Crosstalk on Cell Behavior in Interactive Cocultures of hMSCs with Various Oral Cell Types

Abstract

When prospectively applied for regenerative therapies, human bone-marrow-derived mesenchymal stem cells (hMSCs) interact with the locally residing host cells. With respect to the developmentally particular origin of oral cells, little is known about the putatively discriminative behavioral responses of hMSCs in interaction with various oral cell types, including human alveolar bone osteoblasts (hOAs), periodontal ligament fibroblasts (hPDLs), and gingival fibroblasts (hGFs). To assess the crosstalk between hMSCs and oral cells, interactive cocultures were established by combining well-characterized hMSCs with hOAs, hPDLs, or hGFs, and the behavioral hMSC aspects, that is, proliferation and gene expression, were measured by employing a 5-bromo-2′-deoxyuridine assay and real-time polymerase chain reaction, while apoptosis was quantified by in situ cell death detection kit. hMSCs expressed the typical antigen spectrum lacking CD34, CD45, CD14, CD19, and HLA-DR, while expressing CD73, CD90, and CD105, and could successfully be transformed into adipocytes, osteocytes, and chondrocytes. Monocultured control hMSCs proliferated readily, whereas a general reduction of BrdU-labeled cells was observed in cocultures. Globally, upon extending time periods, interactive coculture combinations of hMSCs with hOAs reduced both osteogenic gene and stem cell marker transcription in hMSCs, a phenomenon appearing less pronounced by combining hMSCs with hPDLs, such that the observed effects in terms of proliferation and gene expression followed the same ranking: hOAs>hGFs>hPDLs. Vice versa, in interactive hMSC cocultures, the cell survival rate was significantly increased, irrespective from the combined coculture cell counterpart. Our results show for the first time that behavior of hMSCs reflected by proliferation and gene expression was governed by interaction with various oral cells in a cell-type-discriminative manner. In addition, hMSC coculture restrains apoptosis, such that influences on cell behavior appear as a crosstalk. In summary, interactive cocultures render the basis for a prospective prediction of mutual cell behavior in hMSC-based oral tissue regeneration disclosing that oral cells shift hMSC behavior from proliferation to differentiation and apoptosis-repressing features.

Introduction

Although intense interdisciplinary working strategies have been developed in regenerative medicine during the last decades, their outcome still seeks improvement. In this context, bone-marrow-derived mesenchymal stem cells (MSCs) symbolize the dawn of clinical regeneration and many exciting findings have already been obtained following MSC implantation in animal models.^1,2 However, the general low survival and poor engraftment rate observed in MSC-based therapies³ are in strong contrast to the paradigm that regeneration merely relies on direct replacement of injured cells by transplanted MSCs.⁴ Nevertheless, the concept that the employment of MSCs promotes the global regeneration outcome is widely accepted, such that for example, MSCs are already engaged by default in regenerating the damaged periodontium.^5–8 These findings are not surprising because MSCs are attractive candidates for hard and soft tissue regeneration due to their inherent mesenchymal nature. Nevertheless, little is known about the precise events of MSC action in this complex multi-tissue environment resulting in new functional tooth attachment. This issue is captivating because oral tissues are suspected to behave exceptionally due to their particular developmental origin: Being ectomesenchyme derivates, oral cells differ completely from cells derived from other body's regions,⁹ such that MSCs may react distinctly to the interaction with oral cells. Regarding oral tissues, especially periodontal regeneration has an enormous clinical significance because periodontitis is widely regarded as the second most common disease worldwide and the most common cause for tooth loss in adults. Thus, our study aimed to shed light on the molecular MSC actions in an oral tissue environment, and to this purpose, we generated interactive coculture systems comprising human MSCs (hMSCs) and various human oral cell types, that is, osteoblasts from the alveolar bone (hOAs), periodontal ligament cells (hPDLs), and gingival fibroblasts (hGFs). Interactive coculture systems are useful tools for the in-depth investigation of cellular crosstalk effects by mimicking tissue-specific cell communication. One of the most prominent examples for the meaningfulness of interactive coculture systems is reflected by the requirement of epithelium–mesenchyme interactions being mandatory for the regular in vitro development of stratified epithelia.¹⁰ In this light, interactive coculture systems provide the opportunity for evaluating the reciprocal influence of hMSCs and the distinct oral cells in real time, thereby having the edge on other in vitro systems that are based on monocultured hMSCs feeded by conditioned media of noncontrollable composition.^11–14 To our best knowledge, this is the first report using interactive coculture systems for a systematic comparison of the discriminative effects of different oral cell types that contribute to the establishment of the hMSC niche when prospectively employed in regenerative therapies. In addition, this study evaluates for the first time the decisive effects of oral cells on hMSC differentiation, which is considered as a complex orchestration of developmental steps closely linked to the nature of their interaction partners. In this light, our study aimed to assess the hMSC performance in dependence on various oral cell types, and putative mutual influences in terms of basic parameters of cell behavior: We asked (1) whether and how divergent oral cell types altered the hMSC proliferation capacity and gene expression and (2) what might be the final outcome in interactive hMSC–oral cell cocultures with regard to global cell survival, as reflected by apoptosis testing.

Materials and Methods

All chemicals were purchased from Sigma Aldrich GmbH, Munich, Germany, unless otherwise indicated. For a detailed description of the experiments performed, see the “Supplementary Materials and Methods” section.

Cell isolation and culture

All experiments have been carried out in accordance to the guidelines of the World Medical Association Declaration of Helsinki and were approved by the Committee of Ethics of the Medical Faculty of the Albert Ludwigs-University Freiburg, Germany (EK-199/09). Primary hMSCs were purchased (PromoCell, Heidelberg, Germany) at passage 2; cultivated in hMSC proliferation medium supplemented with 10% fetal calf serum, 50 ng/mL amphotericin, and 50 μg/mL gentamicin (all from Provitro, Berlin, Germany); and passaged for four to six additional times.

For each oral tissue, three primary cultures derived from operative biopsies of four individual patients (n=12 in total) were established using the explant technique.¹⁵ In brief, tissue specimens were painstakingly separated from contaminating tissue fragments if present, minced, plated as explants, and kept under standard culture conditions in low-glucose Dulbecco's modified Eagle's medium (containing 25 mM HEPES and L-glutamine; PAA, Pasching, Austria) supplemented with 10% fetal bovine serum (FBS; Biochrom, Berlin, Germany), 1% kanamycin, and 1% glutamax (Invitrogen, Darmstadt, Germany) until cell outgrowth. Due to small tissue specimen quantities, the outgrowing cells were expanded in number by subsequent passaging (hOAs: 5–7 passages; hPDLs: 7–12 passages; and hGFs: 7–9 passages) and their ability to synthesize mineralized matrix was routinely checked thereafter.^16,17

Interactive coculture systems

For the generation of interactive coculture systems, hMSCs were established in 96-well plates at a density of 0.15×10⁵ cells/cm² for proliferation assessment and in 12-well plates at a density of 0.25×10⁵ cells/cm² for gene expression analysis. The oral cells were inoculated at 5×10³ cells/96-well insert or 4×10⁵ cells/12-well insert in compatible culture inserts with a porous uncoated poly (ethylene terephthalate) (PET) membrane (3.0 μm), thus facilitating their interactive crosstalk with hMSCs for the respective culture periods and the subsequent examination of BrdU incorporation or mRNA expression in hMSCs devoid of oral cell BrdU or mRNA contamination, respectively. For the generation of cell-containing matrices, hMSCs were suspended in FBS and added to a collagen I solution (4 mg/mL; Life Technologies, Darmstad, Germany) supplemented with 10% Hank's buffered salt saline. The resulting gel was titrated with sodium hydroxide to pH 7.4 and polymerized for 2 h at 37°C, covered with culture medium, supplied with oral cells on the gel surface after 24 h, and lifted to 1.0 μm PET membrane cell culture inserts (Millipore, Schwalbach, Germany) after 48 h in order to guarantee an optimal all-side culture medium nutriment supply during a culture period of 21 days. All cell cultures of each specimen and combination were run in duplicate for all experiments.

Basic cell characterization

Differentiation assays

Medium supplements were used to induce the differentiation of hMSCs into osteoblasts, adipocytes, or chondrocytes. The hMSC-derived adipocytes' lipid vesicles were visualized by Oil Red-O staining and indirect immunofluorescence labeling of lipid-droplet-associated perilipin; the amount of extracellular matrix calcium deposits induced during the osteogenic differentiation was assessed by von Kossa and Alizarin red staining; and the chondrogenic differentiation was evaluated by Safranin-O staining and indirect immunofluorescence labeling of cartilage-specific proteoglycan aggrecan according to established protocols. In addition, the hMSC differentiation was verified by quantitative polymerase chain reaction (qPCR) (for further details, see the “Supplementary Materials and Methods” section).

Flow cytometry

For each run, a total of 1×10⁶ cells were incubated with a final concentration of 5 μg/mL each of fluorescein isothiocyanate–, phycoerythrin-, or allophycocyanin-conjugated mouse anti-human CD34, CD45, CD14, CD19, HLA-DR, CD73, CD90, CD105, and CD166 antibodies (BD Biosciences, Heidelberg, Germany), or isotype-matched controls, respectively (20 min, 4°C in the dark). After intense washing and resuspending in phosphate-buffered saline (PBS), flow cytometric analysis was performed using an FACSCalibur (BD Biosciences) flow cytometer. Data were collected and analyzed with CellQuest software (BD Biosciences). For each run, 10,000 cell events were gated and fluorochrome spectral overlap was checked and compensated whenever required.

Proliferation assay

The BrdU incorporation into newly synthesized DNA in interactive cocultures was measured using a cell proliferation enzyme-linked immunosorbent assay (ELISA) kit (Roche, Mannheim, Germany) according to manufacturer's instructions. In brief, cells were fixed and DNA was denatured with a one-step ready-to-use solution supplied by the kit for 30 min at room temperature (RT). After thoroughly removing the solution, a peroxidase-labeled anti-BrdU antibody was added (1:100, 90 min, RT). The immune complexes were detected by adding tetramethyl benzidine (30 min, RT), the substrate reaction was stopped by adding 1 M sulfuric acid (1:5), and the reaction product was quantified by immediately measuring the absorbance at 450 nm (reference wavelength 690 nm).

Quantitative real-time PCR

After accurately washing the cells with PBS, total cellular RNA was purified separately from both coculture layers using a guanidium–thiocyanate method (RNeasy Mini kit; Qiagen, Hilden, Germany) and stored at −80°C. Genomic DNA contamination was eliminated by DNase digestion according to manufacturer's instructions (RNase free DNase kit; Qiagen). The RNA integrity and quantity were verified using the Experion RNA StdSens chip microfluidic technology and cDNA was synthesized from 500 ng of total RNA by using the iScript reaction mix (Bio-Rad, München, Germany). For real-time PCR, cDNA samples were amplified in duplicate in a 25 μL reaction mixture using pre-validated RT² qPCR primer assays (SABiosciences, Frederick, MD) with a iQ-SYBR Green Super Mix (Bio-Rad). Data were collected and analyzed using the CFX96 Manager Software version 1.0 (Bio-Rad).

Cell survival assay

For in situ cell death detection, cryosections from each interactive coculture gel were stained by using an in situ apoptosis detection kit (Roche). Briefly, DNA strand breaks of apoptotic cells were visualized by terminal deoxynucleotidyl transferase (TdT) dUTP nick end labeling (TUNEL reaction). Specimens were fixed, permeabilized, and incubated with TdT-label solution for 60 min at 37°C in the dark. After counterstaining with 4′,6-diamidin-2-phenylindole (DAPI), specimens were mounted (Fluoromount G; Southern Biotech, Birmingham, AL) and photographed and analyzed with a Biozero BZ-9000 fluorescence microscope and software (BZ-II Analyzer software; Keyence Corp., Neu-Isenburg, Germany).

Statistical analysis

To evaluate the effect of groups and time, a repeated measures analysis was performed for each outcome of interest. To this purpose, one linear model per outcome was fitted within which the structure of data, that is, repeated measurements per group (at several time points), was accounted for. Group effects and differences of least squares mean (LSM) were calculated with their 95% confidence intervals. Multiple comparisons of LSM in group and time combinations were performed and accordingly adjusted p-values were considered significant when p<0.05. All calculations were performed with the statistical software SAS 9.1.2. using PROC MIXED.

Results

hMSCs fulfill the standard characterization criteria

Dealing with MSCs, the current terminology is confusingly heterogeneous.¹⁸ Hence, we employed the standard hMSC characterization criteria of the International Society for Cellular Therapy for hMSCs that were subsequently employed for cocultures (Supplementary Figs. 1A, B and 2A, B; Supplementary Data are available online at www.liebertpub.com/tea).

Expression of surface markers

Flow cytometry screening of four to six times passaged hMSCs revealed that the cells were reasonably homogeneous in size and granularity and positively expressed CD73, CD90, CD105, and CD166, but were neither labeled for CD34 and CD45 (Supplementary Fig. 1A) nor for CD14, CD19, and HLA-DR (Supplementary Fig. 2B), reflecting the typical spectrum of hMSC surface molecules.

Plastic adherence and colony formation

The hMSCs were supplied as adherent cells within a cell culture device and did not hesitate to adopt a polymorphous and elongated polygonal shape and to proliferate well, while emanating from cell colonies at early stages (Supplementary Fig. 1B).

Multilineage differentiation

The hMSCs were successfully transformed into adipocytes, osteoblasts, or chondrocytes. After 21 days of differentiation induction, the abundant number of hMSC-derived adipocytes was easy to identify by Oil Red-O stains of their huge lipid droplets (Supplementary Fig. 1B), quantification of peroxisome-proliferator-activated receptor-γ mRNA (Supplementary Fig. 2A1, A2), and staining of lipid-droplet-associated protein perilipin (Supplementary Fig. 2A3). When induced by transforming growth factor-β, hMSCs expressed remarkable quantities of collagen 2a1 mRNA (Supplementary Fig. 2B1, B2) and their extracellular matrix comprised acid mucopolysaccharides as revealed by Safranin-O and indirect aggrecan immunofluorescence stains (Supplementary Figs. 1B and 2B3), indicating a chondrogenic differentiation. Osteogenic induction resulted in higher amounts of osteopontin (OPN) mRNA expression (Supplementary Fig. 2C1, C2) and huge amounts of extracellular matrix calcium deposits visualized by von Kossa and Alizarin red staining (Supplementary Figs. 1B and 2C3).

Interactive oral cell coculture seriously restricts the proliferation of hMSCs

Behavioral cell reactions to environmental stimuli appear in multiple ways. With respect to the potential in vivo interaction with oral tissue cells, we therefore addressed the proliferation and gene transcription of hMSCs upon interactive coculture with hOAs, hPDLs, and hGFs.

The baseline proliferation of control hMSCs was determined by means of measuring BrdU incorporation at 1, 3, 7, 10, and 14 days of culture. We found that proliferation in solitary grown control hMSCs appeared in a peak-to-valley manner. Herein, compared with day 1, the BrdU incorporation was significantly increased at days 3 and 7 (Fig. 1, solid line), while being declined at days 10 and 14. This hMSC proliferation dynamics suggests a progressive decline in mitotic activity with extending culture periods.

FIG. 1.

Proliferation dynamics of human bone-marrow-derived mesenchymal stem cells (hMSCs) in response to interactive oral cell coculture. Line graphs showing the least squares mean bromodeoxyuridine (BrdU) incorporation with 95% confidence intervals measured as optical density units (y-axis, 450 nm: measuring wavelength, 690 nm: reference wavelength) of control hMSCs (black dots, connected by a black line) and hMSCs cocultured with (A) human alveolar bone osteoblasts (hOAs), (B) human periodontal ligament fibroblasts (hPDLs), and (C) human gingival fibroblasts (hGFs) (each white quadrangles, connected by a dashed line) at different time points (x-axis). In control hMSCs, the BrdU uptake significantly increased until day 3 before it significantly decreased until day 14. (A) Interactive hOA coculture yielded a fairly constant BrdU signal such that the proliferation was significantly lower than in control hMSCs at days 3 and 7. (B) hMSC proliferation was diminished similarly in interaction with hPDLs, and even more pronouncedly with hGFs (C). Intergroup differences in mean BrdU incorporation were considered significant at *p<0.05 and **p<0.01.

Interactive coculture with oral cells generally yielded a striking reduction of hMSC proliferation from the beginning. In detail, in hOA-cocultured hMSCs, the mean BrdU incorporation was significantly reduced at all time points under study (Fig. 1, dashed line), thereby completely leveling out the peak-to-valley dynamics previously described for control hMSCs. An interesting issue concerns the interculture mode comparison, revealing that in cocultures, hMSCs at days 3 and 7 displayed significantly lower levels of proliferating cells [Fig. 1; compare white rectangles (hOA/hMSC cocultures) with black dots (control hMSCs), at days 3 and 7]. This finding may indicate a restriction in hMSC proliferation, due to coculture-innate cell-to-cell interactions. Regarding the obvious proliferation-restrictive effect of a cocultured oral cell counterpart, hMSC proliferation was diminished fairly similarly in interaction with hPDLs (Fig. 1B), and even more pronounced with hGFs (Fig. 1C; compare white rectangles with black dots in Fig. 1B, C, respectively). This pattern points to a ranking in the proliferation restriction, depending on the type respective cocultured cell type by prioritizing hOAs, followed by hGFs and hPDLs.

Interactive coculture with oral cells influences the gene expression in hMSCs

Keeping in mind that oral cells affected hMSC proliferation, we next assessed the interactive coculture impact on hMSC gene expression. For this purpose, the expression of a panel of stemness- and osteogenic differentiation-associated genes was monitored at baseline, day 7, and day 14. Similar to the time course of BrdU incorporation, the gene expression in control hMSCs globally appeared in a peak-to-valley manner or remained unchanged after day 7.

hOA coculture rectifies and hPDL coculture restores the transcription of stemness-associated genes in hMSCs

The baseline transcription of the stemness-associated genes under study tended to decrease with prolonged culture period in solitary grown control hMSCs. Interactive coculture yielded a striking modification of the gene expression gradient, as shown in Figure 2. The most interesting findings with matched controls were that interactive hOA coculture displayed an alternative gene expression slope with respect to stemness-associated gene expression, that is, declining in a continuously linear manner. By contrast, interactive hPDL coculture provoked a peak-to-valley pattern that was even more pronounced than in control hMSCs, while hGF coculture yielded a gene expression pattern similar to control.

FIG. 2.

Expression of stem-cell-associated mRNA in hMSCs relative to interactive oral cell coculture. Least squares mean ΔCt values with 95% confidence interval of control hMSCs (top row) and of interactive hOA (second row), hPDL (third row), and hGF (bottom row) cocultures, respectively, each at day 0 (baseline, represented by a dotted line), day 7, and day 14 (right y-axis, gradient represented by dashed line). The Ct values of the respective gene of interest [GOI: A, POU domain class 5 transcription factor 1 (POU5F1); B, transcription factor homeobox protein Nanog (Nanog); C, ATP-binding cassette sub-family G member 2 (ABCG2); D, tyrosine protein kinase kit (c-kit); E, leukemia inhibitory factor (LIF)] were normalized to the corresponding housekeeping gene β-actin (x-axis: ΔCt_GOI-bAct). In consequence, right-shifted ΔCt values symbolize a lower gene expression. The overall expression of stem-cell-associated mRNAs was significantly downregulated in control hMSCs. Coculture with hOAs yielded a constant downregulation progress while hPDL coculture rather supported a peak-to-valley gene expression progress finally resulting in a reinforced transcription level. The gene transcription in hMSCs cocultured with hGFs resembled the gene expression in control hMSCs. (A) The expression of POU5F1 was continuously but nevertheless significantly downregulated in interactive hOA cocultures (p=0.0003; baseline vs. day 14), but restored in hPDL cocultures. (B) Control hMSCs expressed relatively stable levels of Nanog, while it was decreased by trend in hOA and hGF cocultures, and remarkably increased in hPDL cocultures (p=0.0234; baseline vs. day 14). (C, D) Only interactive hOA coculture yielded a significant downregulation of ABCG2 (p=0.0249) and c-kit expression (p=0.0004; baseline vs. day 14) by leveling out the peak-to-valley expression pattern. (E) Interactive hGF coculture yielded the strongest downregulation of LIF expression (p=0.0030; baseline vs. day 14).

In detail, the expression of POU domain class 5 transcription factor 1 (POU5F1) was significantly lowered at days 7 and 14 in control hMSCs (Fig. 2A). Interactive hOA coculture entailed a delayed (day 7), but finally significant downregulation of the POU5F1 expression (Fig. 2A), while the latter appeared recovered in hPDL coculture (Fig. 2A), but similar to control in hGF coculture (Fig. 2A).

Transcription factor homeobox protein Nanog remained quite constant in control hMSCs (Fig. 2B). In interactive hOA coculture, Nanog expression was downregulated by trend, while in hPDL cultures it appeared significantly increased at day 14, and hGF coculture trended to augment Nanog transiently (Fig. 2B).

The downregulation of ATP-binding cassette sub-family G member 2 (ABCG2) expression appeared in outlines in control hMSCs (Fig. 2C). Only interactive hOA coculture yielded a significant downregulation of ABCG2 transcription at day 14, during a linear gene expression decrease (Fig. 2C). In contrast, interactive hPDL coculture almost recovered ABCG2 expression (Fig. 2C), while it was fairly unaltered in hGF coculture (Fig. 2C).

Stem cell growth factor receptor, also known as tyrosine protein kinase kit (c-kit), was expressed in a peak-to-valley manner being significantly downshifted at day 7 but reinforced at day 14 in control hMSCs (Fig. 2D). Interactive hOA coculture showed abolishment of this alternating trend, by yielding a continuously significant c-kit expression decrease, such that the interculture comparison resulted in a significant diminishment (Fig. 2D). In interactive hPDL culture, peak-to-valley pattern of c-kit appeared re-intensified (Fig. 2D), while hGFs as interactive partner resulted in significant c-kit expression decrease (Fig. 2D).

Similar to POU5F1, the transcription of leukemia inhibitory factor (LIF) was significantly decreased in control hMSCs at days 7 and 14 (Fig. 2E). Precisely, in interaction with hOAs, significant downshifting of LIF expression was prevented, while we observed significant LIF downregulation in hPDL coculture at day 7 and a continuous and the overall strongest LIF downregulation in hGF coculture (Fig. 2E).

These findings may indicate an alteration in the hMSC gene expression profile, due to coculture-dependent cell-to-cell interactions. Intriguingly, the mode of transformation in gene expression of the respective cocultured oral cell type implicates a ranking in gene expression modification with emphasis on hOAs, followed by hGFs and finally hPDLs.

Interactive hOA coculture decreases the expression of osteogenic genes in hMSCs

The transcription of genes correlated with an osteogenic differentiation was either stable or increased with prolonged culture period in solitary grown control hMSCs. Therefore, we were interested in seeking for putative gene transcription modulation under interactive coculture conditions. In fact, we found that the earlier-mentioned trend was mostly counteracted by interactive hOA coculture, which interestingly tended to decrease the osteogenic gene expression, while hPDL and hGF cocultures yielded a virtually unaltered gene expression profile if compared with control hMSCs.

The runt-related transcription factor 2 (RUNX2) was expressed quite constantly with regard to culture duration with a subtle increase at day 7 (Fig. 3A). The interaction with hOAs resulted in a significantly decreased RUNX2 expression (Fig. 3A), while interactive hPDL coculture yielded a marginal increase in RUNX2 expression. Similarly, we observed a transient augmentation in RUNX2 transcription upon hGF coculture (Fig. 3A).

FIG. 3.

Expression of differentiation-associated mRNA in hMSCs interacting with oral cells. Least squares mean ΔCt values with 95% confidence intervals of control hMSCs (top row) and of interactive hOA (second row), hPDL (third row), and hGF (bottom row) cocultures, respectively, each at day 0 (baseline, represented by dotted line), day 7, and day 14 (right y-axis, gradient represented by dashed line). The Ct values of the respective gene of interest [GOI: A, runt-related transcription factor 2 (RUNX2); B, osteopontin (OPN); C, secreted protein acidic and rich in cysteine (SPARC); D, collagen 1 alpha 1 (Col1a1)] were normalized to the corresponding housekeeping gene β-actin (x-axis: ΔCt_GOI-bAct). (A, B) In control hMSCs, RUNX2 and OPN mRNA was quite constantly expressed. Interactive hOA coculture decreased the gene transcription of RUNX2 (p=0.0406; baseline vs. day 14) and OPN (p=0.0416; baseline vs. day 14). In contrast, interactive hPDL and hGF cocultures emphasized the peak-to-valley expression progress of RUNX2 and OPN, respectively. (C, D) SPARC (p<0.0001; baseline vs. day 14) and Col1a1 (p<0.0001; baseline vs. day 14) expression was strongly upregulated in control hMSCs, and interactive coculture effects were only marginal such that the respective gene expression profile resembled control hMSCs.

In control hMSCs, OPN followed a comparable expression time course, showing a gentle and transient increase at day 7 (Fig. 3B). In interaction with hOAs, OPN expression was significantly reduced (Fig. 3B), while coculture neither with hPDLs nor with hGFs remarkably modulated gene expression with matched controls (Fig. 3B).

The expression of secreted protein acidic and rich in cysteine (SPARC) was tremendously increased upon extending culture period in control hMSCs (Fig. 3C). Similar to controls, a significant SPARC increase was detectable in interactive coculture with each hOAs, hPDLs, and hGFs (Fig. 3C) such that it globally failed to strengthen the already strong increase in SPARC expression (Fig. 3C).

Accordingly, collagen 1 alpha 1 (Col1a1) was immensely upregulated in control hMSCs with a strong expression at day 7 (Fig. 3D). This trend was slightly reduced in presence of hOAs as an interactive partner (Fig. 3D), whereas the increase in Col1a1 transcription remained fairly unaffected in hPDL or hGF cocultures (Fig. 3D).

Interactive hMSC cocultures display increase in cell survival rate

Regarding tissue regeneration, it cannot be excluded that supportive effects are reflected by modulation of proliferation and differentiation, but also by an improvement of cell survival. We therefore analyzed apoptosis in solitary oral cell cultures and interactive hMSC cocultures. Taking into account the number of apoptotic events, we observed that irrespective of the oral cell type involved, interactive hMSC coculture significantly increased the cell survival ratio as represented by the ratio total cell number per apoptotic cell nuclei in comparison to hMSC-lacking controls (for comparison: Fig. 4A1 with 4A2, Supplementary Figure 3A1 with 3A2, 3B1 with 3B2, for hOAs; Fig. 4A3 with 4A4, Supplementary Figure 3A3 with 3A4, 3B3 with 3B4, for hPDLs; and Fig. 4A5 with 4A6, Supplementary Figure 3A5 with 3A6, 3B5 with 3B6, for hGFs). In detail, cell survival rates are illustrated in Figure 4B and summarized in Figure 4C. This finding on augmentation of cell survival, obtained from interactive cocultures, points to a possible contribution of hMSCs to an improvement of oral cell survival, or vice versa, concerning putative oral tissue regeneration strategies.

FIG. 4.

Cell survival ratio in interactive oral cell cocultures, as shown by (A) representative immunofluorescence stains of total and apoptotic nuclei, (B) least squares mean cell survival rates with 95% confidence intervals (CIs), and (C) absolute counts each in interactive hOA, hPDL, or hGF cocultures, respectively, in comparison to their solitary grown hOA, hPDL, and hGF control counterparts. Apoptotic events (green, highlighted by white arrows) and total cell nuclei (blue) were visualized in gel sections consisting of (A₁) hOAs combined with hMSCs, (A₂) hOAs only, (A₃) hPDLs in interaction with hMSCs, (A₄) hPDLs only, (A₅) hGFs with hMSCs, and (A₆) hGFs only. As a general result, hMSC cocultures contained only a tiny number of apoptotic events that were scattered throughout the tissue section. In contrast, hMSC-lacking control gels displayed a remarkably higher number of apoptotic cells, such that the cell survival ratio of total nuclei per apoptotic nuclei differed significantly compared with interactive cocultures (B, C). (B) Top row: hOA controls and interactive hOA–hMSC coculture; middle row: hPDL controls and interactive hPDL–hMSC cocultures; bottom row: hGF controls and interactive hGF–hMSC cocultures; x-axis: cell survival rate as calculated by the ratio of total per apoptotic events. (C) Least square means with 95% CIs of total cell count, apoptotic events, and cell survival ratios, which were compared between interactive oral cell–hMSC cocultures and their respective controls by repeated measures analysis and post-test p-value adjustment by the method of Benjamini–Hochberg.

Discussion

In oral regenerative medicine, stem cells face two unique challenges: first, the developmental exceptionality of oral tissues, and second, the necessity to rebuild a complex arrangement of different soft and hard tissue components.¹⁹ In this context, the periodontium represents an ideal model for evaluating the diverse responses of stem cells to the various oral tissue components that originate from neural crest cell induction and hence, from ectodermal–mesenchymal cell interactions.⁹ hMSCs are promising key candidates for periodontal regeneration purposes.^5,7,20,21 Thus, our study has investigated for the first time the interplay of hMSCs with different oral cell types by means of an innovative in vitro approach, based on interactive cocultures facilitating physiologically relevant nondirect contact-based cell-to-cell interactions, also occurring in vivo.

Given that hOAs, hPDLs, and hGFs are derived from distinct provenances as hard or soft tissues, it seems just a logical consequence that hMSCs may react differentially to the diverse cell types. In fact, we have unequivocally shown that hMSCs display discriminative responses to oral cell coculture, thereby in turn allowing for a systematic comparison of oral tissue cells. Interestingly, the performance of hMSCs cocultured with oral cells followed a ranking that was the same for all parameters of hMSC behavior under study, that is, proliferation, gene expression, and apoptosis. The peak-to-valley behavior of control hMSCs was leveled out by hOA coculture, reduced by hGF coculture, and remained fairly unaffected or even re-intensified by hPDLs.

With regard to proliferation, expected growth curves were obtained for control hMSCs, including a strong initial increase and a peak reflecting the transition from a stationary phase to a final decrease. These growth phases are characterized by increasing cell number, and a subsequently ceasing cell growth due to contact inhibition by monolayer formation.^22,23 Since all behavioral hMSC features were modified by oral cells following the same hierarchy, we hypothesize that oral-cell-induced decrease of hMSC proliferation reflects a shift of the hMSC behavior from proliferation to differentiation and environment-supporting actions. Nevertheless, the finding that oral cells reduced the proliferation of hMSCs seems to contrast recent evidence that long-bone osteoblasts^11,24 or hPDLs¹⁴ promote the proliferation of cocultured hMSCs in vitro. However, the reported increase in hMSC proliferation occurred either in response to conditioned medium or to direct cell–cell contacts, which were not existent in our experimental set-up. We speculate that our observations in proliferation may be a result of the above-mentioned more physiological interactions, provided by our coculture approach. In addition, our findings substantiate the hypothesis that hMSCs do react differently in response to femoral than to alveolar osteoblasts.⁹ With respect to connective tissue fibroblasts, as for instance hGFs, their suitability as feeder layer for embryonic stem cells is known for a long time,²⁵ and here we show that they render the same service to hMSCs.

It is tempting to assume that a reduced proliferation comes along with an emphasis in differentiation. In fact, we found that the expression of differentiation-associated genes increased with time in a peak-to-valley manner, in analogy to the observed proliferation profile. Our data clearly show that various oral cells affect the gene transcription in hMSCs in a cell-type-distinctive manner. This distinction is substantiated by hOAs as the main proliferation modulator, which turned out to exert the greatest effects on the expression of hMSC genes under study, followed by hGFs and hPDLs. Nevertheless, hOA-cocultured hMSCs reacted differently from the way their mesoderm-derived counterparts were so far reported, and thus, expected to do.^12,24,26 This was reflected by the finding that transcription of osteogenic genes was not enhanced, but rather decreased in comparison to control hMSCs. Although we admittedly cannot rule out that the expression of further differentiation-associated genes was augmented, we may speculate that the differing reaction relies on the divergent developmental biology nature of alveolar compared with femoral osteoblasts.^9,27,28

These findings may lead to the assumption that in interplay with different oral cells, hMSCs may contribute to tissue regeneration by innate features other than enhancement of proliferation and extensive reinforcement of differentiation. Indeed, we could show that apoptotic events appeared significantly reduced when hMSCs were available as interactive cell counterpart for any of the cocultured oral cell type. This observation strongly suggests that the improvement of general cell survival, including the addressed tissue, may contribute to the supportive effects of hMSCs in the context of tissue regeneration.^29–31

In summary, our results show that interactive hMSC cocultures with various oral cells are characterized by a crosstalk, gradually regulated by the oral cell type. This crosstalk is ratified by government of proliferation and differentiation in hMSCs, and improvement of cell survival, as indicated by global apoptosis reduction. In mechanistic terms, this mutually displayed influence on behavioral cell features under study may be orchestrated by nondirect contact-innate cell-to-cell interactions, provided in our coculture approach. Thus, our results valuably contribute to the understanding of the putative cellular processes during hMSC-based oral tissue regeneration and explicitly outline the urgent need for fine tuning the complex interactions of the different cell types involved. Such knowledge will be precious for the future identification of hMSC-customized therapy concepts, including interactive tissue-based concepts.

Footnotes

Acknowledgments

The authors acknowledge the excellent technical assistance of Yrgalem Abreha and Heike Jahnke who supported the cell culturing and staining. We are grateful to Dr. Brigitte Altmann for fruitful discussions and technical advice. The study was funded by the Research Committee of the Medical Faculty of the University Freiburg Medical Center (PRO723/09) and by the Deutsche Gesellschaft für Zahn-, Mund- und Kieferheilkunde (DGZMK, 11/2010).

Disclosure Statement

The authors declare that they have no competing interests.

References

Jones

, Yang

Mesenchymal stem cells and bone regeneration: current status. Injury, 42:562. 2011.

Koob

, Torio-Padron

, Stark

G.B.

, Hannig

, Stankovic

, Finkenzeller

Bone formation and neovascularization mediated by mesenchymal stem cells and endothelial cells in critical-sized calvarial defects. Tissue Eng Part A, 17:311. 2011.

Satija

N.K.

, Singh

V.K.

, Verma

Y.K.

, Gupta

, Sharma

, Afrin

, Sharma

, Tripathi

R.P.

, Gurudutta

G.U.

Mesenchymal stem cell-based therapy: a new paradigm in regenerative medicine. J Cell Mol Med, 13:4385. 2009.

Bunting

K.D.

, Hawley

R.G.

Integrative molecular and developmental biology of adult stem cells. Biol Cell, 95:563. 2003.

Yang

, Rossi

F.M.

, Putnins

E.E.

Periodontal regeneration using engineered bone marrow mesenchymal stromal cells. Biomaterials, 31:8574. 2010.

, Yan

, Lei

, Li

, Xiao

Application of autologous cryopreserved bone marrow mesenchymal stem cells for periodontal regeneration in dogs. Cells Tissues Organs, 190:94. 2009.

Kawaguchi

, Hirachi

, Hasegawa

, Iwata

, Hamaguchi

, Shiba

, Takata

, Kato

, Kurihara

Enhancement of periodontal tissue regeneration by transplantation of bone marrow mesenchymal stem cells. J Periodontol, 75:1281. 2004.

Weng

, Wang

, Liu

, Hu

, Chai

, Yan

, Zhu

, Cui

, Cao

Repair of experimental alveolar bone defects by tissue-engineered bone. Tissue Eng, 12:1503. 2006.

Luan

, Dangaria

, Ito

, Walker

C.G.

, Jin

, Schmidt

M.K.

, Galang

M.T.

, Druzinsky

Neural crest lineage segregation: a blueprint for periodontal regeneration. J Dent Res, 88:781. 2009.

10.

Roesch-Ely

, Steinberg

, Bosch

F.X.

, Mussig

, Whitaker

, Wiest

, Kohl

, Komposch

, Tomakidi

Organotypic co-cultures allow for immortalized human gingival keratinocytes to reconstitute a gingival epithelial phenotype in vitro. Differentiation, 74:622. 2006.

11.

Heino

T.J.

, Hentunen

T.A.

, Vaananen

H.K.

Conditioned medium from osteocytes stimulates the proliferation of bone marrow mesenchymal stem cells and their differentiation into osteoblasts. Exp Cell Res, 294:458. 2004.

12.

Ilmer

, Karow

, Geissler

, Jochum

, Neth

Human osteoblast-derived factors induce early osteogenic markers in human mesenchymal stem cells. Tissue Eng Part A, 15:2397. 2009.

13.

Kramer

P.R.

, Nares

, Kramer

S.F.

, Grogan

, Kaiser

Mesenchymal stem cells acquire characteristics of cells in the periodontal ligament in vitro. J Dent Res, 83:27. 2004.

14.

Mizuno

, Ozeki

, Shiba

, Kajiya

, Nagahara

, Takeda

, Kawaguchi

, Abiko

, Kurihara

Humoral factors released from human periodontal ligament cells influence calcification and proliferation in human bone marrow mesenchymal stem cells. J Periodontol, 79:2361. 2008.

15.

Tomakidi

, Breitkreutz

, Fusenig

N.E.

, Zoller

, Kohl

, Komposch

Establishment of oral mucosa phenotype in vitro in correlation to epithelial anchorage. Cell Tissue Res, 292:355. 1998.

16.

Reuther

, Kohl

, Komposch

, Tomakidi

Morphogenesis and proliferation in mono- and organotypic co-cultures of primary human periodontal ligament fibroblasts and alveolar bone cells. Cell Tissue Res, 312:189. 2003.

17.

Winter

, Kohl

, Huppertz

, Herold-Mende

, Wiest

, Komposch

, Tomakidi

Expression of mRNAs encoding for growth factors, ECM molecules, and MMP13 in mono-cultures and co-cultures of human periodontal ligament fibroblasts and alveolar bone cells. Cell Tissue Res, 319:467. 2005.

18.

Bianco

, Robey

P.G.

, Simmons

P.J.

Mesenchymal stem cells: revisiting history, concepts, and assays. Cell Stem Cell, 2:313. 2008.

19.

Hughes

F.J.

, Ghuman

, Talal

Periodontal regeneration: a challenge for the tissue engineer?

Proc Inst Mech Eng H, 224:1345. 2010.

20.

Yang

, Hallgrimsson

, Putnins

E.E.

Craniofacial defect regeneration using engineered bone marrow mesenchymal stromal cells. J Biomed Mater Res A, 99:74. 2011.

21.

Zhou

, Shi

, Xie

, Chen

, He

, Guo

, Wen

, Jin

Role of bone marrow-derived progenitor cells in the maintenance and regeneration of dental mesenchymal tissues. J Cell Physiol, 226:2081. 2011.

22.

Schop

, Janssen

F.W.

, van Rijn

L.D.

, Fernandes

, Bloem

R.M.

, de Bruijn

J.D.

, van Dijkhuizen-Radersma

Growth, metabolism, and growth inhibitors of mesenchymal stem cells. Tissue Eng Part A, 15:1877. 2009.

23.

Higuera

, Schop

, Janssen

, van Dijkhuizen-Radersma

, van Boxtel

, van Blitterswijk

C.A.

Quantifying in vitro growth and metabolism kinetics of human mesenchymal stem cells using a mathematical model. Tissue Eng Part A, 15:2653. 2009.

24.

Csaki

, Matis

, Mobasheri

, Shakibaei

Co-culture of canine mesenchymal stem cells with primary bone-derived osteoblasts promotes osteogenic differentiation. Histochem Cell Biol, 131:251. 2009.

25.

Anisimov

S.V.

, Christophersen

N.S.

, Correia

A.S.

, Hall

V.J.

, Sandelin

, Li

J.Y.

, Brundin

Identification of molecules derived from human fibroblast feeder cells that support the proliferation of human embryonic stem cells. Cell Mol Biol Lett, 16:79. 2011.

26.

Kim

, Lee

J.H.

, Suh

Interaction of mesenchymal stem cells and osteoblasts for in vitro osteogenesis. Yonsei Med J, 44:187. 2003.

27.

Quarto

, Wan

D.C.

, Kwan

M.D.

, Panetta

N.J.

, Li

, Longaker

M.T.

Origin matters: differences in embryonic tissue origin and Wnt signaling determine the osteogenic potential and healing capacity of frontal and parietal calvarial bones. J Bone Miner Res, 25:1680. 2010.

28.

Rawlinson

S.C.

, McKay

I.J.

, Ghuman

, Wellmann

, Ryan

, Prajaneh

, Zaman

, Hughes

F.J.

, Kingsmill

V.J.

Adult rat bones maintain distinct regionalized expression of markers associated with their development. PLoS One, 4:e8358. 2009.

29.

Block

G.J.

, Ohkouchi

, Fung

, Frenkel

, Gregory

, Pochampally

, DiMattia

, Sullivan

D.E.

, Prockop

D.J.

Multipotent stromal cells are activated to reduce apoptosis in part by upregulation and secretion of stanniocalcin-1. Stem Cells, 27:670. 2009.

30.

Yagi

, Parekkadan

, Suganuma

, Soto-Gutierrez

, Tompkins

R.G.

, Tilles

A.W.

, Yarmush

M.L.

Long-term superior performance of a stem cell/hepatocyte device for the treatment of acute liver failure. Tissue Eng Part A, 15:3377. 2009.

31.

Walker

P.A.

, Harting

M.T.

, Jimenez

, Shah

S.K.

, Pati

, Dash

P.K.

, Cox

C.S.

Jr . Direct intrathecal implantation of mesenchymal stromal cells leads to enhanced neuroprotection via an NFkappaB-mediated increase in interleukin-6 production. Stem Cells Dev, 19:867. 2010.

Supplementary Material

Please find the following supplemental material available below.

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

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

0.05 MB

0.39 MB

0.22 MB

0.18 MB

0.00 MB