Identification and Characterization of Lineage − CD45 − Sca-1 + VSEL Phenotypic Cells Residing in Adult Mouse Bone Tissue

Abstract

Murine bone marrow (BM)-derived very small embryonic-like stem cells (BM VSELs), defined by a lineage-negative (Lin⁻), CD45-negative (CD45⁻), Sca-1-positive (Sca-1⁺) immunophenotype, were previously reported as postnatal pluripotent stem cells (SCs). We developed a highly efficient method for isolating Lin⁻CD45⁻Sca-1⁺ small cells using enzymatic treatment of murine bone. We designated these cells as bone-derived VSELs (BD VSELs). The incidences of BM VSELs in the BM-derived nucleated cells and that of BD VSELs in bone-derived nucleated cells were 0.002% and 0.15%, respectively. These BD VSELs expressed a variety of hematopoietic stem cell (HSC), mesenchymal stem cell (MSC), and endothelial cell markers. The gene expression profile of the BD VSELs was clearly distinct from those of HSCs, MSCs, and ES cells. In the steady state, the BD VSELs proliferated slowly, however, the number of BD VSELs significantly increased in the bone after acute liver injury. Moreover, green fluorescent protein-mouse derived BD VSELs transplanted via tail vein injection after acute liver injury were detected in the liver parenchyma of recipient mice. Immunohistological analyses suggested that these BD VSELs might transdifferentiate into hepatocytes. This study demonstrated that the majority of the Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells reside in the bone rather than the BM. However, the immunophenotype and the gene expression profile of BD VSELs were clearly different from those of other types of SCs, including BM VSELs, MSCs, HSCs, and ES cells. Further studies will therefore be required to elucidate their cellular and/or SC characteristics and the potential relationship between BD VSELs and BM VSELs.

Introduction

It has been well documented that the fate of adult stem cells (SCs) is to be restricted to their tissue of origin. Surprisingly, it was reported that genetically labeled neural stem cells were found to produce a variety of blood cell types, including myeloid, lymphoid, and early hematopoietic cells [1]. These unexpected observations seemed to be exiting and promising, and may be explained by the term, transdifferentiation or plasticity of these cells. However, Wagers et al. reported an elegant study that a single green fluorescent protein (GFP)-marked hematopoietic stem cell (HSC) transplanted to lethally irradiated nontransgenic recipients robustly reconstituted peripheral blood leukocytes in these animals, but it did not contribute to nonhematopoietic tissues [2]. These data clearly indicated that the transdifferentiation or plasticity of HSCs is an extremely rare event.

Recently, Ratajczak and colleagues proposed an alternative explanation of the “plasticity” of bone marrow (BM)-derived SCs based on the assumption that BM SCs are more heterogeneous than previously thought [3]. Indeed, a number of reported studies identified many kinds of CD45-negative nonhematopoietic SCs in the BMs, including mesenchymal stem/stromal cells (MSCs) [4], muse cells [5], BMSCs [6], MIAMI [7], and MAPCs [8]. As reported, HSCs migrate from the fetal liver to the developing BM tissue by the end of the second trimester of gestation, according to the gradient of the CXCL12 chemokine [9]. It is likely that the BM is also colonized by several other tissue-committed stem cells (TCSCs) that may circulate during ontogenesis/organogenesis, because of their CXCR4 receptor expression [3]. Therefore, the CXCL12-CXCR4 axis plays a pivotal role in the accumulation of TCSCs in the developing BM. These TCSCs may play an important role as a reserved pool of SCs for organ/tissue regeneration during postnatal life.

Supporting this hypothesis, Ratajczak's group has recently identified unique very small embryonic-like (VSEL) SCs in both mouse BM and human cord blood (CB) cells as a putative nonembryonic source of pluripotent stem cells (PSCs) [3, 10 –18]. They isolated mouse BM VSELs as lineage-negative (Lin⁻), CD45-negative (CD45⁻), Sca-1-positive (Sca-1⁺) cells, and also isolated human CB VSELs as Lin⁻CD45⁻CD34⁺ or CD133⁺ or CXCR4⁺ cells by fluorescence-activated cell sorting (FACS) [10 –18]. These mouse BM VSELs and human CB VSELs expressed several markers characteristic for PSCs, including Oct3/4, Nanog, Rex1, and stage-specific embryonic antigen (SSEA)-1 [11,12,18]. Moreover, these mouse BM VSELs were able to differentiate into cardiomyocytes (mesoderm); neurons, astrocytes, and oligodendrocytes (ectoderm); and pancreas (endoderm) in coculture systems [11]. These results demonstrated that mouse BM VSELs can differentiate into cells from all three germ layers.

According to the subsequently reported data [3,10 –18], mouse BM VSELs have been characterized as follows: (1) very small (3–5 μm); (2) express PSC markers, such as Oct3/4, Nanog, SSEA-1, and Rex1; (3) responsive to a SDF-1 (CXCL12) gradient; and (4) possess large nuclei that contain euchromatin. Although the mouse BM VSELs were isolated as Lin⁻CD45⁻Sca-1⁺ cells by FACS, the incidence of BM VSELs in BM-derived nucleated cells (BMNCs) is very low (0.002%–0.006%), as reported [16,17]. Therefore, it is very difficult to isolate BM VSELs efficiently. Many other independent groups have confirmed the existence of mouse BM VSELs and have shown that they could be isolated from various tissues [19 –26]. However, other recent studies have shown that there is a lack of SC characteristics in VSELs isolated from mouse BM [27] or human CB [28]. Moreover, Miyanishi et al. recently reported that they were unable to detect VSELs in mouse BM with any of the reported SC potentials [29]. These studies aroused controversy regarding the existence of pluripotent VSELs in adult tissues.

In this study, we developed a highly efficient method for isolating Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells, which resembled BM VSELs using FACS following enzymatic treatment of murine bone. We designated these cells as bone-derived VSELs (BD VSELs). The present data demonstrate that the majority of the Lin⁻CD45⁻Sca-1⁺ cells reside in the mouse bone tissue rather than the BM. These BD VSELs had cellular characteristics similar to those of BM VSELs. However, the size, the immunophenotype and the gene expression profile of BD VSELs were different from those of original BM VSELs [3,10 –18]. Therefore, it is important to elucidate their cellular and/or SC characteristics, and the potential relationship between BD VSELs and BM VSELs, which appear to be PSCs in postnatal adult tissues.

Materials and Methods

Animals

Five-week-old female C57BL/6 mice and transgenic mice (C57BL/6 strain) that expressed the GFP [C57BL/6-Tg(CAG-EGFP)] were purchased from Shimizu Laboratory Supplies. All mice were handled under sterile conditions and were maintained in germ-free isolators located in the Central Laboratory Animal Facilities of Kansai Medical University. All mice were provided food ad libitum and were maintained in compliance with Kansai Medical University's guidelines for the care and use of laboratory animals in research. The animal experiments were approved by the Animal Care Committees of Kansai Medical University.

Preparation of mouse cells

The tibiae and femurs were removed from these mice when they were 6–8 weeks old. The BM was flushed out from these bones. After cleaning to remove muscle and connective tissue, the remaining bone tissue specimens were washed twice with Ca²⁺ and Mg²⁺-free Dulbecco's Phosphate-Buffered Saline (PBS⁻; Nacalai Tesque) and then crushed in a mortar. The recovered tissues were incubated in cell dissociation buffer containing α-medium (Gibco) supplemented with 5% fetal calf serum (FCS; Thermo Trace Ltd.) containing 1.5 mg/mL collagenase type I (Invitrogen), 2 mg/mL dispase (Invitrogen), and 0.004% DNase I (Takara Shuzo) at 37°C for 1 h. The suspension was filtered with a cell strainer (BD Falcon) to remove debris, and bone-derived nucleated cells (BDNCs) were collected by centrifugation at 1,500 rpm (470G) for 10 min at 4°C. Subsequently, the pellet containing BDNCs was washed twice with α-medium supplemented with 5% FCS, and the suspension was filtered through a cell strainer. Finally, the cells were resuspended in 1 mL of PBS⁻ containing 2% FCS (PBS/FCS). The BM was immersed in hemolysis buffer (BD Pharm Lyse; BD Biosciences) for 5–10 min to lyse the red blood cells, and the BMNCs were collected by centrifugation at 1,500 rpm (470G) for 10 min at 4°C. Subsequently, the pellet containing BMNCs was washed twice with α-medium with 5% FCS. Finally, the cells were resuspended in 1 mL of PBS/FCS.

Antibody staining

The BDNCs and BMNCs were preincubated with mouse BD Fc Block (BD Biosciences) for 20 min at 4°C to reduce nonspecific binding, and then were stained for 30 min at 4°C with the following antibodies: APC-conjugated CD45 (30-F11; eBioscience), Pacific Blue-conjugated Sca-1 (Ly6A/E) (D7; BioLegend), FITC-conjugated CD45 (30-F11; eBioscience), TER119 (TER-119; eBioscience), CD45R/B220 (RA3-6B2; eBioscience), Gr-1 (RB6-8C5; eBioscience), TCRαβ (H57-597; eBioscience), TCRγδ (GL3; eBioscience), CD11b (M1/70; eBioscience), PE-conjugated CD29 (HMb1-1; eBioscience), CD49e (HMa5-1; eBioscience), CD34 (RAM34; BD Biosciences), CD49f (GoH3; eBioscience), CD71 (R17.217.1.4; eBioscience), CD90.2 (53–2.1; BD Biosciences), CD105 (MJ7/18; eBioscience), CD133 (13A4; eBioscience), c-kit (2B8; eBioscience), c-Met (eBioclone7; eBioscience), CXCR4 (2B11; eBioscience), Flk-1 (89B3A5; BioLegend), Notch1 (mN1A; eBioscience), Notch3 (HMN3-133; eBioscience), Notch4 (HMN4-14; eBioscience), PECAM-1 (390; BioLegend), PDGFRα (APA5; eBioscience), PDGFRβ (APB5; eBioscience), SSEA-1 (MC-480; BioLegend), SSEA-4 (MC813-70; BD Biosciences), Tie2 (TEK4; eBioscience), and VE-cadherin (BV13; BioLegend). Subsequently, the cells were washed with Annexin V binding buffer (Beckman Coulter) and stained for 15 min on ice with an FITC-conjugated Annexin V antibody (Beckman Coulter). The 7-amino-actinomycin D (7-AAD; Beckman Coulter) fluorescence was measured, and a live cell gate was defined to exclude the cells positive for 7-AAD. Then, a fraction of the target cells, including BM VSELs and BD VSELs, was sorted as described in the Results section. To detect intracellular Oct3/4, Nanog and Histone H1, the BDNCs were first stained for cell surface markers (Pacific Blue-conjugated Sca-1 and PE-conjugated lineage markers such as CD11b, CD45, CD45R/B220, Gr-1, TCRαβ, TCRγδ, and TER119). These cells were then fixed and permeabilized using a BD Cytofix/Cytoperm kit (BD Bioscience) and stained for the intracellular proteins (Alexa Fluor 488-conjugated Oct3/4; EM92, eBioscience, Alexa Fluor 488-conjugated Nanog; eBioMLC-51, eBioscience or Alexa Fluor 488-conjugated Histone H1; sc-8030; Santa Cruz Biotechnology). Finally, these cells were stained with 7AAD and analyzed using a flow cytometer.

Flow cytometric analysis and FACS

The flow cytometric analysis (FCM) analysis and FACS sorting were performed using a BD FACSCanto II flow cytometer (BD Biosciences) and a BD FACSAria cell sorter (BD Biosciences). A fraction of the VSEL phenotype, Lin⁻CD45⁻Sca-1⁺ cells, were sorted as described in the Results section. As controls, we also sorted Lin⁻CD45⁻Sca-1⁺PDGFRα⁺ cells (MSCs) and Lin⁻CD45⁺ Sca-1⁺c-kit⁺ cells (KSLs). The data were processed using the FlowJo software program (Tree Star). The cell sizes in the FCM and FACS analyses were defined by microspheres (Flow Cytometry Size Calibration Kit; Molecular Probes).

ImageStream analysis

The BDNCs and BMNCs were harvested and stained for cell surface markers (PE-Cy7-conjugated Sca-1, and PE-conjugated lineage markers such as CD11b, CD45, CD45R/B220, Gr-1, TCRαβ, TCRγδ, and TER119). To detect intracellular Oct3/4 and Nanog, the cells stained for surface markers as described above were further stained for the intracellular antigens (Alexa Fluor 488-conjugated Oct3/4 or Alexa Fluor 488-conjugated Nanog) using a BD Cytofix/Cytoperm kit (BD Bioscience). Nuclei were stained with DRAQ5 (Biostatus Ltd.). Cells were analyzed on an ImageStream X mark II imaging cytometer (Amnis Corporation) and the data were processed using the IDEAS image analysis software program (Amnis Corporation).

Microscopic analysis

To observe the morphology of the BD VSELs, cells were sorted onto microscope slides using a BD FACSAria cell sorter with an automatic cell deposition unit. The nuclei were counterstained with Hoechst 33342 (Invitrogen) and then the cells were observed under a Zeiss LSM510 Meta confocal microscope (Carl Zeiss).

RNA preparation and RT-PCR

Total RNAs were isolated with the RNeasy Plus Micro kit (Qiagen) according to the manufacturer's instructions. The cDNAs were synthesized using the iScript DNA synthesis kit (Bio-Rad Laboratories). The cDNAs were amplified using AmpliTaq Gold DNA polymerase (Applied Biosystems). The expression levels of ES cell markers (Oct3/4 and Nanog), MSC markers (Nestin, Angpt1, Cxcl12, and VE-Cadherin), and HSCs markers (C-kit, Gata2, and Tal1) were analyzed by quantitative real-time RT-PCR (RQ-PCR). The RQ-PCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems) in a Rotor-Gene Q (Qiagen), and the data were quantified by the comparative CT method. The primer sets used in the RQ-PCR are shown in Supplementary Table S1(Supplementary Data are available online at www.liebertpub.com/scd).

Cell culture

One hundred sorted BD VSELs, MSCs, and HSCs were seeded per well of 24-well plates (BD Biosciences) and cultured in MSC culture medium (α-MEM supplemented with 20% FCS (Thermo Trace) and penicillin-streptomycin-glutamine (Life Technologies); HSC culture medium (α-MEM supplemented with 30% FCS (Thermo Trace), murine stem cell factor (mSCF, 20 ng/mL), murine interleukin-3 (mIL-3, 10 ng/mL), murine granulocyte-macrophage colony stimulating factor (mGM-CSF, 10 ng/mL), human FMS-like tyrosine kinase 3 ligand (hFlt3 ligand, 50 ng/mL; R&D Systems), human thrombopoietin (hTPO, 20 ng/mL), human erythropoietin (hEPO, 2U), murine insulin growth factor 2 (mIGF2, 20 ng/mL; R&D Systems) and penicillin-streptomycin-glutamine or ES cell culture medium (DMEM (high glucose; Wako Pure Chemicals), Knockout SR (Life Technologies), MEM nonessential amino acids (Life Technologies), sodium pyruvate (Life Technologies), 2-mercaptoethanol (Sigma–Aldrich), murine leukemia inhibitory factor (mLIF, 1,000 U/mL; Wako Pure Chemicals), and penicillin-streptomycin-glutamine). Human TPO and EPO were generous gifts from Kyowa Hakko Kirin Company Ltd. The cell culture plates were incubated at 37°C with 5% CO₂ in air.

Cocultures with OP-9 or C2C12 cells

Freshly sorted GFP mouse-derived Lin⁻CD45⁻Sca-1⁺ BD VSELs were plated over murine OP-9 stroma cells in α-medium with 20% FCS for 5 days. The cells were subsequently trypsinized, washed by centrifugation in α-medium, and replated in methylcellulose culture medium [30] in the presence of a cocktail of cytokines, including mSCF, hTPO, mIL-3, mGM-CSF, hFlt3 ligand, mIGF2, and hEPO, to analyze their hematopoietic differentiation potential, as reported previously [31]. In addition, freshly sorted Lin⁻CD45⁻Sca-1⁺ BD VSELs were plated over murine C2C12 myoblastic cells in DMEM medium with 2% FCS to test whether these cells formed spheres, as reported previously [32].

In vivo cell proliferation analysis using BrdU incorporation

For a short pulse (1 day) of 5-bromo-2-deoxyuridine (BrdU) labeling, mice received two intraperitoneal injections of BrdU (1.0 mg/kg body weight) at the starting point and 3 h before being euthanized. For long-term BrdU labeling, mice received BrdU in their drinking water (0.8 mg/mL in sterilized water) for 4 weeks. BrdU-labeled BDNCs and BMNCs were harvested and stained for cell surface markers (APC-conjugated CD45, Pacific Blue-conjugated Sca-1, and PE-conjugated lineage markers such as CD11b, CD45R/B220, Gr-1, TCRαβ, TCRγδ, and TER119). These cells were then fixed and stained with an FITC-conjugated anti-BrdU antibody using the FITC BrdU Flow Kit (BD Biosciences) according to the manufacturer's instructions. These antibody-stained specimens were analyzed using FCM.

Liver injury model

The mice were intraperitoneally injected with 10 mL of carbon tetrachloride (CC1₄) and olive oil (CC1₄: olive oil = 1:9 mixture) per kg of body weight to induce acute liver injury. The control mice were injected intraperitoneally with olive oil alone. Then, the mice were sacrificed on days 1, 2, 3, or 7 post-CCl₄ administration for the histological and FCM analyses. Mice sera were prepared from peripheral blood, which was collected from the oribital sinus under anesthesia. The aspartate aminotransferase (AST) and alanine aminotransferase (ALT) activities in mouse sera were determined by ALT and AST Activity Assay Kits (Wako Pure Chemicals). For the histological analysis, the liver tissue specimens were immersed in 4% paraformaldehyde in PBS⁻ (pH 7.2) for 48 h. Fixed tissues were dehydrated in a graded ethanol series, cleared in xylene, and embedded in paraffin. Tissue sections were deparaffinized with xylene and rehydrated in a graded alcohol series. The sections were stained with hematoxylin (Wako Pure Chemicals) and eosin (Wako Pure Chemicals) and observed under an OLYMPUS BX-41 microscope (Olympus).

Transplantation of GFP mouse-derived BD VSELs into mice with acute liver injury and analysis of the tissue repair by immunohistochemistry

C57BL/6 mice were injected with CC1₄ to induce acute liver injury as described above (see the Liver injury model section). Twenty-four hours after the injection of CCl₄, 1.2 × 10³ GFP mouse-derived BD VSELs were transplanted via a tail vein into each recipient mouse with liver injury. As controls, GFP mouse-derived 2 × 10³ whole BMNCs and BDNCs were also transplanted. In addition, GFP mouse-derived purified 2 × 10³ BD VSELs were transplanted into each control recipient mouse, which was injected olive oil alone. The liver tissues were analyzed for GFP expression by immunohistochemistry 3 weeks after transplantation. The liver tissues were fixed by perfusion with 4% paraformaldehyde in PBS⁻ (pH 7.2), and sections were prepared for frozen sectioning (7 μm thickness). The liver sections were placed in PBS⁻ containing 0.1% Triton X-100 (Nacalai Tesque) for 5 min to increase the antibody permeability, and then were stained with an anti-GFP primary antibody (rabbit polyclonal; Medical & Biological Laboratories Co. Ltd.) using a tissue staining kit (Cell & Tissue Staining Kit; R&D Systems) according to the manufacturer's instructions. The peroxidase activity was detected with H₂O₂/diaminobenzidine (DAB) substrate solution, and the tissue sections were counterstained with hematoxylin before dehydration and mounting.

Statistical analysis

Statistical analyses were performed to compare the differences in the numbers of Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells from BM and bone, to compare the differences in the numbers of BrdU-positive BD VSELs from the short pulse of BrdU-labeled mice and long term of BrdU-labeled mice, and to compare the differences in the numbers of VSEL phenotypic cells in the bone, BM, and PB from each of the liver injury model mice and control mice. The data are expressed as the means ± SD or SE. Two-tailed Student's t-tests were used to compare the differences in the data. A value of P < 0.05 was considered to be statistically significant.

Results

Isolation and purification of BD VSELs: comparison with BM VSELs

We first tried to isolate murine BM VSELs from BMNCs and BD VSELs from BDNCs by multicolor FACS, targeting a population of Lin⁻CD45⁻Sca-1⁺ cells as shown in Fig. 1. Based on the predicted very small size of the VSELs [10 –17], we performed preliminary experiments to analyze the size of the BD VSELs by the ImageStream system (ISS), as described later. As a result, an analysis of events in the 2–4 μm size range showed that most of the images were cell debris. Based on these ISS data, the R1 gate was set on the FSC channel using 4 and 10 μm synthetic beads (Fig. 1Aa, Ba). Then, the 7-AAD⁻Lin⁻AnnexinV⁻ cells were gated as R2 (Fig. 1Ab, Bb). The 7-AAD⁻Lin⁻AnnexinV⁻CD45⁻Sca-1⁺ cells were then gated as R3 (Fig. 1Ac, Bc). These R3-gated cells were back-gated to FSC/SSC scattergrams. These were plotted in the low FSC/SSC channel positions (Fig. 1Ad, Bd).

FIG. 1.

Isolation and purification of Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells from murine bone tissues. Representative flow cytometric analysis (FCM) profiles of bone-derived nucleated cell (BDNC)-derived VSEL phenotypic cells (BD VSELs) (A) and bone marrow-derived nucleated cell (BMNC)-derived very small embryonic-like stem cells (BM VSELs) (B). (a) The R1 gate was set on the FSC channel using 4 and 10 μm synthetic beads. (b) The hematopoietic (CD11b⁺, CD45R/B220⁺, Gr-1⁺, TCRαβ⁺, TCRγδ⁺, and TER119⁺) cells and Annexin-V⁺ cells (preapoptotic cells), and dead cells (7-AAD⁺) were gated out (R2). The R2-gated cells were further subdivided according to their CD45 and Sca-1 expression. (c) Cells residing in the R3 gate (CD45⁻Sca-1⁺) were classified as VSEL phenotypic cells (BD VSELs and BM VSELs). (d) The R3-gated cells were back- gated to FSC/SSC scattergrams, and their distributions were highlighted by black dots. (C) The incidences of BM VSELs and BD VSELs were approximately 25 and 1500 cells per 10⁶ events, respectively. The bars represent the mean ± SD of seven independent experiments. The data were evaluated by two-tailed Student's t-tests. Asterisks indicate statistical significance (P < 0.01).

The incidences of BM VSELs in the BMNCs and of BD VSELs in the BDNCs were 0.002% ±0.001% (n = 13) and 0.15% ±0.108% (n = 13), respectively. The number of BD VSELs (1507 ± 1082) was significantly larger than that of BM VSELs (25 ± 17) (Fig. 1C). Therefore, the enzymatic treatment of bone tissues provided about 100 times greater efficiency for the isolation of Lin⁻CD45⁻Sca-1⁺ VSEL-phenotypic cells. The BD VSELs were small, but an estimated diameter of approximately 6.5 μm, and possessed a relatively large nucleus surrounded by a narrow rim of cytoplasm (Supplementary Fig. S1). This size of BD VSELs appeared to be larger than the reported size of BM VSELs (2–4 μm) [11,12,17]. However, they were similar to the VSELs identified in other organs, including the brain, pancreas, kidney, and lung [17].

ImageStream analysis of the size and PSC marker expression of BD VSELs and BM VSELs

We next analyzed the size and expression of PSC markers, such as Oct3/4 and Nanog, at the protein level, using the ISS. As shown in Fig. 2Aa–e, the size of the Lin⁻CD45⁻Sca-1⁺ VSEL-phenotypic cells in the BDNCs was widely distributed from 6 to 20 μm. However, there were almost no live cells in the 2–6 μm size range (Fig. 2B). Most of the large cells (over 10 μm) were PDGFRα⁺ MSCs, as described later. Therefore, we analyzed the expression of Oct3/4 and Nanog in small cells distributed between 6 and 9 μm. As expected based on the results of the RQ-PCR analysis (Supplementary Table S2), it was hard to detect Oct3/4⁺ or Nanog⁺ VSEL-phenotypic cells (Supplementary Figs. S2 and S3). These results suggested that the purified population of bone-derived Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells was still heterogeneous in terms of the cell size, and did not express PSC markers, such as Oct3/4 and Nanog. We also analyzed the expression of Oct3/4 and Nanog in BM VSEL phenotypic cells. As shown in Supplementary Fig. S4, the size of these BM VSEL phenotypic cells were 6–9 μm and they did not express these PSC markers.

FIG. 2.

The cell size and morphology of BD VSELs. (A) Representative histograms of BDNCs obtained using an imaging cytometer. (a) The R1 gate (DRAQ5⁺) was set to exclude cell debris. (b) DRAQ5⁺ BDNCs were classified according to the cell size (diameter). The cell sizes were further subdivided into 2–4 μm, 4–6 μm, 6–9 μm, 9–14 μm, and over 14 μm. There were no cells in the extremely small size fraction (<6 μm in diameter). (c) The hematopoietic (CD11b⁺, CD45⁺, CD45R/B220⁺, Gr-1⁺, TCRαβ⁺, TCRγδ⁺, and TER119⁺) cells were gated out (R2). (d) The R2-gated Lin⁻CD45⁻ cells were further subdivided according to their Sca-1 expression and the R3 gate was set on Sca-1⁺ cells. (e) BDNCs-derived DRAQ5⁺Lin⁻CD45⁻Sca-1⁺ cells were classified according to the cell size (diameter). (B) Representative brightfield images of BDNCs-derived DRAQ5⁺Lin⁻CD45⁻Sca-1⁺ cells. These cells were classified according to the cell size (2–4 μm, 4–6 μm, 6–9 μm, 9–14 μm, and over 14 μm). There were no cells in the extremely small size fraction (<6 μm in diameter, shown as “Not detected”). The scale bars are indicated 10 μm, respectively.

Surface immunophenotype of BD VSELs and BM VSELs

We next analyzed the expression of various surface molecules/antigens, including HSC, MSC, endothelial cell, ES cell markers, and others on BD VSELs (Fig. 3). Among the HSC markers, BD VSELs strongly expressed CD34, but weakly expressed CD133, CXCR4 and Tie2. They did not express c-kit, Notch1, Notch3, or Notch4. Interestingly, the BD VSELs strongly expressed MSC markers, including CD90.2, CD29, CD49e, and CD49f. They weakly expressed PDGFRα and CD105. They also moderately to weakly expressed endothelial cell markers, including CD31, VE-cadherin, and Flk-1. However, they did not express ES cell markers, such as SSEA-1 and SSEA-4. Among the other cell markers examined, these cells weekly expressed CD71, but did not express c-Met.

FIG. 3.

The cell surface marker profiles of BD VSELs. The expression of cell surface molecules/antigens, including (A) hematopoietic stem cell (HSC) markers, (B) mesenchymal stem/stromal cell (MSC) markers, (C) endothelial cell markers, (D) ES cell markers, and (E) other markers were analyzed by FCM. Solid histograms, isotype controls; open histograms, specific antibodies, as indicated. All of the experiments were repeated three times and the representative FCM profiles are shown.

Next, we analyzed the expression of above-mentioned molecules/antigens on BM VSEL phenotypic cells (Supplementary Fig. S5). Among HSC markers, BM VSEL phenotypic cells weakly expressed CD34, but they did not express CD133 and CXCR4. In contrast, they distinctly expressed Tie2. In addition, the expression of c-kit, Notch 1, 3, and 4 were not detected. BM VSEL phenotypic cells strongly expressed MSC markers, including CD29, CD49f, and CD105. Interestingly, they weakly to moderately expressed CD31, VE-cadherin, Flk-1, and c-Met. However, they did not express ES cell markers, such as SSEA-1 and SSEA-4. The surface immunophenotype of BD VSELs was therefore quite unique and significantly different from that of BM VSEL phenotypic cells. These results suggest that BD VSELs may have the potentials to differentiate into various kinds of SCs, including HSC, MSC, and endothelial progenitor cells.

Gene expression profiles of BD VSELs: comparison with those of HSCs, MSCs, and ES cells

The gene expression profiles were analyzed using RQ-PCR to evaluate the expression of ES cell markers (Oct3/4, Nanog, Rex1, and Dppa3), HSC (c-kit⁺Sca-1⁺Lin⁻, KSL) markers (C-kit, Tal1, and Gata2) and MSC markers (Nestin, Angpt1, Cxcl12, and VE-Cadherin). Unexpectedly, BD VSELs expressed high levels of Nestin and VE-Cadherin, as shown in Fig. 4 and Supplementary Table S2. The gene expression profile of BD VSELs was unique, and it was different from those of HSCs, MSCs, and ES cells. Moreover, we did not detect PSC markers, such as Oct3/4 and Nanog, by RQ-PCR, as shown by the ISS analyses. These results are consistent with recent reports, in which BM VSELs did not express Oct3/4 [27,29]. Overall, the gene expression profile of the BD VSELs was clearly distinct from those of the well-defined populations of HSCs (KSLs), MSCs, and ES cells.

FIG. 4.

The results of a comparative analysis of the gene expression profiles between BD VSELs, MSCs, HSCs, and ES cells. The gene expression profiles were analyzed using RQ-PCR to evaluate the expression of ES cell markers (Oct3/4, Nanog, Rex1, and Dppa3), HSC (c-kit⁺Sca-1⁺Lin⁻, and KSL) markers (C-kit, Tal1, and Gata2) and MSC (PDGFRα⁺Sca-1⁺, and PαS) markers (Nestin, Angpt1, Cxcl12, and VE-Cadherin). The values of each marker expression in each cell were set as the reference (one-fold) and the characteristics of the gene expression patterns were shown as graphical chart. (A) BD VSELs, (B) MSCs, (C) HSCs, and (D) ES cells. The expression levels (fold) were plotted concentrically. The values of each gene's expression are indicated in the figures, and are the mean value of three independent experiments. More extensive data are shown in Supplementary Table S2.

Separation of BD VSELs and BM VSELs from HSCs and MSCs identified in mouse bones and BMs

As reported in the literature [3 –8], mouse and human BM cells are heterogeneous populations that contain various kinds of SCs, including HSCs, MSCs, and VSELs. Therefore, it is important to accurately separate BD VSELs and BM VSELs to reduce/protect against unexpected contamination with other types of TCSCs.

In our FACS protocols, we first set the R1 gate on the FSC channel using 4 and 10 μm synthetic beads, as shown in Fig. 1. In the data shown in Supplementary Fig. S6A and B, however, we first gated 7-AAD⁻Lin⁻AnnexinV⁻ cells (R1), which contained small and large cells. These R1-gated cells were further separated into four cell fractions according to the expressions of Sca-1 and c-kit. As shown in Supplementary Fig. S6Acii, the bone-derived Lin⁻Sca-1⁺c-kit⁺ (KSL) (R2) cells were all CD45⁺. Almost all of these HSCs (KSLs) were large cells (>10 μm), as shown in the FSC histogram in Supplementary Fig. S6Acii. The Lin⁻c-kit⁻Sca-1⁺ cells (R3) contained CD45⁺ and CD45⁻ cells. As was clearly seen, most of the Lin⁻c-kit⁻Sca-1⁺CD45⁻ cells, containing BD VSELs, were small (<10 μm) (Supplementary Fig. S6Adiii). Therefore, if the R1 gate was first set on the FSC channel using 4 and 10 μm synthetic beads (Fig. 1), we could accurately separate BD VSELs without contamination of HSCs (KSLs). As shown in Supplementary Fig. S6Bdiii, most of the BM VSELs were also small (<10 μm); however, the number of these cells was very limited, as shown in Fig. 1.

We next analyzed whether the contamination of MSCs into BD VSEL fractions could be ruled out (Supplementary Fig. S7). First, the 7-AAD⁻ cells were gated as R1. Among the R1- gated cells, the CD45⁻Lin⁻AnnexinV⁻ cells were gated as R2. Then, these R2-gated cells were further separated into four cell fractions according to the expression of Sca-1 and PDGFRα, which is a marker of MSCs [33]. As shown in Supplementary Fig. 7C and D, almost all of the Lin⁻CD45⁻Sca-1⁺PDGFRα⁺ MSCs (R3) were large cells (mean FSC FI = 519). In contrast, most of the R4-gated Lin⁻CD45⁻Sca-1⁺PDGFRα⁻ cells, containing BD VSELs, were small (mean FSC FI = 160). Therefore, if the R1 gate was first set on the FSC channel using 4 and 10 μm synthetic beads (Fig. 1), we could accurately separate the BD VSELs without contamination of MSCs.

In vitro cultures of BD VSELs

As reported, BM VSELs were able to differentiate into nonhematopoietic tissues, including cardiomyocytes, neurons, and pancreas in the cocultures with freshly isolated BM cells [11,14]. In addition, 5%–10% of purified BM VSELs, if plated over a murine C2C12 myoblastic cell feeder layer, formed spheres that resembled ES cell-derived embryoid bodies [32]. These BM VSEL-derived spheres showed the capacity to differentiate into all three germ-layers after replating over C2C12 cells [32]. Moreover, the BM VSELs differentiated into hematopoietic lineages after being cocultured over OP9 stromal cells [31].

Our identified BD VSELs were somewhat similar to the BM VSELs (Table 1). They expressed various cell surface markers, which were specific for HSCs, MSCs, and endothelial cells (Fig. 3). Therefore, we tried to induce the differentiation of BD VSELs in vitro. First, we simply cultured freshly sorted Lin⁻CD45⁻Sca-1⁺ cells (BD VSELs) in MSC, HSC, and ES cell culture media (Fig. 5A, B). As shown in Fig. 5Ca and b, a portion of the BD VSELs proliferated and showed a colony-like morphology in HSC culture medium on day 5 of culture. Then, we analyzed the expression of hematopoietic cell markers (CD45, Gr-1 (Ly-6G), CD11b c-kit, and Sca-1) on 7-AAD⁻BD VSELs-derived cells by FCM (Fig. 5Da). Most of these cells were larger than the freshly sorted BD VSELs (>10 μm) (Fig. 5Db). They expressed HSC markers, including CD45, Gr-1, CD11b, and c-kit but not Sca-1 (Fig. 5Dc). These results demonstrated that the BD VSELs could differentiate into hematopoietic lineage cells in vitro.

FIG. 5.

Hematopoietic differentiation potential of freshly isolated BD VSELs. (A) Representative FCM profiles of BDNC-derived BD VSELs used in culture. The R1 gate was set on the FSC channel using 4 and 10 μm synthetic beads. The hematopoietic (CD11b⁺, CD45, CD45R/B220⁺, Gr-1⁺, TCRαβ⁺, TCRγδ⁺, and TER119⁺) cells and dead cells (7-AAD⁺) were gated out (R2). Cells residing in the R3 gate (Lin⁻CD45⁻Sca-1⁺ and BD VSELs) did not express c-kit. (B) A schematic diagram of the in vitro cultures of BD VSELs. One hundred sorted BD VSELs (R3-gated cells) were seeded and cultured in MSC, HSC, and ES cell culture media, respectively. (C) The microscopic morphologies of BD VSELs-derived proliferating cells after 5 days of culture in HSC culture medium. (a, b) The cells shown in each photograph were in different wells of cultures. The scale bars indicated 400 μm, respectively. (D) FCM profiles of BD VSELs-derived cells in HSC culture medium. (a) The dead cells (7-AAD⁺) were gated out. (b) The results of a cell size analysis of 7-AAD⁻ BD VSEL-derived cells. Synthetic beads (4 and 10 μm beads) were used to measure the cell size. (c) The expression of hematopoietic cell markers (CD45, Gr-1 (Ly-6G), CD11b c-kit, and Sca-1) on the 7-AAD⁻ BD VSELs-derived cells was analyzed by FCM. Solid and open histograms represent unstained controls cells [shown as Abs (-)] and the experimental group (stained by specific antibodies, shown as the specific antibody), respectively.

Table 1.

Comparison of Cellular and/or Stem Cell Characteristics of Mouse BD VSELs and BM VSELs

	Mouse BD VSELs (this study)	Mouse BM VSELs (reported studies) ^a
Isolation markers	Lin⁻CD45⁻Sca-1⁺	Lin⁻CD45⁻Sca-1⁺
Cell source	Bone	Bone marrow
Cell size	6–9 μm (Fig. 2)	3–5 μm (Ref. 17)
	6.5 μm (Supplementary Fig. S1)	>6 μm (Ref. 19)
Immunophenotypes (Cell surface markers)
HSC markers	CD34⁺, CD133⁺, CXCR4^low, c-kit⁻, Tie2⁺, Notch1⁻, Notch3⁻, Notch4⁻	CD34⁺, CD133⁺, CXCR4⁺
MSC markers	PDGFRα^low, PDGFRβ⁻, CD90⁺, CD29⁺, CD49e⁺ CD49f⁺, CD105⁺	CD29⁻, CD90⁻, CD105⁻
Endothelial cell markers	PECAM-1⁺, VE-cadherin⁺, Flk-1^low	Flk-1⁻
Others	c-Met⁻, CD71^low	c-Met⁺, LIF-R⁺, MHC I⁻, HLA-DR⁻
Pluripotent stem cell markers (Oct3/4, Nanog, SSEA)
RQ-PCR analysis /FCM analysis	Not detected	Strongly expressed (Oct3/4, Nanog)
ISS analysis Immunocytometry	Not detected	Nuclear expression (Oct3/4, Nanog) Cytoplasmic expression (SSEA-1)
Other markers	Nestin⁺, Cxcl12⁺, Tal1⁺	Fetal alkaline phosphatase⁺ High telomerase activity
In vitro differentiation potential	Hematopoietic cell (Liquid culture)	Neuron, Cardiomyocyte, Pancreas (unpurified BM cells coculture)Hematopoietic cell (OP-9 cells coculture)Osteoblast, Endothelial cell, Neural cell, Insulin-producing endodermal cell (C2C12 cells coculture)
In vivo differentiation potential (Transplantation)	Liver cell (Hepatocyte? Cell fusion?)	Bone, Lung epithelial cell, Osteoblast, Endothelial cell, Early neuron, Adipocyte, Hematopoietic cell^b
Mobilization into PB	By liver injury	By tissue injury or by G-CSF administration
In vitro cell proliferation	Unsuccessful (C2C12 cells coculture)	By coculture with C2C12 cells or by addition of sex hormones
In vivo cell proliferation (BrdU incorporation)	Slow cycling	Slow cycling

These data were originally reported in the literatures as reference nos. 3, 10–26, 31, 32, 34, 35 and 37.

In vivo differentiation potential into hematopoietic cell was reported in extremely low level [26].

BD VSELs, bone-derived very small embryonic-like stem cells; BM, bone marrow; FCM, flow cytometric analysis.

We simultaneously cultured BD VSELs in both MSC and ES cell culture media. However, the BD VSELs could not proliferate in these culture media, and they were floating as single cells (Supplementary Fig. S8Ba). On the other hand, BM-derived HSCs (KSL cells) and MSCs could proliferate in HSC and MSC culture media, respectively (Supplementary Fig. S8Ca, b).

Cocultures of BD VSELs with C2C12 or OP-9 cells

Next, we cultured BD VSELs with stromal cells, such as OP9 and C2C12 cells. The GFP mouse-derived BD VSELs could not proliferate in the cocultures with OP-9 cells (Supplementary Fig. S9Aa). Only the BM-derived HSCs (KSL cells) could proliferate in the cocultures with OP-9 cells (Supplementary Fig. S9Ab). BDNC-derived MSCs could adhere to the OP-9 cells; however, they did not proliferate during the observation period (∼20 days) (Supplementary Fig. S9Ac). We replated the recovered cells from cocultures of BD VSELs, HSCs, and MSCs with OP-9 cells in methylcellulose culture medium. As shown in Supplementary Fig. S9B, only the HSCs (KSL cells) formed many hematopoietic colonies after 14 days of culture.

In separate experiments, we also cultured sorted BD VSELs over C2C12 cells to test the sphere formation, as reported previously [32]. However, we could not observe any sphere formation (data not shown).

Cell proliferation status of BD VSELs in vivo

Next, we analyzed the cell proliferation status of BD VSELs in vivo using a BrdU incorporation analysis (Supplementary Fig. S10A, B). In these experiments, the R1 gate was set on 7-AAD⁺ cells. Then, R2-gated BrdU⁺ cells were further subdivided into Lin⁻CD45⁻Sca-1⁺ cells (R4). Finally, we set the R5 gate on the FSC channel using 4 and 10 μm synthetic beads as shown in Supplementary Fig. S10Aa and b. The number of BrdU⁺ cells in the short-term labeling (1 day) experiments was only 56 ± 23/1 × 10⁶ events. In contrast, the number of BrdU⁺ cells observed in the long-term labeling (4 weeks) experiments tended to increase to 442 ± 352/1 × 10⁶ events (Supplementary Fig. S10Ba) (P = 0.1). In addition, approximately 20%–30% of the BD VSELs were proliferating in both the short- and long-term BrdU-labeling experiments (Supplementary Fig. S10Bb). These results suggest that the BD VSELs were slowly cycling in vivo in the steady state.

In vivo kinetics of BD VSELs after acute liver injury

It was previously reported that BM VSELs were mobilized into the peripheral blood (PB) following G-CSF administration, liver injury, cardiac damage, or stoke [22,34,35]. Therefore, we analyzed the in vivo kinetics of BD VSELs after severe acute liver injury induced by carbon tetrachloride (CCl₄) administration. As shown in Fig. 6A, the CCl₄ administration induced massive hepatocyte necrosis, which reached a peak level on day 2 after injection, as determined by a histological examination. Simultaneously, the serum levels of transaminases (AST and ALT) increased and reached a peak level on day 2 (Supplementary Fig. S11). We serially analyzed the numbers of Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells in bone, BM, and PB on days 1, 2, 3, and 7 after the CCl₄ administration by FCM. Interestingly, the number of BD VSELs significantly increased on days 1 and 2 (Fig. 6Ba), while the number of BD VSELs returned to the control level after day 3. The number of VSELs in the BM did not change significantly until day 3; however, it significantly increased on day 7. The number of VSELs in the PB, significantly increased on day 3. Overall, these findings suggest that the BD VSELs were most likely mobilized into the PB through BM (Fig. 6Ba–c).

FIG. 6.

In vivo kinetics of BD VSELs after acute liver injury. (A) Representative tissue images of murine livers after the administration of carbon tetrachloride (CCl₄). Sequential liver tissue images are shown (1, 2, 3, and 7 days after CCl₄ administration). A healthy liver section is shown as the “Control.” The white arrows indicate areas of hepatocyte necrosis. The scale bars indicated 500 μm. (B) The number of Lin⁻CD45⁻Sca-1⁺ small (<10 μm) cells in (a) bone, (b) bone marrow (BM), and (c) peripheral blood (PB) from mice on days 1, 2, 3, and 7 after the CCl₄ administration, as determined by a FCM analysis. The cells isolated from healthy mice are shown as the “Control.” As a comparative experiment against CCl₄ administration, the mice received olive oil alone intraperitoneally (shown as Oil). The numbers of Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells were not significantly different between the healthy mice (Control) and each olive oil-administered mice (Oil). (C) The percentage of BrdU⁺ BD VSELs in total BD VSELs. The ingestion of BrdU in drinking water started 1 week before the administration of CCl₄ or olive oil, and BD VSELs from these mice were collected 2 days after the CCl₄ treatment and analyzed by FCM. The bars represent the mean ± SD of three independent experiments. The data were evaluated by two-tailed Student's t tests. Single and double asterisks indicate statistically significant values (*P < 0.05, **P < 0.01). n.s., not significant.

The percentages of BrdU⁺ cells in the total BD VSELs showed a tendency to increase after acute liver injury (Fig. 6C). These results suggest that BD VSELs actively proliferated after acute liver injury in vivo. There is a possibility that BD VSELs may respond to some soluble factors released from the damaged liver tissue.

Transplantation of GFP mouse-derived BD VSELs into recipient mice after induction of acute liver injury

Finally, we tested whether the GFP mouse-derived BD VSELs can home into liver tissue after acute injury. As shown in Supplementary Fig. S12, 1.2 × 10³ GFP mouse-derived BD VSELs were transplanted via the tail vein into each recipient mouse, which had been injected with CCl₄ 24 h before transplantation. The liver tissues were analyzed 3 weeks after transplantation to detect the presence of GFP expressing cells by immunohistochemistry. As shown in Fig. 7Aa, b and Ba, b, we detected some GFP⁺ multinucleated large cells, suggesting that there was cell fusion of BD VSELs with hepatocytes. In another section, several clusters of GFP⁺ small cells were also identified, suggesting that there was proliferation of injected BD VSELs in the damaged liver tissue (Fig. 7Ca, b). Interestingly, a cluster of GFP⁺ hepatocyte-like cells was identified in the hepatic parenchyma, suggesting that there was transdifferentiation of BD VSELs (Fig. 7Da, b). In the mice that received 2 × 10³ GFP mouse-derived whole BMNCs or BDNCs, no GFP⁺ donor-derived cells were detected in the liver sections (data not shown). Moreover, we did not detect any GFP⁺ cells in the liver sections obtained from mice that received olive oil alone before transplantation (data not shown).

FIG. 7.

Transplantation of green fluorescent protein (GFP) mouse-derived BD VSELs into recipient mice after induction of acute liver injury. GFP mouse-derived BD VSELs were transplanted via a tail vein into recipient mice, which were injected with CCl₄ 24 h before transplantation. The liver tissues were analyzed 3 weeks after transplantation to detect the GFP-expressing cells by immunohistochemistry. (A, B) GFP⁺ multinucleated large cells, (C) several clusters of GFP⁺ small cells and (D) a cluster of GFP⁺ hepatocyte-like cells were identified in the liver sections. These images were magnified with 20× (a) and 40× (b) objective lenses. The scale bars indicate 100 μm. The solid square lines in (a) represent areas showing enlarged images in (b). The black arrows indicate cells of interest.

These results demonstrated that the GFP⁺ BD VSELs were attracted specifically into damaged liver tissue, and also suggest that the injected GFP⁺ BD VSELs may simultaneously proliferate, fuse with hepatocytes, and show transdifferentiation into hepatocytes in the damaged liver tissue. These GFP⁺ cells could only be detected after acute liver injury, suggesting that severe tissue damage may induce the production of some soluble factors to result in the proliferation of BD VSELs in bone tissue and to induce the attraction of BD VSELs into the liver tissue. SDF-1 (CXCL12) was previously reported to play a critical role in recruiting HSCs into wounds after tissue injury [36]. Therefore, it is possible that the BD VSELs mobilized into the PB may also be able to migrate and be attracted into the damaged liver tissue in response to SDF-1.

Discussion

The identification of murine and human VSELs in various tissues, including BM, PB, and CB, suggested that they had promise for use in tissue/organ regeneration [3,10 –26]. However, some recent reports raised questions about their SC characteristics [27,28] and even their existence [29]. In our experience, the number of VSEL phenotypic cells present in human CB is extremely low (Matsuoka and Sonoda, unpublished data). Therefore, it was difficult to accurately isolate/purify these cells because of their rarity, which may be the main cause of the recent controversy surrounding the existence of VSELs.

In contrast, a number of recent studies have reported the existence of VSELs and VSEL-resembling cells in various tissues [19 –26]. Among these studies, Kraus and colleagues reported an elegant study concerning murine BM VSELs [19]. They directly compared the level of BM-derived lung epithelial cells after the transplantation of VSELs, hematopoietic stem/progenitor cells, or other nonhematopoietic non-VSEL-cells purified using a careful sorting strategy. The results were very clear that VSELs had the highest rate of forming type 2 pneumocytes in the lung. Moreover, they clearly demonstrated that this engraftment occurred by differentiation and not cell fusion. This study provides an evidence suggesting that mouse BM VSELs appear to be adult stem cells possessing broad differentiation potential across the various germ layers. This phenomenon can be explained by a hypothesis formulated by Ratajczak et al. [15] that tissues contain some population of PSCs that are deposited in embryogenesis during early gastrulation.

In this study, we developed a highly efficient method for isolating Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells identified in adult mouse bone tissue using enzymatic treatment of mouse bone. We termed these bone-derived Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells as BD VSELs. The BD VSELs resembled with BM VSELs in terms of various cellular and/or SC characteristics as summarized in Table 1. The size of the BD VSELs was approximately 6.5 μm (Supplementary Fig. S1), which was larger than the size of the original BM VSELs [17]. However, Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells were detected in various murine organs, including the brain, lung, pancreas, liver, heart, testes, and so on [12,17]. Interestingly, the sizes of VSELs in these organs were larger (>6 μm) than that of the BM VSELs (3–5 μm) [17].

As clearly shown in this study, the majority of the Lin⁻CD45⁻Sca-1⁺ cells reside in the mouse bone tissue rather than the BM. These results were consistent with a previous report that the majority of the Lin⁻CD45⁻Sca-1⁺ cells reside in the subendosteal region of the mouse bone [37]. The BD VSELs had similar cellular characteristics to the BM VSEL phenotypic cells. However, the immunophenotype and the gene expression profile of BD VSELs were different from those of BM VSEL phenotypic cells (Figs. 3, 4 and Supplementary Figs. S4 and S5). Importantly, our identified BD VSELs did not express PSC markers, such as Oct3/4 and Nanog, by the RQ-PCR and ISS analyses. It is interesting to note that the BD VSELs expressed various MSC and endothelial cell markers, which were not detected in the original BM VSELs, in addition to common HSC markers, including CD34, CD133, and CXCR4 (Fig. 3, Supplementary Fig. S5 and Supplementary Table S1). This unique immunophenotype may suggest their latent differentiation potential.

Our results demonstrated that the surface immunophenotype of our identified BM VSEL phenotypic cells is mostly different from those reported in the literature except for the expression of CD34 and c-Met (Supplementary Fig. S5) (Table 1) [3,10 –18]. However, the size of our identified BM VSEL phenotypic cells (6–9 μm) was much larger than original BM VSELs [3–5 μm] [17]. Therefore, there is a possibility that our identified BM VSEL phenotypic cells may be different from original BM VSELs [3,10 –18]. The BD VSELs weakly expressed CXCR4 (Fig. 3), however, BM VSEL phenotypic cells did not express CXCR4 (Supplementary Fig. S5) at the steady state. According to the previous reports [3, 10 –12,14,15,17], BM VSELs expressed CXCR4, suggesting that the SDF-1/CXCR4 axis plays a role in the homing of BM VSELs into the BM. As reported by Taichman et al. [37], CXCR4 mRNA levels were significantly enhanced in whole BM cells derived from 5-FU treated mice compared with vehicle-treated mice. In this study, we demonstrated that the BD VSELs were most likely mobilized into the PB through BM after severe acute liver injury induced by CCl₄ administration (Fig. 6). Collectively, there is a possibility that 5-FU treatment and severe tissue injury may induce an upregulation of CXCR4 expression on various kinds of TCSCs via some unidentified soluble factors released from the damaged tissue. Alternatively, these results suggest that the SDF-1/CXCR4 axis alone or in combination with other chemoattractants plays a crucial role in the homing of TCSCs.

Moreover, we could not detect PSC markers, including Oct3/4 and Nanog, by FCM and ISS analyses in the BM VSEL phenotypic cells (Supplementary Fig. S4). These results are consistent with previous reports [27,29]. Overall, we could detect a significant number of Lin⁻CD45⁻Sca-1⁺ VSEL-phenotypic cells in mouse BMs. However, their size, immunophenotype and the PSC marker expression were different from those of original reports [3,10 –18]. Therefore, it is important to more precisely elucidate the cellular and/or SC characteristics of the BD VSELs and BM VSELs, in addition to the potential relationship between these two VSEL-phenotypic cells, which appear to be PSCs in postnatal adult tissues.

As described in the Results section, it was difficult to initiate the proliferation of BD VSELs in vitro except in the HSC culture medium. This finding suggests that the BD VSELs remain locked in a dormant state, as was previously reported for BM VSELs [13 –15]. However, as mentioned above, the murine BM VSELs can transdifferentiate into lung epithelial cells in vivo [19]. In addition, Taichman et al. reported that murine subendosteal region-derived Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells had the ability to generate osseous tissues at a low cell density in vivo [37]. Collectively, these findings suggested that murine BM VSELs can proliferate and transdifferentiate into multiple lineages in vivo.

As shown in Fig. 6, the number of BD VSELs in this study significantly increased after acute liver injury, and the cells were then most likely mobilized into the PB. These results are consistent with the previously reported data that BM VSELs were mobilized into the PB after liver injury induced by the administration of CCl₄ [35]. In addition, our data demonstrated that GFP mouse-derived BD VSELs transplanted via the tail vein into mice with acute liver injury were detected in the liver parenchyma. Moreover, our findings suggest that they may simultaneously proliferate, fuse with hepatocytes, and show transdifferentiation into hepatocytes in the damaged liver tissue. Collectively, our findings suggest that severe tissue damage may induce the proliferation of dormant BD VSELs. It is therefore important to clarify the factors/signals required to induce their proliferation to apply BD VSELs for future regenerative medicine.

Millions of patients worldwide suffer from end-stage liver diseases, whose only curative therapy is liver transplantation (LT) [38]. However, the LT is still limited by the donor scarcity, high costs, and the lifelong immunosuppressive therapy needed after LT [38]. Therefore, the development of cell-based therapies for the treatment of end-stage liver diseases is currently under investigation [38,39]. Among the many cell sources, stem cell-based therapy is the most promising [40,41]. In fact, several reports have demonstrated that human BM cells and CB cells could be used for the regeneration of liver tissue [41 –44]. The precise mechanism of liver regeneration has not been fully elucidated. However, HSCs, MSCs, and other TCSCs, including VSEL-like cells, which are contained in BM and CB cells, may contribute to liver regeneration. Our identified BD VSELs may have the potential to repair damaged liver tissue. Further and more in-depth investigations will be required to fully elucidate their functional roles and their potential use for regenerative medicine.

In conclusion, the PSCs in postnatal adult life, including VSELs, can be applied for cell therapies and regenerative medicine. However, the numbers of VSELs in the BM and CB were found to be very limited, which is an obstacle to their use in regenerative medicine. In this study, we identified BD VSELs in mouse bone tissue that resembled BM VSELs. Moreover, the majority of the Lin⁻CD45⁻Sca-1⁺ VSEL phenotypic cells were found to reside in the mouse bone tissue, rather than the BM. As reported in the previous reports [13], BM VSELs are not fully functional, but remain locked in a dormant state. They require activation signals provided by unidentified factors. Our model of acute liver injury suggested that BD VSELs may participate in tissue repair in the damaged liver, and some as-yet unidentified factors may play a role in this process. Therefore, it is important to develop an ex vivo expansion system for BD VSELs to further explore their potential applications for regenerative medicine in the near future.

Footnotes

Acknowledgments

This work was supported by Grants-in-Aid for Scientific Research C (grant nos. 21591251 and 24591432) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan; a grant from the Science Frontier Program of the MEXT; a grant of A-STEP (Adaptable & Seamless Technology Transfer Program through Target-driven R&D) Feasibility Study (FS) Stage (Exploratory Research type) from the Japan Science and Technology Agency (JST) (grant no. AS231Z03109G); a grant from the Strategic Research Base Development Program for Private Universities from the MEXT; the MEXT-Supported Program for the Strategic Research Foundation at Private Universities; a grant from the Promotion and Mutual Aid Corporation for Private Schools of Japan; a grant from the Japan Leukemia Research Foundation; a grant from the Mitsubishi Pharma Research Foundation; a grant from the Takeda Science Foundation; a grant from the Terumo Life Science Foundation and a grant from SENSHIN Medical Research Foundation. All these grants were given to Yo.So. This work was also supported by Grants-in-Aid for Young Scientists (B) (grant no. 23792304) and Scientific Research C (grant no. 26462989) from the Japan Society for the Promotion of Science; a grant from Kansai Medical University (Research grant D1) given to R.N.

Author Disclosure Statement

No competing financial interests exist.

References

Bjornson

CRR

, Rietze

, Reynolds

, Magli

and Vescovi

. (1999). Turning brain into blood: a hematopoietic fate adopted by adult neural stem cells in vivo. Science, 283:534–537.

Wagers

, Sherwood

, Christensen

and Weissman

. (2002). Little evidence for developmental plasticity of adult hematopoietic stem cells. Science, 297:2256–2259.

Kucia

, Reca

, Jala

, Dawn

, Ratajczak

and Ratajczak

. (2005). Bone marrow as a home of heterogeneous populations of nonhematopoietic stem cells. Leukemia, 19:1118–1127.

Pittenger

, Mackay

, Beck

, Jaiswal

, Douglas

, Mosca

, Moorman

, Simonetti

, Craig

and Marshak

. (1999). Multilineage potential of adult human mesenchymal stem cells. Science, 284:143–147.

Kuroda

, Kitada

, Wakao

, Nishikawa

, Tanimura

, Makinoshima

, Goda

, Akashi

, Inutsuka

, et al. (2010). Unique multipotent cells in adult human mesenchymal cell populations. Proc Natl Acad Sci USA, 107:8639–8643.

Yoon

, Wecker

, Heyd

, Park

J-S

, Tkebuchava

, Kusano

, Hanley

, Scadova

, Qin

, et al. (2005). Clonally expanded novel multipotent stem cells from human bone marrow regenerate myocaridium after myocardial infarction. J Clin Invest, 115:326–338.

D'lppolito

, Diabira

, Howard

, Menei

, Roos

and Schiller

. (2002). Marrow-isolated adult multilineage inducible (MIAMI) cells, a unique population of post natal young and old human cells with extensive expansion and differentiation potential. Proc Natl Acad Sci USA, 99:413–418.

Jlang

, Jahaglrdar

, Reinhardt

, Schwartz

, Keene

, Oritz-Gonzalez

, Reyes

, Lenvik

, Lund

, et al. (2002). Pluripotency of mesenchymal stem cells derived from adult marrow. Nature, 418:41–49.

Nagasawa

. (2000). A chemokine, SDF-1/PBSF, and its receptor, CXC chemokine receptor 4, as mediators of hematopoiesis. Int J Hematol, 72:408–411.

10.

Ratajczak

, Kucia

, Reca

, Majka

, Janowska-Wieczorek

and Ratajczak

. (2004). Stem cell plasticity revisited: CXCR4-positive cells expressing mRNA for eraly muscle, liver and neural cells ‘hide out’ in the bone marrow. Leukemeia, 18:29–40.

11.

Kucia

, Reca

, Cambell

, Zuba-Surma

, Majka

, Ratajczak

and Ratajczak

. (2006). A population of very small embryonic-like (VSEL) CXCR4⁺SSEA-1⁺Oct4⁺ stem cells identified in adult bone marrow. Leukemia, 20:857–869.

12.

Zuba-Surma

, Kucia

, Ratajczak

and Ratajczak

. (2009). “Small stem cells” in adult tissues: Very small embryonic-like stem cells stand up!. Cytometry Part A, 75A:4–13.

13.

Shin

, Zuba-Surma

, Wu

, Ratajczak

, Wysoczynski

, Ratajczak

and Kucia

. (2009). Novel epigenetic mechanisms that control pluripotency and quiescence of adult bone marrow-derived Oct4⁺ very small embryonic-like stem cells. Leukemia, 23:2042–2051.

14.

Ratajczak

, Zuba-Surma

, Wysoczynski

, Ratajczak

and Kucia

. (2008). Very small embryonic-like stem cells: Characterization, developmental origin, and biological significance. Exp Hematol, 36:742–751.

15.

Ratajczak

, Machalinski

, Wojakowski

, Ratajczak

and Kucia

. (2007). A hypothesis for an embryonic origin of pluripotent Oct4⁺ stem cells in adult bone marrow and other tissues. Leukemia, 21:860–867.

16.

Suszynska

, Zuba-Surma

, Maj

, Mierzejewska

, Ratajczak

, Kucia

and Ratajczak

. (2014). The proper criteria for identification and sorting of very small embryonic-like stem cells, and some nomenclature issues. Stem Cells Dev, 23:702–713.

17.

Zuba-Surma

, Kucia

, Wu

, Klich

, Lillard

Jr , Ratajczak

and Ratajczak

. (2008). Very small embryonic-like stem cells are present in adult murine organs: ImageStream-based morphological analysis and distribution studies. Cytometry Part A, 73A:1116–1127.

18.

Kucia

, Halasa

, Wysocznski

, Baskiewicz-Masiuk

, Moldenhawer

, Zuba-Surma

, Czajka

, Wajakowski

, Machalinski

and Ratajczak

. (2007). Morphological and molecular characterization of novel population of CXCR4⁺SSEA-4⁺Oct4⁺ very small embryonic-like cells purified from human cord blood - preliminary report. Leukemia, 21:297–303.

19.

Kassmer

, Jin

, Zhang

P-X

, Bruscia

, Heydari

, Lee

J-H

, Kim

and Krause

. (2013). Very small embryonic-like stem cells from the murine bone marow differentiate into epithelial cells of the lung. Stem Cells, 12:2759–2766.

20.

Sauerzweig

, Munsch

, Leβmann

, Reymann

and Braun

. (2009). A population of serumdeprivation-induced bone marrow stem cells (SD-BMSC) express marker typical for embryonic and neural stem cells. Exp Cell Res, 315:50–66.

21.

Shirazi

, Zarnani

, Soleimani

, Abdolvahabi

, Nayernia

and Kashani

. (2012). BMP4 can generate primordial germ cells from bone-marrow-derived pluripotant stem cells. Cell Biol Int, 36:1185–1193.

22.

Sovalat

, Scrofani

, Eidenschenk

, Pasquet

, Rimelen

and Henon

. (2011). Identifiction and isolation from either adult human bone marrow or G-CSF-mobilized peripheral blood of CD34+/CD133+/CXCR4+/Lin-CD45- cells, featuring morphological, molecular, and phenotypic characteristics of very small embryonic-like (VSEL) stem cells. Exp Hematol, 39:495–505.

23.

Bhartiya

, Kasiviswananthan

and Shaikh

. (2012). Cellular origin of testis-derived pluripotent stem cells: A case for very small embryonic-like stem cells. Stem Cells Dev, 21:670–674.

24.

Parte

, Bhartiya

, Telang

, Daithankar

, Salvi

, Zaveri

and Hinduja

. (2011). Detection, characterization, and spontaneous differentiation in vitro of very small embryonic-like putative stem cells in adult mammalian ovary. Stem Cells Dev, 20:1451–1464.

25.

Bhartiya

, Shaikh

, Nagvenkar

, Kasiviswanathan

, Pethe

, Pawani

, Mohanty

, Rao

SGA

, Zaveri

and Hinduja

. (2012). Very small embryonic-like stem cells with maximum regenerative potential get discarded during cord blood banking and bone marrow processing for autologous stem cell therapy. Stem Cells Dev, 21:1–6.

26.

Havens

, Sun

, Shiozawa

, Jung

, Wang

, Mishra

, Jiang

, O'Neill

, Krebsbach

, Rodgerson

and Taichman

. (2014). Human and murine very small embryonic-like cells represent multipotent tissue progenitors, in vitro and in vivo. Stem Cells Dev, 23:689–701.

27.

Szade

, Bukowska-Strakova

, Nowak

, Szade

, Kachamakova-Trojanowska

, Zukowska

, Jozkowicz

and Dulak

. (2013). Murine bone marrow Lin⁻Sca-1⁺CD45⁻ very small embryonic-like (VSEL) cells are heterogeneous population lacking Oct-4A expression. PloS ONE, 8:e63329.

28.

Danova-Alt

, Heider

, Egger

, Cross

and Alt

. (2012). Very small embryonic-like stem cells purified from umbilical cord blood lack stem cell characteristics. PloS ONE, 7:e34899.

29.

Miyanishi

, Mori

, Seita

, Chen

, Karten

, Chan

CKF

, Nakauchi

and Weissman

. (2013). Do pluripotent stem cells exist in adult mice as very small embryonic stem cells?. Stem Cell Rep, 1:1–11.

30.

Sonoda

, Sakabe

, Ohmisono

, Tanimukai

, Yokota

, Nakagawa

, Clark

and Abe

. (1994). Synergistic actions of stem cell factor and other burst-promoting activities on proliferation of CD34⁺ highly purified blood progenitors expressing HLA-DR or different levels of c-kit protein. Blood, 84:4099–4106.

31.

Ratajczak

, Wysoczynski

, Zuba-Surma

, Wan

, Kucia

, Yoder

and Ratajczak

. (2011). Adult murine bone marrow-derived very small embryonic-like stem cells differentiate into the hematopoietic lineage after coculture over OP9 stromal cells. Exp Hematol, 39:225–237.

32.

Kucia

, Wysoczynski

, Ratajczal

and Ratajczak

. (2008). Identification of very small embryonic like (VSEL) stem cells in bone marrow. Cell Tissue Res, 331:125–134.

33.

Morikawa

, Mabuchi

, Kubota

, Nagai

, Niibe

, Hiratsu

, Suzuki

, Miyauchi-Hara

, Nagoshi

, et al. (2009). Prospective identification, isolation, and systemic transplantation of multipotent mesenchymal stem cells in murine bone marrow. J Exp Med, 206:2483–2496.

34.

Paczkowska

, Kucia

, Koziarska

, Halasa

, Safranow

, Masiuk

, Karbicka

, Nowik

, Nowacki

, Ratajczak

and Machalinski

. (2009). Clinical evidence that very small embryonic-loke stem cells are mobilized into peripheral blood in patients after stroke. Stroke, 40:1237–1244.

35.

Kucia

, Wysoczynski

, Wan

, Zuba-Surma

, Ratajczak

and Ratajczak

. (2008). Evidence that very small embryonic-like stem cells are mobilized into peripheral blood. Stem Cells, 26:2083–2092.

36.

Ponomaryov

, Peled

, Petit

, Taichman

, Habler

, Sandbank

, Arenzana-Seisdedos

, Magerus

, Caruz

, et al. (2000). Induction of the chemokine stromal-derived factor-1 following DNA damage improves human stem cell function. J Clin Invest, 106:1331–1339.

37.

Taichman

, Wang

, Shiozawa

, Jung

, Song

, Balduino

, Wang

, Patel

, Havens

, et al. (2010). Prospective identification and skeltal localization of cells capable of multilineage differentiation in vivo. Stem Cells Dev, 19:1557–1570.

38.

Locke

, Shamblott

and Cameron

. (2009). Stem cells and the liver: Clinical applications in transplantation. Adv Surg, 43:35–51.

39.

Fiegel

, Lange

, Kneser

, Lambrecht

, Zander

, Rogiers

and Kluth

. (2006). Fetal and adult liver stem cells for liver regeneration and tissue engineering. J Cell Mol Med, 10:577–587.

40.

Kung

JWC

and Forbes

. (2009). Stem cells and liver repair. Curr Opin Biotech, 20:568–574.

41.

Sancho-Bru

, Najimi

, Caruso

, Pauwelyn

, Cantz

, Forbes

, Roskams

, Ott

, Gehling

, et al. (2009). Stem and progenitor cells for liver repopulation: can we standardise the process from bench to bedside?. Gut, 58:594–603.

42.

Terai

, Ishikawa

, Omori

, Aoyama

, Marumoto

, Urata

, Yokoyama

, Uchida

, Yamasaki

, et al. (2006). Improved liver function in patients with liver cirrhosis after autologous bone marrow cell infusion therapy. Stem Cells, 24:2292–2298.

43.

Ishikawa

, Drake

, Yang

, Fleming

, Minamiguchi

, Visconti

, Crosby

, Agraves

, Harada

, et al. (2003). Transplanted human cord blood cells give rise to hepatocytes in engrafted mice. Ann NY Acd Sci, 996:174–185.

44.

Chen

, Lv

, Dai

, Liu

, Lou

, Chen

and Zou

. (2014). Hepatic regenerative potential of mouse bone marrow very small embryonic-like stem cells. J Cell Physiol. DOI: 10.1002/jcp.24913.

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

0.16 MB

0.12 MB

0.05 MB

0.08 MB

0.02 MB

0.14 MB

0.11 MB

0.18 MB

0.12 MB

0.14 MB

0.09 MB

0.17 MB

0.20 MB