential for glucose sensing and insulin secretion in pancreatic b-cells, and also appears to be vital for functional maturation of b-cells produced by human embryonic stem cell differentiation (Aguayo-Mazzucato et al., 2011; Hang and Stein, 2011; Tsuchiya et al., 2011). Our results of methylation-specific PCR show that MAFA is partly demethylated during hAEC induction (data not shown). In addition, we tested whether a short exposure to a demethylating agent is sufficient to promote INS gene expression in hAECs after induction. For this purpose, hAECs were treated with 5-Aza-2¢-deoxycytidine (Aza) before induction. Aza is a cytidine analog used as a chemotherapeutic agent, and it is also an irreversible inhibitor of DNA methyltransferases. Consistent with our previous hypothesis, our findings suggest that expression of INS in Azatreated hAECs after induction is 4 times higher compared with that in nontreated hAECs. We propose that this plastic epigenetic state of insulin secreting-related genes explains, in part, the ability to induce hEACs to insulin-producing cell clusters here. We speculate that the induction method we used here is conducive to the promotion of hAEC insulin secretion. Further investigation is required to fully elucidate the underlying mechanism of this transdifferentiation, which will ultimately improve the induction efficiency of these hAECs. Conclusion
In conclusion, our results indicated that we isolated pure hAECs from human amnia and induced these isolated cells to form a 3D-structure cell cluster which was similar to islets of the pancreas. Furthermore, we elucidate that DNA methylation changes may contribute to the mechanisms of this transdifferentiation. Acknowledgments
This study was sponsored by the National Natural Science Foundation, No. 30700272; Hunan Provincial Scientific Foundation, No. 2009FJ3180; Changsha Science and Technology Plan, No. K1203005-31; and Ministry of Education New Teacher Foundation, No. 20070533080. Disclosure Statement
No competing financial interests exist. References
Ashcroft, F.M., and Rorsman, P. (2012). Diabetes mellitus and the beta cell: the last ten years. Cell 148, 1160–1171.
598
Albalat, R., Marti-Solans J., and Canestro, C. (2012). DNA methylation in amphioxus: from ancestral functions to new roles in vertebrates. Brief Funct Genomics 11, 142–155. Aguayo-Mazzucato, C., Koh, A., El Khattabi, I., Li, W.C., Toschi, E., Jermendy A., et al. (2011). Mafa expression enhances glucose-responsive insulin secretion in neonatal rat beta cells. Diabetologia 54, 583–593. Bramswig, N.C., Everett, L.J., Schug, J., Dorrell, C., Liu, C., Luo, Y., et al. (2013). Epigenomic plasticity enables human pancreatic alpha to beta cell reprogramming. J Clin Invest 123, 1275–1784. Degerman, E., Ahmad, F., Chung, Y.W., Guirguis, E., Omar, B., Stenson, L., et al. (2011). From PDE3B to the regulation of energy homeostasis. Curr Opin Pharmacol 11, 676–682. Evron, A., Goldman, S., and Shalev, E. (2012). Human amniotic epithelial cells differentiate into cells expressing germ cell specific markers when cultured in medium containing serum substitute supplement. Reprod Biol Endocrinol 10, 108. Fang, C.H., Jin, J., Joe, J.H., Song, Y.S., So, B.I., Lim, S.M., et al. (2012). In vivo differentiation of human amniotic epithelial cells into cardiomyocyte-like cells and cell transplantation effect on myocardial infarction in rats: comparison with cord blood and adipose tissue-derived mesenchymal stem cells. Cell Transplant 21, 1687–1696. Gifford, C.A., Ziller, M.J., Gu, H., Trapnell, C., Donaghey, J., Tsankov, A., et al. (2013). Transcriptional and epigenetic dynamics during specification of human embryonic stem cells. Cell 153, 1149–1463. Gupta, R.K., Vatamaniuk, M.Z., Lee, C.S., Flaschen, R.C., Fulmer, J.T., Matschinsky, F.M., et al. (2005). The MODY1 gene HNF-4a regulates selected genes involved in insulin secretion. J Clin Invest 115, 1006–1015. Hang, Y., and Stein, R. (2011). MafA and MafB activity in pancreatic beta cells. Trends Endocrinol Metab 22, 364–373. Hou, Y., Huang, Q., Liu, T., and Guo, L. (2008). Human amnion epithelial cells can be induced to differentiate into functional insulin-producing cells. Acta Biochim Biophys Sin 40, 830–839. Kai, A.K., Lam, A.K., Chen, Y., Tai, A.C., Zhang, X., Lai, A.K., et al. (2013). Exchange protein activated by cAMP 1 (Epac1)-deficient mice develop beta-cell dysfunction and metabolic syndrome. FASEB J 27, 4122–4135. Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693–705. Krapivner, S., Iglesias, M.J., Silveira, A., Tegner, J., Bjorkegren, J., Hamsten, A., et al. (2010). DGAT1 participates in the effect of HNF4A on hepatic secretion of triglyceriderich lipoproteins. Arterioscler Thromb Vasc Biol 30, 962– 967. Manuelpillai, U., Lourensz, D., Vaghjiani, V., Tchongue, J., Lacey, D., Tee, J.Y., et al. (2012). Human amniotic epithelial cell transplantation induces markers of alternative macrophage activation and reduces established hepatic fibrosis. PLoS One 7, e38631. Miki, T., Lehmann, T., Cai, H., Stolz, D.B., and Strom, SC. (2005). Stem cell characteristics of amniotic epithelial cells. Stem Cells 23, 1549–1559. Miura, A., Yamagata, K., Kakei, M., Hatakeyama, H., Takahashi, N., Fukui, K., et al. (2006). Hepatocyte nuclear factor4alpha is essential for glucose-stimulated insulin secretion by pancreatic beta-cells. J Biol Chem 281, 5246–5257. Moodley, Y., Ilancheran, S., Samuel, C., Vaghjiani, V., Atienza, D., Williams, E.D., et al. (2010). Human amnion epithelial
PENG ET AL.
cell transplantation abrogates lung fibrosis and augments repair. Am J Respir Crit Care Med 182, 643–651. Pennarossa, G., Maffei, S., Campagnol, M., Tarantini, L., Gandolfi, F., and Brevini, T.A. (2013). Brief demethylation step allows the conversion of adult human skin fibroblasts into insulin-secreting cells. Proc Natl Acad Sci U S A 110, 8948–8953. Petronis, A. (2010). Epigenetics as a unifying principle in the aetiology of complex traits and diseases. Nature 465, 721–727. Rastellini, C., Shapiro, R., Corry, R., Fung, J.J., Starzl, T.E., and Rao, A.S. (1997). Treatment of isolated pancreatic islets to reverse pancreatectomy induced and insulin-dependent type i diabetes in humans a 6-year experience. Transplant Proc 29, 746–747. Szukiewicz, D., Pyzlak, M., Stangret, A., Rongies, W., and Maslinska, D. (2010a). Decrease in expression of histamine H2 receptors by human amniotic epithelial cells during differentiation into pancreatic beta-like cells. Inflamm Res 59 Suppl 2, S205–S207. Szukiewicz, D., Szewczyk, G., Mittal, T.K., Rongies, W., and Maslinski, S. (2010b). Involvement of histamine and histamine H2 receptors in nicotinamide-induced differentiation of human amniotic epithelial cells into insulin-producing cells. Inflamm Res 59 Suppl 2, S209–S211. Tsuchiya, M., Tsuchiya, K., Yasuda, K., Fujita, M., Takinishi, A., Furukawa, M., et al. (2011). MafA is a key molecule in glucose and energy balance in the central nervous system and peripheral organs. Int J Biomed Sci 7, 19–26. Wang, H., Maechler, P., Antinozzi, P.A., Hagenfeldt, K.A., and Wollheim, C.B. (2000). Hepatocyte nuclear factor 4alpha regulates the expression of pancreatic beta -cell genes implicated in glucose metabolism and nutrient-induced insulin secretion. J Biol Chem 275, 35953–35959. Wei, J.P., Zhang, T.S., Kawa, S., Aizawa, T., Ota, M., Akaike, T., Kato, K., Konishi, I., and Nikaido, T. (2003) Human amnion-isolated cells normalize blood glucose. Cell Transplant 12, 545–552. Xie, W., Schultz, M.D., Lister, R., Hou, Z., Rajagopal, N., Ray, P., et al. (2013). Epigenomic analysis of multilineage differentiation of human embryonic stem cells. Cell 153, 1134–1148.
Address correspondence to: Jian Wang, PhD Institute of Human Reproduction and Stem Cell Engineering Central South University National Center of Human Stem Cell Research and Engineering Changsha 410078 People’s Republic of China E-mail: [email protected] Guangxiu Lu, MB Institute of Human Reproduction and Stem Cell Engineering Central South University National Center of Human Stem Cell Research and Engineering Changsha 410078 People’s Republic of China E-mail: [email protected] Received for publication February 15, 2014; received in revised form April 11, 2014; accepted April 22, 2014.
Involvement of Gene Methylation Changes in the Differentiation of Human Amniotic Epithelial Cells into Islet-Like Cell Clusters Lin Peng,1 Jian Wang,1,2 and Guangxiu Lu1,2
Insulin-dependent diabetes results from destruction of the insulin-producing b-cells of the pancreas. Islet cell transplantation is a promising cure for diabetes. Here, we induced human amniotic epithelial cells (hAECs) to differentiate into islet-like cell clusters by nicotinamide plus betacellulin in vitro, and further investigated the DNA methylation status by a Nimble MeDIP microarray before and after cell differentiation to shed light on the molecular mechanisms of this differentiation. In addition, 5-Aza-2¢deoxycytidine was used to investigate whether the differentiation of hAECs into islet-like cells occurred through demethylation. Purified hAECs (CK18 + /E-cadherin + /CD29 + /CD90 - /CD34 - /CD45 - ) were isolated from human amnia. After induction, hAECs were found to be insulin positive and sensitive to glucose, indicating successful induction to islet-like cells. The methylation status of cell cytoskeletonrelated genes was down-regulated and that of negative regulation of cell adhesion-related genes was upregulated. The methylation status of pancreas development-related genes such as HNF1a and DGAT1 was decreased in hAECs after induction. After brief demethylation, INS gene expression was up-regulated in islet-like cell clusters, suggesting that DNA methylation changes were associated with the differentiation of hAECs into islet-like cell clusters.
Introduction
D
iabetes currently affects more than 336 million people worldwide (Ashcroft and Rorsman, 2012). Type 1 diabetes is characterized by insulin deficiency due to the destruction of b-cells within pancreatic islets. Increasing the number of insulin-producing b-cells is a promising therapeutic avenue (Rastellini et al, 1997). Embryonic stem (ES) and induced pluripotent stem cells are considered the most promising cell types for clinical applications because of their pluripotent characteristics. However, the use of these cell types is limited by ethical and safety issues. Therefore, more suitable cell sources are desired to satisfy the needs of a clinical setting. Human amniotic epithelial cells (hAECs) are derived from the extra-embryonic tissues delivered at birth, which are typically disposed of as waste. Therefore, hAECs are readily accessible, their use is generally considered ethically sound, and, since hAECs are nontumorigenic when being transplanted into mice, they are considered safe (Moodley et al., 2010; Fang et al, 2012; Manuelpillai et al., 2012). In addition, hAECs have successfully been transdifferentiated into many other kinds of cells (Miki et al., 2005; Hou et al., 2008; Evron et al.,
1 2
2012); thus, hAECs may be considered a suitable cell source for clinical applications. Safety, therapeutic efficiency, and availability/sufficient quantity are the basic requirements of clinically applicable stem cells. However, it is still difficult to harvest abundant amounts of insulin secreting-cells by inducing hAECs to differentiate into pancreatic-related cells. A more thorough understanding of the mechanisms controlling this transdifferentiation is needed before hAECs can be considered for clinical applications. Recently, Gifford et al. (2013) and Xie et al. (2013) suggested that dynamic alterations in DNA methylation play an important role in ES cell differentiation. In general, DNA methylation is negatively associated with the activation status of genes (Albalat et al., 2012). Differentiation of stem cells into specialized cells requires upregulation of genes involved in creation of a specific cell phenotype, and suppression of genes responsible for pluripotency (Kouzarides, 2007). In the present study, we established an induction method using nicotinamide plus betacellulin to obtain three-dimensional (3D) islet-like cell clusters, and further investigated gene methylation changes of cells before and after induction by Roche-NimbleGen MeDIP analysis. These gene methylation changes may be
Institute of Human Reproduction and Stem Cell Engineering, Central South University, Changsha, People’s Republic of China. National Center of Human Stem Cell Research and Engineering, Changsha, People’s Republic of China.
591
592
PENG ET AL.
responsible, at least in part, for the transdifferentiation of hAECs to islet-like cell clusters.
cytometer (BD Biosciences) and analyzed on FACS Diva software (BD Biosciences).
Materials and Methods
Immunofluorescence analysis
hAEC isolation, identification, and induced differentiation to islet-like cell clusters
This study was approved by the local ethics committee of the Maternal and Child Health Hospital of Hunan Province, China. Amnia (n = 12, gestational age = 273 – 4 days) were retrieved from healthy women (excluding HIV and HBV virus infection) aged 28 – 3 years by elective cesarean delivery. Informed consent was obtained from each patient before amnion collection. To isolate the hAECs, amniotic membranes were manually stripped from the chorionic membranes of the placenta immediately after delivery, and amnia were rinsed in Dulbecco phosphate-buffered saline (DPBS; GIBCO) containing 100 U/mL penicillin/streptomycin solution (SigmaAldrich). Tissues were digested twice in 0.05% trypsin/ EDTA (T/E; GIBCO) for 15 min at 37C, then centrifuged at 230 g, and washed twice in Dulbecco modified Eagle medium (DMEM; Hyclone). The isolated cells were then plated in 100-mm cell culture dishes at a density of 0.5 · 105 cells/mL, and cultured in medium containing DMEM supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich) and 10 ng/mL epidermal growth factor (R&D Systems). In addition to hAECs, we also isolated human amniotic mesenchymal stem cells (hAMSCs). This step enabled the comparison of hAECs to hAMSCs in order to eliminate the possibility of mesenchymal cell contamination in our isolated hAECs. To isolate hAMSCs, the remaining amnia were digested overnight in dispase (2 mg/mL), collagenase I (2 mg/mL), and hyaluronic acid (2 mg/mL) (all from SigmaAldrich) at 37C to collect the hAMSCs. The isolated cells were then plated in 100-mm cell culture dishes at a density of 0.5 · 105 cells/mL in culture medium containing DMEM/ F12 (GIBCO) supplemented with 10% FBS and 10 ng/mL fibroblast growth factor-basic (Invitrogen). To induce differentiation, hAECs were passaged once and then plated at a density of 1.5 · 106 cells/petri dish (60-mm) in DMEM supplemented with 1 · N-2 Supplement (GIBCO), 10 mM nicotinamide (Sigma-Aldrich), and 10 ng/mL betacellulin (R&D Systems) for 14 days. The medium was changed every 3 days. Flow cytometry analysis
For analysis of membrane antigens, monoclonal antibodies against PE-conjugated SSEA4, PE-conjugated CD34, FITC-conjugated CD45, PE-conjugated CD29, and PEconjugated CD90 (all from Biolegend) were used, and a monoclonal insulin antibody (Sigma-Aldrich) and speciesappropriate FITC-conjugated secondary antibodies were used. Briefly, cells were digested into a single-cell suspension and then incubated with the monoclonal antibodies mentioned earlier for 45 min, respectively (Cells were incubated overnight with insulin monoclonal antibodies at 4C and then incubated with species-appropriate FITC-conjugated secondary antibodies at room temperature for 40 min protected from light.). Then, cells were washed with PBS twice before analysis. Data were collected using an FACS Aria flow
The isolated cells were immunostained to identify expression of cytokeratin 18 (CK18) and insulin. The process was as follows: First, cells were fixed by immersion in 4% paraformaldehyde (Sigma-Aldrich) at room temperature for 20 min, then permeabilized with 0.1% Triton X-100 (SigmaAldrich) for 25 min, and blocked at room temperature for 45 min (Blocking reagent; East Coast Bio). Second, cells were incubated overnight with a monoclonal CK18 antibody (Santa Cruz Biotechnology) or insulin antibody (Sigma-Aldrich) at 4C, washed twice with DPBS, and then incubated with species-appropriate FITC-, PE-, or CY3conjugated secondary antibodies at room temperature for 40 min protected from light. Cell nuclei were stained with DAPI. Cells were observed under a laser scanning microscope (Nikon TE-2000u) at the corresponding wavelength of FITC, PE, Cy3 (n = 3). RT-PCR and quantitative real-time PCR analysis
Reverse transcription-polymerase chain reaction (RT-PCR) and quantitative real-time PCR (RT-qPCR) were performed to compare transcript levels between freshly isolated hAECs (P0); hAECs before and after induction (n = 3); and hAECs after induction in the presence or absence of 5-Aza-2¢-deoxycytidine (Aza; Sigma-Aldrich; n = 3). Briefly, total RNA was isolated by Trizol reagent (Invitrogen) according to the manufacturer’s protocol, and 1 mg total RNA was reverse transcribed into cDNA using random primers (RevertAid First-Strand cDNA Synthesis Kit #K1622; Formentas). The oligonucleotide primers used for RT-PCR are shown in Table 1. Primers were generated by software Primer V; all primers were designed from two different exons and encompassed at least one intron sequence. RT-qPCR was performed using the Bio-RAD iCycler FAM-490 system using the PCR SuperMix UDG (Invitrogen). The INS primers used were A 5¢-AAGGGCTTTATT CCATCTC-3¢, S 5¢-GAAGCGTGGCATTGTGG-3¢. Levels of mRNA were normalized to ACTB. Glucose stimulated insulin release experiment
After differentiation of hAECs for 14 days, cells were incubated in DPBS for 30 min, then exposed to glucose of either 5.5 or 16.5 mM for 2 h. Insulin concentrations were measured in the collected media by chemiluminescence analysis (Roche). Cell phenotypic shift analysis
To assess cellular morphology, images of cells were taken using a NIKON microscope (TS100; ECLIPSE) every other day after the beginning of the induction. Moreover, a transmission electron microscope (TEM; JEOL-1230) was used to search for secretion granules in the cytoplasm. Treatment of hAECs with 5-Aza-2 ¢-deoxycytidine
Cells were exposed to 10 mM Aza for 72 h. Concentration and duration of exposure were selected according to
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
593
Table 1. Primer Sequences and PCR Conditions for RT-PCR Experiments Genes
Base pair
Primers
Temperature (C)
Cycles
CDH1
169
60C
35
CLDN3
107
60C
35
CLDN4
137
60C
35
SNAI1
174
60C
35
Pou5f1
168bp
64C
30
Nanog
387bp
64C
30
TRIMI7
159bp
5¢-GTTCCGCTCTGTCTTTGG-3¢ 5¢-TGGCTTCCCTCTTTCATC-3¢ 5¢-GCAGCGAGTCGTACACCTT-3¢ 5¢-CATCGGCAGCAACATCAT-3¢ 5¢-CCACCACTGCCCAAACCT-3¢ 5¢-GTGCCTTGCTCACCGAAAC-3¢ 5¢-AGGGACATTCGGGAGA-3¢ 5¢-CTGGCTGCTACAAGGC-3¢ 5¢-CTTGCTGCAGAAGTGGGTGGAGGAA-3¢ 5¢-CTGCAGTGTGGGTTTCGGGCA-3¢ 5¢-ACTGTCTCTCCTCTTCCCTCCTCC-3¢ 5¢-GTAGAGGCTGGGGTAGGTAGGTG-3¢ 5¢-GCAACAGCGCAGAGGCTATTATT-3¢ 5¢-AGGGAAGTTGGGCTCAGGACTGG- 3¢ 5¢-AGCCTTTGTGAACCAACACC-3¢ 5¢-GCTGGTAGAGGGAGCAGATG-3¢ 5¢-GGATGAAGTCTACCAAAGCTCACGC-3¢ 5¢-CCAGATCTTGATGTGTCTCTCGGTC-3¢ 5¢-CAATCGAATGCACAACCTCA-3¢ 5¢-GGGAGACTGGGGAGTAGAGG-3¢ 5¢-AGGCAGACCCACTCAGTGA-3¢ 5¢-AACAATGGCGACCTCTTCTG-3 5¢-GTTCCTCCTCCTCCTCTTCCTC-3¢ 5¢-AAGATCTGCTGTCCGGAAAAAG-3¢ 5¢-GCCTTTTATTGTGAGAGTGG-3¢ 5¢-CTCACACATCCGTTGGACAC-3¢ 5¢-CGCACCACTGGCATTGTCAT-3¢ 5¢-TTCTCCTTGATGTCACGCAC-3¢
64C
35
60C
35
60C
35
62C
35
62C
35
60C
35
60C
35
62C
30
INS
244
PDX1
217
NEUROG3
253
GCG
307
NKX6-1
381
PAX6
256
ACTB
200
PCR, polymerase chain reaction; RT-PCR, reverse transcription-polymerase chain reaction.
previous studies (Pennarossa et al., 2013) and based on preliminary experiments in which different doses and different incubation times were tested. At the end of the 72-h exposure, cells were harvested, rinsed thrice, and then plated in 60-mm petri dishes for 14 days in induction medium. NimbGen MeDIP analysis
Samples were hybridized to promoter tiling microarrays (MM9_min_promoter_array, NimbleGen Systems, Inc.) representing 20,404 putative promoters and 15,980 CpG islands. Sample labeling, hybridization, and array scanning were performed by NimbleGen Systems, Inc. according to standard procedures. For DNA methylation experiments, we considered weak and strong CpG island promoters with a log2 > 0.2 to be DNA hypermethylated. This cutoff was verified by PCR. We compared differentially methylated promoters between hAECs after induction and hAECs before induction. Here, a ratio less than or equal to 0.5 was identified as significantly hypomethylated, whereas ratios equal to or greater than 2 were identified as significantly hypermethylated. Gene ontology analysis and statistical analysis
Gene ontology (GO) analysis was used to analyze the main functions of the differentially expressed genes according to the functional classifications of NCBI. Enrichment values provide a measure of the significance of the function: as the enrichment value increases, the corresponding function is more specific, which helps us find those
GOs with more concrete function descriptions in the experiment. Within the significant category, the enrichment Re was given by Re = (nf /n)/(Nf /N), where nf is the number of differential genes within the particular category, n is the total number of genes within the same category, Nf is the number of differential genes in the entire microarray, and N is the total number of genes in the microarray. Enrichment values equal to or greater than 2 are considered significant. Statistical analysis
Data are expressed as the mean – SEM. Values between two groups were compared using two-tailed Student’s ttests. A probability value of < 0.05 was considered statistically significant. Results The isolated hAECs possessed epithelial characteristics
Cell isolation was performed on 12-term amnia that were dissected into about 10 square centimeters. The isolations of all of the amnia were successful. By using the two-step T/E enzymatic method, an average of 1.13 · 108 – 0.23 · 108 epithelial cells/dm2 could be isolated from each amnion. hAECs proliferated robustly and formed a confluent monolayer of cobblestone-shaped epithelial cells (Fig. 1A) in culture. The isolated cells were confirmed to be epithelial cells by immunofluorescence analysis for antibodies against CK18 and E-cadherin. The positive stain rate of cells to the two antigens was 100% (Fig. 1B, C), indicating a homogenous
594
PENG ET AL.
FIG. 1. Gene expression and immunocytofluorescence of isolated amniotic cells. (A) The morphology of freshly isolated hAECs (200 · ). (B, C) Immunofluorescence analysis of the epithelial cell markers CK18 and E-cadherin in isolated cells (B: fluorescence microscopy 300 · ; C: confocal microscopy 400 · ). (D) mRNA expression of the epithelial-related genes CDH1, CLDN3, and CLDN4 in isolated cells. (E, F) The flow cytometric analysis showed the SSEA4 + cells in the isolated cells. (G) mRNA expression of the stem cell totipotent-related genes Pou5f1, Nanog, and TRIMI7 in isolated cells (The freshly isolated cells were defined as P0; cells passaged once in vitro were defined as P1; and those that passaged twice in vitro were defined as P2). hAECs, human amniotic epithelial cells. Color images available online at www.liebertpub.com/dna population of the isolated cells. The isolated cells were also positive for the expression of epithelial cell-related genes, such as CDH1, CLDN3, and CLDN4 (Fig. 1D), and stem cell totipotent-related genes, such as Pou5f1, Nanog, and TRIMI7 (Fig. 1G). Flow cytometry results displayed that almost 46.03% – 2.3% isolated cells were positive for SSEA4 (Fig. 1E, F), suggesting that cells isolated were hAECs and these cells retained stem cell properties. Approximately all of
the cells isolated from the amnia were negative for CD34 (0.27% – 0.15%) and CD45 (0.27% – 0.12%), which were used as hematopoietic markers. hAECs showed high expression of CD29 (97% – 2.19%) and very low expression of CD90 (3.8% – 0.15%) (Fig. 2A–E), while the hAMSCs showed high expression of both CD29 and CD90, about 97.9% – 2.29% and 97.3% – 0.71%, respectively (Fig. 2F–J). The absence of CD34, CD45, and CD90-positive cells in
FIG. 2. The expression of CD34, CD45, CD29, and CD90 in isolated cells from amnia. (A–E) The expression of CD34, CD45, CD29, and CD90 in hAECs. (F–J) The expression of CD34, CD45, CD29, and CD90 markers in hAMSCs.
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
hAECs indicated that these isolated hAECs were not contaminated with other stem cells such as umbilical cord blood cells or embryonic fibroblasts. The phenotypic shift of hAECs during differentiation in vitro
During induction, hAEC cultures lost the typical epithelial characteristics, including cobblestone cellular morphology, but gained a spindle-shaped morphology and finally migrated into cell aggregations. These results were observed every other day using an optical microscope. Representative pictures are shown in Fig. 3A–C. The cells were seeded in Petri dishes and grew into monolayers; then, by day 8 of induction, cells had morphed into spindle-shaped cells. As induction progressed, cells migrated closely together and aggregated on day 12. Finally, the cells transitioned into spherical-like cell aggregations on day 14. Using the optical microscope ruler, most of the spheroid-like cell aggregates were 119.2 – 56.8 mm in diameter on average.
595
The hAECs retained the ability to differentiate into islet-like cells
To assess the potential of hAECs to differentiate into islet-like cells, we detected the expression of pancreatic developmentrelated genes, such as pancreatic duodenum homeobox-1 (PDX1), nk homeobox factor 6.1 (NKX6-1), neurogenin 3 (NEUROG3), paired-box gene 6 (PAX6), and the mature hormones INS and GCG. The RT-PCR results showed negative expression of these genes before induction, but positive expression after hAECs was induced for 14 days (Fig. 3D). Through immunofluorescence, we proved that the spherical cell aggregations formed by hAECs after induction stained positive for insulin (Fig. 3E–I). Meanwhile, TEM analysis of the islet-like cell clusters revealed many granules in the cytoplasm, which are likely endocrine granules (Fig. 3J–K). Glucose-stimulated insulin secreting analysis was performed to further evaluate the capacity of the induced cells to secrete insulin in response to changes in glucose concentrations. The results demonstrate that insulin secretion increases
FIG. 3. The cell phenotype and gene expression changes of hAECs during the islet-like cell cluster differentiation. (A–C) Phenotype of cells during differentiation at day 8, 10, and 14, respectively. (D) The expression of pancreatic developmentrelated genes of cells induced for 14 days. (E) The immunocytofluorescence images taken by confocal microscopy showed that, after induction, cell clusters contained insulin-positive cells (Green; 600 · ). (F) The immunocytofluorescence images of insulin + cells in induced cell clusters after digestion (Green; 200 · ). (G–I) The flow cytometric analysis showed the insulin + cells in cell clusters after induction. ( J, K) Electronic microscope images showed that the granules in the cytoplasm were close to the cell membrane of hAECs after induction (10,000 · ; black arrows). (L) Glucose-stimulated insulinsecreting tests were performed to investigate the ability of the induced cells to respond to glucose fluctuations. Induced cells showed the ability to secrete insulin when exposed to medium containing 16.5 mM glucose (about 59.5 – 3.99 uU/mL; n ‡ 3). When the difference between the test group and control group was statistically significant (p < 0.05), it was marked as *. Color images available online at www.liebertpub.com/dna
596
PENG ET AL.
to 59.5 – 3.99 uU/mL when the cells are exposed to 16.5 mM glucose for 2 h (Fig. 3L). These results indicate that induced cells can correctly transcribe and dock insulin.
sults showed that the expression of INS was 4 times higher in hAECs treated with Aza compared with nontreated cells (Fig. 4H, I).
GO analysis and gene methylation changes
Discussion
Roche-NimbleGen MeDIP results showed the ratio of HNF1a was down to 0.24 and DGAT1 was down to 0.19, while the ratio of PDE3B was approximately 3.34 in hAECs after induction compared with original hAECs (Fig. 4A–C). In addition, the ratios of MAST1, FGF3, and NOTCH1 genes were decreased remarkably in induced cells compared with noninduced cells (Fig. 4D–F). The ratio was the methylation status ratio of hAECs before and after induction. GO analysis showed that endoderm development and insulin secreting-related GOs were down-regulated, and establishment of cell polarity and negative regulation of cell adhesion-related GOs were up-regulated in hAECs after induction (Fig. 4G). Genes involved in each pathway were listed in Table 2.
Currently, islet transplantation is an effective method for treating insulin-dependent diabetes mellitus, which is apt to supersede pancreas transplantation. The diameter of human islets is *75–500 mm on average, while the clusters we obtained after hAECs induction were 119.2 – 56.8 mm in diameter on average. In addition, the clusters we obtained were a 3D structure that may be more similar to in vivo islets in structure and function, not like single-layer adherent cells obtained by previous methods (Wei et al., 2003; Miki et al., 2005; Hou et al., 2008). Meanwhile, our microarray results revealed that FGF3, MAST1, and NOTCH1 genes were hypomethylated after induction. These genes are associated with cell migration and cytoskeleton organization; thus, we speculate that they may facilitate the cellular morphology changes observed during hAEC differentiation. There are a few studies that have investigated the molecular mechanisms of hAEC differentiation into other cell types. Epigenetic modifications not only play important roles in the differentiation and functional maintenance of cell lineages, but also contribute to the development of complex diseases (Petronis, 2010). Szukiewicz et al. (2010a) suggested that histamine H2 may be involved in the
Brief demethylation step elevated the INS gene expression in hAEC-derived islet-like cell clusters
Furthermore, to assess whether demethylation would promote the differentiation of hAECs into islet-like cells or not, we measured INS gene expression in noninduced hAECs with or without treatment with Aza. The qPCR re-
FIG. 4. Gene ontology (GO) analysis and methylation profiles of genes in hAECs before and after induction. (A–C) The methylation status of pancreatic endocrine development and insulin secretion-related genes. (D–F) The methylation status of cell cytoskeleton organization-related genes. The methylation array profiles are shown in these pictures; blue areas denote the location of transcriptional start sites of the genes analyzed. (G) The GO analysis of samples after induction compared with those before induction. The X-axis is enrichment. Enrichment provides a measure of the significance of the function: as the enrichment increases, the corresponding function is more specific. Enrichment values equal to or greater than 2 are defined as meaningful, down GO enrichment was shown as negative number, while up GO was shown as a positive number. (H) The expression of insulin in hAECs before and after brief demethylation by Aza treatment. (I) The fold changes of INS gene (n ‡ 3). When the difference between the test group and control group was statistically significant (p < 0.05), it was marked as *.
METHYLATION CHANGES IN DIFFERENTIATION OF HAECS
Table 2. Up- and Down-GOs and Genes in hAECs After Induction Compared with hAECs Before Induction GO term Up-GOs and genes Negative regulation of cell adhesion
Establishment of cell polarity
Down-GOs and genes Endoderm development cAMP catabolic process Calcium iondependent exocytosis
Query ID
Enrichment
PDE3B RASA1 TGFBI LAMB1 DSCAM DSCAML1 CDH13 RAB11FIP2 KIF26B ARHGEF11 IGF1R EPHB1 PTK2
3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 3.58403492360462 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779 4.87910636658779
WNT9B SMAD3 ARC HNF1B PDE3B PDE4D ARHGAP17 RIMS1
2.96419616888978 2.96419616888978 2.96419616888978 2.96419616888978 4.28161668839635 4.28161668839635 3.50314092686974 3.50314092686974
hAECs, human amniotic epithelial cells; GO, gene ontology.
early stages of nicotinamide-induced differentiation of hAEC into pancreatic b-like cells (Szukiewicz et al., 2010a, 2010b). The studies of Bramswig et al. (2013) showed that a- to b-cell reprogramming could be promoted by manipulating the histone methylation signature of human pancreatic islets (Bramswig et al., 2013). Here, we found that the state of DNA methylation was different in hAECs before and after induction, suggesting a role for DNA methylation in the mechanism of hAEC differentiation into islet-like cell clusters. In humans and mice, impaired acute secretion of insulin by pancreatic b-cells is a characteristic of hepatocyte nuclear factor HNF1a deficiency (Gupta et al., 2005). Studies in mice have shown that a heterozygous mutation of the HNF1a gene causes maturity onset diabetes of the young (MODY1). The primary cause of MODY1 in these genetically altered mice is impaired acute insulin secretion by pancreatic b-cells in response to glucose loading, indicating that the loss of HNF1a leads to abnormal insulin secretion by these cells (Wang et al., 2000; Miura et al., 2006). Diacylglycerol acyltransferase 1 (DGAT1) can catalyze the final reaction of triglyceride synthesis. Triglyceride accumulation is associated with obesity and type 2 diabetes (Krapivner et al., 2010). Thus, DGAT1 is considered a potential therapeutic target for treating obesity and related metabolic disorders. Our NimbleGen MeDIP results revealed that the methylation status of HNF1a and DGAT1 genes in hAECs was twice higher than that in induced hAECs. On the other hand, the methylation status of phosphodiesterase 3B (PDE3B) was higher in hAECs after induction. PDE3B is expressed in cells that are important in the regulation of energy homeostasis, including adipocytes,
597
hepatocytes, hypothalamic cells, and b-cells. This protein is one of the several enzymes that hydrolyzes cAMP—a key mediator in the regulation of lipolysis, glycogenolysis, gluconeogenesis, and pancreatic b-cell insulin secretion (Degerman et al., 2011). Increased cAMP levels potentiate insulin secretion via the combined action of PKA and Epac2 (Kai et al., 2013). Though the NimbleGen MeDIP results did not show methylation status differences of insulin and PDX1 genes in hAECs before and after induction, the pancreatic b-cell development-related gene MAFA was partly hypomethylated in hAECs after induction. MAFA regulates many genes that are essential for glucose sensing and insulin secretion in pancreatic b-cells, and also appears to be vital for functional maturation of b-cells produced by human embryonic stem cell differentiation (Aguayo-Mazzucato et al., 2011; Hang and Stein, 2011; Tsuchiya et al., 2011). Our results of methylation-specific PCR show that MAFA is partly demethylated during hAEC induction (data not shown). In addition, we tested whether a short exposure to a demethylating agent is sufficient to promote INS gene expression in hAECs after induction. For this purpose, hAECs were treated with 5-Aza-2¢-deoxycytidine (Aza) before induction. Aza is a cytidine analog used as a chemotherapeutic agent, and it is also an irreversible inhibitor of DNA methyltransferases. Consistent with our previous hypothesis, our findings suggest that expression of INS in Azatreated hAECs after induction is 4 times higher compared with that in nontreated hAECs. We propose that this plastic epigenetic state of insulin secreting-related genes explains, in part, the ability to induce hEACs to insulin-producing cell clusters here. We speculate that the induction method we used here is conducive to the promotion of hAEC insulin secretion. Further investigation is required to fully elucidate the underlying mechanism of this transdifferentiation, which will ultimately improve the induction efficiency of these hAECs. Conclusion
In conclusion, our results indicated that we isolated pure hAECs from human amnia and induced these isolated cells to form a 3D-structure cell cluster which was similar to islets of the pancreas. Furthermore, we elucidate that DNA methylation changes may contribute to the mechanisms of this transdifferentiation. Acknowledgments
This study was sponsored by the National Natural Science Foundation, No. 30700272; Hunan Provincial Scientific Foundation, No. 2009FJ3180; Changsha Science and Technology Plan, No. K1203005-31; and Ministry of Education New Teacher Foundation, No. 20070533080. Disclosure Statement
No competing financial interests exist. References
Ashcroft, F.M., and Rorsman, P. (2012). Diabetes mellitus and the beta cell: the last ten years. Cell 148, 1160–1171.
598
Albalat, R., Marti-Solans J., and Canestro, C. (2012). DNA methylation in amphioxus: from ancestral functions to new roles in vertebrates. Brief Funct Genomics 11, 142–155. Aguayo-Mazzucato, C., Koh, A., El Khattabi, I., Li, W.C., Toschi, E., Jermendy A., et al. (2011). Mafa expression enhances glucose-responsive insulin secretion in neonatal rat beta cells. Diabetologia 54, 583–593. Bramswig, N.C., Everett, L.J., Schug, J., Dorrell, C., Liu, C., Luo, Y., et al. (2013). Epigenomic plasticity enables human pancreatic alpha to beta cell reprogramming. J Clin Invest 123, 1275–1784. Degerman, E., Ahmad, F., Chung, Y.W., Guirguis, E., Omar, B., Stenson, L., et al. (2011). From PDE3B to the regulation of energy homeostasis. Curr Opin Pharmacol 11, 676–682. Evron, A., Goldman, S., and Shalev, E. (2012). Human amniotic epithelial cells differentiate into cells expressing germ cell specific markers when cultured in medium containing serum substitute supplement. Reprod Biol Endocrinol 10, 108. Fang, C.H., Jin, J., Joe, J.H., Song, Y.S., So, B.I., Lim, S.M., et al. (2012). In vivo differentiation of human amniotic epithelial cells into cardiomyocyte-like cells and cell transplantation effect on myocardial infarction in rats: comparison with cord blood and adipose tissue-derived mesenchymal stem cells. Cell Transplant 21, 1687–1696. Gifford, C.A., Ziller, M.J., Gu, H., Trapnell, C., Donaghey, J., Tsankov, A., et al. (2013). Transcriptional and epigenetic dynamics during specification of human embryonic stem cells. Cell 153, 1149–1463. Gupta, R.K., Vatamaniuk, M.Z., Lee, C.S., Flaschen, R.C., Fulmer, J.T., Matschinsky, F.M., et al. (2005). The MODY1 gene HNF-4a regulates selected genes involved in insulin secretion. J Clin Invest 115, 1006–1015. Hang, Y., and Stein, R. (2011). MafA and MafB activity in pancreatic beta cells. Trends Endocrinol Metab 22, 364–373. Hou, Y., Huang, Q., Liu, T., and Guo, L. (2008). Human amnion epithelial cells can be induced to differentiate into functional insulin-producing cells. Acta Biochim Biophys Sin 40, 830–839. Kai, A.K., Lam, A.K., Chen, Y., Tai, A.C., Zhang, X., Lai, A.K., et al. (2013). Exchange protein activated by cAMP 1 (Epac1)-deficient mice develop beta-cell dysfunction and metabolic syndrome. FASEB J 27, 4122–4135. Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693–705. Krapivner, S., Iglesias, M.J., Silveira, A., Tegner, J., Bjorkegren, J., Hamsten, A., et al. (2010). DGAT1 participates in the effect of HNF4A on hepatic secretion of triglyceriderich lipoproteins. Arterioscler Thromb Vasc Biol 30, 962– 967. Manuelpillai, U., Lourensz, D., Vaghjiani, V., Tchongue, J., Lacey, D., Tee, J.Y., et al. (2012). Human amniotic epithelial cell transplantation induces markers of alternative macrophage activation and reduces established hepatic fibrosis. PLoS One 7, e38631. Miki, T., Lehmann, T., Cai, H., Stolz, D.B., and Strom, SC. (2005). Stem cell characteristics of amniotic epithelial cells. Stem Cells 23, 1549–1559. Miura, A., Yamagata, K., Kakei, M., Hatakeyama, H., Takahashi, N., Fukui, K., et al. (2006). Hepatocyte nuclear factor4alpha is essential for glucose-stimulated insulin secretion by pancreatic beta-cells. J Biol Chem 281, 5246–5257. Moodley, Y., Ilancheran, S., Samuel, C., Vaghjiani, V., Atienza, D., Williams, E.D., et al. (2010). Human amnion epithelial
PENG ET AL.
cell transplantation abrogates lung fibrosis and augments repair. Am J Respir Crit Care Med 182, 643–651. Pennarossa, G., Maffei, S., Campagnol, M., Tarantini, L., Gandolfi, F., and Brevini, T.A. (2013). Brief demethylation step allows the conversion of adult human skin fibroblasts into insulin-secreting cells. Proc Natl Acad Sci U S A 110, 8948–8953. Petronis, A. (2010). Epigenetics as a unifying principle in the aetiology of complex traits and diseases. Nature 465, 721–727. Rastellini, C., Shapiro, R., Corry, R., Fung, J.J., Starzl, T.E., and Rao, A.S. (1997). Treatment of isolated pancreatic islets to reverse pancreatectomy induced and insulin-dependent type i diabetes in humans a 6-year experience. Transplant Proc 29, 746–747. Szukiewicz, D., Pyzlak, M., Stangret, A., Rongies, W., and Maslinska, D. (2010a). Decrease in expression of histamine H2 receptors by human amniotic epithelial cells during differentiation into pancreatic beta-like cells. Inflamm Res 59 Suppl 2, S205–S207. Szukiewicz, D., Szewczyk, G., Mittal, T.K., Rongies, W., and Maslinski, S. (2010b). Involvement of histamine and histamine H2 receptors in nicotinamide-induced differentiation of human amniotic epithelial cells into insulin-producing cells. Inflamm Res 59 Suppl 2, S209–S211. Tsuchiya, M., Tsuchiya, K., Yasuda, K., Fujita, M., Takinishi, A., Furukawa, M., et al. (2011). MafA is a key molecule in glucose and energy balance in the central nervous system and peripheral organs. Int J Biomed Sci 7, 19–26. Wang, H., Maechler, P., Antinozzi, P.A., Hagenfeldt, K.A., and Wollheim, C.B. (2000). Hepatocyte nuclear factor 4alpha regulates the expression of pancreatic beta -cell genes implicated in glucose metabolism and nutrient-induced insulin secretion. J Biol Chem 275, 35953–35959. Wei, J.P., Zhang, T.S., Kawa, S., Aizawa, T., Ota, M., Akaike, T., Kato, K., Konishi, I., and Nikaido, T. (2003) Human amnion-isolated cells normalize blood glucose. Cell Transplant 12, 545–552. Xie, W., Schultz, M.D., Lister, R., Hou, Z., Rajagopal, N., Ray, P., et al. (2013). Epigenomic analysis of multilineage differentiation of human embryonic stem cells. Cell 153, 1134–1148.
Address correspondence to: Jian Wang, PhD Institute of Human Reproduction and Stem Cell Engineering Central South University National Center of Human Stem Cell Research and Engineering Changsha 410078 People’s Republic of China E-mail: [email protected] Guangxiu Lu, MB Institute of Human Reproduction and Stem Cell Engineering Central South University National Center of Human Stem Cell Research and Engineering Changsha 410078 People’s Republic of China E-mail: [email protected] Received for publication February 15, 2014; received in revised form April 11, 2014; accepted April 22, 2014.
