TY - JOUR
AU - Crooks, Gay M.
AB - Abstract Although oval cells are postulated to be adult liver stem cells, a well-defined phenotype of a bipotent liver stem cell remains elusive. The heterogeneity of cells within the oval cell fraction has hindered lineage potential studies. Our goal was to identify an enriched population of bipotent oval cells using a combination of flow cytometry and single cell gene expression in conjunction with lineage-specific liver injury models. Expression of cell surface markers on nonparenchymal, nonhematopoietic (CD45−) cells were characterized. Cell populations were isolated by flow cytometry for gene expression studies. 3,5-Diethoxycarbonyl-1,4-dihydrocollidine toxic injury induced cell cycling and expansion specifically in the subpopulation of oval cells in the periportal zone that express CD133. CD133+CD45− cells expressed hepatoblast and stem cell-associated genes, and single cells coexpressed both hepatocyte and cholangiocyte-associated genes, indicating bilineage potential. CD133+CD45− cells proliferated in response to liver injury. Following toxic hepatocyte damage, CD133+CD45− cells demonstrated upregulated expression of the hepatocyte gene Albumin. In contrast, toxic cholangiocyte injury resulted in upregulation of the cholangiocyte gene Ck19. After 21–28 days in culture, CD133+CD45− cells continued to generate cells of both hepatocyte and cholangiocyte lineages. Thus, CD133 expression identifies a population of oval cells in adult murine liver with the gene expression profile and function of primitive, bipotent liver stem cells. In response to lineage-specific injury, these cells demonstrate a lineage-appropriate genetic response. Disclosure of potential conflicts of interest is found at the end of this article. CD133, Oval cells, Adult stem cells, Cell surface markers, Fluorescence-activated cell sorting analysis Hepatic stem cells, Liver regeneration Introduction The contribution of liver stem cells to liver regeneration in different forms of liver injury remains unclear. The two basic epithelial cell types in the adult liver are hepatocytes and cholangiocytes. Hepatocytes compromise the vast majority of liver cell mass and have many metabolic roles, including glucose and amino acid metabolism [1]. Cholangiocytes are specialized duct cells that line bile ducts and are important in modifying bile composition [2]. During embryonic liver development, these two cell types arise from a common progenitor, fetal hepatoblasts [3]. In postnatal life, the ability of mature hepatocytes to regenerate themselves during acute liver injury or partial hepatectomy has been well-established [4]. However, the contribution of bipotent adult liver stem cells to the hepatocyte and cholangiocyte populations during repair after liver injury remains uncertain [5]. Current research describes the liver stem cell, or oval cell (OC), as a small cell, within the nonparenchymal (NP) fraction of the liver, that resides near the terminal bile ducts at the hepatocyte-cholangiocyte interface [6, 7]. Thus, OCs have traditionally been identified using cell morphology and histologic location [8, 9]. Different toxic liver injury models have been created to induce OC proliferation in rodents, including choline-deficient diets, 2-acetylaminofluorene/partial hepatectomy, and 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) [10, 11]. These chronic injury models often involve a block in the normal regenerative capacity of hepatocytes [12]. Prior studies of the lineage potential of OCs have focused on the role of these cells to differentiate into hepatocytes [13, 14]; the cholangiocyte potential of OCs has not been well studied [15]. Furthermore, the lack of clonal in vitro assays or in vivo transplantation models that allow bilineage differentiation has made true evidence of OC bipotency elusive. Various reports have described the expression of cell surface markers on murine OCs, including CD34, c-kit, Sca-1, A6, and CD45 [11, 13, 16–18]. Many of these markers are associated with hematopoietic stem cells [19–21], which have size and morphology comparable to those of OCs [22, 23]. One report demonstrated that OCs efflux Hoechst dye with the transmembrane pump ABCG2, thus allowing their identification as a so-called “side population” and the suggestion that they may have a relative survival advantage in the face of toxic injury [24]. However, an immunophenotype of OCs that identifies a population capable of both hepatocyte and cholangiocyte regeneration remains elusive. Given this background, our goal was to identify a reliable OC immunophenotype, which would allow further exploration of its bilineage potential. In the process of testing the previous OC markers, a population of CD133-expressing cells was consistently identified within the nonparenchymal, nonhematopoietic, liver cell population and found to expand in response to liver injury. CD133, or Prominin1, is a membrane protein found on several types of adult stem cells, including hematopoietic and neural stem cells [25, 26]. CD133 represents a 120-kDa cholesterol-binding glycoprotein with five transmembrane domains that is concentrated in the apical membrane of epithelial cells and is highly expressed in many fetal epithelial tissues [27]. Although the exact function of CD133 in these tissues is unknown, it appears to play a role in membrane protrusions [28]. Using serial enrichment techniques, we identified a population of CD133+CD45− cells within the oval cell fraction, with a pattern of gene expression consistent with primitive stem cells. Single-cell analysis of CD133+CD45− OCs demonstrated bilineage (i.e., both hepatocyte- and cholangiocyte-specific gene expression). Furthermore, CD133+CD45− OCs upregulated lineage-specific genes in response to either hepatocyte or cholangiocyte injury. Functional analysis demonstrated the presence of hepatocyte growth factor (HGF)- and epithelial growth factor (EGF)-responsive cells and that these cells were able to generate bilineage progeny after 3 weeks of in vitro culture. This is the first report to demonstrate on a single-cell level the presence of primitive oval cells within the adult liver with bilineage gene expression, which expand in number after damage and respond differentially at a molecular level to cholangiocyte and hepatocyte damage. Materials and Methods Animals were housed in a temperature-controlled animal facility with a 12-hour light/dark cycle. Procedures were approved by the Institutional Animal Care and Use Committee. Liver Injury Models DDC is a model of OC induction [11, 13, 29]. DDC was administered at 0.1% (wt/wt) in mouse chow (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) and fed to 3-month-old C57BL/6 mice (The Jackson Laboratory, Bar Harbor, ME, http://www.jax.org) for 4–8 weeks. α-Naphthylisothiocyanate (ANIT) (Sigma-Aldrich) is a model of cholangiocyte-specific injury and causes cholangiocyte necrosis and cholestasis [30]. ANIT was dissolved in sterile filtered corn oil and administered to 3-month-old mice at a dose of 100 mg/kg by gastric gavage each week for 4 consecutive weeks. Carbon tetrachloride (CCl4) (Sigma-Aldrich) is a hepatocyte toxin that causes centrilobular necrosis. CCl4 was dissolved in sterile filtered corn oil and administered to 3-month-old mice by i.p. injection at a dose of 0.4 ml of CCl4 per kilogram. Mice were anesthetized using 3% isoflurane anesthesia prior to i.p. injection. CCl4 was administered weekly for 4 weeks. Select animals from each damage model received bromodeoxyuridine (BrdU) (BD Pharmingen, San Diego, http://www.bdbiosciences.com/pharmingen) i.p. 12 hours prior to sacrifice. Immunohistochemistry Fluorescent immunohistochemistry (FIHC) was performed on perpendicular wedges to ensure that central and peripheral sections were analyzed as described [31]. Liver tissue was fixed in formalin and paraffin-embedded. After antigen retrieval (Vector Unmasking Solution; Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com), slides were incubated with Tris-buffered saline (TBS) (pH 7.5, 100 mM Tris, 150 mM NaCl) + 0.1% Triton X-100 solution (Sigma-Aldrich). Slides were blocked for nonspecific binding using TBS + 0.1% Tween 20 (TBST) (Sigma-Aldrich) with 1% bovine albumin (Sigma-Aldrich) and 5% normal donkey serum (Jackson Immunoresearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com). Slides were incubated with primary antibody overnight at 4°C (concentration in TBST 1% bovine serum albumin [BSA]). Primary antibodies included rabbit anti-cow pan-cytokeratin (pan-CK) (1:400; DAKO, Carpinteria, CA, http://www.dako.com) [14], sheep anti-BrdU (1:200; Abcam, Cambridge, U.K., http://www.abcam.com), rabbit anti-mouse albumin (1:200; Accurate Chemical and Scientific, Westbury, NY, http://www.accuratechemical.com), and rat anti-mouse A6 (1:10, generous gift of Dr. V. Factor). Slides were washed twice and incubated with secondary antibodies at room temperature for 2 hours. Secondary antibodies (concentration in TBST 1% BSA) were as follows: donkey anti-rabbit fluorescein isothiocyanate (FITC) or Cy3 (1:200; Abcam), donkey anti-sheep Cy3 (1:200; Abcam), and donkey anti-rat Cy3 (1:200; Abcam). Slides were washed in TBST and TBS, dried, and mounted with coverslips using Vectashield mounting medium with 4,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories). All FIHC experiments were conducted with negative control that involved no primary antibody. Alternatively, liver sections were frozen in liquid nitrogen embedded in Tissue-Tek OTC compound (Sakura, Torrance, CA, http://www.sakura-americas.com). Five-micrometer sections were cut at −20°C (CM1900; Leica, Wetzlar, Germany, http://www.leica.com) and mounted on slides (Fisher Scientific International, Pittsburgh, http://www.fisherscientific.com). Slides were air-dried for 20 minutes, fixed for 5 minutes in 100% acetone, and washed. Slides were blocked with TBST 5% normal donkey serum for 30 minutes at room temperature. The following primary antibodies were incubated for 90 minutes at room temperature: rat anti-mouse CD133 (1:50; eBioscience, San Diego, http://www.ebiosciences.com), rabbit anti-mouse albumin (1:200), and rabbit anti-cow pan-cytokeratin (1:400) [14]. Slides were washed twice and incubated with secondary antibody at room temperature for 1 hour. Secondary antibodies (concentration in TBST 1% BSA) were as follows: donkey anti-rabbit Cy3 and donkey anti-rat Cy3 (all 1:200; Abcam). Slides were washed and mounted with coverslips using Vectashield mounting medium with DAPI (Vector Laboratories). Microscopy Images were viewed with a Leica DMRA microscope using a Plan Apo ×40/1.25 NA phase III DIC or Plan Apo ×63/1.32 oil immersion objective lens. The microscope was equipped with a Shutter LS175W ozone-free xenon arc lamp (Sutter Instruments, Novato, CA, http://www.sutter.com). Images were acquired with an Applied Spectral Imaging Sky Vision-2/VDS camera from EasyFISH software (Applied Spectral Imaging, Vista, CA, http://www.spectral-imaging.com) and printed using Photoshop (Adobe Systems Inc., San Jose, CA, http://www.adobe.com). Parenchymal and Nonparenchymal Cell Separation Protocol for the digestion and centrifugation of the liver cells was modified from Shimano et al. [24]. Liver tissue was minced with a razor and digested with 0.1% collagenase, 0.1% pronase, and 0.01% DNase (all from Sigma-Aldrich) for 45 minutes at 37°C. Cell suspensions were filtered through a 70-μm-pore filter (BD Biosciences, Franklin Lakes, http://www.bd.com) and suspended in 10% fetal calf serum (Omega Scientific, Tarzana, CA, http://www.omegascientific.com) in phosphate-buffered saline (PBS) (Mediatech, Herndon, VA, http://www.cellgro.com). Cells were centrifuged at 50g for 1 minute (GS-6R centrifuge; BD Biosciences). The pellet was saved as the parenchymal cell fraction. The supernatant was recentrifuged at 50g for 1 minute, and the supernatant was saved for a third centrifugation at 50g for 1 minute. The final supernatant was centrifuged at 180g for 8 minutes, with the pellet representing the NP fraction. The final NP cell pellet was resuspended in 1× ammonium chloride red blood cell (RBC) lysis buffer (PharMLyse; BD Biosciences) and washed, and the NP cells were subjected to CD45 depletion using Miltenyi Biotec magnetic bead depletion per the manufacturer's protocol (Miltenyi Biotec, Auburn, CA, http://www.miltenyibiotec.com). Fluorescence-Activated Cell Sorting Analysis of the Oval Cell-Enriched NP Fraction RBC-depleted, CD45-depleted liver NP cells (1 × 106) were resuspended in PBS. Following Fc blocking, combinations of the following fluorescence-activated cell sorting (FACS) antibodies were added and incubated at 4°C for 30 minutes: CD45 FITC, phycoerythrin (PE), and allophycocyanin (APC); CD34 FITC and PE; Thy 1.2 FITC and PE; c-Kit FITC, Sca-1 FITC, and PE (all from BD Pharmingen); and CD133 FITC and PE (eBioscience). Cells were washed twice with PBS prior to analysis using a FACSCalibur (BD Biosciences). Cells were isolated on a FACSVantage instrument (BD Biosciences). Compensation for FITC, PE, and APC was performed using compensation beads (BD Pharmingen). Analysis was done using the FlowJo program (Tree Star, Ashland, OR, http://www.flowjo.com). Positive and negative gates were determined using IgG-stained and unstained controls. To determine the relative cell size of the parenchymal and NP fractions, cells were analyzed using size calibration beads of 4, 6, 10, and 15 μm (Molecular Probes Inc., Eugene, OR, http://www.probes.invitrogen.com) and 30 μm (G Kisker, Steinfurt, Germany, http://www.kisker-biotech.com) in diameter. Hematopoietic Repopulation Assay Bone marrow was harvested from C57BL/6-Tg (ACTbGFP) 1Osb/5 mice (Jackson Laboratory, Bar Harbor, ME, http://www.jax.org) as described in Wang et al. [31]. These transgenic mice express enhanced green fluorescent protein (eGFP) in all tissues. CD133+CD45−eGFP+ liver NP cells were isolated by FACS from the same strain of mouse after 2 months of DDC 0.1% diet. Cells were counted with trypan blue exclusion to determine numbers of live cells and resuspended in PBS for transplantation at a concentration of 1 × 106 live cells per 50 μl. Six-week-old immune-deficient NOD/SCID/Gama-chain-null mice were transplanted as described [32]. Mice received 270 rads of sublethal irradiation with attenuator and turntable 1 hour pretransplant. Positive control mice were transplanted with 1 × 106 bone marrow cells from eGFP transgenic mice. Blood mononuclear cells were analyzed by FACS for eGFP expression 6 weeks post-transplant. FACS Analysis for BrdU-Labeled Cell Cycling Animals were pulsed with BrdU 12 hours prior to sacrifice. OC-enriched populations (RBC-lysed, CD45-depleted, NP fraction) were labeled with CD133 PE and CD45 APC antibodies as described above. The cells were washed and labeled with rat anti-BrdU FITC-conjugated antibodies (BrdU Flow Kit; BD) per the manufacturer's protocol. Reverse Transcription-Polymerase Chain Reaction Analysis of CD133+CD45− NP Cells CD133+CD45− NP liver cells were isolated on a BD FACSVantage instrument. Cells were pelleted at 200g for 5 minutes, and total RNA was extracted using RNA STAT60 (Tel-Test, Friendswood, TX, http://www.tel-test.com) using chloroform (Sigma-Aldrich) followed by isopropanol (Sigma-Aldrich), per the manufacturers' protocols. RNA was quantified using an ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, http://www.nanodrop.com). Five nanograms of purified RNA per 20-μl reaction volume was used in the synthesis of first-strand cDNA using oligo(dT) reverse transcriptase kit at 37°C for 60 minutes (Invitrogen). Polymerase chain reaction (PCR) was conducted using primers for mouse β-Actin (5′-TGTTACCAACTGGGACGACA-3′ and 5′-GGGGTGTTGAAGGTCTCAAA-3′), Albumin (5′-CATGCCAAATTAGTGCAGGA-3′ and 5′-GCTGGGGTTGTCATCTTTGT-3′), Tyrosine aminotransferase (5′-CCCTACT-GTGTTTGGGAACC-3′ and 5′-GGAGCCTCAGGACAGTGGTA-5′), α1-antitrypsin (5′-GGGTGCTGCTGATGGATTAC-3′ and 5′-GGACAGTCTGGGGATATGGA-3′), Cytokeratin 19 (5′-TGCTGGATGAGCTGACTCTG-3′ and 5′-AATCCACCTCCACACTGACC-3′), Biliary glycoprotein (5′-CACAAGGAGGCCTCTCAGAT-3′ and 5′-GCTGAGGGTTTGTGCTCTGT-3′), Abcg2 transmembrane pump (5′-AGCAGCAAGGAAAGATCCAA-3′ and 5′-GGAAGTCGAAGAGCTGCGA-3′), β-Catenin (5′-TGACACCTCCCAAGTCCTTT-3′ and 5′-CATGCCCTCATCTAGCGTCT-3′), Cyclin D1 (5′-TTGACTGCCGAGAAGTTGTG-3′ and 5′-CTGGCATTTTGGAGAGGAAG-3′), Survivin (5′-CTGATTTGGCCCAGTGTTTT-3′ and 5′-CTTGGCTCTCTGTCTGTCCA-3′), Hnf4α (5′-ACTACGGAGCC-TCGAGCTGT-3′ and 5′-AGCCCGGAAGCACTTCTTA-3′), c-Met (hepatocyte growth factor receptor) (5′-TCTCGAACAGCACACCTCAC-3′ and 5′-AGAGGCACTGACTGCAGGAT-3′), α-Fetoprotein (5′-TCAAGAACTCACCCCAACCT-3′ and 5′-GGCTCTCCTCGATGTGTTTC-3′), α-Smooth muscle actin (5′-GCCGAGATCTCACCGACTAC-3′ and 5′-CTTCTCCAGGGAGGAAGAGG-3′), Desmin (5′-TCGCGGCTAAGAACATCTCT-3′ and 5′-GCATCAATCTCGCAGGTGTA-3′), and CD45 (5′-TCACAAGCATGCATCCATCC-3′ and 5′-TTCCAAGAGATTGAACAAGGCA-3′). All primers were selected in two separate exons to distinguish cDNA from possible contaminating genomic DNA. PCR conditions consisted of 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds for 35 cycles, with a final elongation step at 72°C for 15 minutes. PCR products were run on 2% agarose gel with ethidium bromide and visualized using the Eagle Eye Gel reader (Stratagene, La Jolla, CA, http://www.stratagene.com). Single-Cell Gene Expression Analysis Single CD133+CD45− OCs (n = 45) and single cells from the parenchymal fraction (n = 50) were isolated using a FACSVantage set for single-cell purity using the protocol modified by Hamrouni et al. [33]. Cells were isolated directly into ice-cold RNA Stat 60 in a 1.7-ml microcentrifuge tube. RNA was immediately extracted per the manufacturer's protocol for low cell number. Following cDNA synthesis and reverse transcription (RT), PCR was conducted using primers for β-Actin, Albumin, and Ck19 using identical denaturing, annealing, and elongation conditions as above, for 40 cycles. Experiments included a negative control fraction of water that was loaded at the end of the experiment to assess for contamination, as well as cDNA synthesized from 1 × 106 whole bone marrow cells, which provided a negative control for liver-specific genes. For statistical analysis of the single-cell experiments, each sample was divided into five categories: (a) bipotent cells with Albumin and Ck19 expression; (b) hepatocyte-like, with Albumin expression; (c) cholangiocyte-like, with Ck19 expression; (d) support cells, with β-Actin only; and (e) negative for all three primers. Forty percent of attempts to detect RNA after sorting failed to demonstrate expression of any genes, representing the limits of cell isolation and RNA extraction from a single cell; these events were not included in further statistical analysis. There were 32 cells in each group in which at least one gene was detectable. For statistical analysis, percent bipotent CD133+CD45− cells were counted as the experimental group, and parenchymal cells were counted as the control group. Real-Time PCR of Distinct Damage Models Total RNA was extracted from 1 × 106 CD133+CD45− NP FACS-isolated cells. Ten nanograms of purified mRNA per 20-μl reaction volume was used to construct cDNA. Real-time experiments were conducted by use of an ABI Prism 7700 Thermal Cycler and TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com). Primers for β-Actin were used as loading control, and the level of expression for the genes Albumin, Ck19, Hnf4α, c-Met, and Cyclin D1 was assessed (ready-made real-time primer/probe sets; Applied Biosystems). Amplification efficiency was determined by the ΔCt method of the real-time PCR amplification plots [23]. Fold increase or decrease in gene expression was calculated using undamaged control as a baseline. Cell Culture CD45− cells were isolated from the NP fraction using magnetic bead separation as described above. Once the NP CD45− cell fraction was isolated, the cells were labeled with CD133 magnetic bead antibodies (Miltenyi Biotec). CD133+CD45− cells were isolated using positive cell selection per the manufacturer's protocol. This enriched population of CD133+CD45− cells was plated in six-well BD BioCoat laminin-coated culture plates (BD Biosciences) at a density of 1 × 104 cells per cm2. Culture medium was modified from Suzuki et al. [15]. Medium contained Dulbecco's modified Eagle's medium:Ham's F-12 medium at 1:1 (Sigma-Aldrich) with 10% fetal calf serum, heat-inactivated (Omega Scientific), and the following additives: insulin (1 μg/ml), dexamethasone (1 × 10−7 mol/l), nicotinamide (10 mmol/l), Hepes (5 mmol/l), and penicillin/streptomycin (1% vol/vol) (all from Sigma-Aldrich). Recombinant HGF (50 ng/ml) and EGF (20 ng/ml) (both from Sigma-Aldrich) were added on day 1. Cells cultured without HGF and EGF did not grow. Medium was changed every 3 days. For alkaline phosphatase analysis, cells were fixed in 90% methanol/10% formalin (Sigma-Aldrich) for 1 minute after TBST wash. Cells were stained using the Alkaline Phosphatase Detection Kit (Chemicon, Temecula, CA, http://www.chemicon.com) per the manufacturer's protocol. RNA was extracted directly from the culture well using RNA Stat 60 per the manufacturer's protocol, and RT-PCR was conducted using the same conditions and primers as described above, for 35 cycles. Statistical Analyses Paired, two-tailed Student's t test was used when comparing two groups. p values were generated by comparing the specific value in the undamaged control to the value in the damage model, and p < .05 was considered significant. Fisher's exact test was used to determine significance in single-cell analysis with power .80, and p < .05 was considered significant. Results Oval Cell Enrichment Through DDC Injury and Hematopoietic Cell Depletion As OCs exist at a very low frequency in the normal liver, the DDC model of liver injury, which has been reported to induce OC proliferation, was used to facilitate the identification of candidate OC populations for further analysis [13, 29]. DDC-damaged liver demonstrated periportal expansion, portal zone infiltration, and an increase in OC number (Fig. 1A, 1B) [29], as well as increased expression of A6, a marker of OCs (Fig. 1C, 1D). In DDC-damaged livers, A6 expression was similar to pan-CK antibody staining, another marker of epithelial cells (data not shown) [14]. BrdU cycling analysis by histochemistry demonstrated a significant increase in the proliferation of pan-CK+ cells compared with undamaged controls (22% ± 5% pan-CK+ cells were BrdU+ in DDC model vs. 0.4% ± 0.2% in undamaged control; p < .05). Thus, DDC administration induces cycling and an increase in frequency of OCs. Figure 1. Open in new tabDownload slide 3,5-Diethoxycarbonyl-1,4-dihydrocollidine (DDC) exposure as a model of oval cell proliferation. (A): H&E staining of undamaged control liver with normal bile ducts (arrows). (B): H&E staining of DDC-damaged liver with periportal expansion and oval cell proliferation (box). Fluorescent immunohistochemistry A6 staining in red, nuclei in blue, from undamaged control (C) and DDC-damaged livers (D) with increase in frequency of A6+ oval cells in the DDC model (×20 objective). (E): Table summarizing oval cell markers in CD45+ and CD45− fractions of NP cells (p < .05 for each marker; n = 3 slides per group). Abbreviations: αFP, α-fetoprotein; NP, nonparenchymal. Figure 1. Open in new tabDownload slide 3,5-Diethoxycarbonyl-1,4-dihydrocollidine (DDC) exposure as a model of oval cell proliferation. (A): H&E staining of undamaged control liver with normal bile ducts (arrows). (B): H&E staining of DDC-damaged liver with periportal expansion and oval cell proliferation (box). Fluorescent immunohistochemistry A6 staining in red, nuclei in blue, from undamaged control (C) and DDC-damaged livers (D) with increase in frequency of A6+ oval cells in the DDC model (×20 objective). (E): Table summarizing oval cell markers in CD45+ and CD45− fractions of NP cells (p < .05 for each marker; n = 3 slides per group). Abbreviations: αFP, α-fetoprotein; NP, nonparenchymal. To further explore the immunophenotype of the oval cell population, a technique of liver digestion and density centrifugation was used to separate hepatocytes (the parenchymal fraction) from the smaller cells (the NP fraction) (Fig. 2). The NP fraction was heterogeneous and included hematopoietic cells (Kupffer cells, leukocytes, and RBCs), stellate cells, cholangiocytes, and OCs. CD45+ hematopoietic cells made up approximately 50% of the NP fraction. FACS analysis of the NP fraction for coexpression of CD45 and previously reported OC markers demonstrated that the vast majority of Sca-1-, Thy-1-, CD34-, and c-kit-expressing cells were also CD45+ (data not shown). For further enrichment of the nonhematopoietic cells in the NP fraction, hematopoietic CD45+ cells were removed from NP fraction using magnetic separation. Figure 2. Open in new tabDownload slide Experimental design with enrichment of CD133+ oval cells. Prior to density centrifugation, 45- oval cells were less than 0.05% of the total cell population. By removing the 45+ hematopoietic cells from the NP fraction, the 45- oval cells became 3% of the remaining population in undamaged controls. In the DDC model of liver damage, the 45- oval cells increased from 3% of the enriched NP fraction to 35%, a 600-fold increase from the unfractionated control. Abbreviations: 45-, CD45−; 45+, CD45+; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; FACS, fluorescence-activated cell sorting; NP, nonparenchymal; P, parenchymal. Figure 2. Open in new tabDownload slide Experimental design with enrichment of CD133+ oval cells. Prior to density centrifugation, 45- oval cells were less than 0.05% of the total cell population. By removing the 45+ hematopoietic cells from the NP fraction, the 45- oval cells became 3% of the remaining population in undamaged controls. In the DDC model of liver damage, the 45- oval cells increased from 3% of the enriched NP fraction to 35%, a 600-fold increase from the unfractionated control. Abbreviations: 45-, CD45−; 45+, CD45+; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; FACS, fluorescence-activated cell sorting; NP, nonparenchymal; P, parenchymal. Cytospin analysis of the CD45+ and CD45− fraction from DDC-damaged livers demonstrated a significant increase in the number of cells expressing the OC markers A6, c-Met, and α-fetoprotein in the CD45-depleted fraction compared with the CD45+ cells (Fig. 1E) [34, 35]. Thus, by depleting the hematopoietic fraction, we were able to enrich the NP fraction for OCs. CD133-Expressing Cells Within the Oval Cell Population Most prior studies of OCs have relied upon immunohistochemistry for characterization of OCs [14, 36]. Flow cytometry (FACS) has been successfully used to identify rare populations of hematopoietic stem cells by combinations of markers. We thus applied FACS analysis to explore the different subpopulations present in the NP fraction of liver. Candidate OC populations were analyzed within an “oval cell gate” based on cell size, set by size calibration beads between 6 and 15 μm, a range that includes OCs and excludes most hepatocytes [13]. FACS analysis consistently identified a subpopulation within the CD45− NP fraction that expressed the marker CD133. The frequency of CD133+CD45− NP cells was increased from a baseline of 3.5% ± 1.7% in undamaged controls (Fig. 3B) to 33.5% ± 9.4% in DDC-damaged livers (Fig. 3C) (p < .05). Figure 3A demonstrates isotype control staining within the CD45-depleted NP fraction. Figure 3D summarizes the results of six separate FACS isolation experiments using undamaged controls and DDC-damaged livers. Thus, the overall enrichment process was 600-fold, from <0.05% of unfractionated, undamaged liver to 33.5% of the CD45− NP fraction from the DDC-damaged liver (Fig. 2). Figure 3. Open in new tabDownload slide Characterization of CD133+CD45− oval cells. Representative fluorescence-activated cell sorting plots of CD45-depleted NP cells. (A): Isotype-stained sample, undamaged control. CD133 versus CD45 in undamaged control (B) and 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC)-damaged liver (C), with significant increase in CD133+CD45− NP cells in DDC damage model. (D): The population of CD133+CD45− NP cells was significantly increased in the DDC damage model compared with undamaged controls (p < .05). (E, F): Fluorescent immunohistochemistry demonstrating CD133 staining (red) and nuclei (blue) in undamaged control (E) and DDC model (F). In the DDC model, there was CD133 staining (red) with albumin costaining (green) (G) or pan-cytokeratin costaining (green) (H); many cells with coexpression (orange) were identified (×20 objective). Abbreviations: FITC, fluorescein isothiocyanate; NP, nonparenchymal; PE, phycoerythrin. Figure 3. Open in new tabDownload slide Characterization of CD133+CD45− oval cells. Representative fluorescence-activated cell sorting plots of CD45-depleted NP cells. (A): Isotype-stained sample, undamaged control. CD133 versus CD45 in undamaged control (B) and 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC)-damaged liver (C), with significant increase in CD133+CD45− NP cells in DDC damage model. (D): The population of CD133+CD45− NP cells was significantly increased in the DDC damage model compared with undamaged controls (p < .05). (E, F): Fluorescent immunohistochemistry demonstrating CD133 staining (red) and nuclei (blue) in undamaged control (E) and DDC model (F). In the DDC model, there was CD133 staining (red) with albumin costaining (green) (G) or pan-cytokeratin costaining (green) (H); many cells with coexpression (orange) were identified (×20 objective). Abbreviations: FITC, fluorescein isothiocyanate; NP, nonparenchymal; PE, phycoerythrin. To define the intrahepatic location of CD133+ cells, FIHC was performed. Liver sections from control animals demonstrated sparse CD133+ cells, the majority of which were periportal (Fig. 3E). After DDC treatment, CD133+ cells were detected extending out from the portal areas in groups (Fig. 3F). Coexpression of albumin (Fig. 3G) and duct epithelial cytokeratins (Fig. 3H) was demonstrated in some of the CD133+ cells. Cell cycling analysis after a 12-hour BrdU pulse showed that nuclear incorporation of BrdU was significantly increased in CD133+ cells from DDC-damaged livers compared with CD133+ cells from undamaged controls (22.3% vs. 2.7%; p < .05) (Fig. 4A, 4B). Thus, the population of CD133+ cells had the expected location and morphology of OCs and were increased in number by cell cycling. Figure 4. Open in new tabDownload slide BRDU cycling and gene expression analysis of CD133+CD45− oval cells (OCs). (A, B): CD133+CD45− OCs from control undamaged (solid red) and DDC damage model were gated for BRDU analysis. It was seen that 2.7% of CD133+CD45− OCs from control and 22.3% from DDC damage model demonstrated BRDU staining; thus, there are more CD133+CD45− OCs cycling in response to chronic injury. (C, D): Reverse transcription-polymerase chain reaction (PCR) from CD133+CD45− nonparenchymal (NP) cells: undamaged control (C) and DDC damage model (D). Bilineage gene expression with upregulation of OC genes was demonstrated in CD133+CD45− NP cells after DDC damage. (E): Real-time PCR showing upregulation of gene expression in CD133+CD45− OCs from DDC damage model compared with uninjured control: Alb, CK19, HNF4, Cyc, and Cmet (Act housekeeping gene). Real-time data are expressed as the mean from three isolation experiments, each test performed in triplicate, with fold increase in gene expression of CD133+CD45− OCs from the DDC model over undamaged controls, when expression was normalized to Act. Abbreviations: A1AT, α1-Antitrypsin; Act, β-Actin; AFP, α-Fetoprotein; Alb, Albumin; APC, allophycocyanin; Bcat, β-Catenin; BG, Biliary glycoprotein; Bp, base pairs; BRDU, bromodeoxyuridine; CK19, Cytokeratin 19; Cmet, Hepatocyte growth factor receptor; Cyc, Cyclin D1; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; FITC, fluorescein isothiocyanate; G6P, Glucose 6-phosphatase; HNF4, Hepatocyte nuclear factor 4α; Max, maximum; PE, phycoerythrin; TAT, Tyrosine aminotransferase. Figure 4. Open in new tabDownload slide BRDU cycling and gene expression analysis of CD133+CD45− oval cells (OCs). (A, B): CD133+CD45− OCs from control undamaged (solid red) and DDC damage model were gated for BRDU analysis. It was seen that 2.7% of CD133+CD45− OCs from control and 22.3% from DDC damage model demonstrated BRDU staining; thus, there are more CD133+CD45− OCs cycling in response to chronic injury. (C, D): Reverse transcription-polymerase chain reaction (PCR) from CD133+CD45− nonparenchymal (NP) cells: undamaged control (C) and DDC damage model (D). Bilineage gene expression with upregulation of OC genes was demonstrated in CD133+CD45− NP cells after DDC damage. (E): Real-time PCR showing upregulation of gene expression in CD133+CD45− OCs from DDC damage model compared with uninjured control: Alb, CK19, HNF4, Cyc, and Cmet (Act housekeeping gene). Real-time data are expressed as the mean from three isolation experiments, each test performed in triplicate, with fold increase in gene expression of CD133+CD45− OCs from the DDC model over undamaged controls, when expression was normalized to Act. Abbreviations: A1AT, α1-Antitrypsin; Act, β-Actin; AFP, α-Fetoprotein; Alb, Albumin; APC, allophycocyanin; Bcat, β-Catenin; BG, Biliary glycoprotein; Bp, base pairs; BRDU, bromodeoxyuridine; CK19, Cytokeratin 19; Cmet, Hepatocyte growth factor receptor; Cyc, Cyclin D1; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; FITC, fluorescein isothiocyanate; G6P, Glucose 6-phosphatase; HNF4, Hepatocyte nuclear factor 4α; Max, maximum; PE, phycoerythrin; TAT, Tyrosine aminotransferase. CD133+CD45− Oval Cells Demonstrate No Significant Hematopoietic Repopulating Capacity eGFP+CD133+CD45− OCs were isolated by FACS from eGFP transgenic mice and injected into sublethally irradiated, immune-deficient mice as a means of assaying hematopoietic repopulating capacity. As expected, eGFP+CD133+CD45− OC possessed no significant repopulating capacity. In contrast, 92% ± 3% of blood mononuclear cells were of donor origin in control mice transplanted with the same number of eGFP+ bone marrow cells. (p < .05) (supplemental online Fig. 1). CD133+CD45− Oval Cells Exhibit Bilineage and Stem Cell Gene Expression Profile RNA was extracted from FACS-isolated CD133+CD45− OCs from DDC-injured and uninjured control animals to determine the gene expression profile of this specific population. Postisolation, FACS analysis confirmed a high cell purity of the isolated fraction (CD133+CD45− OCs were 97.5% ± 3% of isolated fraction). CD133+CD45− OCs isolated by FACS were studied for expression of hepatocyte genes (Albumin, α1-antitrypsin, Glycogen-6-phosphatase, Tyrosine-aminotransferase) [37], cholangiocyte genes (Ck19, Biliary glycoprotein) [38], and other genes associated with primitive cells capable of tissue regeneration, such as Hnf4α and αfp. Six separate isolation experiments using liver cells from undamaged controls and DDC damage models were conducted; the CD133+CD45− OCs demonstrated gene expression profiles consistent with a primitive population of stem or progenitor cells with hepatocyte and cholangiocyte lineage gene expression. For example, DDC damage induced expression of Hnf4α, a transcription factor of fetal hepatoblast differentiation, and αfp, a marker of liver regeneration [39]. Furthermore, elements of the wnt/β-catenin signal pathway (β-catenin, Cyclin D1, and Survivin) [40] were upregulated in the CD133+ OCs isolated from the DDC model, as was Abcg2, a transmembrane pump associated with the side-population cells [24]. In addition, DDC damage upregulated expression of c-Met, the hepatocyte growth factor receptor [41], in the CD133+ OCs (Fig. 4C, 4D). CD133+ OCs did not express the stellate cell markers α-Smooth muscle actin and Desmin (supplemental online Fig. 2). Real-time PCR confirmed the upregulation of selected genes observed in CD133+ OCs from the DDC damage model. The CD133+ OCs demonstrated a 6.5-fold increase in Albumin expression and 10.5-fold increase in Ck19 expression when isolated from the DDC treated animals, compared with undamaged controls. Furthermore, this population demonstrated threefold increase in Cyclin D1 expression, eightfold increase in Hnf4α expression, and sixfold increase in C-Met expression (Fig. 4E). Gene expression analysis of CD133+CD45− OCs indicated that this population has bilineage gene expression, with upregulation of primitive liver and stem cell-associated genes. Assessment of Previously Published Oval Cell Markers The CD45-depleted NP fraction was labeled with fluorescent antibodies previously reported to mark OCs: Thy-1, Sca-1, CD34, and c-kit (Fig. 2). The percentage of Sca-1+CD45− liver cells increased significantly after 4 weeks of DDC damage (7% in DDC vs. 2.5% in undamaged controls; p < .05); Thy-1 and CD34 markers were also slightly increased: Thy-1+CD45− cells, 2.2% to 1.2% (not significant); and CD34+CD45− cells, 2.4% to 0.9% (not significant). The percentage of c-kit+CD45− cells was very low and did not change with DDC treatment. CD34+, Sca-1+, and Thy-1+ cell populations were isolated by FACS from the CD45− NP population for gene expression analysis. None of these cell population consistently demonstrated both hepatocyte and cholangiocyte gene coexpression. Furthermore, these populations failed to consistently demonstrate expression of other genes associated with oval cells, such as αfp and Hnf4α (data not shown). Single-Cell Analysis Confirms Bilineage Gene Expression The use of single-cell analysis is necessary to document that populations of cells with a common immunophenotype are homogeneous [19]. Thus, gene expression on single cells isolated by FACS was used to investigate whether CD133+CD45− OCs represented a homogeneous population of bipotent liver progenitors, rather than two distinct populations of cholangiocyte- or hepatocyte-specific progenitors [33, 42]. Of single CD133+CD45− OCs isolated from the DDC model, 68% demonstrated transcripts of both a hepatocyte gene (Albumin) and a cholangiocyte gene (Ck19) (Fig. 5A, lanes 2, 3, 4, 6, 10, 11, and 12; Fig. 5D, summary table). In contrast, single cells isolated from the parenchymal fraction, the majority of which is hepatocytes, were predominately unipotent, with 66% expressing only Albumin (Fig. 5B, lanes H3, H5–H12; Fig. 5D, summary table) (n = 32 cells of both CD133+CD45− OCs and parenchymal cells isolated from four animals; p < .05, power >98 using Fisher's exact test). In the majority of cases, β-actin message was coexpressed with other genes; however, in rare cells (Fig. 5B, hepatocyte H5 or H12) β-Actin was not detected despite the presence of Albumin or Ck19 message, representing the limits of detection of RNA from single cells. Rare single parenchymal cells demonstrated Ck19 expression, indicating that the cell isolated from this heterogeneous population was likely a cholangiocyte (Fig. 5B, H1). Thus, single-cell analysis demonstrated that the CD133+CD45− OC population is significantly enriched for cells with bilineage gene expression. Figure 5. Open in new tabDownload slide Bilineage gene expression of single CD133+CD45− oval cells (OCs). Shown are representative reverse transcription-polymerase chain reaction (RT-PCRs) from single CD133+CD45− OCs isolated from the 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) model (A) and single cells isolated from the parenchymal fraction of DDC model (B). (C): Water was used as negative control, whole liver was positive control, and bone marrow was negative control for liver-specific genes. (D): Summary of single-cell RT-PCR experiments: 68.75% of CD133+ OCs demonstrated bilineage gene expression compared with 6.25% of isolated parenchymal cells (p < .05, power 98, using Fisher's exact test). Any cell with Alb expression ± Act expression was tabulated under Alb, and likewise for Ck-19. Cells were counted as negative for both lineages if they demonstrated only Act expression. Any cell with no detectable gene expression in all three genes (Act, Alb, or Ck-19) was discarded from the analysis (40% for CD133+ OCs and 45% for parenchymal cells). Abbreviations: Act, β-Actin; Alb, Albumin; Ck-19, Cytokeratin 19; Kb, kilobases. Figure 5. Open in new tabDownload slide Bilineage gene expression of single CD133+CD45− oval cells (OCs). Shown are representative reverse transcription-polymerase chain reaction (RT-PCRs) from single CD133+CD45− OCs isolated from the 3,5-diethoxycarbonyl-1,4-dihydrocollidine (DDC) model (A) and single cells isolated from the parenchymal fraction of DDC model (B). (C): Water was used as negative control, whole liver was positive control, and bone marrow was negative control for liver-specific genes. (D): Summary of single-cell RT-PCR experiments: 68.75% of CD133+ OCs demonstrated bilineage gene expression compared with 6.25% of isolated parenchymal cells (p < .05, power 98, using Fisher's exact test). Any cell with Alb expression ± Act expression was tabulated under Alb, and likewise for Ck-19. Cells were counted as negative for both lineages if they demonstrated only Act expression. Any cell with no detectable gene expression in all three genes (Act, Alb, or Ck-19) was discarded from the analysis (40% for CD133+ OCs and 45% for parenchymal cells). Abbreviations: Act, β-Actin; Alb, Albumin; Ck-19, Cytokeratin 19; Kb, kilobases. CD133+CD45− Oval Cell Response to Lineage-Specific Liver Injury Models The oval cell proliferation seen with DDC administration is associated with both hepatocyte and cholangiocyte damage [43]. In view of the bilineage gene expression pattern noted in the CD133+CD45− OCs isolated from the DDC injured liver, we hypothesized that CD133+CD45− OCs may have the potential to differentiate preferentially into either hepatocytes or cholangiocytes in response to liver injury specific to each lineage. Chronic CCl4 administration induced cycling specifically in hepatocytes (Fig. 6B, 6D). Using quantitative RT-PCR, CD133+CD45− OCs isolated from the CCl4 injury model demonstrated a 36-fold increase in Albumin expression compared with undamaged controls (Fig. 6E). Chronic ANIT administration induced cycling in cholangiocytes (Fig. 6H, 6I). Using quantitative RT-PCR, CD133+CD45− OCs isolated from the ANIT-damaged livers demonstrated a sixfold increase in Ck19 expression, with no associated increase in Albumin expression compared with undamaged controls (Fig. 6J). These quantitative data from two distinct liver damage models demonstrate that CD133+CD45− OCs provide an in vivo lineage-specific response to hepatocyte or cholangiocyte injury. Figure 6. Open in new tabDownload slide Lineage-specific gene expression after liver damage. (A–C): Fluorescent immunohistochemistry (FIHC) with BRDU cycling hepatocytes in undamaged control (A), CCl4 (B), and ANIT injury (C) models, anti-albumin (green), anti-BRDU antibody (red), and nuclei (blue). There was a significant increase in cycling hepatocytes in CCl4 model (D) compared with control. (E): Real-time polymerase chain reaction (PCR) data of RNA from isolated CD133+CD45− oval cells (OCs) demonstrates upregulation of Albumin expression in hepatocyte injury model (CCl4). (E–G): FIHC with BRDU cycling cholangiocytes in undamaged control (E), CCl4 (F), and ANIT (G); cholangiocytes with anti-pan-cytokeratin (green), anti-BRDU antibody (red), and nuclei (blue). (All images, ×20 objective). (H): Significant increase in cycling cholangiocytes in ANIT model. (E, H): ∗︁ indicates p < .05, n = 3 animals for each group, five sections per animal. (J): Real-time PCR data of RNA from isolated CD133+CD45− OCs demonstrated upregulation Ck19 expression in cholangiocyte injury model (ANIT). (E, J): Fold increase for real-time PCR expression relative to undamaged control liver, n = 3, each performed in triplicate, normalized to β-Actin expression. Abbreviations: ANIT, α-naphthylisothiocyanate; BRDU, bromodeoxyuridine; CCl4, carbon tetrachloride. Figure 6. Open in new tabDownload slide Lineage-specific gene expression after liver damage. (A–C): Fluorescent immunohistochemistry (FIHC) with BRDU cycling hepatocytes in undamaged control (A), CCl4 (B), and ANIT injury (C) models, anti-albumin (green), anti-BRDU antibody (red), and nuclei (blue). There was a significant increase in cycling hepatocytes in CCl4 model (D) compared with control. (E): Real-time polymerase chain reaction (PCR) data of RNA from isolated CD133+CD45− oval cells (OCs) demonstrates upregulation of Albumin expression in hepatocyte injury model (CCl4). (E–G): FIHC with BRDU cycling cholangiocytes in undamaged control (E), CCl4 (F), and ANIT (G); cholangiocytes with anti-pan-cytokeratin (green), anti-BRDU antibody (red), and nuclei (blue). (All images, ×20 objective). (H): Significant increase in cycling cholangiocytes in ANIT model. (E, H): ∗︁ indicates p < .05, n = 3 animals for each group, five sections per animal. (J): Real-time PCR data of RNA from isolated CD133+CD45− OCs demonstrated upregulation Ck19 expression in cholangiocyte injury model (ANIT). (E, J): Fold increase for real-time PCR expression relative to undamaged control liver, n = 3, each performed in triplicate, normalized to β-Actin expression. Abbreviations: ANIT, α-naphthylisothiocyanate; BRDU, bromodeoxyuridine; CCl4, carbon tetrachloride. Functional Analysis of CD133+ Oval Cells In Vitro To further characterize lineage potential, CD133+CD45− OCs were isolated from control undamaged liver and the livers of mice treated with DDC, ANIT, and CCl4, using a two-step enrichment process of magnetic filtration. After enrichment, CD133+CD45− OCs were plated in six-well, laminin-coated tissue culture plates using a fetal hepatoblast-derived medium [15]. On day 1, small round cells were observed adhering to the plate (Fig. 7A, ×20 inset). After 3 weeks in culture, clusters of cells were observed growing from noninjured liver cells and all three liver injury models. The cells isolated from the livers without injury demonstrated the least proliferation and overall diversity of cell type, with the majority of the cells having a flat, round appearance (Fig. 7B). The cells from the DDC injury model demonstrated the most rapid initial growth and had morphology ranging from flat, cuboidal cells in sheets to more elongated cells with projections (Fig. 7B). Cells from the ANIT group demonstrated a similar range of cell types after 21 days, with the majority having a more elongated appearance (Fig. 7B). Lastly, cells isolated from the CCl4 damage model demonstrated more cuboidal or round appearance, with tight clusters of cells (Fig. 7B). Representative images shown in Figure 7B demonstrate the range of cell type, although each cell type shown was observed in each model. Cells cultured in identical conditions without HGF and EGF failed to form colonies and remained as single cells (data not shown). Figure 7. Open in new tabDownload slide Culture of CD133+ oval cells (OCs) generated cells with expression of hepatocyte and cholangiocyte markers. CD133+CD45− nonparenchymal cells from control undamaged livers, DDC model, CCl4 model, and ANIT model. (A): Phase images from day 1 showed similar appearance of round cells adherent to the culture well (×10 objective; inset, ×20). (B): Phase images from day 21 with a variety of cell phenotypes (×10). (C): Alkaline phosphatase staining demonstrated activity in individual and clusters of cells with biliary-epithelial phenotype (×20). (D): Cells with c-Met surface expression found in cells from all cultures (×20). (E): After 21 days in vitro, gene expression analysis demonstrated bilineage and OC-associated genes in each culture, with the absence of stellate cell-associated gene expression. Abbreviations: α1AT, α1-Antitrypsin; αFP, α-Fetoprotein; αSMA, α-Smooth muscle actin; ACT, β-Actin; ALB, Albumin; ANIT, α-naphthylisothiocyanate; BG, Biliary glycoprotein; Ck19, Cytokeratin 19; c-Met, C-Met; Cont, control; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; DES, Desmin; G6P, Glucose 6-phosphatase; Hnf4a, Hepatocyte nuclear factor 4α; RT-PCR, reverse transcription polymerase chain reaction; TAT, Tyrosine aminotransferase. Figure 7. Open in new tabDownload slide Culture of CD133+ oval cells (OCs) generated cells with expression of hepatocyte and cholangiocyte markers. CD133+CD45− nonparenchymal cells from control undamaged livers, DDC model, CCl4 model, and ANIT model. (A): Phase images from day 1 showed similar appearance of round cells adherent to the culture well (×10 objective; inset, ×20). (B): Phase images from day 21 with a variety of cell phenotypes (×10). (C): Alkaline phosphatase staining demonstrated activity in individual and clusters of cells with biliary-epithelial phenotype (×20). (D): Cells with c-Met surface expression found in cells from all cultures (×20). (E): After 21 days in vitro, gene expression analysis demonstrated bilineage and OC-associated genes in each culture, with the absence of stellate cell-associated gene expression. Abbreviations: α1AT, α1-Antitrypsin; αFP, α-Fetoprotein; αSMA, α-Smooth muscle actin; ACT, β-Actin; ALB, Albumin; ANIT, α-naphthylisothiocyanate; BG, Biliary glycoprotein; Ck19, Cytokeratin 19; c-Met, C-Met; Cont, control; DDC, 3,5-diethoxycarbonyl-1,4-dihydrocollidine; DES, Desmin; G6P, Glucose 6-phosphatase; Hnf4a, Hepatocyte nuclear factor 4α; RT-PCR, reverse transcription polymerase chain reaction; TAT, Tyrosine aminotransferase. In terms of functional analysis, alkaline phosphatase staining demonstrated that there were single cells and clusters of cells with alkaline phosphatase activity in all groups (Fig. 7C). Alkaline phosphatase staining has been used to identify primitive embryonic stem cells, as well as biliary epithelium [44, 45]. Individual cells and clusters of cells with c-Met expression were demonstrated in all groups, indicating the presence of HGF-responsive cells (Fig. 7D). To further characterize the cells growing after 21 days, gene expression analysis was conducted. RT-PCR demonstrated that each culture contained cells from both hepatocyte and cholangiocyte lineages (Fig. 7E). OC markers such as αfp and Hnf4α were also expressed in each culture, as was c-Met (Fig. 7E). The cultured cells failed to demonstrate any expression of stellate cell markers (α-Smooth muscle actin and Desmin; Fig. 7E). Therefore, CD133+ OCs proliferate in vitro in the presence of HGF and EGF and generate both cholangiocyte and hepatocyte lineages. Discussion The role of adult liver stem cells in liver regeneration is currently not known. The ability to identify and isolate adult liver stem cells is an essential first step in studying their role in liver regeneration and their potential for use in cell transplantation and as targets of other therapeutic modalities [46]. Reliance on histologic localization to identify OC risks inclusion of hematopoietic and other cells in any analysis. This is particularly a problem in models of liver damage, such as DDC, that induce marked infiltration of inflammatory cells [47]. The surface markers previously attributed to oval cells in rodents and humans, such as c-kit, CD34, Thy-1, and Sca-1, are traditionally assigned to hematopoietic stem cells [11, 17, 48, 49]. Moreover, one study reported that oval cells express CD45, a pan-leukocyte antigen normally used to discriminate hematopoietic from nonhematopoietic cells [11]. Thus, definitive OC identification requires not only the traditional fractionation of cells based on size but also the exclusion of contaminating hematopoietic cells. Single-cell clonal studies in fetal liver have identified fetal hepatoblasts that are capable of both cholangiocyte and hepatocyte differentiation [15, 32]. In adult animals, OCs are often described as bipotent progenitors; however, no reports to date have shown that OCs in the adult liver have bilineage potential as single cells. One study demonstrated that OCs, fractionated on the basis of size, were able to reconstitute hepatocytes in fumarylacetoacetate hydrolase−/− mice, a model of the disease tyrosinemia [13]. In this model, wild-type adult hepatocytes were more efficient than wild-type OCs in their contribution of hepatocyte regeneration; cholangiocyte regeneration was not reported. By identifying a more defined subpopulation of OCs using CD133 surface expression, we reasoned that it is possible to identify rarer progenitor cells with a wider range of lineage potential from the NP liver fraction. CD133, or Prominin 1, was originally described in 1997 as being localized to the apical membrane protrusions of murine neuroepithelial cells [50]. CD133 is also expressed on stem cells of the central nervous system and both normal and malignant stem cells of epithelial origin [25]. A marker with similar homology was described later in 1997 on the surface of human hematopoietic stem cells [51]. Both human and mouse CD133 have five membrane-spanning domains and share 60% homology at the protein level [52]. CD133 mRNA was detected in rat oval cells, isolated by size, as part of a microarray screen [53]. CD133 was recently reported by Kordes et al. to be expressed on hepatic stellate cells [54], which, when cultured with certain growth factors, generated progeny that expressed hepatocyte markers. However, in that study, analysis was performed on all CD133+ cells in the nonparenchymal fraction of rat liver and thus included hematopoietic cells. Our recently published work demonstrated that up to 50% of the NP fraction of the DDC-damaged liver contains CD45+ hematopoietic cells [47]. Although the exact origin of stellate cells remains controversial, recent work indicates that CD45+ fibrocytes in the liver are derived from bone marrow [55]. Recently, our group has defined a unique relationship between fibroblast growth factor 10-expressing stellate cells and fibroblast growth factor receptor 2b+ hepatoblasts within the embryonic liver [56], suggesting a role for stellate cells as part of the microenvironment required for proliferation of hepatoblasts. The current report demonstrates that CD133 identifies a subpopulation of nonparenchymal cells in the adult liver with the traditionally accepted morphology, size, and location of oval cells. CD133+CD45− OCs proliferated and significantly expanded in number after DDC-induced damage, a model commonly accepted to induce oval cell proliferation. The CD133+CD45− OC population expressed markers associated with fetal hepatoblasts and several independent markers of cholangiocyte and hepatocyte lineages. Of note, the CD133+CD45− oval cells did not express the stellate cell markers (αSma and Desmin) either in the freshly isolated state or after 21 days of culture, further supporting the conclusion that the population described here is distinct from that reported by Kordes et al. [54]. Consistent with the concept of organ-specific adult stem cells, liver injury induced both proliferation and differentiation signals in CD133+CD45− OCs. BrdU uptake showed a marked increase in cell cycling within the CD133+CD45− OC population. β-Catenin regulation of liver regeneration has been well described in hepatocytes [57]. Upregulation of the β-catenin signal pathway genes Cyclin D1 and Survivin in CD133+CD45− OCs during DDC induced injury indicates a potential mechanism of cell proliferation during liver injury. CD133+CD45− OCs upregulated the HGF receptor c-Met in response to liver damage. HGF/c-Met signaling has been shown to stimulate expression of HNF4α in models of oval cell proliferation [41], and HGF/c-Met signaling has been implicated in noncanonical activation of the β-catenin pathway [58]. Thus, HGF/c-Met may mediate the upregulation of HNF4α, Cyclin D1, and Survivin in CD133+CD45− OCs. Furthermore, HNF4α has been well described as a master regulator of hundreds of fetal hepatoblast differentiation genes, including Tyrosine aminotransferase and Glucose 6-phosphatase, two genes also upregulated in CD133+CD45− OCs after liver damage [59, 60]. This gene expression profile showing upregulation of Cyclin D1, Survivin, and Hnf4α is similar to the profile identified by previous studies using transcriptome analysis of whole liver after OC proliferation [22, 23]. However, these prior studies did not isolate specific populations of OCs prior to analysis, and thus gene expression data reflected net changes within the many different hematopoietic and nonhematopoietic cells of the liver. Analysis of single CD133+CD45− OCs demonstrated that the bilineage gene expression occurred clonally rather than from two unipotent subpopulations. Functional analysis of the CD133+CD45− OCs demonstrated that this population is capable of growth and differentiation in vitro. Furthermore, CD133+CD45− OCs demonstrated a lineage-appropriate response to specific types of liver injury. These data demonstrate that the CD133+CD45− OCs are not unique to the DDC injury model and respond to other toxic models of liver injury. In summary, we have dissected the immunophenotypes of the NP fraction of the liver using flow cytometry and identified the molecular phenotype of highly purified cells using single-cell RT-PCR. Cell surface markers previously associated with OCs were found to be expressed largely on contaminating hematopoietic cells. Expression of the epithelial stem cell antigen CD133 consistently identified a nonhematopoietic subpopulation of OCs with characteristics consistent with a bipotent liver progenitor. Disclosure of Potential Conflicts of Interest The authors indicate no potential conflicts of interest. Acknowledgements We acknowledge Neil Kaplowitz, M.D., for insight into different mechanism of liver injury; the Animal Care Facility at The Saban Research Institute, Children's Hospital Los Angeles (CHLA); The Flow Cytometry Core of the Gene, Immunology, and Stem Cell Program, CHLA; and George McNamara, Ph.D. and the Congressman Dixon Imaging Core at CHLA. C.B.R. is a National Institute of Child Health and Human Development (NICHD) Fellow of the Pediatric Scientist Development Program (NICHD Grant Award K12-HD00850) and a recipient of the AGA/AstraZeneca Fellow/Faculty Transition Award. This work was supported by Pilot and Feasibility Grant DK48522 from the University of Southern California Research Center for Liver Disease (to G.M.C.) References 1 Fausto N , Lemire JM, Shiojiri N. Cell lineages in hepatic development and the identification of progenitor cells in normal and injured liver . Proc Soc Exp Biol Med 1993 ; 204 : 237 – 241 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Glaser S , Francis H, Demorrow S et al. Heterogeneity of the intrahepatic biliary epithelium . World J Gastroenterol 2006 ; 12 : 3523 – 3536 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Lemaigre F , Zaret KS. Liver development update: New embryo models, cell lineage control, and morphogenesis . Curr Opin Genet Dev 2004 ; 14 : 582 – 590 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Sell S . Heterogeneity and plasticity of hepatocyte lineage cells . Hepatology 2001 ; 33 : 738 – 750 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Fausto N , Campbell JS, Riehle KJ. Liver regeneration . Hepatology 2006 ; 43 : S45 – S53 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Theise ND . Gastrointestinal stem cells. III. Emergent themes of liver stem cell biology: Niche, quiescence, self-renewal, and plasticity . Am J Physiol Gastrointest Liver Physiol 2006 ; 290 : G189 – G193 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Crosby HA , Strain AJ. Adult liver stem cells: Bone marrow, blood, or liver derived? Gut 2001 ; 48 : 153 – 154 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Yasui O , Miura N, Terada K et al. Isolation of oval cells from Long-Evans Cinnamon rats and their transformation into hepatocytes in vivo in the rat liver . Hepatology 1997 ; 25 : 329 – 334 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 9 Petersen BE , Zajac VF, Michalopoulos GK. Hepatic oval cell activation in response to injury following chemically induced periportal or pericentral damage in rats . Hepatology 1998 ; 27 : 1030 – 1038 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Arai M , Yokosuka O, Fukai K et al. Gene expression profiles in liver regeneration with oval cell induction . Biochem Biophys Res Commun 2004 ; 317 : 370 – 376 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Petersen BE , Grossbard B, Hatch H et al. Mouse A6-positive hepatic oval cells also express several hematopoietic stem cell markers . Hepatology 2003 ; 37 : 632 – 640 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Newsome PN , Hussain MA, Theise ND. Hepatic oval cells: Helping redefine a paradigm in stem cell biology . Curr Top Dev Biol 2004 ; 61 : 1 – 28 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Wang X , Foster M, Al-Dhalimy M et al. The origin and liver repopulating capacity of murine oval cells . Proc Natl Acad Sci U S A 2003 ; 100 (suppl 1) : 11881 – 11888 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 14 Kofman AV , Morgan G, Kirschenbaum A et al. Dose- and time-dependent oval cell reaction in acetaminophen-induced murine liver injury . Hepatology 2005 ; 41 : 1252 – 1261 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Suzuki A , Zheng Y, Kondo R et al. Flow-cytometric separation and enrichment of hepatic progenitor cells in the developing mouse liver . Hepatology 2000 ; 32 : 1230 – 1239 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Petersen BE , Goff JP, Greenberger JS et al. Hepatic oval cells express the hematopoietic stem cell marker Thy-1 in the rat . Hepatology 1998 ; 27 : 433 – 445 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Crosby HA , Kelly DA, Strain AJ. Human hepatic stem-like cells isolated using c-kit or CD34 can differentiate into biliary epithelium . Gastroenterology 2001 ; 120 : 534 – 544 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Engelhardt NV , Factor VM, Medvinsky AL et al. Common antigen of oval and biliary epithelial cells (A6) is a differentiation marker of epithelial and erythroid cell lineages in early development of the mouse . Differentiation 1993 ; 55 : 19 – 26 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Krause DS , Theise ND, Collector MI et al. Multi-organ, multi-lineage engraftment by a single bone marrow-derived stem cell . Cell 2001 ; 105 : 369 – 377 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Grompe M . The role of bone marrow stem cells in liver regeneration . Semin Liver Dis 2003 ; 23 : 363 – 372 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 21 Menthena A , Deb N, Oertel M et al. Bone marrow progenitors are not the source of expanding oval cells in injured liver . Stem Cells 2004 ; 22 : 1049 – 1061 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Batusic DS , Cimica V, Chen Y et al. Identification of genes specific to “oval cells” in the rat 2-acetylaminofluorene/partial hepatectomy model . Histochem Cell Biol 2005 ; 124 : 245 – 260 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Cimica V , Batusic D, Chen Y et al. Transcriptome analysis of rat liver regeneration in a model of oval hepatic stem cells . Genomics 2005 ; 86 : 352 – 364 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Shimano K , Satake M, Okaya A et al. Hepatic oval cells have the side population phenotype defined by expression of ATP-binding cassette transporter ABCG2/BCRP1 . Am J Pathol 2003 ; 163 : 3 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Uchida N , Buck DW, He D et al. Direct isolation of human central nervous system stem cells . Proc Natl Acad Sci U S A 2000 ; 97 : 14720 – 14725 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Corbeil D , Roper K, Hellwig A et al. The human AC133 hematopoietic stem cell antigen is also expressed in epithelial cells and targeted to plasma membrane protrusions . J Biol Chem 2000 ; 275 : 5512 – 5520 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Roper K , Corbeil D, Huttner WB. Retention of prominin in microvilli reveals distinct cholesterol-based lipid micro-domains in the apical plasma membrane . Nat Cell Biol 2000 ; 2 : 582 – 592 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Giebel B , Corbeil D, Beckmann J et al. Segregation of lipid raft markers including CD133 in polarized human hematopoietic stem and progenitor cells . Blood 2004 ; 104 : 2332 – 2338 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Cochon AC , Gonzalez N, San Martin de Viale LC. Effects of the porphyrinogenic compounds hexachlorobenzene and 3,5-diethoxycarbonyl-1,4-dihydrocollidine on polyamine metabolism . Toxicology 2002 ; 176 : 209 – 219 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Jean PA , Roth RA. Naphthylisothiocyanate disposition in bile and its relationship to liver glutathione and toxicity . Biochem Pharmacol 1995 ; 50 : 1469 – 1474 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Wang X , Ge S, Gonzalez I et al. Formation of pancreatic duct epithelium from bone marrow during neonatal development . Stem Cells 2006 ; 24 : 307 – 314 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Suzuki A , Zheng YW, Kaneko S et al. Clonal identification and characterization of self-renewing pluripotent stem cells in the developing liver . J Cell Biol 2002 ; 156 : 173 – 184 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Hamrouni A , Aublin A, Guillaume P et al. T cell receptor gene rearrangement lineage analysis reveals clues for the origin of highly restricted antigen-specific repertoires . J Exp Med 2003 ; 197 : 601 – 614 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Dumble ML , Croager EJ, Yeoh GC et al. Generation and characterization of p53 null transformed hepatic progenitor cells: Oval cells give rise to hepatocellular carcinoma . Carcinogenesis 2002 ; 23 : 435 – 445 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Suzuki A , Zheng YW, Fukao K et al. Liver repopulation by c-Met-positive stem/progenitor cells isolated from the developing rat liver . Hepatogastroenterology 2004 ; 51 : 423 – 426 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 36 Paku S , Nagy P, Kopper L et al. 2-Acetylaminofluorene dose-dependent differentiation of rat oval cells into hepatocytes: Confocal and electron microscopic studies . Hepatology 2004 ; 39 : 1353 – 1361 . Google Scholar Crossref Search ADS PubMed WorldCat 37 Runge D , Michalopoulos GK, Strom SC et al. Recent advances in human hepatocyte culture systems . Biochem Biophys Res Commun 2000 ; 274 : 1 – 3 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Saxena R , Theise N. Canals of Hering: Recent insights and current knowledge . Semin Liver Dis 2004 ; 24 : 43 – 48 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Cavin LG , Venkatraman M, Factor VM et al. Regulation of alpha-fetoprotein by nuclear factor-kappaB protects hepatocytes from tumor necrosis factor-alpha cytotoxicity during fetal liver development and hepatic oncogenesis . Cancer Res 2004 ; 64 : 7030 – 7038 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Monga SP , Monga HK, Tan X et al. Beta-catenin antisense studies in embryonic liver cultures: Role in proliferation, apoptosis, and lineage specification . Gastroenterology 2003 ; 124 : 202 – 216 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Hasuike S , Ido A, Uto H et al. Hepatocyte growth factor accelerates the proliferation of hepatic oval cells and possibly promotes the differentiation in a 2-acetylaminofluorene/partial hepatectomy model in rats . J Gastroenterol Hepatol 2005 ; 20 : 1753 – 1761 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Correia-Neves M , Waltzinger C, Wurtz JM et al. Amino acids specifying MHC class preference in TCR V alpha 2 regions . J Immunol 1999 ; 163 : 5471 – 5477 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 43 Cole SP , Marks GS. Ferrochelatase and N-alkylated porphyrins . Mol Cell Biochem 1984 ; 64 : 127 – 137 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Resnick JL , Bixler LS, Cheng L et al. Long-term proliferation of mouse primordial germ cells in culture . Nature 1992 ; 359 : 550 – 551 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Kanno N , LeSage G, Glaser S et al. Regulation of cholangiocyte bicarbonate secretion . Am J Physiol Gastrointest Liver Physiol 2001 ; 281 : G612 – G625 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Walkup MH , Gerber DA. Hepatic stem cells: In search of . Stem Cells 2006 ; 24 : 1833 – 1840 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Rountree CB , Wang X, Ge S et al. Bone marrow fails to differentiate into liver epithelium during murine development and regeneration . Hepatology 2007 ; 45 : 1250 – 1260 . Google Scholar Crossref Search ADS PubMed WorldCat 48 Weiss MC , Strick-Marchand H. Isolation and characterization of mouse hepatic stem cells in vitro . Semin Liver Dis 2003 ; 23 : 313 – 324 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 49 Omori N , Omori M, Evarts RP et al. Partial cloning of rat CD34 cDNA and expression during stem cell-dependent liver regeneration in the adult rat . Hepatology 1997 ; 26 : 720 – 727 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Weigmann A , Corbeil D, Hellwig A et al. Prominin, a novel microvilli-specific polytopic membrane protein of the apical surface of epithelial cells, is targeted to plasmalemmal protrusions of non-epithelial cells . Proc Natl Acad Sci U S A 1997 ; 94 : 12425 – 12430 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Yin AH , Miraglia S, Zanjani S et al. AC133, a novel marker for human hematopoietic stem and progenitor cells. Blood 90 : 5002 – 5012 . Crossref Search ADS PubMed WorldCat 52 Shmelkov SV , St Clair R, Lyden D et al. AC133/CD133/Prominin-1 . Int J Biochem Cell Biol 2005 ; 37 : 715 – 719 . Google Scholar Crossref Search ADS PubMed WorldCat 53 Yovchev MI , Grozdanov PN, Joseph B et al. Novel hepatic progenitor cell surface markers in the adult rat liver . Hepatology 2007 ; 45 : 139 – 149 . Google Scholar Crossref Search ADS PubMed WorldCat 54 Kordes C , Sawitza I, Muller-Marbach A et al. CD133+ hepatic stellate cells are progenitor cells . Biochem Biophys Res Commun 2007 ; 352 : 410 – 417 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Kisseleva T , Uchinami H, Feirt N et al. Bone marrow-derived fibrocytes participate in pathogenesis of liver fibrosis . J Hepatol 2006 ; 45 : 429 – 438 . Google Scholar Crossref Search ADS PubMed WorldCat 56 Berg T , Rountree CB, Lee L et al. FGF10 is critical for liver growth during embryogenesis and controls of hepatoblast survival via b-catenin activation. Hepatology 2007 ; in press. Google Scholar OpenURL Placeholder Text WorldCat 57 Monga SP , Pediaditakis P, Mule K et al. Changes in WNT/beta-catenin pathway during regulated growth in rat liver regeneration . Hepatology 2001 ; 33 : 1098 – 1109 . Google Scholar Crossref Search ADS PubMed WorldCat 58 Monga SP , Mars WM, Pediaditakis P et al. Hepatocyte growth factor induces Wnt-independent nuclear translocation of beta-catenin after Met-beta-catenin dissociation in hepatocytes . Cancer Res 2002 ; 62 : 2064 – 2071 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 59 Battle MA , Konopka G, Parviz F et al. Hepatocyte nuclear factor 4alpha orchestrates expression of cell adhesion proteins during the epithelial transformation of the developing liver . Proc Natl Acad Sci U S A 2006 ; 103 : 8419 – 8424 . Google Scholar Crossref Search ADS PubMed WorldCat 60 Parviz F , Matullo C, Garrison WD et al. Hepatocyte nuclear factor 4alpha controls the development of a hepatic epithelium and liver morphogenesis . Nat Genet 2003 ; 34 : 292 – 296 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2007 AlphaMed Press This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - A CD133-Expressing Murine Liver Oval Cell Population with Bilineage Potential
JF - Stem Cells
DO - 10.1634/stemcells.2007-0176
DA - 2007-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/a-cd133-expressing-murine-liver-oval-cell-population-with-bilineage-IpwUINkPYZ
SP - 2419
EP - 2429
VL - 25
IS - 10
DP - DeepDyve
ER -