Published online before print July 22, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Institute for Advanced Medical Research;
Department of Internal Medicine; and
Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan
1 Correspondence: Institute for Advanced Medical Research, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. E-mail: kuwanam{at}sc.itc.keio.ac.jp
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: fibronectin • lineage • mesenchymal stem cell • plasticity
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Many adult tissues contain populations of stem cells that can self-replicate and give rise to daughter cells that undergo an irreversible terminal differentiation [8 ]. The best-characterized are hematopoietic stem cells and their progeny, but a variety of postnatal stem cells have been identified and characterized [9 10 11 ]. Mesenchymal stem cells (MSCs) are identified as adherent fibroblast-like cells in the bone marrow that can differentiate into mesenchymal tissues, including bone, cartilage, fat, muscle, and bone marrow stroma [9 ]. Recently, mesenchymal progenitors having morphologic and phenotypic features and differentiation potentials similar to MSCs have been reported at extremely low frequencies in umbilical cord blood [12 ] as well as in fetal [13 ] and adult peripheral blood [14 ]. MSCs and circulating MSC-like cells do not express hematopoietic markers or the stem cell/endothelial marker CD34 [9 , 12 13 14 ].
In this study, we have identified a human cell population with fibroblast-like morphology that has the unique phenotype CD14+CD45+CD34+type I collagen+. This cell type is generated in vitro from peripheral blood mononuclear cells (PBMCs) cultured on fibronectin (FN)-coated plastic plates. These cells are derived from circulating CD14+ monocytes but can differentiate into several distinct mesenchymal cell types, similar to MSCs. We named this novel cell type the monocyte-derived mesenchymal progenitor (MOMP). Our findings represent the first evidence that circulating CD14+ monocytes include progenitors that can differentiate into a variety of mesenchymal cell types in addition to those for phagocytes.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To examine the origin of MOMPs, PBMCs depleted of CD14+, CD34+, or CD105/SH2+ cells were cultured on FN-coated plates for 7 days. The depletion of CD14+ or CD34+ cells was performed by using an anti-CD14 or anti-CD34 monoclonal antibody (mAb) coupled to magnetic beads (DynaBeads; Dynal, Oslo, Norway) followed by magnetic separation according to the manufacturer’s protocol. CD105+ cell-depleted PBMCs were prepared by incubating PBMCs with anti-CD105 mAb (Immunotech, Marseille, France) and subsequently with goat anti-mouse IgG antibody coupled to magnetic beads (Dynal). Mock-treated PBMCs incubated with isotype-matched mouse mAb and bead-conjugated anti-mouse IgG antibody were also prepared as a control. The proportion of CD14+ cells in the CD14+ cell-depleted PBMC fraction was consistently <0.5%, and the proportions of CD34+ cells in the CD34+ cell-depleted fraction and of CD105+ cells in the CD105+ cell-depleted fraction were <0.01% by flow cytometry. The number of attaching cells per cm3 was counted, and the results were expressed as the ratio to the untreated PBMC culture.
In some experiments, circulating CD14+ monocytes and CD14- cells were separated from PBMCs by using anti-CD14 mAb-coupled magnetic beads (CD14 MicroBeads; Miltenyi Biotech, Bergisch Gladbach, Germany) followed by MACS column separation according to the manufacturer’s protocol. Flow cytometric analysis revealed that monocyte and CD14- cell fractions contained >98% and <0.5% CD14+ cells, respectively. Monocytes were labeled with PKH67 green (Sigma) and cultured with unlabeled CD14- cells (ratio of 1:4) on FN-coated or uncoated plastic plates for 7 days. PKH67-labeled monocytes were also cultured alone on FN-coated plates in regular medium or in CD14- cell-conditioned medium, which was prepared by culturing CD14- cells on FN-coated plates for 3 days.
Preparation of macrophages and dendritic cells
Macrophages were prepared by culturing adherent PBMCs on plastic plates in Medium 199 (Sigma) supplemented with 20% FBS and 4 ng/ml macrophage-colony stimulating factor (M-CSF; R&D Systems, Minneapolis, MN) for 7 days [5
]. Mature monocyte-derived dendritic cells were obtained from plastic adherent PBMCs by maturation using a series of culture conditions [15
]. Briefly, adherent cells were cultured in RPMI1640 with 10% FBS containing 50 ng/ml granulocyte/macrophage-colony stimulating factor (GM-CSF) and 50 ng/ml interleukin (IL)-4 (all from PeproTech, Rocky Hill, NJ) for 7 days. The immature dendritic cells were subsequently incubated with 50 ng/ml tumor necrosis factor-
(PeproTech) for 3 days. Flow cytometric analysis revealed that the macrophage fraction contained >98% CD14+CD80+ cells, and the dendritic cell fraction contained >95% CD83+HLA-DR+ cells and <1% CD14+ cells.
Cell lines
Primary cultures of human dermal fibroblasts and myoblasts were established from biopsies of healthy donors and maintained in low-glucose DMEM with 10% FBS. The human osteosarcoma cell line MG-63, the rhabdomyosarcoma cell line RD, and the chondrosarcoma cell line OUMS-27 were obtained from Health Science Research Resources Bank of Japan (Osaka, Japan), and maintained in low-glucose DMEM or minimal essential medium containing 10% FBS.
In vitro differentiation of MOMPs into mesenchymal cells
MOMPs, which were either freshly generated from PBMCs, cultured for several passages, or cryo-preserved, were replated on new FN-coated plastic plates or chamber slides in high-glucose DMEM with 10% FBS (Hyclone Laboratories, Logan, UT) and grown to semi-confluence. MOMPs generated from purified CD14+ monocytes cultured in CD14- cell-conditioned medium were also applied to this assay. The cells were then cultured under conditions known to induce MSCs to differentiate into various mesenchymal cell types [9
, 16
17
18
19
20
]. Monocytes, macrophages, and dermal fibroblasts were cultured under identical conditions as controls.
Osteogenesis
The adherent cells were cultured in Osteogenesis induction medium (Clonetics; San Diego, CA) containing 100 nM dexamethasone, 10 mM ß-glycerophosphate, and 50 µM ascorbic acid. The medium was changed twice a week for 3 weeks.
Myogenesis
The adherent cells were treated with 10 µM 5-azacytidine (Sigma) for 24 h. The cells were washed with Hanks’ balanced salt solution and cultured in DMEM containing 10% FBS, 5% horse serum (Life Technologies, Grand Island, NY), 50 mM hydrocortisone (Sigma), and 4 ng/ml basic fibroblast growth factor (Sigma). The medium was changed twice a week for 3 weeks.
Chondrogenesis
The adherent monolayer cells were cultured for 3 weeks in serum-free medium in the presence of TGF-ß1 (R&D Systems), which was added to the culture medium every other day at a final concentration of 10 ng/ml. The cells were also cultured in micromass cultures, either in a droplet micromass on the plate or a pelleted micromass in a conical tube [9
, 18
, 19
], in the presence of TGF-ß1.
Adipogenesis
The cells were incubated with adipogenic induction medium (DMEM containing 10% FBS, 1 µM dexamethasone, 0.5 mM methyl-isobutylxantine, 10 µg/ml insulin, and 100 mM indomethacin; all from Sigma). After 72 h, the medium was changed to maintenance medium (DMEM containing 10% FBS and 10 µg/ml insulin) for 24 h. The cells were treated three times with adipogenic induction medium and maintained in the maintenance medium for one additional week.
Flow cytometric analysis
For fluorescent cell staining, the adherent cells were detached from the plastic plates by incubation with 2 mM ethylene diamine tetraacetate on ice and blocked with normal mouse serum for 10 min. The cells were stained with the following mouse mAbs, which were either unconjugated or conjugated to FITC, phycoerythrin (PE), or PC5: anti-HLA-DR, anti-CD11c (BD PharMingen, San Diego, CA), anti-CD11b/Mac-1, anti-CD14, anti-CD29, anti-CD34, anti-CD44, anti-CD83, anti-CD105/SH2, anti-CD117/c-kit (Immunotech), anti-CD34, anti-CD133 (Miltenyi Biotech), anti-HLA class I, anti-HLA-DR, anti-CD31/PECAM-1, anti-Flt-1/VEGFR1, anti-Flk-1/VEGFR2 (Sigma), anti-CD40, anti-CD54, anti-CD80, anti-CD86 (Ancell, Bayport, MN), anti-CD144/VE-cadherin, or anti-type I collagen (Chemicon International, Temecula, CA). When unconjugated mAbs were used, goat anti-mouse IgG F(ab’)2 conjugated to FITC or PE (Immunotech) was used as a secondary antibody. For intracellular staining, the cells were permeabilized and fixed by using IntraPrepTM permeabilization reagent (Immunotech). Cells were analyzed on a FACS® Calibur flow cytometer (Becton Dickinson) by using the CellQuest software. Viable cells were identified by gating on forward and side scatters, and the data are shown as logarithmic histograms or dot-plots.
Immunohistochemistry
Slides were coated with monocytes, macrophages, or dendritic cells by using a cytospin technique, and the remaining cell types were cultured on FN-coated chamber slides, except for samples to be used for FN staining, for which type I collagen-coated slides were used instead. The cells were fixed with 10% formalin, and the endogeneous peroxidase activity was quenched with 0.3% peroxide for 5 min. Slides were incubated for 30 min with one of the following mouse mAbs: anti-CD45, anti-vimentin, anti-skeletal muscle-specific actin (SkM-actin) (Dako, Carpinteria, CA), anti-CD34 (Calbiochem-Novabiochem, San Diego, CA), anti-type I collagen (Chemicon), anti-type III collagen, anti-fibronectin (Sigma), anti-type II collagen (ICN Biomedicals, Aurora, OH), or anti-skeletal muscle-specific myosin heavy chain (SkM-MHC) (Zymed Laboratories, San Francisco, CA). The slides were then incubated with biotin-labeled anti-mouse IgG antibody (Nichirei, Tokyo, Japan). The antibody-biotin conjugates were detected with a streptavidin-horseradish peroxidase complex (Nichirei) applied for 10 min at room temperature by using 3,3'-diaminobenzidine as the substrate. Nuclei were counterstained with hematoxylin. The negative controls were cells incubated with normal mouse IgG instead of the primary antibody. To enumerate the proportion of cells staining positive for a given marker, at least 300 cells per culture were evaluated.
For fluorescence double-staining, fixed cells were incubated with goat anti-Cbfa1/Runx2 or anti-Sox-9 polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) followed by incubation with AlexaFluor® 568 goat-specific IgG (Molecular Probes) and then with FITC-conjugated mouse anti-CD45 mAb (Dako). Similarly, the cells were stained with mouse anti-MyoD (Dako) or anti-peroxisome proliferation-activated receptor 
(PPAR) mAb (Santa Cruz) followed by incubation with tetramethylrhodamine isothiocyanate isomer R-labeled mouse-specific IgG (Dako) and subsequently with FITC-conjugated anti-CD45 mAb. The cells were examined with a confocal laser fluorescence microscope (LSM5 PASCAL; Carl-Zeiss, Göttingen, Germany).
Uptake of acetylated low-density lipoprotein (Ac-LDL)
The adherent cells were cultured with 2.5 µg/ml Dil-Ac-LDL (Molecular Probes) for 1 h, and Ac-LDL uptake was evaluated by flow cytometry.
Alkaline phosphatase staining
The cells were fixed with 10% formalin and subsequently incubated in a solution containing 0.2 mg/ml naphthol AS-TR phosphate and 0.5 mg/ml Fast Red RC (all from Sigma) for 10 min. The cells were counterstained with hematoxylin.
Intracellular calcium detection
To detect intracellular calcium deposits, the cells were fixed with 10% formalin and stained with 2% alizarin red S (Sigma) for 3 min, followed by extensive wash with distilled water and subsequent staining with hematoxylin. The intracellular calcium concentration was measured by using a commercially available kit (Sigma) [20
]. The protein content in cell extract was also determined by using the Bradford protein assay kit (Bio-Rad Laboratories, Hercules, CA) by using bovine serum albumin as a standard, and the calcium concentration was expressed as microgram per microgram of protein content.
Oil-red O staining
The cells were fixed with 0.2% glutaraldehyde for 5 min, rinsed in 60% isopropanol, and covered with 0.1% oil-red O for 10 min. After rinsed in 60% isopropanol and subsequently in distilled water, the cells were counterstained with hematoxylin.
Transmission electron microscopy
Cultured MOMPs were immediately fixed with 2.5% glutaraldehyde, post-fixed in 2% osmium tetroxide, dehydrated in a series of graded ethanol solutions and propylene oxide, and embedded in Epoxy resin. The cells were thin-sectioned on an LKB ultratome with a diamond knife. Sections in the range of gray to silver were collected on 150-mesh grids, stained with uranyl acetate and lead citrate, and examined under a JEOL-1200 EXII electron microscope (Jeol, Tokyo, Japan).
Cell proliferation studies
Proliferating MOMPs were detected by BrdU-labeling as described previously [21
]. Briefly, MOMPs were cultured in the presence of 10 µM BrdU (Sigma) for 2 h before staining. After a 20-min fixation in Carnoy’s fixative (methanol/acetic acid) at -20°C, the cells were air-dried, treated with 2N HCl for 1 h to denature DNA, and then neutralized by 0.1 M borate, pH 8.5. The cells were then incubated with mouse anti-BrdU mAb (Chemicon International) in the presence of 0.05% Tween 20, followed by biotin-streptavidin-peroxidase complex staining. Nuclei were counterstained with hematoxylin. Negative controls were the cells incubated with isotype-matched mouse control mAb instead of the primary antibody. Apoptotic cells were detected by incubating unfixed cells with propidium iodide (PI; Sigma) for 30 min, 4°C and observed under a fluorescent microscope.
For cell-division studies, the cells were labeled with 5-carboxyfluorescein diacetate succinimidyl ester (CFSE) as described previously [22 ]. CFSE-labeled monocytes were cultured with unlabeled CD14- cells on FN-coated plates for 1, 3, and 5 days, and the adherent cells were harvested and stained with PC5-labeled anti-CD14 mAb. CFSE-labeled MOMPs were also cultured for 1, 3, and 5 days. The intensity of CFSE labeling was evaluated by flow cytometry.
Analysis of mRNA expression
The expression of lineage-specific mRNA was examined by using reverse transcription (RT)–polymerase chain reaction (PCR). Total RNA was extracted from MOMPs that had or had not been induced to differentiate by using the RNeasy kit (Qiagen, Valencia, CA). Total RNA was also extracted from monocytes, macrophages, dendritic cells, dermal fibroblasts, myoblasts, osteosarcoma, rhabdomyosarcoma, and chondrosarcoma. Human muscle and fat tissue total RNAs were purchased from Clontech Laboratories (Palo Alto, CA). First-strand cDNA was synthesized from the total RNA by using Molony murine leukemia virus reverse transcriptase (Takara, Kyoto, Japan) with oligo-dT priming. The cDNA (50 ng total RNA equivalent) was then subjected to PCR amplification by using the panel of specific primers listed in Table 1
. The PCR products were resolved by electrophoresis on 2% agarose gels and visualized by staining with ethidium bromide.
|
View this table: [in a new window] |
Table 1. Primers used for RT-PCR analysis and the expected sizes of the products
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (128K): [in a new window] |
Figure 1. Morphology of MOMPs. MOMPs were generated by culturing PBMCs on FN in low-glucose DMEM with 10% FBS for 7 days. MOMPs were re-plated on a new FN-coated plate and cultured for 24 h. Phase-contrast images at Day 6 (A), Day 7 (B), and after the first passage (C). Electron microscopic pictures after the first passage (D–G). A bundle of intermediate filaments (arrow) and a structure similar to a rod-shaped microtubulated body (arrowhead) are shown. L, labyrinth-like endocytic vesicle; LD, lipid droplet; N, nucleus; and PS, pseudopodia. Original magnification: a, ’80; b, c, ’40; d, ’5000; e, ’10,000; and f, g, ’30000. Results shown are representative of 50 cells prepared in three independent experiments.
|
 View larger version (52K): [in a new window] |
Figure 2. Flow cytometric analysis of MOMPs. PBMCs were cultured on FN-coated plastic plates, and the adherent cells were harvested on Day 7. The cells were stained with a series of mAbs as indicated and analyzed by flow cytometry. The expression of the molecules of interest is shown as shaded histograms. Open histograms represent staining with isotype-matched control mAbs. The results shown are representative of at least three experiments.
|
Origin of MOMPs in circulation
The adherent cells obtained in the PBMC culture on FN-coated plates were serially examined for their surface expression of CD14 and CD34 by flow cytometry (Fig. 3A
). The majority of cells attached to the plates at 1 h were CD14+ and CD34-, but CD34 expression appeared to be up-regulated on the adherent CD14+ cells. Nearly all the adherent cells were positive for both CD14 and CD34 after 7 days in culture. Because peripheral blood was reported to contain CD34+ endothelial progenitors [23
] and CD105/SH2+ mesenchymal progenitors [14
], we examined the effect of depleting PBMCs of cells positive for CD14, CD34, or CD105 on the generation of MOMPs. As shown in Fig. 3B
, the appearance of MOMPs was almost completely inhibited by the depletion of CD14+ monocytes, whereas depletion of CD34+ or CD105+ cells showed no effect. To further confirm that MOMPs originated from circulating monocytes, CD14+ monocytes were isolated from PBMCs, labeled with PKH67, and cultured with unlabeled CD14- cells on FN-coated plates. As shown in Fig. 3C
, PKH67-labeled monocytes expressed CD34 after 7 days of the culture. Together, these findings indicate that the MOMP population originated from circulating CD14+ monocytes. However, non-adherent cells collected on Day 3 contained a significant proportion of CD14+ cells, which suggests that only a subset of monocytes can attach to FN and differentiate into MOMPs.
 View larger version (39K): [in a new window] |
Figure 3. MOMPs originated from circulating CD14+ monocytes. (A) PBMCs were cultured on FN for 1 h, and 1, 3, 5, 7, 10, and 14 days. The adherent cells were harvested and stained with FITC-conjugated anti-CD14 and PC5-conjugated anti-CD34 mAbs, and analyzed by flow cytometry. The results shown are representative of four independent experiments. (B) PBMCs depleted of CD14+ cells, CD34+ cells, or CD105+ cells, or mock-treated PBMCs were cultured on FN for 7 days. The number of attaching cells per cm3 was counted, and the results are expressed as the ratio of attached cells to the number of attached cells in the untreated PBMC culture. The results shown are the mean and standard deviation of PBMCs from three independent donors. The asterisk indicates a significant difference compared with mock-treated PBMC cultures, which was determined by using the Student’s t-test. (C) MACS-sorted CD14+ monocytes were stained with PKH67 and cultured with or without unlabeled CD14- cells on FN-coated or uncoated plastic plates for 7 days. The adherent cells were harvested, stained with PC5-conjugated anti-CD34 mAb, and analyzed by flow cytometry. The results shown are representative of three experiments. (D) MACS-sorted CD14+ monocytes were labeled with CFSE and cultured with unlabeled CD14- cells on FN for 1, 3, and 5 days. The adherent cells were harvested and stained with PC5-conjugated anti-CD14 mAb. The cells were subsequently analyzed by flow cytometry. The results shown are representative of three experiments.
|
To evaluate whether monocytes proliferate during MOMP differentiation, CD14+ monocytes were labeled with CFSE and cultured with unlabeled CD14- cells on FN. Adherent cells were serially harvested and examined for CFSE intensity and CD14 expression by flow cytometry (Fig. 3D) . Nearly all FN-attached monocytes proliferated during the first 24 h of culture, and proliferated slowly later in culture. Adherent cells were almost exclusively CFSE-labeled CD14+ monocytes, and the expansion of adherent cells from the CD14- cell fraction was not obvious, indicating that generation of MOMPs did not result from the proliferation of specific precursors in the CD14- cell population.
Phenotype of MOMPs
The protein expression profile of MOMPs was examined by flow cytometry and immunohistochemistry (Fig. 2
and Fig. 4
), and compared with that of monocytes, macrophages, and dendritic cells (Table 2
). MOMPs expressed hematopoietic and monocyte lineage markers, but lacked expression of dendritic cell markers. The expression of HLA-DR and some costimulatory molecules on MOMPs suggests they may be able to induce antigen-specific T cell activation as antigen-presenting cells. MOMPs expressed CD34 and CD105/SH2 [25
], but lacked expression of CD117/c-kit and CD133. MOMPs were positive for the endothelial markers CD144/VE-cadherin and Flt-1/VEGFR1. MOMPs were also positive for type I and III collagen, fibronectin, and vimentin, which are typically produced by cells of mesenchymal origin. These protein expression profiles did not change for up to 5 passages. These findings clearly show that MOMPs are phenotypically distinct from monocytes and monocyte-derived phagocytes. In particular, the simultaneous expression of stem cell markers (CD34 and CD105/SH2), endothelial markers (CD144/VE-cadherin and Flt-1/VEGFR1), and mesenchymal markers (type I and III collagen, and fibronectin) is unique to MOMPs.
 View larger version (101K): [in a new window] |
Figure 4. Immunohistochemical analysis of MOMPs. MOMPs generated by culturing PBMCs on FN-coated plates for 7 days were passaged and replated onto FN-coated chamber slides. After 24 h of culture, the slides were fixed with 10% formalin and stained with mAbs as indicated. Cells that were incubated with normal mouse IgG were used as controls. The nuclei were counterstained with hematoxylin. Magnification is original  00. The results shown are representative of at least five independent experiments.
|
|
View this table: [in a new window] |
Table 2. Protein expression profiles in monocytes, macrophages, dendritic cells, and MOMPs
|
 View larger version (30K): [in a new window] |
Figure 5. Proliferative capacity of MOMPs. (A) The number of MOMPs during cultures. The number of attaching cells recovered at each passage was counted, and the results are expressed as the ratio to the number of cells at the first passage. The results shown are the mean and standard deviation of MOMPs from five independent donors. (B) MOMPs generated by culturing PBMCs on FN-coated plates for 7 days were replated on FN-coated chamber slides. MOMPs were further cultured for 1, 3, 5, 7, and 10 days, and labeled with BrdU. At least 200 cells were counted for the BrdU staining experiment, and the proportion of BrdU+ cells was calculated for individual slides cultured for 1, 3, 5, 7, and 10 days. The results shown are the mean and standard deviation of five independent experiments. Representative pictures at Days 1 and 5 are shown in an inset. The nuclei were counterstained with hematoxylin. Arrows show nuclei positive for BrdU staining. (C) MOMPs were labeled with CFSE and were not cultured (0 days) or were cultured on new FN-coated plates for 1, 3, or 5 days. The non-cultured and cultured adherent cells were analyzed by flow cytometry. The results shown are representative of three independent experiments.
|
Differentiation of MOMPs along mesenchymal cell lineages
Because the MOMPs had morphologic and phenotypic properties of mesenchymal cells and expressed several stem cell markers, we hypothesized that they might be induced to differentiate along mesenchymal lineages, although MOMPs did not differentiate spontaneously into mature mesenchymal cells without specific treatments. Therefore, MOMPs were cultured under conditions known to induce differentiation of MSCs into osteogenic, myogenic, chondrogenic, or adipogenic cells.
MOMPs treated with the osteogenic induction procedure underwent a change in their morphology from spindle-shaped to cuboidal. Almost every adherent cell formed calcium deposits, which stained with alizarin red (Fig. 6A ), and could be stained for alkaline phosphatase (Fig. 6B) . This process was associated with a marked increase in the intracellular calcium content (Fig. 6C) . Undifferentiated MOMPs faintly expressed mRNAs for the bone-specific transcription factor osterix [26 ], but its expression was markedly up-regulated after the induction treatment. mRNAs for bone sialoprotein II and osteocalcin, specifically produced by mature osteocytes [27 ], were detected after induction; whereas CD34, CD45, and CD14 expression was lost (Fig. 6D) .
 View larger version (71K): [in a new window] |
Figure 6. Osteogenic, myogenic, chondrogenic, and adipogenic differentiation of MOMPs. MOMPs were prepared by culturing PBMCs on FN-coated plates. MOMPs before and after three weeks of osteogenic induction were stained with Alizarin red (A) or for alkaline phosphatase (B, original magnification 00). The intracellular calcium concentration was measured in MOMPs and dermal fibroblasts before and after osteogenic induction, and expressed as microgram per microgram protein content (C). Expression of mRNAs for osterix, bone sialoprotein II (BSP II), osteocalcin, CD34, CD45, CD14, and GAPDH was examined in MOMPs before and after the osteogenic induction, and in an osteosarcoma cell line (D). MOMPs before and after 3 weeks of myogenic induction were stained for SkM-actin (E, original  00) or SkM-MHC (F, original 00). Expression of mRNAs for myogenin, SkM-MHC, CD34, CD45, CD14, and GAPDH was examined in MOMPs before and after the myogenic induction, and in myoblasts, muscle tissue, and a rhabdomyosarcoma cell line (G). MOMPs before and after 3 weeks of chondrogenic induction were stained for type II collagen (H, original  0). Expression of mRNAs for 1(II) and 1(X) collagen, CD34, CD45, CD14, and GAPDH was examined in MOMPs before and after the chondrogenic induction, and in a chondrosarcoma cell line (I). MOMPs before and after 3 weeks of adipogenic induction were stained with Oil red-O (J, original 00). The induced cell with lipid vacuoles is shown in an inset (original x200). Expression of the mRNAs for PPAR, aP2, CD34, CD45, CD14, and GAPDH was examined in MOMPs before and after the adipogenic induction, and in fat tissue (K). The results shown are representative of at least five experiments.
|
20% of the cells that did not express SkM-MHC (data not shown). CD34 was also detectable in cultured myoblasts and muscle tissue, and even in a rhabdomyosarcoma cell line, consistent with the expression of CD34 in a subset of primitive muscle cells [28
]. MOMPs were cultured in several different micromass cultures in the presence of TGF-ß1, a standard method to induce chondrocyte differentiation in MSCs [9 , 18 , 19 ], but they died within 1 week. Therefore, we cultured monolayer MOMPs in the presence of TGF-ß1 for 3 weeks. The cells proliferated during the induction treatment. Type II collagen, which is typical of articular cartilage, was weakly expressed in untreated MOMPs, but its expression was markedly up-regulated in all adherent cells after the induction treatment (Fig. 6H) . RT-PCR further demonstrated the up-regulated expression of chondrocyte-specific type X collagen after the induction treatment (Fig. 6I) . The expression of CD45 and CD14 was lost, but CD34 expression was retained after the treatment, and CD34 was also expressed in a chondrosarcoma cell line.
Electron microscopic examination revealed small lipid droplets in MOMPs (Fig. 1)
, but lipid vacuoles were apparent after the induction treatment and increased over time in both size and number until they underwent coalescence (Fig. 6J
, inset). Such lipid vacuoles were stained with oil red O (Fig. 6J)
. This induction treatment resulted in 50–80% of the adherent cells being committed to this lineage, depending on the sample. mRNAs for PPAR and the fatty acid binding protein aP2 were weakly expressed in undifferentiated MOMPs, but the expression of these genes was markedly up-regulated after the induction treatment (Fig. 6K)
. Expression of mRNAs for CD45 and CD14 was lost, but CD34 expression was retained and CD34 expression was also detected in fat tissue.
The differentiation into mesenchymal cells was observed in MOMPs that were freshly generated from PBMCs, cultured for up to five passages or were cryo-preserved before use in this experiment. In addition, MOMPs from 10 independent donors showed similar differentiation potential. Two strains of human dermal fibroblasts, which are mature mesenchymal cells as well as freshly isolated CD14+ monocytes and macrophages, were also cultured under the identical induction conditions. The dermal fibroblasts did not undergo any such differentiation when cultured for 3 weeks, although the cells appeared healthy in these conditions. Circulating monocytes and macrophages subjected to these culture conditions detached from the plates within 1 week without apparent differentiation.
To investigate whether the induction treatments specifically induced the MOMPs to adopt different lineages, cultures from each treatment were cross-stained with alizarin red, oil red O, or were immunostained for SkM-MHC or type II collagen. Cultures were positive only for the markers specific to the expected lineage, and the other stainings were all negative.
The number of cells exhibiting lineage-specific differentiation at 3 weeks of culture was markedly decreased compared with the initial MOMP number after osteogenic (8–15%), myogenic (4–17%), and adipogenic (18–35%) induction treatment, indicating that only a subset of MOMPs differentiated into the induced lineages. In contrast, the number of differentiated cells was two- or threefold greater than the initial MOMP number after chondrogenic differentiation.
To exclude the possibility that the differentiated mesenchymal cells were derived from cells with MSC capacity that contaminated the MOMP fraction, CD14+ cells were positively purified from the MOMPs by using the MACS separation system before the induction treatment. As expected, the differentiation into the osteogenic, myogenic, chondrogenic, and adiopogenic lineages was observed as for the unselected MOMP cultures. In addition, the depletion of cells expressing CD34 or CD105/SH2 at the initiation of the PBMC cultures did not affect the mesenchymal cell development. MOMPs generated from purified CD14+ monocytes in culture with CD14- cell-conditioned medium exhibited similar mesenchymal cell differentiation.
We further examined the expression of lineage-specific transcription factors, Cbfa1/Runx2 [29
], MyoD [30
], Sox-9 [31
], and PPAR [32
], in MOMPs after 1 week of induction treatment. MOMPs that underwent osteogenic, myogenic, chondrogenic, or adipogenic induction treatment were double-stained for CD45 and individual transcription factors. As shown in Fig. 7
, MOMPs that underwent osteogenic differentiation for 1 week showed membranous and cytoplasmic expression of CD45 and nuclear expression of Cbfa1/Runx2. Similarly, the simultaneous expression of CD45/MyoD and CD45/Sox-9 was detected in MOMPs that underwent induction of the myogenic and chondrogenic lineages, respectively, for 1 week. PPAR was weakly expressed in the nuclei of some untreated MOMPs, but after adipogenic induction treatment for 1 week, the expression of PPAR was increased and CD45 expression became faint. The expression of lineage-specific transcription factors was specific to the expected lineage (data not shown). These findings indicate CD45+ hematopoietic cells underwent lineage-specific differentiation under specific permissive conditions. The expression of these lineage-specific transcription factors, except PPAR, was lost at 3 weeks of the induction treatment, as determined by immunohistochemistry and RT-PCR.
 View larger version (30K): [in a new window] |
Figure 7. Coexpression of CD45 (green) and tissue-specific transcription factors (red) in MOMPs that underwent 1 week of mesenchymal differentiation. MOMPs were prepared by culturing PBMCs on FN-coated plates. MOMPs before induction treatment (A–D) and MOMPs treated for osteogenic (E), myogenic (F), chondrogenic (G), or adiopogenic (H) induction for 1 week were examined for the immunohistochemical localization of CD45 in combination with Cbfa1/Runx2 (A and E), MyoD (B and F), Sox-9 (C and G), or PPAR (D and H). The cells were observed with confocal laser fluorescence microscopy (original magnification 00). The results shown are representative of three experiments.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The MOMPs described here exhibit mixed morphologic and phenotypic features of monocytes and mesenchymal and endothelial cells, all of which are of mesodermal origin. In addition, MOMPs express some stem cell markers and have the ability to self-replicate with limited cell-doublings and to undergo terminal differentiation. These characteristics are distinct from those of monocytes and monocyte-derived phagocytes. Several different precursors that potentially differentiate into mesenchymal cell types have been reported in human postnatal peripheral blood, including precursors for endothelial cells [23 ], smooth muscle [34 ], and mesenchymal cells [14 ]. In vitro expansion of endothelial and smooth muscle progenitors requires a combination of exogenous growth factors in addition to FBS [23 , 34 ]. Mesenchymal progenitors can be expanded in culture with FBS alone, but their development in PBMC cultures was reported to be unaffected by eliminating CD14+ cells [14 ]. These data, taken together with the characteristic MOMP phenotype, which is positive for CD14, CD45, CD34, and type I collagen, indicate that MOMPs can be easily discriminated from these progenitors.
We have provided several lines of evidence to demonstrate that MOMPs originate from circulating CD14+ monocytes. First, MOMPs were positive for the monocyte lineage markers CD14, CD11b/Mac-1, and CD11c. Second, serial phenotypic analyses of adherent peripheral blood cells cultured on FN showed increased expression of CD34 on the adherent CD14+ cells. Third, the development of MOMPs was almost completely inhibited by depleting the PBMCs of CD14+ cells, but not by depleting the PBMCs of cells positive for CD34 or CD105. MOMPs can be generated from purified circulating CD14+ monocytes by culturing them in CD14- cell-conditioned medium. Finally, studies using vital dye-labeled CD14+ cells revealed up-regulated expression of CD34 on CD14+ monocytes and lack of proliferation in the CD14- cell popula-tion. The differentiation of monocytes into distinct phagocytes is regulated by signals from different sets of growth factors: M-CSF for macrophages, GM-CSF and IL-4 for dendritic cells, and the receptor activator of the NF-
B ligand (RANKL) and M-CSF for osteoclasts [2
, 3
, 35
]. In this regard, the differentiation of monocytes into MOMPs requires soluble factor(s) derived from circulating CD14- cells and ligation to FN, probably through 4ß1 or 5ß1 integrin.
MOMPs can differentiate into mesenchymal cell types in specific permissive conditions principally developed for MSCs. Moreover, the differentiation of MOMPs into individual mesenchymal cells follows the steps observed in MSC differentiation, in terms of the timing of lineage-specific transcription factor expression. For example, the expression of MyoD precedes the expression of SkM-actin and myosin [30 ]. These findings suggest that differentiation processes leading to the adoption of individual mesenchymal lineages are shared by MOMPs and MSCs. However, chondrogenic differentiation was not observed in MOMPs when they were cultured at high density. This might reflect a difference in the sensitivity of mesenchymal and hematopoietic cells to the compressive hydrostatic pressure required for chondrogenic differentiation [36 ].
The possibility of contaminating pluripotent or monopotent non-hematopoietic precursors for mesenchymal cells in the MOMP population cannot be entirely excluded. However, we believe this to be unlikely. First, differentiation of MOMPs into mesenchymal cells was observed when purified CD14+ cells in the MOMP population or MOMPs generated from purified circulating CD14+ monocytes in the presence of CD14- cell-conditioned medium were used in the induction cultures. The depletion of potential endothelial and mesenchymal progenitors at the initiation of PBMC cultures did not affect the mesenchymal cell development. Second, the analysis of cell proliferation by using the vital dye CFSE revealed homogeneous growth in adherent CD14+ cells in the primary PBMC cultures, as well as in MOMP cultures after several passages, indicating that the differentiation potential did not result from the proliferation of a trace number of specific precursors. Finally, the co-expression of CD45 and lineage-specific transcription factors early in the differentiation process of MOMPs further supports the idea that we were observing the differentiation of hematopoietic cells into individual mesenchymal lineage cell types.
It is not yet clear whether the MOMP population contains multipotent cells or a mixture of committed progenitors with restricted potential. Because only a subset of MOMPs differentiated into the induced lineages, MOMPs are likely to be a heterogeneous cell population consisting of various committed progenitors. In this regard, the proportion of stem cells with multilineage differentiation potential is very low, even in MSCs [37 ]. The cloning of MOMPs at the single-cell level will be necessary to clarify this point, but it is technically difficult at the moment.
Although in vivo development appears to follow a sequential pathway of progressive fate restriction, several lines of evidence suggest that differentiation is not irreversible [38 ]. Such plasticity has been described in tissue-specific stem cells and intermediate precursors. Our findings indicate that a subset of circulating monocytes, which are committed hematopoietic cells, can differentiate into mesenchymal cells. This phenomenon can be explained by dedifferentiation, which means that cells revert to an earlier, more primitive phenotype that has a wider differentiation potential [39 ]. Alternatively, monocytes may acquire the capacity to differentiate into a novel phenotype without going through a more developmentally immature phenotype, a process called transdifferentiation [39 ]. Therefore, circulating monocytes may be more multipotent than previously thought, or retain the ability to dedifferentiate.
Another possibility is that monocytes may be competent to differentiate into mesenchymal cells by first becoming MOMPs, but this competence has been overlooked. This fate may not be expressed during normal development in the absence of cues, but may be readily revealed by altering the environment. The differentiation of CD14+ monocytes into MOMPs requires binding to FN and soluble factor(s) from CD14- blood cells. Circulating monocytes may encounter these signals at the site of tissue injury and inflammation. In this scenario, circulating monocytes infiltrate into the site of tissue injury and are exposed to FN and soluble factor(s) produced by infiltrating inflammatory cells, resulting in their differentiation into MOMPs. The MOMPs subsequently differentiate into tissue-specific cells in response to organ-specific cues provided by the surrounding cells. By this process, monocytes may participate in tissue homeostasis by replacing differentiated cells lost to physiologic turnover, injury, and senescence.
Because of the potential to differentiate into several distinct mesenchymal lineages, the monocyte is a possible cell source for tissue reconstitution therapy in trauma as well as congenital or degenerative disorders. In this regard, various adult tissue-specific stem cells and embryonic stem (ES) cells are currently being considered as candidate sources for future therapeutic intervention for tissue regeneration [8
]. It has been shown that MSCs engraft in many organs and differentiate along tissue-specific lineages upon transplantation in animal models [40
, 41
], as well as in patients with osteogenesis imperfecta [42
]. However, MSCs are rare in the adult human bone marrow (1 MSC per 105 stromal cells), and expansion to the number of cells required for regeneration is technically difficult, expensive, and time-consuming [37
]. ES cells are undifferentiated totipotent cells derived from blastocytes that can be propagated indefinitely in vitro and induced to differentiate to most cell types in vivo [43
]. Although ES cells have been isolated from humans, their use in research as well as therapeutics is en-cumbered by ethical and political considerations [44
]. In contrast, cellular therapy using circulating monocytes has consid-erable advantages over currently proposed strategies using tissue-specific stem cells and ES cells. Circulating monocytes could be a relatively easy source of autologous cells, as a large number of monocytes can be obtained from living individuals without using invasive procedures. Furthermore, the generation of MOMPs from monocytes is technically easy and quick, and the ethical dilemma of using ES cells can be bypassed.
Our observation challenges the traditional view of the biology of the monocyte/phagocyte system. We believe that it will lead to further progress in the understanding of the differentiation potential of monocytes and the roles they play in the physiology of health and disease.
Received April 22, 2003; revised June 13, 2003; accepted June 19, 2003.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
: a nuclear regulator of metabolism, differentiation, and cell growth J. Biol. Chem. 276,37731-37734This article has been cited by other articles:
 |
|
 |
|
  S. Apostolakis, G. Y.H. Lip, and E. Shantsila Monocytes in heart failure: relationship to a deteriorating immune overreaction or a desperate attempt for tissue repair? Cardiovasc Res, March 1, 2010; 85(4): 649 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Strieter, E. C. Keeley, M. A. Hughes, M. D. Burdick, and B. Mehrad The role of circulating mesenchymal progenitor cells (fibrocytes) in the pathogenesis of pulmonary fibrosis J. Leukoc. Biol., November 1, 2009; 86(5): 1111 - 1118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Yasuoka, Y. Yamaguchi, and C. A. Feghali-Bostwick The Pro-Fibrotic Factor IGFBP-5 Induces Lung Fibroblast and Mononuclear Cell Migration Am. J. Respir. Cell Mol. Biol., August 1, 2009; 41(2): 179 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Tasso, A. Augello, M. Carida', F. Postiglione, M. G. Tibiletti, B. Bernasconi, S. Astigiano, F. Fais, M. Truini, R. Cancedda, et al. Development of sarcomas in mice implanted with mesenchymal stem cells seeded onto bioscaffolds Carcinogenesis, January 1, 2009; 30(1): 150 - 157. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Swaminathan and S. V. Shah New Insights into Nephrogenic Systemic Fibrosis J. Am. Soc. Nephrol., October 1, 2007; 18(10): 2636 - 2643. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Thomas-Ecker, A. Lindecke, W. Hatzmann, C. Kaltschmidt, K. S. Zanker, and T. Dittmar Alteration in the gene expression pattern of primary monocytes after adhesion to endothelial cells PNAS, March 27, 2007; 104(13): 5539 - 5544. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Ogawa, M. S. Razzaque, K. Kameyama, G. Hasegawa, S. Shimmura, M. Kawai, S. Okamoto, Y. Ikeda, K. Tsubota, Y. Kawakami, et al. Role of Heat Shock Protein 47, a Collagen-Binding Chaperone, in Lacrimal Gland Pathology in Patients with cGVHD Invest. Ophthalmol. Vis. Sci., March 1, 2007; 48(3): 1079 - 1086. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. C. Johnson, J. A. Leopold, and J. Loscalzo Vascular Calcification: Pathobiological Mechanisms and Clinical Implications Circ. Res., November 10, 2006; 99(10): 1044 - 1059. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Tan, R. Steiner, S. P. Hoerstrup, and W. Weder Tissue-engineered trachea: History, problems and the future Eur. J. Cardiothorac. Surg., November 1, 2006; 30(5): 782 - 786. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.J.M. Loomans, H. Wan, R. de Crom, R. van Haperen, H.C. de Boer, P.J.M. Leenen, H.A. Drexhage, T.J. Rabelink, A.J. van Zonneveld, and F.J.T. Staal Angiogenic Murine Endothelial Progenitor Cells Are Derived From a Myeloid Bone Marrow Fraction and Can Be Identified by Endothelial NO Synthase Expression Arterioscler Thromb Vasc Biol, August 1, 2006; 26(8): 1760 - 1767. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ogawa, H. Kodama, K. Kameyama, K. Yamazaki, H. Yasuoka, S. Okamoto, H. Inoko, Y. Kawakami, and M. Kuwana Donor Fibroblast Chimerism in the Pathogenic Fibrotic Lesion of Human Chronic Graft-Versus-Host Disease Invest. Ophthalmol. Vis. Sci., December 1, 2005; 46(12): 4519 - 4527. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Romagnani, F. Annunziato, F. Liotta, E. Lazzeri, B. Mazzinghi, F. Frosali, L. Cosmi, L. Maggi, L. Lasagni, A. Scheffold, et al. CD14+CD34low Cells With Stem Cell Phenotypic and Functional Features Are the Major Source of Circulating Endothelial Progenitors Circ. Res., August 19, 2005; 97(4): 314 - 322. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. A. Hruska, S. Mathew, and G. Saab Bone Morphogenetic Proteins in Vascular Calcification Circ. Res., July 22, 2005; 97(2): 105 - 114. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kuwana, H. Kodama, Y. Okazaki, T. Satoh, T. Inoue, Y. Kawakami, and Y. Ikeda Multi-Lineage Potential of Human Monocyte-Derived Mesenchymal Progenitors (MOMPs). Blood (ASH Annual Meeting Abstracts), November 16, 2004; 104(11): 3595 - 3595. [Abstract]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
