Prospective Isolation and Characterization of Bipotent Progenitor Cells in Early Mouse Liver Development

Abstract

Outgrowth of the foregut endoderm to form the liver bud is considered the initial event of liver development. Hepatic stem/progenitor cells (HSPCs) in the liver bud are postulated to migrate into septum transversum mesenchyme at around embryonic day (E) 9 in mice. The studies of liver development focused on the mid-fetal stage (E11.5–14.5) have identified HSPCs at this stage. However, the in vitro characteristics of HSPCs before E11.5 have not been elucidated. This is probably partly because purification and characterization of HSPCs in early fetal livers have not been fully established. To permit detailed phenotypic analyses of early fetal HSPC candidates, we developed a new coculture system, using mouse embryonic fibroblast cells. In this coculture system, CD13⁺Dlk⁺ cells purified from mouse early fetal livers (E9.5 and E10.5) formed colonies composed of both albumin-positive hepatocytic cells and cytokeratin (CK) 19-positive cholangiocytic cells, indicating that early fetal CD13⁺Dlk⁺ cells have properties of bipotent progenitor cells. Inhibition of signaling by Rho-associated coiled-coil containing protein kinase (Rock) or by nonmuscle myosin II (downstream from Rock) was necessary for effective expansion of early fetal CD13⁺Dlk⁺ cells in vitro. In sorted CD13⁺Dlk⁺ cells, expression of the hepatocyte marker genes albumin and α-fetoprotein increased with fetal liver age, whereas expression of CK19 and Sox17, endodermal progenitor cell markers, was highest at E9.5 but decreased dramatically thereafter. These first prospective studies of early fetal HSPC candidates demonstrate that bipotent stem/progenitor cells exist before E11.5 and implicate Rock-myosin II signaling in their development.

Introduction

L iver development is regulated by hormonal factors as well as by cell–cell interaction. In the beginning of liver development, around embryonic day (E) 9 in mice, hepatic stem/progenitor cells (HSPCs) are believed to differentiate from foregut endoderm and to expand in the early fetal liver bud. Fibroblast growth factor derived from the cardiac mesoderm and bone-morphogenetic proteins from septum transversum mesenchyme are important for differentiation of foregut endodermal cells into hepatic-lineage cells [1,2]. During mid- to late-fetal development (around E11.5 to E16.5), hematopoietic stem cells originating from the aorta-gonad-mesonephros region migrate into the fetal liver. This implies significant change in the microenvironment of the liver from early- to late-fetal liver development. Oncostatin M secreted from hematopoietic cells induces hepatoblasts, hepatic progenitor cells in mid-fetal livers, to differentiate into mature hepatocytes [3,4]. Hepatoblasts can proliferate at high rate and possess bipotency, the ability to differentiate into both hepatocytes (hepatic parenchymal cells) and cholangiocytes [5]. Several groups have purified hepatoblasts derived from E11.5 to E14.5 livers. We have shown that a sorted individual cell derived from mid-fetal liver gives rise to a relatively large colony after 5 days of culture on extracellular matrix-coated dishes [6]. A member of this class of cells is designated as a “hepatic colony-forming unit in culture (H-CFU-C).” Utilizing H-CFU-C culture, delta-like 1 homolog (Dlk), Liv2, CD13, and CD133 were identified as cell surface markers for hepatoblasts [7 –9]. H-CFU-C culture and reclone sorting have demonstrated that hepatoblasts in mid-fetal livers have self-renewal capacity and bipotency [10]. Several transcription factors, such as Prox1, Tbx3, and Sall4, regulate proliferation and differentiation of hepatoblasts in liver development [11 –13].

In contrast to hepatoblasts in mid-fetal livers, few studies have been done with early fetal HSPCs in E9.5 and 10.5 liver buds (Fig. 1A), because no suitable culture system for these cells has been established. HSPCs in early fetal livers thus remain largely uncharacterized. Explant culture systems have been used to study early fetal liver cells, and the effects of fibroblast growth factor secreted from cardiac mesoderm on early fetal livers were found using explanted-liver organ culture [1,2]. However, as explanted early fetal livers do not consist solely of HSPCs, to establish a culture system for purified progenitor cells in such livers is crucial for analyses of the initial steps of liver development. In this study, we found that cells expressing CD13, Dlk, and Liv2 exist during early- to mid-fetal liver development. We established a new culture system for in vitro expansion of these cells, candidate HSPCs, at the single-cell level using mouse embryonic fibroblasts (MEFs) as feeder cells. CD13⁺Dlk⁺ cells derived from early fetal liver showed bipotency and could proliferate to form large colonies in this culture system. Inhibition of Rho-associated coiled-coil–containing protein kinase (Rock) or myosin II activity using, respectively, Y-27632 or blebbistatin significantly enhanced colony-forming activities of early fetal CD13⁺Dlk⁺ cells. This study, the first investigation of purified HSPCs derived from early fetal livers, demonstrates that these progenitor cells in early fetal livers have properties distinct from those in mid-fetal livers.

FIG. 1.

Expression of cell surface markers in early- to mid-fetal liver cells. (A) Studies of hepatic stem/progenitor cells (HSPCs) during embryonic development. Bipotent progenitor cells in E11–14 mid-fetal livers (hepatoblasts) have been well characterized using H-CFU-C assays. In contrast, progenitor cells in early fetal livers (E9–10) have not been previously purified. (B) Cells from E9.5, E10.5, E11.5, and E13.5 fetal livers were stained with antibodies to various markers against mid-fetal hepatoblasts (CD13 and Dlk) and hematopoietic cells (CD45 and Ter119). Some populations of CD45⁻Ter119⁻ nonhematopoietic cells in early- and mid-fetal livers expressed both CD13 and Dlk. Nonstained cells (Nonstained sample), cells stained with isotype antibodies (Isotype control), or cells stained with anti-mouse CD13 and Dlk antibodies (Sample) were analyzed using flow cytometer. Ratios of CD13-single positive, Dlk-single positive, and CD13⁺Dlk⁺ cells in the CD45⁻Ter119⁻ fraction are shown. Numbers of CD13⁺Dlk⁺ cells in each of the CD45⁻Ter119⁻ fractions are as follows: 652 in 36,796 (E9.5), 1,630 in 25,481 (E10.5), 2,764 in 17,429 (E11.5), and 526 in 6,587 (E13.5). H-CFU-C, hepatic colony-forming unit in culture.

Materials and Methods

Materials

C57BL/6NCrSlc, C3H, and green fluorescent protein (GFP)-transgenic mice (Nihon SLC, Shizuoka, Japan) were used in this study. All animals were treated under guidelines of the Institute of Medical Science, The University of Tokyo. Reagents and commercial suppliers were Dulbecco's modified Eagle's medium (DMEM), DMEM/Ham's F12 half medium, penicillin/streptomycin/l-glutamine (100×), dexamethasone, dimethyl sulfoxide, nicotinamide, and gelatin from porcine skin (Sigma, St. Louis, MO); insulin–transferrin–selenium X, nonessential amino acid solution, KnockOut Serum Replacement (KSR), and HEPES buffer solution (Invitrogen, Carlsbad, CA); fetal bovine serum (FBS; Nichirei Biosciences, Tokyo, Japan); hepatocyte growth factor and epidermal growth factor (PeproTech, Rocky Hill, NJ); collagen type I (Nitta Gelatin, Osaka, Japan); PD0325901 and CHIR99021 (Axon Biochemicals, Groningen, The Netherlands); Y-27632 (Wako Pure Chemical Industries, Osaka, Japan); A-83-01 (Tocris Bioscience, Bristol, United Kingdom); and (S)-(−)-blebbistatin (Toronto Research Chemicals, Inc., Toronto, ON). Phycoerythrin (PE)-conjugated anti-CD13 (Pharmingen, San Jose, CA), fluorescein isothiocyanate (FITC)-conjugated anti-Dlk (Medical and Biological Laboratories, Nagoya, Japan), allophycocyanin (APC)-conjugated anti-Ter119, and APC-conjugated anti-CD45.2 (eBioscience, San Diego, CA) antibodies were used for cell staining.

Flow cytometric analysis

Minced liver tissues from E9.5 through E13.5 mice were dissociated with 0.05% collagenase solution. Dissociated cells were washed with phosphate-buffered saline (PBS) supplemented with 3% FBS and incubated with antibodies against cell surface markers for 60 min at 4°C. After washing with PBS supplemented with 3% FBS and staining of dead cells with propidium iodide, the cells were analyzed and sorted using a MoFlo™ fluorescence-activated cell sorter (DAKO, Glostrup, Denmark). Results with isotype control antibodies were shown as negative control.

Analysis of Liv2 expression using a fluorescence-activated flow cytometer

Dissociated cells were incubated with rat anti-Liv2 antibody [8] for 30 min on ice. After washing with PBS supplemented with 3% FBS, cells were stained with anti-rat immunoglobulin G (IgG)-Alexa647 (Invitrogen) for 30 min on ice and were washed with PBS supplemented with 3% FBS. Cells were further stained with FITC-conjugated anti-Dlk, PE-conjugated anti-CD13, PE-Cy7-conjugated anti-CD45, and PE-Cy7-conjugated anti-Ter119 (eBioscience) antibodies for 30 min on ice. After washing with PBS supplemented with 3% FBS and staining of dead cells with propidium iodide, the cells were analyzed and sorted using a MoFlo fluorescence-activated cell sorter.

Preparation of MEF

E13.5 ICR mouse embryos (Nihon SLC) were dissected and the head and internal organs were completely removed. The torso was minced and dissociated in 0.05% trypsin–EDTA (Sigma) for 30 min. After washing with PBS, cells were expanded in DMEM with 10% FBS. To halt cell proliferation, these MEFs were treated with mitomycin C (Wako Pure Chemical Industries) at 37°C for 2 h and used as feeder cells.

Colony formation assay

CD13⁺Dlk⁺ and other types of cells in the nonhematopoietic cell fraction were plated onto type I collagen or gelatin-coated 35-mm tissue culture dishes at a low density (25 cells/cm²) or into type I collagen-coated 96-well plates at one cell per well. For colony formation assay with feeder cells, mitomycin C-treated feeder cells (MEF or other cell lines) were plated onto 0.1% gelatin-coated 12-well plates (2×10⁵ cells per well). After 24 h of culture, cells in the nonhematopoietic cell fraction were sorted onto feeder cells at a low density (25 cells/cm²).

Our standard culture medium is a 1:1 mixture of H-CFU-C medium (DMEM/Ham's F12 half medium with 10% FBS or 10% KSR, 1×Insulin–Transferrin–Selenium X, 10 mM nicotinamide, 10^–7 M dexamethasone, 2.5 mM HEPES, 1×penicillin/streptomycin/l-glutamine, and 1×nonessential amino acid solution) and conditioned medium derived from E14.5 hepatic cells [13]. In several experiments, we used a 1:1 mixture of H-CFU-C medium and fresh DMEM with 10% KSR as a culture medium for coculture colony assays. Cells were cultured for 6 days, in the presence of 40 ng/mL hepatocyte growth factor and 20 ng/mL epidermal growth factor, in either standard culture medium or 1:1 mixture of H-CFU-C medium and fresh DMEM with 10% KSR. To count colonies derived from individual single cells, we used an ArrayScan VTI HCS Reader (Thermo Scientific, Waltham, MA) following immunostaining. Albumin-positive colonies were counted.

To analyze the effects of Rock-myosin II pathway using inhibitors, CD45⁻Ter119⁻Dlk⁺ cells derived from GFP transgenic mouse embryos were cocultured with MEF in the presence of either Y-27632 or blebbistatin. GFP-positive colonies were counted. To analyze the effects of soluble factors derived from MEF, MEF-conditioned medium was harvested from 2-day confluent cultures of MEF cultured in a 1:1 mixture of H-CFU-C medium and fresh DMEM with 10% KSR. We cultured early fetal cells on collagen-coated dishes or on MEF feeder cells in MEF-conditioned medium.

Messenger RNA detection by reverse transcription–polymerase chain reaction

Total RNA was extracted from CD13⁺Dlk⁺ cells in the nonhematopoietic cell fraction using the RNeasy Micro Kit (Qiagen, Venlo, The Netherlands). First-strand cDNA synthesized using the Primescript first strand cDNA synthesis kit (Takara, Otsu, Japan) was used as a template for quantitative reverse transcription–polymerase chain reaction (RT-PCR) amplification. The cDNA samples were normalized by number of glyceraldehyde 3-phosphate dehydrogenase copies using quantitative RT-PCR with the TaqMan probe (Applied Biosystems, Foster City, CA). Universal Library (Roche Diagnosis, Basel, Switzerland) was used to quantify the copy numbers of albumin, α-fetoprotein (AFP), cytokeratin (CK) 19, c-Met, E-cadherin, and Sox17 transcripts. Intron-spanning primer sequences and probe number for each gene are shown in Supplementary Table S1 (Supplementary Data are available online at www.liebertonline.com/scd).

For hepatic gene expression analyses in HSPC colonies, E10.5 and E13.5 CD13⁺Dlk⁺ cells derived from GFP-transgenic mice were cultured with MEF for 3 days. Colonies were dissociated in 0.05% trypsin–EDTA and GFP⁺ cells were purified using a MoFlo fluorescence-activated cell sorter. Total RNA was extracted and first-strand cDNA was synthesized using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Expression of hepatic genes was detected using quantitative RT-PCR with the TaqMan probe: Cyp1A1, Mm00487218_m1; Cyp1A2, Mm00487224_m1; Cyp3A11, Mm00731567_m1; Cyp3A13, Mm00484110_m1; Cyp7a1, Mm00484152_m1; glucose-6-phosphatase, Mm00839363_m1; tyrosine aminotransferase, Mm01244282_m1; albumin, Mm00802090_m1; CK19, Mm00492980_m1.

Immunostaining

Cultured cells were washed with PBS and were fixed with 4% paraformaldehyde/PBS. After washing 3 times with PBS, cells were permeabilized with 0.5% Triton/PBS for 10 min, washed with PBS, and incubated with 5% donkey serum/PBS for 1 h at room temperature. They were incubated with diluted primary antibodies overnight at 4°C. Goat anti-albumin antibody (Bethyl, Montgomery, TX) and rabbit anti-CK19 antibody (A gift from Prof. A. Miyajima, University of Tokyo, Tokyo, Japan) were used as primary antibodies [7]. The cells next were washed with PBS and were incubated for 1 h at room temperature with anti-rabbit IgG-Alexa488 and anti-goat IgG-Alexa546 antibodies (Invitrogen). The cells were washed with PBS and their nuclei were stained with 40,6-diamidine-20-phenylindole dihydrochloride (Sigma).

For the analyses of proliferation, colonies were stained with goat anti-albumin antibody and rabbit anti-Ki67 antibody (Abcam, Cambridge, United Kingdom). After washing with PBS, cells were stained with anti-rabbit IgG-Alexa555 and anti-goat IgG-Alexa488 antibodies (Invitrogen). Intensities of Ki67 in individual albumin-positive colonies were analyzed with an ArrayScan VTI HCS Reader.

In-droplet cell-staining methods

To quantify albumin and CK19 expression in individual cells, in-droplet staining methods were used [14]. CD13⁺Dlk⁺ cells derived from fetal livers were sorted onto slide glasses. After fixation with 2% paraformaldehyde/PBS and permeabilization with 0.5% Triton/PBS, cells were incubated in 5% donkey serum/PBS for 30 min at room temperature. They then were incubated with goat anti-albumin and rabbit anti-CK19 antibodies overnight at 4°C, washed with PBS, and incubated with secondary antibodies (anti-rabbit IgG-Alexa488 and anti-goat IgG-Alexa546 antibodies) for 40 min at room temperature. The cells were washed with PBS and their nuclei were stained with 40,6-diamidine-20-phenylindole dihydrochloride. For each analysis, addition of an appropriate immune serum provided a negative control. Antibody fluorescence intensity was measured using the ArrayScan Reader.

Statistics

We used Microsoft Excel 2004 for Mac, Version 11.6.2 (Microsoft, Redmond, WA) to calculate standard deviations (SDs) and statistically significant differences between samples using Student's 2-tailed t-test.

Results

Early fetal liver contains cells expressing mid-fetal hepatoblast cell-surface markers

Several studies show that CD13 and Dlk are cell surface markers of hepatoblasts in mid-fetal livers (E11.5 to E14.5) and hepatoblasts exist in the CD13⁺Dlk⁺ fraction [7,15]. We assessed whether early fetal livers (E9.5 and E10.5) contain cells expressing these cell-surface markers (Fig. 1B). Livers derived from E9.5 to E13.5 mouse embryos were dissected and dissociated using collagenase. Cells were stained with antibodies against hematopoietic cell surface markers (CD45 and Ter119) as well as CD13 and Dlk. CD13⁺Dlk⁺ double-positive cells were found in the CD45⁻Ter119⁻ nonhematopoietic cell fraction derived from both early- and mid-fetal livers, although the expression level of CD13 and Dlk was low at E9.5 and increased during liver development. Liv2 is another cell surface molecule expressed on hepatic progenitor cells; numbers of Liv2-positive cells increase during E9.5 to E12.5 [8]. We found that CD13⁺Dlk⁺ cells in E9.5 fetal liver also expressed Liv2, indicating that cells expressing several hepatoblast cell-surface markers existed during early- to mid-fetal liver development (Supplementary Fig. S1). These results suggested that cell surface markers of hepatoblasts are commonly encountered during fetal liver development.

Phenotypic differences between early- and mid-fetal CD13⁺Dlk⁺ cells

To further characterize CD13⁺Dlk⁺ cells from each stage, expression of endodermal, hepatocyte, and cholangiocyte marker genes was analyzed using real-time RT-PCR. Strong expression of albumin and AFP was detected in CD13⁺Dlk⁺ cells derived from E13.5 mid-fetal liver (Fig. 2A). In contrast, these hepatocyte marker genes were detected at low levels in CD13⁺Dlk⁺ cells derived from E9.5 early fetal liver. Interestingly, E9.5 CD13⁺Dlk⁺ cells expressed CK19 and the early endodermal cell marker Sox17 at high levels (Fig. 2B). Although CK19 is known as a marker for cholangiocytes, it is also expressed in primitive gut endoderm at an earlier stage, including E9.0 [16]. Synthesis of albumin and CK19 was also analyzed using an in-droplet staining method (Fig. 2C). Although albumin levels were barely detectable in E9.5 CD13⁺Dlk⁺ cells, but they were accumulated during liver development. In contrast, CK19 expression levels in CD13⁺Dlk⁺ cells derived from E9.5 and E10.5 livers were higher than those in CD13⁺Dlk⁺ cells derived from E11.5 and E13.5 livers. These results indicate that gene expression patterns in early fetal CD13⁺Dlk⁺ cells are distinct from those in mid-fetal CD13⁺Dlk⁺ cells.

FIG. 2.

Gene expression profiles of CD45⁻Ter119⁻CD13⁺Dlk⁺ cells during fetal liver development. (A, B) Expression of several hepatocyte (Met, albumin, AFP, and E-cadherin), primitive gut endoderm and cholangiocyte (CK19), and endodermal progenitor (Sox17) markers was analyzed using quantitative reverse transcription–polymerase chain reaction. CD45⁻Ter119⁻CD13⁺Dlk⁺ cells derived from E9.5, E10.5, E11.5, and E13.5 fetal livers were purified using flow cytometry. Expression level of marker genes in E9.5 cells was set to 1.0. Results are represented as relative expression±SD (triplicate samples). (C) Protein expression (albumin and CK19) in CD45⁻Ter119⁻CD13⁺Dlk⁺ cells derived from E9.5, 10.5, 11.5, and E13.5 fetal livers. CD45⁻Ter119⁻CD13⁺Dlk⁺ cells were directly sorted onto slide glasses and were stained with primary antibodies (goat anti-albumin and rabbit anti-CK19 antibodies) and secondary antibodies (anti-rabbit IgG-Alexa488 and anti-goat IgG-Alexa546 antibodies). Results are represented as intensities of fluorescence of randomly picked-up cells (n=50, **P<0.01). CK, cytokeratin; IgG, immunoglobulin G; AFP, α-fetoprotein.

Proliferative capacity of CD13⁺Dlk⁺ cells derived from early- and mid-fetal livers

To analyze whether early fetal CD13⁺Dlk⁺ cells contain phenotypes of hepatic progenitor cells (exhibiting high proliferative potential and bipotency), we analyzed these cells using single-cell colony assays. Sorted single cells were inoculated into individual wells of 96-well collagen type I-coated culture plates. Cells derived from mid-fetal livers (E11.5 and E13.5) formed several small colonies (50–100 cells) and large colonies (over 100 cells) after 6 days of culture [13]. In contrast, cultures derived from E9.5 and E10.5 livers yielded few colonies or none (Supplementary Fig. S2). The nonhematopoietic cell fraction of E9.5 fetal livers comprises almost 1% CD13⁺Dlk⁺ cells and 99% CD13⁻ or Dlk⁻ cells (Supplementary Fig. S3A). To exclude the possibility that early fetal progenitor cells are in the Dlk⁻ or CD13⁻ fraction, we sorted 10,000 E9.5 fetal liver cells (100 CD13⁺Dlk⁺ cells and 9,900 CD13⁻ or Dlk⁻ cells). Cells from neither fraction could form colonies in H-CFU-C culture medium on collagen-coated dishes. In addition, no colonies were detected in cultures derived from 10,000 nonsorted E9.5 liver cells, indicating that low colony-forming activities of E9.5-derived cells were not due to damage sustained during flow cytometry (Supplementary Fig. 3B). Nonsorted cells derived from E10.5 livers could form only a few small colonies but not large colonies (Supplementary Fig. 3C). Few colony formation of early fetal livers was also detected on gelatin-coated dishes (data not shown). These results suggested that conventional H-CFU-C culture on collagen- and gelatin-coated dishes is not suitable for early fetal (E9.5–10.5) HSPCs.

Early fetal liver CD13⁺Dlk⁺ cells require both feeder cell interaction and the addition of ROCK inhibitor for their optimal expansion

Several studies have suggested that cell–cell interactions are important for proliferation and differentiation of somatic stem cells and progenitor cells. In early fetal liver development, the interaction between endodermal and mesenchymal populations is important for proper liver bud growth. Transcription factor Hlx, expressed in the septum transversum mesenchyme, is essential for proliferation of early fetal hepatic cells [17]. Therefore, we inferred that E9.5 CD13⁺Dlk⁺ cells need to interact with other cell populations to propagate both in vivo and in vitro. To mimic the interaction of hepatic and mesenchymal cells, CD13⁺Dlk⁺ cells were cocultured with MEF as mesenchymal feeder cells. We sorted either CD13⁺Dlk⁺ cells or other cells (Dlk⁻ or CD13⁻ cells) in the nonhematopoietic cell fraction of E9.5 and E13.5 fetal livers. After 6 days of coculture with MEF, a few large colonies (containing both albumin⁺ hepatocytic cells and CK19⁺ cholangiocytic cells) were detected in culture of E9.5 CD13⁺Dlk⁺ cells (Fig. 3A). Some early fetal CD13⁺Dlk⁺ cells possess phenotypes of hepatic progenitor cells, viz., high proliferative activity and bipotency. In contrast to mid-fetal hepatoblasts, early fetal HSPCs required interaction with mesenchymal fibroblasts for in vitro expansion. After 3 days of colony formation culture, early fetal CD13⁺Dlk⁺ cells expressed several hepatic genes and their expression levels were lower than those in mid-fetal CD13⁺Dlk⁺ cells, suggesting that early fetal cells are more immature types of progenitors (Supplementary Table S2). KSR is routinely employed in serum-free embryonic stem-cell culture protocols. We found that numbers of large and small colonies derived from E9.5 CD13⁺Dlk⁺ cells detected in KSR supplemented culture increased compared with those detected in FBS-supplemented culture (Fig. 3B). To explore signaling pathways regulating proliferation of early fetal liver cells, Y-27632 (a Rock inhibitor), PD0325901 (a MEK inhibitor), CHIR99021 (a GSK3β inhibitor), and A-83-01 (a transforming growth factor β type I receptor inhibitor) were added to cell cultures. The morphology of colonies was not changed by the addition of these inhibitors (Fig. 4A). Although PD0325901, CHIR99021, and A-83-01 did not change the number of large colonies formed by E9.5 CD13⁺Dlk⁺ cells, Y-27632 significantly increased the number of large colonies formed by these cells, indicating that inhibition of Rock is important for proliferation of E9.5 CD13⁺Dlk⁺ cells (Fig. 4B). The addition of Y-27632 induced colony formation of E10.5 CD13⁺Dlk⁺ cells (Fig. 4C). In contrast, the addition of Y-27632 could not induce proliferation of E13.5 CD13⁺Dlk⁺ cells, as previously shown (Fig. 4B) [9].

FIG. 3.

Early fetal liver CD45⁻Ter119⁻CD13⁺Dlk⁺ cells could form large colonies in the coculture with MEF. (A) Representative view of a colony formed from a single E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cell cocultured with MEF. (B) KSR induced colony formation by E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells. E13.5 and E9.5 fetal liver cells were sorted and cocultured with MEF for 6 days. Cells were cultured in H-CFU-C medium supplemented with either fetal bovine serum or KSR. Small colonies (gray bars) consisting of 50–100 cells and large colonies (white bars) consisting of >100 cells were counted. Results are represented as mean colony count±SD (triplicate samples; * and ** denote P<0.05 and P<0.01). MEF, mouse embryonic fibroblast; KSR, KnockOut Serum Replacement.

FIG. 4.

Addition of Rock inhibitor is important for colony formation of early fetal liver CD45⁻Ter119⁻CD13⁺Dlk⁺ cells. (A) Representative view of colonies formed from a single E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cell in the presence of signaling inhibitors. E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells were cocultured with MEF for 6 days in the presence of either Y-27632 (a Rock inhibitor), PD0325901 (a MEK inhibitor), CHIR99021 (a GSK3β inhibitor), or A-83-01 (a transforming growth factor β type I receptor inhibitor). DW: distilled water (DW) added (control for Y-27632); DMSO: 0.01% DMSO added (control for PD0325901, CHIR99021, and A-83-01). (B) E9.5 and E13.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells were cocultured with MEF for 6 days in the presence of inhibitors shown in A. The number of colonies derived from E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells was significantly increased by culture in medium containing Y-27632, whereas the number of colonies derived from E13.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells did not increase with Y-27632 exposure. (C) E10.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells were cocultured for 6 days with MEF in the presence of Y-27632. DW: DW added (control). (B, C) Small colonies (gray bars) consisting of 50–100 cells and large colonies (white bars) consisting of >100 cells were counted. Results are represented as mean colony count±SD (triplicate samples; * and ** denote P<0.05 and P<0.01). DMSO, dimethyl sulfoxide.

We used conditioned medium derived from E14.5 liver cells in conventional H-CFU-C culture system for mid-fetal liver hepatoblasts [10,13]. When E9.5 CD13⁺Dlk⁺ cells were cocultured with MEF, however, the addition of fetal liver cell-conditioned medium decreased the number of large and small colonies (data not shown). The number of proliferative cells in individual colonies was not significantly changed (Fig. 5A), suggesting that cell apoptosis might be involved in the inhibition of colony formation by the addition of fetal liver cell-conditioned medium. In consequence, we cultured early fetal liver cells in the following experiments without using fetal liver cell-conditioned medium. We also found that these isolation and culture methods could induce proliferation of early fetal HSPCs derived from C3H mice, in addition to C57BL6 mice (Supplementary Fig. S4A, B).

FIG. 5.

A Rock inhibitor, Y-27632, and a myosin II synthetic inhibitor, blebbistatin, induced colony formation by early fetal liver CD45⁻Ter119⁻CD13⁺Dlk⁺ cells. (A) Proliferation of early fetal HSPCs regulated by ROCK inhibitor. Intensities of Ki67 in individual albumin-positive colonies were analyzed. DW: distilled water (DW) added (control for Y-27632); DMSO: 0.01% DMSO added (control for PD0325901, CHIR99021, and A-83-01); Fetal-liver CM: fetal-liver conditioned medium added. (B) CD45⁻Ter119⁻Dlk⁺ cells were cocultured for 6 days with MEF in the presence of either Y-27632 or blebbistatin. DW: DW added (control for Y-27632); DMSO: 0.1% DMSO added (control for blebbistatin); Y-27632 to DW: first 3 days of culture with Y-27632 followed by 3 days of culture without Y-27632; DW to Y-27632: first 3 days of culture without Y-27632 followed by 3 days of culture with Y-27632; Blebbistatin to DMSO: first 3 days of culture with blebbistatin followed by 3 days of culture with DMSO; DMSO to blebbistatin: first 3 days of culture with DMSO followed by 3 days of culture with blebbistatin. Small colonies (gray bars) consisting of 50–100 cells and large colonies (white bars) consisting of >100 cells were counted. Results are represented as mean colony count±SD (triplicate samples; * and ** denote P<0.05 and P<0.01).

In addition to CD13⁺Dlk⁺ cells, the CD13^midDlk^mid cells (the intermediate fraction) existed in E10.5 and E11.5 livers. We asked whether the intermediate fractions also contain HSPCs. However, significant colony formation by these fractions was not detected compared with the CD13⁺Dlk⁺ fractions (Supplementary Fig. S4C, D), indicating that most progenitor cells exist in the CD13⁺Dlk⁺ fraction during early- to mid-fetal liver development. These results suggested that the addition of ROCK inhibitor was required for clonal expansion of early fetal CD13⁺Dlk⁺ progenitor cells but not of mid-fetal CD13⁺Dlk⁺ hepatoblasts.

Inhibition of the Rock-myosin II pathway induced colony formation of early fetal liver CD13⁺Dlk⁺ cells

We varied length of exposure to Y-27632 in E9.5 fetal liver cell culture. Short-time exposure to Y-27632 (culture days 0–3 or 3–6) partially induced progression of colony formation. Interestingly, early-stage addition of Y-27632 (days 0–3) significantly induced formation of large colonies compared with late-stage addition of Y-27632 (days 3–6). Thus, inhibition of Rock is particularly important for the early stage of colony formation by E9.5 cells (Fig. 5B). Rock induces phosphorylation of several substrates, leading to various cellular responses [18]. The inactivation of myosin phosphatase target subunit, which is induced by Rock, protects the phosphorylated form of myosin regulatory light chain. This phosphorylation keeps myosin II in its active form. Blebbistatin, which specifically inhibits myosin II, has an effect similar to that of Y-27632. It inhibits apoptosis of single-suspended human embryonic stem cells [19,20]. Blebbistatin, like Y-27632, significantly induced colony formation by E9.5 cells (Fig. 5B). We analyzed proliferation of colonies in the presence of several inhibitors and found that a number of colonies expressed high levels of Ki67 proliferation marker in the culture stimulated with Y-27632 (Fig. 5A). These results suggest that inhibition of the Rock-myosin II pathway is important in expansion of early fetal HSPCs.

Soluble factors derived from MEF partly induced proliferation of early fetal CD13⁺Dlk⁺ cells

We then assessed whether soluble factors derived from MEF are involved in expansion of E9.5 CD13⁺Dlk⁺ cells. Confluent MEFs were cultured for 2 days in H-CFU-C medium. This medium (now “MEF conditioned”) was used as medium for various colony formation assays (Fig. 6A). When cells were cocultured with MEF, use of MEF-conditioned medium made no difference in the efficiency of colony formation by E9.5 CD13⁺Dlk⁺ cells. However, when E9.5 CD13⁺Dlk⁺ cells were cultured on collagen type I, small and large colonies were detected only when MEF-conditioned media were used. Fresh medium not conditioned with MEF did not support colony formation on collagen-coated dishes. These data suggest that expansion of early fetal progenitor cells was partly supported by soluble factors derived from MEF.

FIG. 6.

Soluble factors and cell–cell/extracellular matrix interactions are important for MEF-induced expansion of early fetal cells. (A) E9.5 CD45⁻Ter119⁻CD13⁺Dlk⁺ cells were cultured for 6 days with MEF-conditioned medium or fresh medium in the presence of Y-27632. Cells were sorted onto collagen-coated dishes or dishes containing MEF. Small colonies (gray bars) consisting of 50–100 cells and large colonies (white bars) consisting of >100 cells were counted. Results are represented as mean colony count±SD (triplicate samples; **P<0.01). (B) Schema of phenotypes of progenitor cells during fetal liver development. CD13 and Dlk are surface markers common to early fetal progenitor cells and mid-fetal hepatoblasts. However, expression of several genes (albumin, AFP, CK19, and Sox17) differed significantly between early fetal progenitor cells and mid-fetal hepatoblasts. N.D., not detected.

Discussion

In this report, we showed that early fetal (E9.5 and E10.5) liver-derived CD13⁺Dlk⁺ cells have characteristics of hepatic progenitor cells: They have a high proliferative capacity and the ability to differentiate into both albumin-positive hepatocytic cells and CK19-positive cholangiocytic cells. In contrast to mid-fetal hepatoblasts, early fetal HSPCs require interaction with MEF to expand clonally. Hlx is a transcription factor expressed in septum transversum mesenchyme and fetal liver expansion is severely deficient in Hlx-knockout mice [17]. Therefore, at an early fetal liver developmental stage, interaction with Hlx-positive mesenchymal cells is important for proliferation of hepatoblasts in vivo. Under our culture conditions, MEF, which express Hlx (data not shown), supported proliferation of early fetal progenitor cells. MEF-conditioned medium partially supported clonal expansion of E9.5 CD13⁺Dlk⁺ cells, suggesting that soluble factors derived from MEF are at least partly necessary for the survival or growth of E9.5 CD13⁺Dlk⁺ cells. Other cell–cell and cell–matrix interactions also appear important for proliferation of early fetal progenitor cells, because large colonies derived from E9.5 CD13⁺Dlk⁺ cells cocultured with MEF were significantly more numerous than when the same population of cells was cultured on collagen-coated dishes in the presence of MEF-conditioned medium.

Not only coculture with MEF but also inhibition of Rock or myosin II activity remarkably improved clonal expansion of E9.5 CD13⁺Dlk⁺ cells. Rock inhibitor Y-27632 promotes the survival and growth of various other types of cells, including human embryonic stem cells and adult liver-derived progenitor cells [9,21]. The molecular mechanisms by which Rock and myosin II inhibitors promoted the colony forming efficiency of early fetal liver cells in our culture system are unknown and await further investigation. As hepatic progenitor cells differentiated from foregut endoderm in E9.5 embryos into the septum transversum mesenchyme, they start to lose epithelial properties and to acquire mesenchymal properties [22]. An epithelium-specific property is that of polarity, established by the segregation of apical and basolateral domains. Epithelial shape is regulated by apical constriction, a process dependent on activated myosin II [23,24]. Rock also participates in apical constriction [25]. Dissociated early fetal progenitor cells, deprived of cell–cell contact and of traction from adjacent cells, may not be able to tolerate the force generated by apical constriction and thus undergo apoptosis. Inhibition of Rock or myosin II activity might thus rescue these cells from apoptosis through inhibition of excessive apical constriction.

We found that E9.5 to E13.5 CD13⁺Dlk⁺ cells in fetal livers can serve as bipotent progenitor cells. However, expression of several hepatic and endodermal genes differed remarkably between early- and mid-fetal CD13⁺Dlk⁺ cells. Levels of albumin mRNA, but not CK19 mRNA, were high in E13.5 CD13⁺Dlk⁺ cells. In contrast, E9.5 CD13⁺Dlk⁺ cells scarcely expressed mRNA of hepatic genes (AFP, albumin, and c-met) but exhibited high Sox17 and CK19 mRNA levels (Fig. 2A, B). Sox17 is expressed in definitive endodermal progenitor cells and CK19, a cholangiocytic marker gene in mid-fetal livers, is also expressed in primitive gut endoderm [16]. Therefore, E9.5 CD13⁺Dlk⁺ cells, which differentiate into mid-fetal hepatoblasts during liver development, seem to possess the properties of endodermal progenitor cells (Fig. 6B).

In the present study, we showed that CD13⁺Dlk⁺ cells derived from early fetal livers have high proliferative capacity and can differentiate into both albumin-positive cells and CK19-positive cells, suggesting that at the single-cell level CD13 and Dlk are markers for bipotent progenitor cells in the early fetal liver developmental stage. These cells show gene expression patterns distinct from those of hepatoblasts in the mid-fetal liver. Signaling pathways regulating the proliferative capacity of CD13⁺Dlk⁺ hepatic progenitor cells in vitro also differ between cells derived from early fetal livers and those derived from mid-fetal livers. These findings highlight a biologically important and potentially therapeutic role for mesenchymal cells and for the Rock-myosin II signaling pathway in the differentiation and expansion of hepatic progenitor cells derived from pluripotent stem cells.

Footnotes

Acknowledgments

This work was supported in part by grants-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology in Japan and Core Research for Evolutional Science and Technology from Japan Science and Technology Agency.

Author Disclosure Statement

There is no conflict of interest to disclose.

References

Jung

, Zheng

, Goldfarb

, Zaret

. 1999. Initiation of mammalian liver development from endoderm by fibroblast growth factors. Science, 284:1998–2003.

Rossi

, Dunn

, Hogan

, Zaret

. 2001. Distinct mesodermal signals, including BMPs from the septum transversum mesenchyme, are required in combination for hepatogenesis from the endoderm. Genes Dev, 15:1998–2009.

Kamiya

, Kinoshita

, Ito

, Matsui

, Morikawa

, Senba

, Nakashima

, Taga

, Yoshida

, Kishimoto

, Miyajima

. 1999. Fetal liver development requires a paracrine action of oncostatin M through the gp130 signal transducer. EMBO J, 18:2127–2136.

Kinoshita

, Sekiguchi

, Xu

, Ito

, Kamiya

, Tsuji

, Nakahata

, Miyajima

. 1999. Hepatic differentiation induced by oncostatin M attenuates fetal liver hematopoiesis. Proc Natl Acad Sci USA, 96:7265–7270.

Dabeva

, Petkov

, Sandhu

, Oren

, Laconi

, Hurston

, Shafritz

. 2000. Proliferation and differentiation of fetal liver epithelial progenitor cells after transplantation into adult rat liver. Am J Pathol, 156:2017–2031.

Suzuki

, Zheng

, Kondo

, Kusakabe

, Takada

, Fukao

, Nakauchi

, Taniguchi

. 2000. Flow-cytometric separation and enrichment of hepatic progenitor cells in the developing mouse liver. Hepatology, 32:1230–1239.

Tanimizu

, Nishikawa

, Saito

, Tsujimura

, Miyajima

. 2003. Isolation of hepatoblasts based on the expression of Dlk/Pref-1. J Cell Sci, 116:1775–1786.

Watanabe

, Nakagawa

, Ohata

, Kitagawa

, Nishitai

, Seo

, Tanemura

, Shimizu

, Kishimoto

et al. 2002. SEK1/MKK4-mediated SAPK/JNK signaling participates in embryonic hepatoblast proliferation via a pathway different from NF-kappaB-induced anti-apoptosis. Dev Biol, 250:332–347.

Kamiya

, Kakinuma

, Yamazaki

, Nakauchi

. 2009. Enrichment and clonal culture of progenitor cells during mouse postnatal liver development in mice. Gastroenterology, 137:1114–1126, 1126.e1111–1114.

10.

Suzuki

, Zheng

, Kaneko

, Onodera

, Fukao

, Nakauchi

, Taniguchi

. 2002. Clonal identification and characterization of self-renewing pluripotent stem cells in the developing liver. J Cell Biol, 156:173–184.

11.

Suzuki

, Sekiya

, Buscher

, Izpisua Belmonte

, Taniguchi

. 2008. Tbx3 controls the fate of hepatic progenitor cells in liver development by suppressing p19ARF expression. Development, 135:1589–1595.

12.

Oikawa

, Kamiya

, Kakinuma

, Zeniya

, Nishinakamura

, Tajiri

, Nakauchi

. 2009. Sall4 regulates cell fate decision in fetal hepatic stem/progenitor cells. Gastroenterology, 136:1000–1011.

13.

Kamiya

, Kakinuma

, Onodera

, Miyajima

, Nakauchi

. 2008. Prospero-related homeobox 1 and liver receptor homolog 1 coordinately regulate long-term proliferation of murine fetal hepatoblasts. Hepatology, 48:252–264.

14.

Yamazaki

, Iwama

, Takayanagi

, Morita

, Eto

, Ema

, Nakauchi

. 2006. Cytokine signals modulated via lipid rafts mimic niche signals and induce hibernation in hematopoietic stem cells. EMBO J, 25:3515–3523.

15.

Kakinuma

, Ohta

, Kamiya

, Yamazaki

, Oikawa

, Okada

, Nakauchi

. 2009. Analyses of cell surface molecules on hepatic stem/progenitor cells in mouse fetal liver. J Hepatol, 51:127–138.

16.

Tamai

, Ishikawa

, Bosl

, Mori

, Nozaki

, Baribault

, Oshima

, Taketo

. 2000. Cytokeratins 8 and 19 in the mouse placental development. J Cell Biol, 151:563–572.

17.

Hentsch

, Lyons

, Li

, Hartley

, Lints

, Adams

, Harvey

. 1996. Hlx homeo box gene is essential for an inductive tissue interaction that drives expansion of embryonic liver and gut. Genes Dev, 10:70–79.

18.

Riento

, Ridley

. 2003. Rocks: multifunctional kinases in cell behaviour. Nat Rev Mol Cell Biol, 4:446–456.

19.

Harb

, Archer

, Sato

. 2008. The Rho-Rock-Myosin signaling axis determines cell-cell integrity of self-renewing pluripotent stem cells. PLoS One, 3:e3001.

20.

Ohgushi

, Matsumura

, Eiraku

, Murakami

, Aramaki

, Nishiyama

, Muguruma

, Nakano

, Suga

et al. 2010. Molecular pathway and cell state responsible for dissociation-induced apoptosis in human pluripotent stem cells. Cell Stem Cell, 7:225–239.

21.

Watanabe

, Ueno

, Kamiya

, Nishiyama

, Matsumura

, Wataya

, Takahashi

, Nishikawa

, Muguruma

, Sasai

. 2007. A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat Biotechnol, 25:681–686.

22.

Bort

, Signore

, Tremblay

, Martinez Barbera

, Zaret

. 2006. Hex homeobox gene controls the transition of the endoderm to a pseudostratified, cell emergent epithelium for liver bud development. Dev Biol, 290:44–56.

23.

Martin

, Kaschube

, Wieschaus

. 2009. Pulsed contractions of an actin-myosin network drive apical constriction. Nature, 457:495–499.

24.

Lecuit

, Lenne

. 2007. Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis. Nat Rev Mol Cell Biol, 8:633–644.

25.

Nakajima

, Tanoue

. 2010. Epithelial cell shapae is regulated by Lulu proteins via myosin-II. J Cell Sci, 123:555–566.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.41 MB

Prospective Isolation and Characterization of Bipotent Progenitor Cells in Early Mouse Liver Development

Abstract

Introduction

Materials and Methods

Materials

Flow cytometric analysis

Analysis of Liv2 expression using a fluorescence-activated flow cytometer

Preparation of MEF

Colony formation assay

Messenger RNA detection by reverse transcription–polymerase chain reaction

Immunostaining

In-droplet cell-staining methods

Statistics

Results

Early fetal liver contains cells expressing mid-fetal hepatoblast cell-surface markers

Phenotypic differences between early- and mid-fetal CD13+Dlk+ cells

Proliferative capacity of CD13+Dlk+ cells derived from early- and mid-fetal livers

Early fetal liver CD13+Dlk+ cells require both feeder cell interaction and the addition of ROCK inhibitor for their optimal expansion

Inhibition of the Rock-myosin II pathway induced colony formation of early fetal liver CD13+Dlk+ cells

Soluble factors derived from MEF partly induced proliferation of early fetal CD13+Dlk+ cells

Discussion

Footnotes

Acknowledgments

Author Disclosure Statement

References

Supplementary Material

Phenotypic differences between early- and mid-fetal CD13⁺Dlk⁺ cells

Proliferative capacity of CD13⁺Dlk⁺ cells derived from early- and mid-fetal livers

Early fetal liver CD13⁺Dlk⁺ cells require both feeder cell interaction and the addition of ROCK inhibitor for their optimal expansion

Inhibition of the Rock-myosin II pathway induced colony formation of early fetal liver CD13⁺Dlk⁺ cells

Soluble factors derived from MEF partly induced proliferation of early fetal CD13⁺Dlk⁺ cells