ARTICLE pubs.acs.org/jpr
Use of Proteomic Differential Displays to Assess Functional Discrepancies and Adjustments of Human Bone Marrow- and Wharton Jelly-Derived Mesenchymal Stem Cells Hsing-Chun Kuo,†,‡, Chi-Chin Chiu,†,|| Wan-Ching Chang,|| Jiunn-Ming Sheen,|| Chia-Yu Ou,§ Ho-Chang Kuo,|| Rong-Fu Chen,^ Te-Yao Hsu,§ Jen-Chieh Chang,,^ Chang-Chun Hsaio,,^ Feng-Sheng Wang,,^ Chung-Cheng Huang,# Hsuan-Ying Huang,z and Kuender D. Yang*,||,^, ‡
)
Institute of Nursing and Department of Nursing, Chang Gung Institute of Technology Chia-Yi Campus, Taiwan Departments of Pediatrics, §Obstetrics, ^Medical Research, #Radiology and zPathology, Chang Gung Memorial Hospital-Kaohsiung Medical Center; Clinical Genomic And Proteomic Core Laboratory, Chang Gung Memorial Hospital-Kaohsiung Medical Center, Chang Gung University, Kaohsiung, Taiwan ABSTRACT: Mesenchymal stem cells (MSCs) from bone marrow are suitable for the reconstruction of connective tissues and even brain tissue but have limitations in terms of cell expansion and fully specific differentiation. In our current study, we have attempted to adjust and improve the cell expansion and differentiation properties of human MSCs from different tissues. MSCs from normal bone marrow and Wharton jelly were subjected to proteomic differential displays, followed by functional adjustments based on these displays. Bone marrow MSCs expressed more transgelin-2 and differentiated more rapidly into bone nodules but showed a slower growth rate. A knockdown of transgelin-2 expression by specific small interfering RNA (siRNA) significantly increased the growth rate of these cells, the G1/S phase cell cycle transition, and the interaction of cyclin D1 with cdk2. Wharton jelly MSCs expressed the chaperone protein HSP90β at higher levels and differentiated slowly toward an osteogenic lineage. However, the knockdown of HSP90β expression significantly increased bone nodule formation, inhibited cell growth, decreased the number of cells in the G1/S phase of the cell cycle, and decreased the interaction of cyclin D1 with cdk2 and of cyclin E with cdk2. These results were validated by the in vivo repair of segmental bone defects in a mouse model with severe combined immunodeficiency. We thus demonstrate an improvement in the cell expansion and tissue regeneration properties of human MSCs through specific adjustments. KEYWORDS: mesenchymal stem cells (MSCs), proteomics, regeneration, Wharton jelly, bone marrow, osteogenesis
’ INTRODUCTION Animal studies have shown that mesenchymal stem cells (MSCs) derived from bone marrow can be used to reconstruct bone, muscle, composite tissue, and even brain tissue.1,2 MSCs reside in several human tissues including bone marrow, adipose tissue, umbilical cord blood (UCB), and umbilical cord Wharton jelly.3-5 MSCs from different human tissues possess different capacities and various potencies to fully or partially differentiate into mature bone, muscle, adipose tissue, or neurons. For example, MSCs from human UCB do not differentiate into mature myotubes in vitro unless exposed to the in vivo muscle environment.3 In addition, MSCs from bone marrow and adipose tissue share a set of common genes that function in the early differentiation into osteogenic, chondrogenic, and adipogenic lineages but use different sets of genes for late mature differentiation.6 Hence, the functional adjustment r 2010 American Chemical Society
of the cell expansion and differentiation characteristics among human MSCs from different tissues is of critical importance to enable the future clinical application of human MSC cell therapy. Many studies have been performed to characterize the differentiation capacity of MSCs derived from bone marrow, adipose tissue, and UCB.5,7,8 Very few studies, however, have been conducted to adjust and improve cell expansion and osteogenic repair by human MSCs based on functional proteomic differential displays. Several lines of evidence have demonstrated that bone marrow MSCs are not fully suitable for ex vivo cell expansion or in vivo tissue regeneration therapy. First, the proliferation and multipotent differentiation capacity of human bone marrow Received: October 20, 2010 Published: December 14, 2010 1305
dx.doi.org/10.1021/pr101057w | J. Proteome Res. 2011, 10, 1305–1315
Journal of Proteome Research MSCs decreases dramatically during ex vivo expansion and in aging hosts.9,10 Second, the evidence accumulated from hematopoietic reconstruction with human UCB transplantation indicates that using UCB cells is advantageous for this procedure due to their higher proliferative capacity and fewer number of graft-versus-host reactions, although less stem cells are present in the UCB.11,12 Third, MSCs from fetal tissues tend to grow faster and reserve their pluripotency,13 and this can be enhanced by placental growth factor to promote mesenchymal cell proliferation and differentiation.14 Taken together, the results of these earlier studies suggest that a better cell expansion and improved potential plasticity of human MSCs could be achieved through a chimeric combination of functions from both human bone marrow and umbilical cord sources of these cells. MSCs present in human UCB are capable of differentiating into bone nodules in cell culture and into bony calli in mice with severe combined immunodeficiency (SCID).15 However, the occurrence of MSCs in UCB is less than 1%. Fortunately, large numbers of human MSCs reside in the Wharton jelly of human umbilical cords,16 and these MSCs are capable of differentiating into bone, muscle, and neural tissues.17-19 Because Wharton jelly MSCs in umbilical cords are younger or more naïve than MSCs in bone marrow, they may possess a greater capacity for ex vivo expansion. Wharton jelly is therefore a good source of multi- or pluripotent MSCs for clinical applications. The molecular or functional differences that exist between MSCs from bone marrow and from Wharton jelly with regard to cell expansion and tissue regeneration are unknown, however. To elucidate the different functions related to the varied protein expression profiles between Wharton jelly and bone marrow MSCs in our current study, we employed a gel-based proteomic differential display. We thereby identified the different protein profiles between these two MSC subtypes, followed by loss-offunction analyses to validate how the differential protein expression patterns are involved in osteogenic differentiation of human MSCs from bone marrow and Wharton jelly in vitro and in vivo. Using these models, we demonstrate that the targeting of differentially expressed proteins can improve in vitro cell expansion of bone marrow MSCs and accelerate in vivo differentiation and regeneration of Wharton jelly MSCs.
’ EXPERIMENTAL SECTION Harvest and Culture of Wharton Jelly MSCs
Human Wharton Jelly MSCs were prepared from fresh human umbilical cords obtained during normal spontaneous deliveries after informed consent was obtained. The human umbilical cords were placed in Hanks’ balanced salt solution (Gibco) before proceeding to the harvesting of the MSCs. The umbilical arteries and vein were first removed. The remaining Wharton Jelly was then diced into small explants in Dulbecco’s Modified Eagle’s Medium (DMEM; low glucose) and transferred to 10-cm dishes containing regular DMEM in a 37 C incubator at 5% CO2. The dishes were left undisturbed for 5-7 days to allow migration of MSCs from the explants.16 These studies were approved by the Institutional Review Board of the Chang Gung Memorial Hospital. Preparation and Culture of Bone Marrow MSCs
After informed consent had been obtained, human MSCs were isolated from bone marrow aspirates of the iliac crest of four pediatric patients (aged from 1 to 6 years) who had febrile illnesses and findings on normal bone marrow examinations. An additional portion of the 2-mL bone marrow aspirates was also
ARTICLE
obtained. The MSCs from the bone marrow aspirates were prepared as described previously.20 The immunological characteristics of MSCs were assessed by flow cytometry. 2D SDS-PAGE Analysis of MSC Proteomic Profiles
Bone marrow and Wharton jelly MSCs (2 107 cells) were harvested at between 3 and 6 cell culture passages for total protein extraction using cell lysis solution (PRO-PREP, Intron Inc., Seoul, Korea). After cell lysis, the total cell protein was precipitated by 10% trichloroacetate in acetone. Protein samples (100 μg) were suspended in rehydration solution and subjected to isoelectric focusing (IEF) on 13-cm, nonlinear, pH 3-10 immobilized-gradient strips (Immobiline DryStrips, Amersham Biosciences, Uppsala, Sweden) using the Ettan IPGphor II apparatus (Amersham Biosciences). The second dimension electrophoresis was carried out using 10% SDS-PAGE gels. Gels were fixed and then subjected to silver staining, and the reaction was finally stopped with 3.65 g of EDTA in 250 mL of ddH2O.
Differential Displays of Total Cell Proteins and Analysis of the Peptide Fingerprints of Bone Marrow and Wharton Jelly MSCs
Four pairs of silver-stained 2D SDS-PAGE gels in which total cell proteins of bone marrow and Wharton jelly MSCs had been resolved were scanned using ImageMaster 2D Platinum Software 6.0 (Amersham Biosciences). The protein profiles of each pair of silver-stained bone marrow and Wharton jelly MSC gels were recorded and compared as previously described.21 Only the protein spots that differed by at least 3-fold (P < 0.05) between all four pairs of 2D gels were subjected to in-gel digestion for matrixassisted laser resorption ionization-time-of flight/time-of-flight (MALDI-TOF/TOF) mass spectrometric analysis.22 The gel pieces were then dehydrated and subjected to trypsin digestion. Mass spectra were acquired as the sum of the ion signals that were generated by irradiation of the target with a mean of 300 laser pulses using the FlexAnalysis system (Bruker-Franzen Analytik, Bremen, Germany). Peptide fingerprints were selected in the mass range of 1250-2500 Da and analyzed using Mascot software. Western Blot Confirmation of Differential Protein Displays between Bone Marrow and Wharton Jelly MSCs
Protein samples (20 μg) were denatured and subjected to 12% SDS-PAGE, followed by transfer of gel proteins onto a nitrocellulose membrane for Western blot analysis as previously described.15,23 The protein blots were then probed with primary antibodies, including rabbit antihuman HSP90β antibody (Chemicon, Temecula, CA), mouse antihuman TCP-1R antibody (Chemicon), mouse antihuman transgelin antibody (Santa Cruz Biotechnology, Santa Cruz, CA), and mouse antihuman vimentin at a 1:200 dilution (Chemicon). Cell Proliferation Assay and Cell Viability Test
The proliferation of MSCs was assessed using the Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kamimashiki-gun, Kumamoto, Japan). Briefly, cell suspensions were seeded onto 96-well plates at 1 103 cells/well and incubated for 1, 2, and 3 days. The 96-well plate was then subjected to absorbance measurements using a microplate reader.24 The cell viability was greater than 98%, as assessed by the trypan blue exclusion test in an equal volume of 0.25% trypan blue cell suspension.25 Osteogenic Differentiation of MSCs
Bone marrow and Wharton jelly MSCs at between 3 and 6 passages were cultured with or without conditional factors for 1306
dx.doi.org/10.1021/pr101057w |J. Proteome Res. 2011, 10, 1305–1315
Journal of Proteome Research
ARTICLE
osteogenic differentiation containing 50 μg/mL ascorbate-2 phosphate (Sigma, St Louis, MO), 10-8 M dexamethasone (Sigma), and 10 mM β-glycerophosphate (Calbiochem, San Diego, CA). Osteogenic differentiation was determined by RT-PCR analysis of the osteogenic-specific transcription factor, RUNX2 and osteocalcin (OCN).15 The MSCs showing osteogenic differentiation were also stained with von Kossa stain (3% silver nitrite; Sigma) to determine the degree of bone nodule formation.15,23 The area of mineralized nodules in the cultures was determined by measuring the numbers of pixels using ImageGauge 3.46 software (Fujifilm, Inc.). The alkaline phosphatase (ALP) activity and DNA content were measured from the cell layer.26 ALP activity was corrected for the amount of DNA in the culture. Functional Adjustment of MSCs Proliferation, Cell Cycle and Differentiation by Knockdown of HSP90β and Transgelin-2 with Small Interfering RNA (siRNA)
Wharton jelly or bone marrow MSCs were subcultured into a 6-well culture plate overnight before being subjected to siRNA transfection. HSP90, TCP-1 transgelin-2, vimentin siRNA, and their nonspecific analogs as a control (Ambion, Applied Biosystems), at 300 nM were used to transfect MSCs (2 105 cells) in oligofectamine (Invitrogen, Grand Island, NY). The bone marrow MSCs were transfected with the specific transgelin-2 siRNA (UCCCUAUAAAUUAAGUUCCTG) and vimentin siRNA (UCACGAUGACCUUGAAUAA), while the Wharton jelly MSCs were transfected with the specific HSP90β siRNA (UCAGUCAGUGAUAUUAGCCTT) and TCP-1R siRNA (UCACGAUGACCUUGAAUAA). To confirm the efficient knockdown of transgelin-2 and HSP90β, the primers for detection of transgelin-2 mRNA expression were forward 50 -GACGCGAGAACTTCCAGAAC-30 and reverse 50 -ACACAGGCCATGTTCTTTCC-30 . The primers for the detection of HSP90β mRNA expression were forward 50 -TGGGAAGAGGTTCCAGAATG-30 and reverse 50 -GTTGCCAGACCATCCGTACT30 . The primers for the detection of vimentin mRNA expression were forward 50 -GAACGCCAGATGCGTGAAATG-30 and reverse 50 -CCAGAGGGAGTGAATCCAGATTA-30 . The primers for the detection of TCP-1R mRNA expression were forward 50 CCTTTGTCCGTGTTCGGTGA-30 and reverse 50 -GCAGGATGTTCTACCTCCAGTA-30 . The 18S rRNA house-keeping gene was used as the internal control. Redistribution of cell cycle was assessed by flow cytometric analysis and by the Western blot analysis of cell cycle proteins. Protein aliquots (300 μg protein) of cell extracts from bone marrow and Wharton jelly MSCs were incubated for 4 h at 4 C with 1.0 μg of primary cdk2 antibody (Santa Cruz Biotechnology). Immune complexes were precipitated with 20 μL of prewashed protein A/G PLUS Agarose beads (Santa Cruz Biotechnology), shaken gently, and washed four times. Western blotting for the cdk2-associated cyclin D1 or cyclin E partner was then performed. In Vivo Adjustment of Segmental Bone Defect Repair by Knockdown of HSP90β and Transgelin-2 with siRNA
A segmental bone defect model in SCID mice (NOD mice, National Animal Center, Taiwan) with a 4-mm gap in the right femur was developed as described previously with slight modifications.15 To determine whether tissue regeneration in the segmental bone defect was derived from human MSCs, PCR for human-specific Alu DNA expression was performed using specific primers.27 We also detected human HLA-ABC antigens
Figure 1. WJ-MSCs show faster growth but a slower rate of osteogenesis than BM-MSCs. (A) WJ-MSCs grow at a significantly faster rate than BM-MSCs. (B) Microscopically, WJ-MSCs exhibited a higher cell density than BM-MSCs. (C) WJ-MSCs show lower levels of osteogenic transcription factor (RUNX2) and osteocalcin (OCN) mRNA expression than BM-MSCs. (D) von-Kossa staining revealed less osteogenic nodule formation in WJ-MSCs than in BM-MSCs. The data presented were calculated from three replicate experiments. The asterisk indicates a significant difference (P < 0.01) between the cell growth of WJ-MSCs and BM-MSCs. Scale bar, 100 μm.
(Dako, Glostrup, Denmark) in the regenerative tissue by immunohistochemical staining to confirm the regeneration of human MSCs. The MSCs (2 105 cells) transfected with specific or control (nonspecific) siRNA as described above were seeded into the segmental defect of the right femur bone with the support of fibrin glue prepared using human thrombin. Four SCID mice each were implanted with Wharton jelly or bone marrow MSCs; the fifth group underwent a sham procedure. Callus formation in the osteotomy gap was evaluated using a mammography system (20 kV, 8 mA, film to focus distance 50 cm; Lorad M-IV; Varian, Salt Lake City, UT). Bridging and closing of the osteotomy gap was radiographically evaluated for bone union by a radiologist in a blind manner.15 The evaluation was based on a scoring system of segmental bone healing visualized by mammography.26,28 The regeneration of callus tissue was determined by conventional hematoxylin-eosin staining on a series of 5-μm tissue sections. Distribution of fibrous tissue, osteochondrotic (cartilage) tissue, and osteogenic (bone) tissue was measured in three random microscopic fields for each tissue section.28,29 Data Management and Statistics
Four pairs of total cell protein extracts from bone marrow MSCs and Wharton jelly MSCs were used for differential proteomic display using an ImageMaster scanner (Amersham Biosciences). Differences in cell proliferation, differentiation and tissue regeneration parameters between bone marrow MSCs and Wharton jelly MSCs were statistically analyzed using the MannWhitney-U test. 1307
dx.doi.org/10.1021/pr101057w |J. Proteome Res. 2011, 10, 1305–1315
Journal of Proteome Research
ARTICLE
Figure 2. Differential proteomic displays between Wharton jelly MSCs and bone marrow MSCs. Four pairs of 2-dimensional SDS-PAGE proteomic profiles for (A) Wharton jelly MSCs and (B) bone marrow MSCs were compared using ImageMaster 2D software. (C) Western blot validation of differential proteomic displays of Wharton jelly (WJ) and bone marrow (BM) MSCs. The Western blot data shown are derived from a representative experiment, and (D) dot density comparisons of protein expression were calculated from three replicate experiments.
’ RESULTS MSC Marker CD105 and CD166 Expression
MSCs usually express specific mesenchymal markers, but not hematopoietic cell markers. In our current study, we found that MSCs from Wharton jelly cultured in vitro for seven days and allowed to spontaneously migrate out of the jelly expressed the mesenchymal cell markers CD44, CD166, and CD105, but not the hematopoietic cell markers CD14 and CD133. Bone marrow MSCs cultured in a similar manner also expressed CD44, CD166, and CD105 (data not shown).15,16,20 Differences in Cell Growth and Osteogenic Differentiation
Experiments were next performed to explore functional discrepancies between the bone marrow and Wharton jelly MSCs. The Wharton jelly MSCs, which expressed higher levels of cell recycling proteins, and showed a faster growth rate over three days than the bone marrow MSCs, which expressed higher levels of mesenchymal structural proteins (P = 0.003; Figure 1A and B). The faster growing Wharton jelly MSCs showed slower osteogenic differentiation, as indicated by the lower levels of RUNX2 and OCN mRNA expression after five days of osteogenesis induction
(Figure 1C). In contrast, the bone marrow MSCs, which expressed fewer cell recycling proteins, showed a slower cell expansion rate faster osteogenic differentiation, as evidenced by the higher levels of RUNX2 mRNA expression and greater formation of a mineralized bone matrix after 14 days of osteogenic induction (Figure 1D). Differences in Protein Expression Profiles
The number of protein spots on bone marrow MSCs total protein displays ranged from 778 to 1009 and the number of spots on the Wharton jelly MSCs total protein displays ranged from 689 to 896. The expression level for 38 protein spots was at least 3-fold or greater in bone marrow MSCs than in Wharton jelly MSCs on four pairs of total protein displays. In contrast, the expression levels for 26 protein spots were at least 3-fold or greater in Wharton jelly MSCs than in bone marrow MSCs (Figure 2A and B). The protein spots were next subjected to in-gel digestion and MALDI-TOF/ TOF analysis of peptide fingerprints. Differentially Displayed Cell Structure and Proliferation Proteins
For the peptide fingerprinting by MALDI-TOF/TOF mass spectrometry, 23 peptides from protein gels of Wharton jelly
1308
dx.doi.org/10.1021/pr101057w |J. Proteome Res. 2011, 10, 1305–1315
Journal of Proteome Research
ARTICLE
Table 1. Thirty-two Differentially Displayed Proteins between WJ and BM MSCs #
categories of proteins
3 fold or greatera
P-value
accession no.
MALDI-MS/MS:
(Swiss-Prot)
MOWSE-score
MW/pI
Proliferation 1
Transcripition factor BTF3
WJ > BM
0.000
BTF3_HUMAN
70
22/9.4
2
HSP90
WJ > BM
0.018
ENPL_HUMAN
71
92/6.2
3
Brefeldin A-inhibitd GEP 2
WJ > BM
0.000
BIG2_HUMAN
55
4
PPlase A
WJ > BM
0.000
PPIA_HUMAN
77
5
TCP-1-zeta
WJ > BM
0.000
TCPZ_HUMAN
74
58/6.3
6
TCP-1-alpha
WJ > BM
0.000
TCPA_HUMAN
115
60/5.8
7 8
TCP-1-epsilon hnRNP A2/hnRNP B1
WJ > BM BM > WJ
0.013 0.000
TCPE_HUMAN ROA2_HUMAN
96 65
60/5.5 37/9.4
9
Zinc finger protein 585A
BM > WJ
0.028
Z585A_HUMAN
56
88/6.9
10
Rsu-1
BM > WJ
0.017
RSU1_HUMAN
72
31/8.6
11
HSP70
BM > WJ
0.000
HSP7C_HUMAN
66
71/5.4
12
protein CXorf22
WJ > BM
0.001
CX022_HUMAN
64
110/8.4
13
superoxide dismutase C Mn) mitochondrial precursor
WJ > BM
0.000
SODM_HUMAN
55
25/8.4
14 15
Glucosidase II beta subunit Pyruvate kinase isozyymes M1M2
WJ > BM WJ > BM
0.000 0.024
GLU2B_HUMAN KPYM_HUMAN
75 195
59/4.3 58/7.7
16
Thioredoxin domain-containing protein 4
BM > WJ
0.032
TXND4_HUMAN
72
47/5.6
17
Beta-ketothiolase
BM > WJ
0.011
THIM_HUMAN
117
42/8.3
18
Phosphoglycreate kinase 1
BM > WJ
0.000
PGK1_HUMAN
81
44/8.3
19
50 (30 )-deoxyribonucleotidase, mitochondria precursor
BM > WJ
0.000
NT5M_HUMAN
63
26/8.8
20
Chloride intracellular channel protein 1
BM > WJ
0.032
CLIC1_HUMAN
180
27/5.1
21
Procollagen-lysine,2-oxoglutarate 5-dioxygenase 2 precursor
WJ > BM
0.000
PLOD2_HUMAN
70
85/6.6
22 23
Actin-related protein 3 Microtubule-actin cross-linking factor 1
WJ > BM WJ > BM
0.000 0.000
ARP3_HUMAN MACF4_HUMAN
118 47
47/5.6 67/4.0
24
Syndecan-binding protein beta subunit
WJ > BM
0.000
SDCB1_HUMAN
61
32/7.1
25
Collagen alpha-2(VI) chain precursor
BM > WJ
0.017
CO6A2_HUMAN
76
109/6.4
26
Collagen alpha-2(I) chain precursor
BM > WJ
0.000
CO1A2_HUMAN
27
Vimentin
BM > WJ
0.029
VIME_HUMAN
28
SUMO-3
BM > WJ
0.002
29
PEBP-1
BM > WJ
30 31
Galectin-3,calregulin Transgelin
32
Transgelin2
201/6 18/8.4
Metabolism
Structure protein
66
129/9.1
223
53/5.1
SUMO3_HUMAN
70
12/5.3
0.000
PEBP1_HUMAN
33
21/7.4
BM > WJ BM > WJ
0.000 0.000
LEG3_HUMAN TAGL_HUMAN
60 93
26/8.6 22/9.0
BM > WJ
0.045
TAGL2_HUMAN
74
22/8.8
a
List of proteins consistent 3-fold or greater difference and p-value
