Differential Effect of Water-Soluble Chitin on Collagen Synthesis of Human Bone Marrow Stem Cells and Human Periodontal Ligament Stem Cells

Isolation and culture of hBMSCs and hPDLSCs

Fresh human bone marrow (BM) was harvested from the vertebrae of five volunteers (one man and four women; median age, 46 years; age range, 41–58 years) during orthopedic surgery (Naeun Hospital, Anyang, Korea). The patients provided informed consent for their tissue being used in accordance with guidelines approved by the Institutional Review Board, College of Dentistry, Yonsei University (IRB No. 2-2012-0050). The mononuclear cell fraction was separated from the BM aspirates for the isolation of MSCs by conventional density-gradient centrifugation with a density gradient solution (Ficoll-Paque Plus; GE Healthcare Bio-Sciences) using a slightly modified version of a previously described protocol. ²⁹

The hPDLSC culture protocol which was used is described by Seo et al., ³⁰ and it is a slightly modified version of that used in our previous study. ³¹ The experimental protocol was approved by the Ethics Committee of the College of Dentistry, Yonsei University (IRB No. 2-2012-0049), and written informed consent was obtained from all subjects before they were enrolled in the study. hPDLSCs were isolated from healthy teeth that had been extracted from three adults aged 11–19 years receiving orthodontic treatment.

hBMSCs (10–100×10⁶) and hPDLSCs (5×10⁵) were seeded into T75 cell culture flasks with a growth medium and incubated at 37°C in 5% CO₂. Basically, the basic methods to isolate hBMSCs and hPDLSCs are different, and, therefore, it is difficult to regulate the initial seeding numbers. In addition, initial fresh BM cells are fully heterogeneous, including a large number of red blood cells or other hematopoietic cells as well as MSCs. Therefore, there was a greater initial cell seeding number of hBMSCs. However, after passage of 2–4, both hBMSCs and hPDLSCs used for this study were seeded under the same cell number. Colony formation in the culture was observed using a light microscope (CK40; Olympus Optical). Adherent MSC-like cells were obtained after 2 weeks of culture.

For the purposes of this study, the cells were further expanded in the presence of 0.3 mg/mL WSC for WSC-treated hBMSCs and hPDLSCs; for control (i.e., nontreated) groups, hBMSCs and hPDLSCs were expanded using a conventional method that did not involve WSC. WSC was manufactured using the method described by Kurita et al., ³² as described in our previous study. ²⁸ The expansion medium was exchanged every 3–4 days, and the cells were harvested when they reached confluence. Cells from passage 2 (P2) to P4 were used for this study. We have previously provided full details of the methods used to characterize these cultured MSCs.^31,33

Evaluation of surface-marker expression: fluorescence-activated cell sorting

Cell-surface-marker characterization of the isolated cells in the four experimental groups (i.e., nontreated hBMSCs, WSC-treated hBMSCs, nontreated hPDLSCs, and WSC-treated hPDLSCs) was performed using flow cytometry analysis (fluorescence-activated cell sorting [FACS]), the procedure that is described elsewhere.^29,31,33 The cells were then incubated with specific fluorescein-isothiocyanate- or phycoerythrin-conjugated mouse monoclonal antibodies raised against human CD105, CD73, CD90, CD44, CD29, CD14, CD34, CD45, CD146, and STRO-1 for 20 min at room temperature. The cells were washed twice and subjected to flow cytometry analysis using a flow cytometer (Beckman Coulter), and analyzed using FACScan software (FACS Calibur; BD Biosciences).

Morphology of WSC-treated hBMSCs and hPDLSCs

WSC-treated hBMSCs and hPDLSCs were continuously expanded in the presence of 0.3 mg/mL WSC, while those in the nontreated groups were expanded in the usual expansion medium. Cell morphologies were observed at P2 using a light microscope (CK40; Olympus Optical).

Assessment of cell proliferation and doubling time

The proliferation rates were assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay at P4. Cells (1×10³) were plated onto 48-well plates, cultured, and then subjected to MTT assay after 1, 3, and 5 days. After incubation with 10% MTT solution (5 mg/mL; Amresco) at 37°C at 5% CO₂ for 2 h, the supernatant was discarded and dimethylsulfoxide (Amresco) was added to each well. The absorbance was measured at 540 nm using a microplate spectrophotometer (Benchmark Plus enzyme-linked immunosorbent assay [ELISA] reader; Bio-Rad).

After confirming the MTT assay, the expressions of mRNAs related to proliferation were observed by real-time polymerase chain reaction (PCR) for the gene encoding Ki-67. ³⁴ Information regarding specific primers are given in Table 1. The real-time PCR protocol is described in our previous study. ²⁸

The doubling time was calculated using the following equation: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} T{\rm d} = (t2 - t1) \times \log (2) / \log ( q2 / q1 ) ,\end{align*} \end{document}

where Td is the doubling time, q1 is the quantity grown at time t1, and q2 is the quantity grown at time t2.

Clonogenic assays

The number of colony-forming units (CFUs) was evaluated for all four experimental groups at P2 using a light microscope (CK40; Olympus Optical). The cells were plated at a density of 1×10² cells/mL onto 100-mm culture dishes with growth medium and allowed to grow for 14 days. The dishes were then fixed with 10% neutral-buffered formalin and stained with crystal violet (Sigma-Aldrich). The number of CFUs at P2 was used to determine the colony-forming efficiency (CFE). The colony with the diameter of more than 2 mm wide was counted, and the lightly stained one was included if it was wider than 2 mm.

Cell adhesion activity

Double-label immunofluorescence was used to observe differences in cellular morphology related to the hBMSCs and hPDLSCs adhesion activity for each subpopulation of cells.^35–37 The cells were seeded in two-well chamber slides at a density of 1×10⁴ cells per well and expanded for 5 days. The cell, which was in a one-well chamber, was permeabilized with 100% methanol and blocked with 1% bovine serum albumin and 4% goat serum before incubation with the primary antibody raised against integrin alpha 5 (diluted 1:100, ab6131; Abcam) for 30 min at 37°C. After washing with washing buffer, an Alexa-fluor 488-conjugated got anti-mouse lgG (FITC, diluted 1:750, ab6785; Abcam) was used as a secondary antibody for 1 h at room temperature.

The cell was fixed with 10% neutral-buffered formalin and blocked using 10% serum for 30 min at 25°C in another chamber. The primary antibody raised against E-cadherin (diluted 1:25, ab15148; Abcam) was added and incubated for 1 h at 25°C. After incubation, an Alexa-fluor 555-conjugated anti-rabbit lgG (TRITC, diluted 1:300, ab6718; Abcam) was used as a secondary antibody for 1 h at room temperature. The chamber slides were washed thrice and then mounted with mounting medium.

The samples were observed using an Olympus BX41 microscope fitted with an Olympus DP71 camera and a U-RFL-T burner connected to a personal computer running DP-BSW software, and the mean fluorescence intensity of integrin alpha 5 and E-cadherin was analyzed using Adobe Photoshop CS4 software (Adobe Systems).

Induction of osteogenic, adipogenic, and chondrogenic differentiation

Osteogenic and adipogenic differentiation was induced as in the previous study. ³¹ The cells (P4) were seeded into six-well plates at a density of 1×10⁵ cells/well, and cultured until they reached subconfluence. The culture medium used for osteogenic differentiation was refreshed at 3-day intervals. After 4 weeks of induction, the cells were stained with alizarin red. For adipogenic differentiation, the cells were subjected to three cycles of adipogenic induction/maintenance once confluence had occurred. The adipogenic differentiation media were purchased from Lonza. Mature adipocytes were detected by Oil Red O staining at the end of 21 days of culture. The total areas of mineralized nodule formation and lipid vacuoles were measured using an automated image-analysis system (Image-Pro Plus; Media Cybernetics).

Chondrogenic differentiation was assessed by conducting a pellet culture using a slightly modified version of a previously described method. ³⁸ The cultured cell pellets were left free floating for 21 days and then fixed in 10% neutral-buffered formalin. Paraffin-embedded, 5-μm-thick section were cut and immunostained with Alcian blue to enable the visual assessment of chondrogenic differentiation.

After confirming the differentiation potential of each group of cells, samples were collected for quantitative osteogenic, adipogenic, and chondrogenic marker expressions (Table 1). The expressions of each of the differentiation markers were analyzed using the real-time PCR according to the manufacturer's instructions for each product. The real-time PCR protocol is described in our previous study. ³¹ Each PCR was performed in triplicate with the same total RNA.

In vitro collagen and insoluble synthesis assay

hBMSCs and hPDLSCs were cultured in culture medium with or without 0.3 mg/mL WSC for 7 days, and the total levels of soluble and insoluble collagen in the culture were measured using the Sircol collagen assay kit (Biocolor). The cell supernatants in the four experimental groups were collected and measured using a previously described soluble collagen assay method, ²⁸ and the insoluble collagen assay was performed using the scraped contents from the culture dishes as also described in the previous study. ²⁸ Then, the content of measured soluble and insoluble collagen was normalized to the final cell number.

Analysis of hydroxyproline in soluble and insoluble collagen

Hydroxyproline in the culture supernatant and the collagenous ECM that had accumulated on the culture dishes were quantified using a slightly modified version of the method of Reddy and Enwemeka.^28,39 In brief, after 7 days of culture with or without WSC, the cell supernatants in the four experimental groups were collected and a 200-μL aliquot of the supernatant was placed into a 2-mL clear wide-opening crimp-top vial (Agilent Technologies) and closed with an 11-mm crimped cap (Agilent Technologies). The ECM that had accumulated on the cultured dishes was collected into 100 μL of dH₂O and 100 μL of concentrated HCl (Duksan Pure Chemicals) in a same vial. The content of hydroxyproline in soluble and insoluble collagen in the culture was measured using the hydroxyproline assay kit (Biovision), and the content of hydroxyproline was normalized at the final cell number to be measured in the same way as the sircol assay.

Expression of collagen, hyaluronic acid synthase, and lysyl oxidase-family-related genes

mRNA was collected from cultured cells and then the expressions of mRNAs related to collagen formation were observed by real-time PCR. The quantitative real-time expressions of the genes encoding collagen types I, II, III, IV, and V, hyaluronic acid synthase 1 (HAS1), HAS2, HAS3, lysyl oxidase (LOX), lysyl oxidase-like 1 (LOXL1), LOXL2, and LOXL3 were detected. Information regarding each specific primer is given in Table 1.

In vivo collagen synthesis assay (transplantation of hBMSCs and hPDLSCs with HA into immunodeficient mice)

The collagen synthesis of the WSC-treated and nontreated hPDLSCs and hBMSCs was evaluated using an in vivo assay model. ³¹ For a single transplantation, 6×10⁶ cells were mixed with 0.15 mL of HA filler (ReDexis; Prollenium Medical Technologies) and then injected into subcutaneous pockets that had been made on the dorsal surface of 5-week-old immunocompromised mice. The mice were allocated to one of five groups (one nude mouse with three ectopic transplantations per group) with the following specific transplant conditions: (1) HA carrier only (negative control), (2) nontreated hBMSCs with HA carrier, (3) WSC-treated hBMSCs with HA carrier, (4) nontreated hPDLSCs with HA carrier, and (5) WSC-treated hPDLSCs with HA carrier. The mice were killed at 8 weeks post-transplantation, and the implants were harvested. All animal experiments were carried out in accordance with the Guidelines and Regulations for the Use and Care of Animals of the Animal Care Committee of the Medical College, Yonsei University.

In preparation for frozen sectioning, tissue samples were weighed on a balance (Adventurer; Ohaus) and the tissue was then frozen and sectioned serially at 6 μm. The obtained sections were mounted onto glass slides and then stained with Masson's trichrome and immunohistochemically against human-specific mitochondria (hMito), collagen types I and III, and hydroxyproline: human-specific mitochondrial ribosomal protein (diluted 1:50, ab74285; Abcam), collagen type I (diluted 1:200, ab7778; Abcam), collagen type III (diluted 1:1000, ab292; Abcam), and hydroxyproline (diluted 1:500, bs-7512R; Bioss). After washing the sections thrice, immunodetection was performed using a commercially available kit (EnVision Detection System; Dako REAL) according to the manufacturer's instructions. The sections were then counterstained with hematoxylin.

Statistical analysis

All in vitro experiments were performed at least thrice. Means and standard deviations were calculated, and unpaired t-testing was used to analyze the differences between two groups. For multiple analyses, repeated-measures ANOVA was performed followed by Scheffé's comparison. Differences were considered statistically significant at p<0.05.

Comparison of the effects of WSC on the characteristics of hBMSCs and hPDLSCs

The FACS analysis of nontreated hBMSCs and hPDLSCs revealed the expressions of representative MSC markers, testing positive for CD29, CD44, CD73, CD90, and CD105, and negative for CD14, CD34, and CD45. WSC appeared to have little effect on the level of surface-marker expression in either hBMSCs or hPDLSCs (Fig. 1A). Cells that expanded in the presence of WSC appeared to be more vigorous and longer than those grown in normal medium, producing a greater cell yield in the same-size culture dishes (Fig. 1B).

OPEN IN VIEWER

FIG. 1.

Influence of water-soluble chitin (WSC) on the characteristics of human bone marrow stem cells (hBMSCs) and human periodontal ligament stem cells (hPDLSCs). (A) Immunophenotype analysis of nontreated hBMSCs, WSC-treated hBMSCs, nontreated hPDLSCs, and WSC-treated hPDLSCs by flow cytometry. Similar cell-surface-marker expressions of CD29, CD44, CD73, CD90, CD105, CD14, CD34, and CD45 were observed in all of the mesenchymal stem cells (MSCs), regardless of the treatment condition. (B) During expansion, imaging of morphology differences revealed that the cells appeared slimmer and more extended in the WSC-treated groups than in the nontreated groups (magnification×100, scale bar=500 μm). (C) 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays were conducted to test the proliferation of the MSCs over 5 days. Treatment with WSC significantly increased the proliferative potentials of both hBMSCs and hPDLSCs (*p<0.05). The expression of mRNA for Ki-67 was significantly increased in WSC-treated hMSCs compared with nontreated hMSCs (*p<0.05). (D) Isolated nontreated hBMSCs, WSC-treated hBMSCs, nontreated hPDLSCs, and WSC-treated hPDLSCs formed colonies and were stained with crystal violet after 14 days. More colonies were formed in the WSC-treated groups than in the nontreated groups (*p<0.05). The doubling time was gradually increased in later passages of all groups, and treatment with WSC decreased the doubling time in both hBMSCs and hPDLSCs. (E) Double-label immunofluorescence staining was performed to test the cell adhesion activity of the MSCs. The intensity of integrin alpha 5 and E-cadherin expression were greater for WSC-treated hMSCs than for nontreated hMSCs and implied that the cell adhesion was increased by the treatment of WSC (magnification ×200, *p<0.05). Color images available online at www.liebertpub.com/tea

The proliferative potential of the WSC-treated cells was assessed relative to that of nontreated cells using the MTT assay and statistically significant differences were found among the four cell groups investigated, particularly at day 5. WSC significantly increased the proliferative potentials of both hBMSCs and hPDLSCs. The real-time PCR performed to determine the changes in proliferation expressed at the mRNA level revealed that the expression of mRNA for Ki-67 was significantly increased in WSC-treated hMSCs compared with nontreated hMSCs (p<0.05; Fig. 1C). The effect of WSC expansion was clearly distinguishable in the CFU assay. The CFE was significantly increased in WSC-treated hMSCs compared with their nontreated counterparts. The doubling time was gradually increased in later passages of all groups, and treatment with WSC decreased the doubling time in both hBMSCs and hPDLSCs (p<0.05; Fig. 1D). Double-label immunofluorescence staining was performed to test the cell adhesion activity of the MSCs. The intensity of integrin alpha 5 and E-cadherin expression was greater for WSC-treated hMSCs than for nontreated hMSCs and implied that the cell adhesion was increased by the treatment of WSC (p<0.05; Fig. 1E).

Multilineage differentiation of hBMSCs and hPDLSCs, and the effect of WSC application

Calcium deposits were observed after 4 weeks of osteogenic induction in all cell groups, as visualized by alizarin-red staining (Fig. 2A). The total area of mineralized nodules was significantly larger for hPDLSCs than for hBMSCs, and WSC-treated hBMSCs and hPDLSCs exhibited a further significant increase in nodule formation (p<0.05; Fig. 2B). This finding was corroborated by an additional assessment of osteogenic differentiation by exploring the representative mRNA expression of runt-related transcription factor 2 (Runx-2) using quantitative real-time PCR (p<0.05; Fig. 2C).

OPEN IN VIEWER

FIG. 2.

Comparison of in vitro osteogenic, adipogenic, and chondrogenic differentiation of hBMSCs and hPDLSCs, and the effect of WSC on these properties. (A) Mineralized nodules formed after 4 weeks of osteogenic induction (alizarin-red stain; magnificatio ×400, scale bar=100 μm). (B) The total area of mineralized nodules was significantly smaller for hBMSCs than for hPDLSCs (*p<0.05), and was increased by the application of WSC to hBMSCs and hPDLSCs (*p<0.05). (C) mRNA expression for the gene encoding runt-related transcription factor 2 (Runx-2) confirmed the mineralized nodule formation pattern (*p<0.05). (D) Lipid vacuoles were observed by Oil Red O staining after 4 weeks of adipogenic induction (magnification ×400, scale bar=100 μm). (E) The lipid vacuole formations were significantly lower for hPDLSCs than for hBMSCs (*p<0.05), and the treatment of WSC to MSCs was increased in lipid vacuole. (F) The adipogenic differentiation was revealed by the mRNA of expression of the gene encoding peroxisome proliferator-activated receptor γ2 (PPARγ2), the expression of which increased explosively in WSC-treated hBMSCs (*p<0.05). (G) The effect of WSC on MSC chondrogenesis was examined after 21 days and visualized by Alcian blue staining. The positively stained area was larger in WSC-treated hMSCs than in their nontreated counterparts (magnification ×1000, scale bar=50 μm). (H, I) The mRNA expressions of the genes encoding the transcription factor SOX-9 (SOX-9) and ACAN (cartilage-specific proteoglycan core protein; Aggrecan) were significantly greater in WSC-treated hBMSCs than in WSC-treated hPDLSCs (*p<0.05). Color images available online at www.liebertpub.com/tea

Adipogenic potential was also evaluated by measuring the area of stained lipid vesicles that formed after adipogenic stimulation (Fig. 2D). The total area of lipid vacuoles was significantly larger for hBMSCs than for hPDLSCs (p<0.05), and treatment of WSC further increased the area of lipid vacuoles in these two cell types (p<0.05; Fig. 2E). The expression of peroxisome proliferator-activated receptor γ2 (PPARγ2), which is considered a master gene for adipogenesis, was also observed, with an explosive increase in WSC-treated hBMSCs compared with WSC-treated hPDLSCs (p<0.05; Fig. 2F).

The degree of proteoglycan synthesis was greater for WSC-treated cells than for their nontreated counterparts after 21 days of chondrogenesis (Fig. 2G). Moreover, the expressions of SOX9 and aggrecan (ACAN) mRNA in hBMSCs and hPDLSCs under chondrogenic induction were significantly increased in the presence of WSC (p<0.05; Fig. 2H, I).

Comparison of the effect of WSC on in vitro collagen synthesis by hBMSCs and hPDLSCs

The Sircol assay for soluble and insoluble collagen revealed that, as shown previously, the amount of soluble collagen synthesized was significantly lower in WSC-treated hPDLSCs than in the nontreated group. ²⁸ In contrast, the synthesis of soluble collagen was increased in WSC-treated hBMSCs. The insoluble collagen deposits on the WSC-treated hBMSC and hPDLSC culture dishes were increased compared with their nontreated counterparts (p<0.05; Fig. 3A). The content of hydroxyproline exhibited a similar pattern as found for soluble and insoluble collagen by the Sircol assay (p<0.05; Fig. 3B). Each content was normalized according to the final cell number. The real-time PCR performed to determine the changes in collagen expression at the mRNA level revealed that the expressions of mRNA for collagen types I–V, HAS1, HAS2, and HAS3, and the LOX family were higher in WSC-treated hPDLSCs than in the nontreated group. In contrast, the mRNA expressions of collagen types I and IV were decreased in WSC-treated hBMSCs. In this study, the expressions of HAS2 and HAS3 were significantly greater in WSC-treated hBMSCs than in their nontreated counterparts. Treatment with WSC increased the expressions of LOX family genes in both hBMSCs and hPDLSCs (p<0.05; Fig. 4).

OPEN IN VIEWER

FIG. 3.

Enhanced collagen synthesis in WSC-treated hBMSCs compared with WSC-treated hPDLSCs in vitro. (A) The Sircol collagen detection assay revealed that the concentration of soluble collagen was significantly lower in WSC-treated hPDLSCs than in nontreated hPDLSCs. In contrast, the treatment of hBMSCs with WSC increased the formation of soluble collagen. After 7 days of incubation, the collagenous tissue on the bottom of the culture dishes was measured for both hMSC types. The amount of insoluble collagen deposition was increased with WSC treatment. (B) The pattern of hydroxyproline content reflected similar patterns of soluble and insoluble collagen levels found in the Sircol assay (*p<0.05).

OPEN IN VIEWER

FIG. 4.

Results of the real-time polymerase chain reaction (PCR) assay. The mRNA expressions of collagen types I–V, hyaluronic acid synthase 1 (HAS1), HAS2, HAS3, lysyl oxidase (LOX), LOX-like 1 (LOXL1), LOXL2, and LOXL3 were significantly enhanced by treatment of hPDLSCs with WSC. In contrast, the mRNA expressions of collagen types I and IV were decreased in WSC-treated hBMSCs. Each experiment was performed at least in triplicate (*p<0.05).

Comparison of the effect of WSC on the in vivo collagen tissue formation

Histology analysis was performed at 8 weeks after transplantation of either WSC-treated or nontreated MSCs. Each transplant maintained its shape, with no signs of inflammation at the point of injection or within the transplants after removal from the mice after 8 weeks (Fig. 5A). The explants from the WSC-treated groups proved to be slightly heavier than those from the nontreated groups, and the difference was not statistically significant (Fig. 5B). Masson's trichrome staining revealed that the amount and density of collagen fibers and collagenous ECM synthesized by the WSC-treated hBMSCs and hPDLSCs were significantly greater than those synthesized by their nontreated counterparts (p<0.05; Fig. 5C, D).

OPEN IN VIEWER

FIG. 5.

Regeneration of collagen fibers by WSC-treated hBMSCs at 8 weeks post-transplantation compared with WSC-treated hPDLSCs. (A) Implants from mice at 8 weeks post-transplantation. (B) The weight of the explants is illustrated. Although explants from WSC-treated groups appeared to be both heavier and harder for both hBMSCs and hPDLSCs, there were no significant differences between any of the groups. (C) The area of collagen fibers. (D) Masson's trichrome staining revealed that significantly more collagenous tissues (lower, black arrowheads) were generated after treating MSCs with WSC. (upper row: magnification ×200, scale bar=200 μm; lower row: magnification ×400, scale bar=100 μm; inset box in lower-magnification view indicates area shown at higher magnification; *p<0.05). The large red dots are diethylaminoethyl cellulose (DEAE) sephadex A 25, which is one of the HA filler constructs to maintain structures. Color images available online at www.liebertpub.com/tea

Quantitative analysis of the collagen synthesized by the MSCs was achieved via immunohistochemistry. Confirmation that the synthesized tissues originated from the transplanted hBMSCs and hPDLSCs and not from host cells came from the finding that the explant tissue in all except the HA carrier-only (negative control) group was stained darkly for hMito. Exposure of the newly formed collagen tissue to types I and III anticollagen antibodies in vivo revealed that the cells in the WSC-treated groups were more positively stained than those in the nontreated groups. The expression pattern of hydroxyproline was similar to that of collagen types I and III, and it was expressed along the newly formed collagen tissue (Fig. 6A). The hPDLSCs in the WSC-treated group were more positively stained than the nontreated group as previously shown. ²⁸ However, the number of collagen type I-positive cells was decreased in WSC-treated hBMSCs (Fig. 6B), which was corroborated by the in vitro data of real-time PCR.

OPEN IN VIEWER

FIG. 6.

Immunohistochemistry results for collagen types I and III, and hydroxyproline. (A) Immunohistochemistry with an anti-human-specific mitochondrial antibody confirmed that the regenerated collagenous tissues originated from the newly transplanted human cells (black arrowheads). WSC-treated groups showed more positively stained area (black arrowheads) than nontreated groups against type I and III collagens and hydroxyproline (magnification ×400, scale bar=100 μm). (B) The hPDLSCs in the WSC-treated group were more positively stained than the nontreated group. In contrast, the number of collagen type I-positive cells was decreased in WSC-treated hBMSCs (*p<0.05), which was corroborated by the in vitro data of real-time PCR. Color images available online at www.liebertpub.com/tea