The authors have declared that no competing interests exist.
Conceived and designed the experiments: HZ NS JCW DMP. Performed the experiments: HZ NS SRY RN JS. Analyzed the data: HZ NS SRY RN JS. Wrote the paper: HZ JCW DMP.
Induced pluripotent stem (iPS) cells are a valuable resource for discovery of epigenetic changes critical to cell type-specific differentiation. Although iPS cells have been generated from other terminally differentiated cells, the reprogramming of normal adult human basal prostatic epithelial (E-PZ) cells to a pluripotent state has not been reported. Here, we attempted to reprogram E-PZ cells by forced expression of Oct4, Sox2, c-Myc, and Klf4 using lentiviral vectors and obtained embryonic stem cell (ESC)-like colonies at a frequency of 0.01%. These E-PZ-iPS-like cells with normal karyotype gained expression of pluripotent genes typical of iPS cells (Tra-1-81, SSEA-3, Nanog, Sox2, and Oct4) and lost gene expression characteristic of basal prostatic epithelial cells (CK5, CK14, and p63). E-PZ-iPS-like cells demonstrated pluripotency by differentiating into ectodermal, mesodermal, and endodermal cells in vitro, although lack of teratoma formation in vivo and incomplete demethylation of pluripotency genes suggested only partial reprogramming. Importantly, E-PZ-iPS-like cells re-expressed basal epithelial cell markers (CD44, p63, MAO-A) in response to prostate-specific medium in spheroid culture. Androgen induced expression of androgen receptor (AR), and co-culture with rat urogenital sinus further induced expression of prostate-specific antigen (PSA), a hallmark of secretory cells, suggesting that E-PZ-iPS-like cells have the capacity to differentiate into prostatic basal and secretory epithelial cells. Finally, when injected into mice, E-PZ-iPS-like cells expressed basal epithelial cell markers including CD44 and p63. When co-injected with rat urogenital mesenchyme, E-PZ-iPS-like cells expressed AR and expression of p63 and CD44 was repressed. DNA methylation profiling identified epigenetic changes in key pathways and genes involved in prostatic differentiation as E-PZ-iPS-like cells converted to differentiated AR- and PSA-expressing cells. Our results suggest that iPS-like cells derived from prostatic epithelial cells are pluripotent and capable of prostatic differentiation; therefore, provide a novel model for investigating epigenetic changes involved in prostate cell lineage specification.
Induced pluripotent stem (iPS) cells generated by forced expression of certain transcription factors including Oct4, Klf4, c-Myc, and Sox2 resemble embryonic stem cells (ESCs) in morphology, gene expression, and ability to differentiate into any somatic cell type
iPS cells provide a valuable resource for identifying epigenetic changes that occur during cell differentiation because reprogramming reverses the process of cell specification through epigenetic modification, erasing tissue-specific DNA methylation and re-establishing the embryonic-like methylome
Little is known about the epigenetic changes underlying prostate differentiation, partly because of the lack of suitable models. While cell cultures have been a valuable resource for discovery of epigenetic changes occurring during differentiation, these are largely limited to tumor cell lines or transformed derivatives that carry genetic and epigenetic artifacts of accommodation to cell culture
The prostatic epithelium is composed of two compartments of basal and luminal (secretory) epithelial cells. The lineage relationship between these two types of cells is controversial. While most studies suggest that prostate stem cells exist in the basal epithelial compartment and give rise to both basal and secretory cells
All animal studies were approved by the Stanford Administrative Panel on Laboratory Animal Care (APLAC) and done in compliance with the regulations for animal studies at Stanford University. Primary cultures of normal human prostatic epithelial (E-PZ) cells were established from radical prostatectomy specimens obtained immediately after surgery under a protocol approved by the Stanford Institutional Review Board. The participants provided their written informed consent to participate in this study.
E-PZ cells were established and characterized as previously described
293FT cells (Invitrogen) were inoculated at ∼80% confluency on 100-mm dishes in DME medium supplemented with 10% FBS and transfected with 12 µg of each lentiviral vector (Oct4, Sox2, Klf4, c-Myc) plus 8 µg packaging plasmids and 4 µg VSVG plasmids using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. The cell culture medium was collected 48 h after transfection, filtered through a 0.45-µm pore-size cellulose acetate filter (Whatman, Piscataway, NJ), and mixed with PEG-it Virus Concentration Solution (System Biosciences, Mountain View, CA) overnight at 4°C. Viruses were precipitated at 1,500 g the next day and resuspended in Opti-MEM medium (Invitrogen).
Cells/spheres were fixed with 2% paraformaldehyde in phosphate-buffered saline (PBS) for 2 min, permeabilized with 0.5% Triton X-100 in PBS for 10 min, and blocked with 10% horse serum in PBS for 1 h. Cells were then stained with appropriate primary antibodies and AlexaFluor-conjugated secondary antibodies (Invitrogen). All antibodies are listed in
Mice kidneys carrying iPS cell grafts were fixed in 10% buffered formalin overnight and embedded in paraffin. Five-micron sections were cut from the blocks. Immunohistochemistry was performed as previously described
Total RNA from E-PZ and E-PZ-iPS-like cells was isolated using Trizol (Invitrogen) and reverse transcribed using SuperScript™ III Reverse Transcriptase (Invitrogen) according to the manufacturer's instructions. cDNA product was then mixed with SYBR® GreenER™ qPCR super mix (Invitrogen) and primers of interest in the subsequent PCR using a M×3005P® QPCR System (Strategene, La Jolla, CA). Transcript levels in each sample were determined in triplicate to minimize the experimental variation (standard deviation was calculated for each reaction). The transcript level of TATA box binding protein (TBP) was assayed simultaneously as an internal control. The primer sequences used in this study are listed in
Genomic DNA was isolated from E-PZ and E-PZ-iPS-like cells using AllPrep DNA/RNA mini kit (Qiagen, Germantown, MD) and submitted to EpigenDx Inc. (Hopkinton, MA) for bisulfite modification, PCR reactions, and pyrosequencing analysis. For the Nanog promoter, the methylation status at six CpG sites that are 431 bp, 512 bp, 517 bp, 519 bp, 556 bp, and 565 bp upstream of ATG was examined. For the Oct4 promoter, the methylation status at six CpG sites that are 14 bp, 24 bp, and 50 bp upstream of ATG and 80 bp, 97 bp, 70 bp, and 5 bp downstream of ATG was examined.
Colcemid (Sigma-Aldrich, St. Louis, MO) was added to the culture medium at a final concentration of 1 ug/ml and cells were incubated at 37°C for 1 hour. Cells were then detached and resuspended in 5 ml of ice cold 0.56% KCl solution. After incubation at room temperature (RT) for 6 minutes, cells were fixed in 5 ml of methanol: glacial acetic acid (3∶1) solution. Cells were then dropped on clean slides and air dried. Slides were then stained with Geimsa dye and chromosomes were counted under microscopy. Fifty chromosome spreads were counted per cell line.
The following inductive differentiation systems were used: 1) For neural induction, E-PZ-iPS-like cells cultured on Matrigel were digested with collagenase type IV (Invitrogen) and seeded at a density of 1.0×105 cells/cm2 on a poly-HEMA-coated dish (BD Biosciences). After culture for 7 d in Neurobasal medium (Invitrogen) supplemented with 1× B-27 (Invitrogen), 30 ng/mL bFGF (PeproTech) and 30 ng/mL epidermal growth factor (PeproTech) to induce sphere formation, spheres were then transferred onto poly-L-lysine-coated dishes (BD Biosciences) and cultured for 10 d in α-MEM supplemented with 2% FBS, 25 ng/mL bFGF, and 25 ng/mL BDNF (Peprotech); 2) Adipocyte and osteocyte inductions were performed according to the Human Mesenchymal Stem Cell Functional Identification Kit (R&D Systems, Minneapolis, MN); 3) For hepatocyte induction, cells at a density of 2.0×104 cells/cm2 were cultured on collagen-coated dishes for 14 d in DMEM supplemented with 10% FBS, 1× insulin-transferrin-selenium (Gibco, Grand Island, NY), 10 nM dexamethasone (Sigma-Aldrich), 100 ng/mL hepatocyte growth factor (PeproTech), and 50 ng/mL FGF-4 (R&D Systems); 4) For prostate sphere differentiation: E-PZ-iPS-like cells cultured on Matrigel were digested with collagenase type IV (Invitrogen) and transferred to ultra-low attachment plates (Corning Life Sciences, Tewksbury, MA) for suspension culture for 8 days in DMEM/F12 (1∶1) medium supplemented with 5 ng/ml bFGF (PeproTech), 10 ng/ml LIF (Santa Cruz Biotechnology Inc.), and 5% knockout FBS (Invitrogen). For basal epithelial cell differentiation, E-PZ-iPS-like spheres were then cultured in Complete PFMR-4A medium
Rat urogenital mesenchymal (UGM) cells were isolated as previously described
Genomic DNA was isolated from E-PZ-iPS-like cells cultured in control or differentiation-inducing media as described above and submitted to the Stanford Functional Genomic Facility for DNA methylation profiling using Infinium HumanMethylation450 BeadChip containing >485,000 methylation sites at single-nucleotide resolution (Illumina, San Diego, CA). Beta values representing the methylation levels of individual sites were extracted using Illumina GenomeStudio software. Statistical analyses were performed using Excel. Raw data were deposited in Gene Expression Omnibus (GEO) with accession number GSE45469.
We chose to reprogram two E-PZ cultures, E-PZ-1 and E-PZ-2, derived from normal peripheral zone prostatic tissues of two men aged 56- and 44-years old, respectively. These primary cultures are a mixture of basal and transit amplifying cells since they express basal epithelial cell markers including cytokeratin 14 (CK14), cytokeratin 5 (CK5) and p63, but not secretory epithelial cell markers such as androgen receptor (AR) and prostate-specific antigen (PSA)
E-PZ cells at passage 3 were transduced with individual lentiviruses containing human Oct4, Sox2, Klf4, and c-Myc at a 1∶1∶1∶1 ratio on day 0. On day 3 after the transduction, cells were transferred onto mouse embryonic fibroblast (MEF) feeder layers, with the culture medium switched from Complete MCDB 105 medium to human embryonic stem cell (ESC) growth medium mTeSR-1. To improve the efficiency of reprogramming, we cultured cells in low oxygen (5% O2) and in the presence of 2 µM SB431542, 0.5 µM PD0325901, and 0.5 µM Thiazovivin starting from day 7. Previous studies have shown that these three compounds and low oxygen enhance the efficiency of reprogramming
(A) A diagram of the experimental design. The three compounds added to the medium at day 7 were 2 µM SB431542, 0.5 µM PD0325901, and 0.5 µM Thiazovivin. (B) Representative images of colonies derived from two E-PZ cultures at different time points.
We next attempted to expand the E-PZ-iPS-like cells that were isolated from the large ESC-like colonies in mTeSR-1 medium with feeder layers. We detached and broke the colonies into small pieces and transferred them onto new feeder layers. The cells attached and the colonies grew in size, but lost the ESC-like morphology over time. For example, the defined boundaries of the colonies became disrupted as the cells at the edge started to become loosely associated rather than tightly packed in the colonies (
Colonies lost ESC-like morphology in mTeSR-1 medium on feeder layers (A); however, they maintained ESC-like morphology in DMEM/F12 (1∶1) supplemented with 5 ng/ml bFGF, 10 ng/ml LIF and 5% knock out FBS on feeder layers (B) or Matrigel-coated plates (C). E-PZ-iPS-like cells formed spheres when cultured on ultra-low attachment plates in DMEM/F12 (1∶1) supplemented with 5 ng/ml bFGF, 10 ng/ml LIF and 5% knock out FBS (D).
We next characterized the E-PZ-iPS-like cells that were isolated from the large ESC-like colonies. We initially picked seven single colonies from reprogramming of E-PZ-1, and five from E-PZ-2. We characterized two cell lines from each individual in detail and obtained similar results from all 4 cell lines (E-PZ-1-iPS-like-4 and -7, E-PZ-2-iPS-like-1 and -5). Characteristics of clones 4 and 7, derived from E-PZ-1, are shown here as typical of the E-PZ-iPS-like cell lines. Immunofluorescence staining of E-PZ-1-iPS-like-4 cells with antibodies against TRA-1–81 (
E-PZ-1-iPS-like-4 cells showed strong cell membrane staining of TRA-1–81 (A) and SSEA-3 (D), and nuclear staining of Nanog (G), Sox2 (J), Oct4 (M), and c-Myc (P). These cells displayed strong nuclear staining of Ki67 (S) and the human nuclear antigen Ku70 (V). They did not express basal cell marker CK5 (Y) except in a few cells mostly located along the edge of the colonies. (B), (E), (H), (K), (N), (Q), (T), (W), and (Z) are DAPI staining of the nuclei. (C), (F), (I), (L), (O), (R), (U), (X), and (AA) are merged images of staining of DAPI and antibodies against specific markers.
The expression level of several pluripotency genes in E-PZ-iPS-like cells was also analyzed by qRT-PCR (
mRNA levels of Nanog (A), Rex1 (B), Oct4 (C), Klf4 (D), c-Myc (E), Sox2 (F), and CD133 (G) were measured by qRT-PCR and normalized against TBP. Methylation of Nanog (H) and Oct4 (I) promoters were determined by bisulfite pyrosequencing. The Y-axis is the fold-level of gene expression or promoter methylation in E-PZ-1-iPS-like cells compared to those in E-PZ-1 cells, which were set as 1. (J) and (K) were histograms of the number of chromosomes in 100 E-PZ-1-iPS-4 and -7 cells, respectively, determined by metaphase chromosome counting.
The promoter regions of pluripotency genes in reprogrammed somatic cells are often demethylated, causing increased expression of downstream genes. We determined the methylation level of the promoter regions of Nanog and Oct4 in E-PZ-1 and E-PZ-1-iPS-like-4 and -7 cells by quantitative bisulfite pyrosequencing. Of the 6 CpG sites examined in the Nanog promoter, 3 showed demethylation in E-PZ-1-iPS-like cells compared to parent cells, while the other 3 didn't show significant changes in methylation (
To test the pluripotency of E-PZ-iPS-like cells, we first performed in vitro differentiation assays. Two individual lines of E-PZ-1-iPS-like cells each exhibited the capability of differentiating into derivatives of the three embryonic germ layers in vitro when subjected to conditions that induced differentiation into neural cells (ectoderm), adipocytes (mesoderm), osteoblasts (mesoderm), or hepatocytes (endoderm) (
E-PZ-1-iPS-like-4 and -7 cells were subjected to conditions that induced differentiation into neural cells (ectoderm), adipocytes (mesoderm), osteoblasts (mesoderm), or hepatocytes (endoderm). Osteoblast induction produced cells positive for osteocalcin (A and B). Neural induction generated cells positive for the neural cell marker MAP-2 (D and E). Adipocyte induction produced cells with lipid droplets that stained with oil red O (G and H). Hepatocyte induction generated cells positive for human α-fetoprotein (α-FP) (J and K). (C), (F) and (L) are negative controls that were stained with secondary antibodies only. (I) is a negative control without oil red staining. Some spheres derived from E-PZ-iPS-like cells and cultured in E-PZ medium expressed basal prostatic epithelial cell markers including CD44 (M), MAO-A (N), and p63 (O). In addition, some spheres expressed CK18 (P) and PCNA (Q). The spheres also expressed AR (R) in the presence of R1881. When co-cultured with rat UGS, a subset of the spheres expressed PSA (S, inserts showing UGS negative for PSA). In the absence of UGS, no PSA expression was detected (T).
We also determined whether E-PZ-iPS-like cells could be directed to differentiate into prostatic epithelial cells in vitro. Spheres derived from E-PZ-1-iPS-like-4 cells and cultured in Complete PFMR-4A medium expressed basal prostatic epithelial cell markers including CD44 (
To determine the differentiation potential of E-PZ-iPS-like cells in vivo, we injected E-PZ-iPS-like cells under the renal capsule of immunodeficient mice. After 6–8 weeks, tissue masses formed under the renal capsule with 100% frequency. We did not observe teratoma-like histology in the tissue masses derived from E-PZ-iPS-like cells. The cells expressed human-specific nuclear antigen Ku70 (
E-PZ-1-iPS-like cells injected under the renal capsule of immunodeficient mice expressed human-specific nuclear antigen Ku70 (A) and basal prostatic epithelial markers including p63 (B) and CD44 (C). A subset of cells was positive for the transit amplifying epithelial cell marker CK18 (D) but not the secretory cell markers AR (E) or PSA (F). When combined with UGM, E-PZ-1-iPS-like cells gave rise to tissue expressing Ku70 (G) and CK18 (J). In addition, p63 was expressed only by cells at the edge of the tissue (H) and CD44 expression was reduced (I). Although the cells were negative for PSA (L), they expressed an intermediate level of AR in the nuclei (K). White dotted lines mark the boundary of grafts derived from E-PZ-1-iPS-like cells and mouse kidney. All magnifications are 40×.
We attempted to induce further differentiation of the E-PZ-iPS-like cells toward the secretory epithelial phenotype by combining the cells with rat urogenital mesenchymal (UGM) cells, which have been shown previously to induce prostatic differentiation in vivo
As a comparison, we performed the same in vivo differentiation experiments using F-iPS cells. F-iPS cells without UGM injected under the renal capsule of immunodeficient mice formed teratomas 6–8 weeks after injection. Histological analysis showed the presence of cartilage (
F-iPS cells injected under the renal capsule of immunodeficient mice showed histological characteristics typical of teratoma including cartilage (A), gut-like epithelium (B), muscle (C), adipose tissue (D), pigmented cells (E), and neuroepithelial rosettes (F). Many glands were positive for CK18 (G), but not p63 (H) or PSA (I). When combined with UGM, F-iPS cells gave rise to cell clusters expressing both CK18 (J) and p63 (K), but not PSA (L). All magnifications are 20×.
To demonstrate the utility of E-PZ-iPS-like cells as a model in elucidating the mechanisms of prostate differentiation, we analyzed temporal epigenetic changes occurring during the induction of secretory prostatic differentiation of E-PZ-iPS-like cells using DNA methylation profiling. DNA methylation levels of >485,000 sites were measured in spheres generated from E-PZ-1-iPS-like-4 cells cultured either in iPS cell medium as control, or in Complete PFMR-4A medium with 10 nM R1881 to induce AR expression for 1 or 3 days. In addition, spheres cultured in Complete PFMR-4A medium with 10 nM R1881 in the presence of UGS to induce PSA expression were harvested at 1, 3, or 5 days to capture methylation changes during mature secretory cell differentiation. In all, 5 pairs of samples were compared in the study, i.e., AR day 1 vs. control, AR day 3 vs. control, PSA day 1 vs. control, PSA day 3 vs. control, and PSA day 5 vs. control. Changes in methylation levels were examined in two ways. First, fold-change was calculated for each pair of samples as methylation level in induced cells divided by that in corresponding control cells. We focused on genes whose methylation levels increased or decreased by at least 50% in induced cells compare to corresponding control cells in at least 3 pairs of samples. Second, student's t-test was performed between the 5 control and 5 induced cells as two groups. Only genes with significant differential methylation levels in control vs. induced groups were selected. After filtering data with these two criteria, we identified 398 genes and 250 genes that were consistently and significantly hyper- or hypo-methylated in induced cells compared to control, respectively (
(A) Genes that showed significantly higher (in red) or lower (in green) levels of methylation in cells cultured under AR- or PSA- inducing conditions compared to control across different time points of the induction process. (B) Canonical pathways identified by IPA that are enriched with genes hypermethylated in AR and PSA induced cells at late time points compared to early time point. Red arrows point out key pathways known to be involved in prostatic cell differentiation.
We further compared methylation levels in cells cultured under AR induction conditions for 1 vs. 3 days, and PSA induction conditions for 1 vs. 5 days. Genes that showed >4-fold higher or lower methylation levels were selected for further analysis (
We have generated iPS-like cells from human basal prostatic epithelial cells by forced expression of four reprogramming factors including Oct4, Sox2, Klf4 and c-Myc. These cells have several characteristics of iPS cells derived from other types of somatic cells, including: 1) expression of pluripotency markers such as Tra-81, Nanog, and SSEA-3 in vitro; 2) activation of the endogenous copy of pluripotent genes such as Oct4; 3) hypomethylation in the promoter regions of Oct4 and Nanog; and 4) differentiation into derivatives of the three embryonic germ layers in vitro under induction. However, they are different from iPS cells derived from other types of somatic cells such as F-iPS cells in several ways, including: 1) lack of formation of teratomas in vivo; 2) ability to differentiate into prostatic epithelial cells in vitro and in vivo; and 3) induction by UGM toward further prostatic differentiation in vivo. The differences are likely due to incomplete reprogramming of the epigenome, which has been observed in human iPS cells derived from a variety of cell types
Although in theory F-iPS cells should give rise to mature prostate glands as do ESC
Differentiated somatic cells are replete with epigenetic regulatory devices that lock characteristic gene expression patterns into place. In general, such cells are quite refractory to reprogramming, as indicated by the very low efficiency of iPS generation, and lack of complete reprogramming to the ESC state. The reprogramming efficiency observed for E-PZ cells was similar to that reported for other terminally differentiated cells. Unlike previous studies reporting improved reprogramming efficiency when p53 expression is decreased or Glis1 expression is increased
Many somatic cells, including terminally differentiated cells, can be reprogrammed to an ESC-like state
E-PZ-iPS-like cells retained differentiation capability toward the parental cell lineage, i.e., prostatic epithelial cells, consistent with the theory of “epigenetic memory”. It has been shown that iPS cells derived by factor-based reprogramming of adult murine tissues harbor residual DNA methylation signatures characteristic of their somatic tissue of origin, which favors their differentiation along lineages related to the donor cell, while restricting alternative cell fates
Perhaps the most interesting finding of our study is that E-PZ-iPS-like cells are capable of prostatic secretory differentiation in vitro. Molecular mechanisms and pathways involved in prostatic differentiation, especially secretory cell differentiation, are poorly understood due to the lack of suitable models. Current cell culture-based models only attain limited AR and PSA expression
As proof-of-principle, we carried out a comprehensive epigenetic characterization of E-PZ-iPS-like cells at different time points of differentiation. DNA methylation profiling captured key genes and pathways that have been implicated in prostatic differentiation including Wnt5a, PTEN, and Nkx3.1. In addition, new genes and pathways have been selected as possible regulators of secretory differentiation. For example, many target genes of miRNAs let-7, mir-1, and mir-145 were hypermethylated in cells cultured under AR-inducing conditions for 3 days compared to 1 day (
(TIF)
(TIF)
(TIF)
(TIF)
(TIF)
(TIF)
(TIF)
(DOCX)
(DOCX)
(XLS)
(XLS)
(XLS)
(XLS)