TY - JOUR
AU - Hellwig, Elmar
AB - Abstract Human mesenchymal stem cells (hMSCs) are promising candidates for regenerative periodontal strategies, due to the broad spectrum of supportive effects on cells and tissues at the site of application. Although positive effects are visible, the understanding of their underlying mechanisms still requires further elucidation. Recently, we have shown that hMSCs are capable to prompt osteogenic differentiation of alveolar osteoblasts, thereby presumably contributing to alveolar bone regeneration. Another issue that is critical in this context is the attraction of hard tissue-forming cells to regeneration sites, but it is an open question whether hMSCs can afford this. In the present manuscript, we show by life cell imaging that in interactive cocultures, hMSCs successfully trigger osteoblast chemotaxis. Gene expression analysis for hMSC-innate chemoattractive biomolecules, orchestrating this process, revealed vascular endothelial growth factor (VEGF), PgE synthase, osteoprotegerin (OPG), monocyte colony-stimulating factor, and transforming growth factor β1, which was confirmed for VEGF and OPG on the protein level. Noteworthy, we showed that only corresponding levels of VEGF but not OPG attracted alveolar osteoblasts similar to hMSC coculture, while VEGF inhibitor abolished both the VEGF and the hMSC-triggered chemoattraction. In summary, we have identified secreted OPG and VEGF proteins as potential chemoattractants, of which further characterization yielded VEGF as a causative for hMSC-directed osteoblast chemotaxis. With respect to the better understanding of potential hMSC-based periodontal regeneration strategies, we propose hMSC-derived VEGF release as a mechanism in the recruitment of hard tissue-forming cells to alveolar bone sites in need of regeneration. Stem Cells 2015;33:3114—3124 Mesenchymal stem cells (MeSH ID D059630), Osteoblasts (MeSH ID D010006), Chemotaxis (MeSH ID D002633), Vascular endothelial growth factor (MeSH ID D042461) Significance Statement In the context of tissue specific stem cell interactions with host tissue cells, the present manuscript is to our best knowledge the first report which provides mechanistic evidence that human bone marrow-derived mesenchymal stem cells (hMSCs) act as trophic mediators and interact with alveolar osteoblasts via their VEGF release, the latter inducing osteoblast recruitment to the lesion site. This functional property of hMSCs renders a possible explanation of the improved tissue function, observed following hMSC transplantation. In our opinion, this knowledge is precious for the conceptual predictability and decision-making regarding the prospective use of hMSCs in alveolar bone regeneration strategies. Introduction Alveolar ridge preservation and regeneration are key concerns in dental, oral, and maxillofacial medicine. New therapy strategies aim at regenerating the alveolar bone by means of tissue engineering and cell administration purposes. In this context, human mesenchymal stem cells from the bone marrow (hMSCs) are promising [1] because they hold the inherent potential for bone formation [2, 3] and are widely accessible in clinical relevant lot sizes [4] thereby efficiently preventing the drawbacks of extensive up-scaling and in vitro manipulation [5]. While it was earlier believed that the therapeutic outcome mainly relies on hMSC differentiation, evidence is now rising that the frequency of these events is low [6]; in other words, functional improvement was observed in injured tissues without significant hMSC engraftment [7] or persistence, respectively [8]. Thus, other mechanisms are suggested to account for the observed tissue regeneration after hMSC administration, including immunomodulation [9, 10] and angiogenesis [11], both resulting from the production of multiple paracrine factors by hMSCs [6, 7]. The increasing appreciation of hMSCs as trophic mediators [12] manifests in recent concepts which even imply to abandon hMSC administration in favor of using hMSC-conditioned culture medium for the purpose of bone regeneration [13-15]. The most fundamental mechanism by which hMSC-derived paracrine factors are supposed to affect immune and endothelial cells is the attraction of cells to the lesion site [16]. Nevertheless, little is known about the underlying mechanisms, and it is almost surprising that only little attention has been paid so far to the reactions of the recipient site-specific cells to the presence of hMSCs. In the context of alveolar bone regeneration, it seems hence obvious to hypothesize that hMSCs release paracrine factors that may act on alveolar osteoblasts similar to the effects described for immune or endothelial cells. The fact that the dynamic processes of bone regeneration require a concerted cellular interplay involving an initial recruitment of osteoblasts demonstrates that the latter are required to feature both an outstanding signal sensing and migration capacity [17, 18]. The assumption that hMSCs interact with alveolar bone cells implies that these very fundamental characteristics of osteoblasts, that is, approaching the lesion site and trigger bone formation, supposedly are modified by hMSCs [19]. To our best knowledge, this is the first report assessing whether hMSCs do express biomolecules that may potentially attract alveolar osteoblasts and whether the latter do follow the chemoattraction triggered by hMSCs. Candidate biomolecules that attract osteoblasts or their progenitors include vascular endothelial growth factor (VEGF), bone morphogenic proteins (BMPs), transforming growth factor β (TGFβ), platelet-derived growth factor, or constituents of the bone matrix [20-23], of which the mainly the first ones were also described to be released by hMSCs [6, 16]. Hence, we screened the cells for a broad panel of putative osteoblast chemoattractants [24, 25], which interestingly enough are so far rather established in the context of differentiation induction [26, 27] or in the framework of angiogenic–osteogenic coupling [28]. Next, we checked whether the biomolecules with robust gene expression levels were actually released as proteins by hMSCs. The identification of strongly released biomolecules in turn enabled the evaluation of their role for osteoblast chemotaxis in interactive hMSC cocultures. This experimental design valuably contributes to the understanding of the hMSC communication with recipient site tissue cells, represented by alveolar osteoblasts, and from the mechanistic viewpoint allowed for identifying the VEGF as a key player for hMSC-related chemoattraction of alveolar osteoblasts. Materials and Methods Cell Isolation, Culture, and Characterization All experiments were 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 (vote number EK-516/12). Human bone marrow-derived mesenchymal stem cells (hMSCs) were obtained from pelvic bone aspirate remnants (bone marrow aspiration pack, Harvest Technologies Corp., Plymouth, MA) of healthy patients undergoing hMSC-based sinus floor augmentation. The plastic-adherent cells were cultivated in NH expansion medium supplemented with CytoMix (both Miltenyi, Bergisch Gladbach, Germany), passaged up to 2–4 times (P2-4), and stored in liquid nitrogen until usage. Osteoblasts were derived from alveolar bone biopsies of patients undergoing corrective osteotomy surgery. Trabecular bone parts were removed to avoid bone marrow cell contamination, and compact bone was minced. Tissue fragments were plated as explants in Dulbecco's modified Eagle's medium (DMEM) (Lifetechnologies, Darmstadt, Germany) supplemented with 10% fetal bovine serum (Biochrom, Berlin, Germany), 1% glutamax (Lifetechnologies, Darmstadt, Germany), and 1% antibiotics, which was exchanged every 2–3 days until cell outgrowth. Upon confluency, cells were trypsinized, expanded by splitting up to 4–6 times (P4-6). For cell characterization, the hMSC-inherent clonogenicity, multilineage potential and surface marker expression was assessed as reported previously [19]. In brief, expression of CD14, CD19, HLA-DR, CD34, CD44, CD45, CD73, CD90, CD105, and CD166 was checked by flow cytometry using a FACSCalibur (BD Biosciences, Heidelberg, Germany). For each run, 20,000 cell events were gated and fluorochrome spectral overlap was checked and compensated whenever required. Data were analyzed using the CellQuest software (BD Biosciences, Heidelberg, Germany). Cell colonies were photographed using a Leica DMIL inverted microscope (Leica Microsystems, Wetzlar, Germany) connected to a Leica D-Lux3 CCD camera (Leica Camera, Solms, Germany). Multilineage differentiation was induced using StemMACS OsteoDiff, AdipoDiff, and ChondroDiff media (Miltenyi, Bergisch Gladbach, Germany) according to manufacturer's instructions for 21 days. Osteogenic differentiation was visualized by Alizarin Red staining. Adipogenic differentiation was verified by oil red O staining, and chondrogenic differentiation was evaluated by indirect immunofluorescence staining of aggrecan (mouse anti-human aggrecan IgG, Merck Chemicals, Schwalbach, Germany, and rhodamine-labeled rabbit anti-mouse IgG, Dianova, Hamburg, Germany) using a Zeiss Axio Imager M1 and the AxioVision SE64 Rel4.9.3 software (Zeiss, Oberkocheln, Germany). For measuring the cytokine gene expression and release, two independent cell culture replicates were run for hMSCs derived from n = 5 donors each. Cells were seeded at 1 × 105 cells per T25, cultured to 80%–90% confluence in NH expansion medium without supplements, washed in phosphate-buffered saline and serum-deprived for t = 1, 2, 3, 5, and 7 days. Quantitative Real-Time Polymerase Chain Reaction Total cellular RNA was purified using a guanidium–thiocyanate method (RNeasy Mini kit; Qiagen, Hilden, Germany) and stored at −80°C. The RNA integrity and quantity were verified using the Experion RNA StdSens chip microfluidic technology according to manufacturer's instructions (Bio-Rad, Munich, Germany), and cDNA was synthesized from 200 ng of total RNA each by using the RT2 First Strand cDNA synthesis kit (Qiagen, Hilden, Germany), preceded by a genomic DNA elimination step at 42°C for 5 minutes, in a C1000 Thermal Cycler (Bio-Rad, Munich, Germany) at 42°C for 30 minutes followed by 95°C for 5 minutes. cDNA was amplified in duplicate each using prevalidated RT2 Primer assays (Supporting Information Table S1) in a 25 µl reaction volume, and cycling conditions were 95°C for 10 minutes followed by 40 cycles of 95°C for 15 seconds and 60°C for 60 seconds, using SYBR Green mastermix in a CFX96 cycler (both Bio-Rad, Munich, Germany), according to manufacturer's instructions. The products' specificity of each amplicon was checked by examining the melting temperatures (heating at 0.05°C/s to 95°C). Negative reverse transcription and negative template controls were included in all polymerase chain reaction (PCR) runs. Data were collected with CFX96 Manager Software version 1.0 (Bio-Rad, Munich, Germany), and relative quantities of the respective genes of interest were normalized to the relative quantity of ACTB, RPL13A and HMBS as references, which were validated for Ct-value consistency (inclusion criteria: ΔCt < 0.5 irrespective of group and culture condition). Data were analyzed and plotted using the RT2 Profiler PCR array data analysis template, that is, fold-change (2^(-delta delta Ct)) described as fold-regulation, representing fold-change results in a biologically meaningful way (http//pcrdataanalysis.sabiosciences.com/pcr/arrayanalysis). Enzyme-Linked Immunosorbent Assays Supernatants were collected, ultrafiltrated using Amicon Ultra-15 devices (Merck Millipore, Darmstadt, Germany), aliquoted and stored at −80°C until usage. Cytokine release was quantified using human TGFβ1, VEGF, prostaglandin E2 (PTGE2), osteoprotegerin (OPG), monocyte colony-stimulating factor (MCSF) (abcam, Cambridge, UK), and receptor activator of nuclear factor-kappaB ligand (RANKL) (Abnova, Heidelberg, Germany) sandwich enzyme-linked immunosorbent assay (ELISA) kits according to manufacturer's instructions as follows: standard/sample incubation for 2.5 hours, biotinylated anti-TGFβ1, anti-VEGF, anti-PgE2, anti-OPG, anti-MCSF, and anti-RANKL detection antibody, respectively, for 1 hour, horseradish peroxidase (HRP) solution for 30 minutes, and detection solution for 30 minutes in the dark. All incubation periods were alternated with thorough plate washing. Absorbance was read at 450 nm in a microplate reader, and data were collected and analyzed using the Magellan v6.2 software (Tecan, Crailsheim, Austria). The biomolecule release was normalized to the respective total protein levels measured by a modified Bradford method using the Pierce660 assay kit (Thermo Fisher, Schwerte, Germany). Briefly, protein quantities were measured at 660 nm in triplicates, and protein concentrations were estimated by reference to absorbance values obtained from a standard series. Chemotaxis Assay For chemotaxis analysis, alveolar osteoblasts were established in the observation area of 3D Chemotaxis µ-slides (ibidi, Planegg, Germany) in bovine collagen 1-gels for hMSC cocultures and without the use of a gel for chemokine analysis, respectively, according to manufacturer's instructions. Briefly, 1.5 × 106 osteoblasts per milliliter were seeded or inoculated in bovine collagen 1-gels (5 mg/ml, Gibco, Darmstadt, Germany) containing 10× DMEM (c.c.pro, Oberdorla, Germany), sodium bicarbonate (Sigma Aldrich, Munich, Germany), NH medium (Miltenyi, Bergisch Gladbach, Germany), and sodium hydroxide. After osteoblast adhesion for 30–60 minutes at 37°C, the slide reservoirs were filled with nonsupplemented alphaMEM without (−/−, negative controls) or with 5.5 ng/ml VEGF or 0.6 ng/ml OPG (both Sigma Aldrich, Munich, Germany) at one side (+/−, test groups), or both sides, respectively (+/+, positive controls). For hMSC cocultures, slide reservoirs were filled with NH medium without (−/−, negative controls), with 1 × 104 hMSCs at one side (+/−, test groups), or both sides, respectively (+/+, positive controls), after gel polymerization for 30 minutes at 37°C (Supporting Information Fig. S1). VEGF signaling was inhibited using SU4312 (Sigma Aldrich, Munich, Germany) at 2 µM added both to the 3D gel and the reservoirs. OPG was neutralized using an anti-human OPG antibody (R&D Systems, Wiesbaden, Germany) at 0.5 µg/ml, while a normal goat IgG antibody served as control. Life cell imaging was performed by time-lapse imaging at 37°C with 5% CO2 and humid atmosphere using an inverted wide field microscope equipped with a motorized x-/y-table and a CCD camera (Zeiss, Oberkochen, Germany). Phase contrast images were taken every 10 minutes for 24 hours at 5× magnification which allowed for entirely imaging each observation area. Data were collected using the ZenBlue software (Zeiss, Oberkochen, Germany) and n = 40 cell tracks per observation area were recorded using the ImageJ plug-in for manual tracking (http://imagej.nih.gov/ij/plugins/track/track.html). Migration parameters, that is, parallel and perpendicular orientation, directness, path length, and velocity were analyzed using the ibidi chemotaxis software tool (http://www.ibidi.de/applications/ap_chemo.html) with forward migration index (FMI)‖ representing the y- or parallel-directed cell migration and FMI⊥ as a measure of x- or perpendicular-directed cell migration. FMI‖ and FMI⊥ were calculated as follows [29]: FMI‖=1n∑i=1nYi,enddi,accumFMI⊥=1n∑i=1nXi,enddi,accum Cells were considered to migrate in a hMSC-directed manner, that is, performing chemotaxis, if |FMI‖| > |FMI⊥| and p < 0.05 in Rayleigh distribution testing for statistical distribution of cell end points. Statistics For chemoattractant mRNA expression data, p values were calculated based on a Student's t test of the replicate fold change values for each gene in the treatment group and control groups. To evaluate the hMSC chemoattractant release, a repeated measure analysis was performed with a linear mixed model for each outcome of interest (biomolecules) and time (1, 2, 3, 5, and 7 days). The group effects and differences of least square (ls) means are calculated with their 95% confidence intervals. Several multiple comparisons of ls means in groups were performed and p values were adjusted for multiple testing by the method of Tukey-Kramer. Calculations have been done using PROC MIXED from the statistical software SAS 9.1.2 (Cary, NC). Results Human Mesenchymal Stem Cells Fulfil the Standard Characterization Criteria To verify the stem cell character [30] of the cells employed in this study, hMSC from n = 5 donors were characterized according to the minimal criteria proposed by the International Society for Cellular Therapy (ISCT) [31]. The mononuclear cells derived from bone marrow aspirates readily adhered to the cell culture device and rapidly adopted a polymorphous and elongated polygonal shape while growing from colonies (Supporting Information Fig. S2A). Colony-derived cells were successfully transformed into osteoblasts, adipocytes or chondrocytes as visualized by alizarin Red, oil red O and Aggrecan immunofluorescence stainings (Supporting Information Fig. S2B). Flow cytometry screening revealed that the cells were reasonably homogeneous in size and granularity and expressed CD73, CD90, CD105, and CD166, but were negative for CD14, CD19, CD34, and CD45 (Supporting Information Fig. S2C). The expression of HLA-DR was detected on hMSCs from two of five donors (35.17% and 59.55%, respectively), while CD44 and CD166 were expressed as further mesenchymal cell-associated markers (Supporting Information Fig. S2C). Human Mesenchymal Stem Cells Express mRNAs for Bone Cell Chemoattractants To identify the biomolecules which may contribute to the hMSC action as trophic mediators in the osteoblast context, the cells were screened for mRNA transcription of a panel of multitasking biomolecules which had previously been reported to putatively stimulate alveolar bone regeneration [23, 32, 33], and support osteoblast migration [24, 25, 27, 34]. In detail, the mRNA expression of the following cytokines and growth factors was screened twice in hMSCs from all donors: VEGF, fibroblast growth factor 2 (FGF2), insulin-like growth factor 1 (IGF1), Wnt10B, growth/differentiation factor 5 and 10 (GDF5 and GDF10), BMPs 2, 4, and 6 (BMP2, BMP4, and BMP6), prostaglandin E synthase (PTGES), TGF β1 and 3 (TGFβ1 and TGFβ3), monocyte colony-stimulating factor (MCSF), OPG, RANKL, and interleukin 6 (IL6). Noteworthy, serum-deprivation served for closely imitating clinical conditions that apply both for the production of conditioned medium and for the global intention to keep hMSCs as native as possible, that is, only short-term in vitro cultivation and the avoidance of serum use. In general, hMSCs expressed all of the aforementioned genes. The heatmap analysis of gene expression levels for each donor and time point, respectively, revealed a panel of biomolecules that were highly expressed during the whole investigation time, of which the most prominent was VEGF (Fig. 1A). The latter was expressed constitutively with high intensity irrespective of donor and time point of investigation. Interestingly, also OPG showed elevated expression levels for all donors and time points, and obviously, it was much stronger transcribed than RANKL (Fig. 1A). Furthermore, PTGES was robustly expressed, while showing some discriminative transcription patterns with regard to the respective donor (Fig. 1A, compare donor 4 with donor 5). Similar to PTGES, the expression of MCSF was sizable but showed some donor-related differences (Fig. 1A, compare donor 1 with donor 4). TGFβ1 was intensively expressed and displayed slight time-dependent differences in mRNA levels (Fig. 1A). Compared with VEGF, OPG, PTGES, MCSF, and TGFβ1, the expression levels of all other multitasking biomolecule genes under study were quite low. Open in new tabDownload slide Human mesenchymal stem cell (hMSC) expression and release of bone cell chemoattractant biomolecules. (A): Gene expression magnitude heatmap for all donors and each time points from two independent experiments with green indicating low gene expression levels relative to the red-colored high gene expression magnitudes. Of note, the expression of most of the genes under study was relatively small compared with vascular endothelial growth factor (VEGF), prostaglandin E synthase (PTGES), osteoprotegerin (OPG), and monocyte colony-stimulating factor (MCSF), which revealed high expression levels. Interestingly, only VEGF and OPG were found to be transcribed unanimously at high expression magnitudes for all donors. (B): Mean biomolecule levels relative to total protein amount with standard deviation (y-axis, biomolecule [pg]/total protein [µg]) in hMSC culture supernatants at 1, 2, 3, 5, and 7 days (d, x-axis) of serum deprivation. Compared with the other biomolecules under study, the initial quantity of VEGF was high and was further significantly increased until d7. The levels of OPG exceeded receptor activator of nuclear factor-kappa B ligand significantly for each time point under study, the latter being additionally decreased at d5 and d7. For all other molecules, the amount relative to total protein levels was small, but modulated by time: PgE2 was further decreased at d3, d5, and d7, transforming growth factor β1 (TGFβ1) peaked at d2, and MCSF was continuously increased until d7. Differences in biomolecule per total protein levels were considered significant at *, p < 0.05; **, p < 0.01; and ***, p < 0.001. Abbreviations: MCSF, monocyte colony-stimulating factor; OPG, osteoprotegerin; PgE2, prostaglandin E2; RANKL, receptor activator of nuclear factor-kappaB ligand; TGFβ1, transforming growth factor β1; VEGF, vascular endothelial growth factor. Open in new tabDownload slide Human mesenchymal stem cell (hMSC) expression and release of bone cell chemoattractant biomolecules. (A): Gene expression magnitude heatmap for all donors and each time points from two independent experiments with green indicating low gene expression levels relative to the red-colored high gene expression magnitudes. Of note, the expression of most of the genes under study was relatively small compared with vascular endothelial growth factor (VEGF), prostaglandin E synthase (PTGES), osteoprotegerin (OPG), and monocyte colony-stimulating factor (MCSF), which revealed high expression levels. Interestingly, only VEGF and OPG were found to be transcribed unanimously at high expression magnitudes for all donors. (B): Mean biomolecule levels relative to total protein amount with standard deviation (y-axis, biomolecule [pg]/total protein [µg]) in hMSC culture supernatants at 1, 2, 3, 5, and 7 days (d, x-axis) of serum deprivation. Compared with the other biomolecules under study, the initial quantity of VEGF was high and was further significantly increased until d7. The levels of OPG exceeded receptor activator of nuclear factor-kappa B ligand significantly for each time point under study, the latter being additionally decreased at d5 and d7. For all other molecules, the amount relative to total protein levels was small, but modulated by time: PgE2 was further decreased at d3, d5, and d7, transforming growth factor β1 (TGFβ1) peaked at d2, and MCSF was continuously increased until d7. Differences in biomolecule per total protein levels were considered significant at *, p < 0.05; **, p < 0.01; and ***, p < 0.001. Abbreviations: MCSF, monocyte colony-stimulating factor; OPG, osteoprotegerin; PgE2, prostaglandin E2; RANKL, receptor activator of nuclear factor-kappaB ligand; TGFβ1, transforming growth factor β1; VEGF, vascular endothelial growth factor. As a side note, we found that only genes at low expression levels were subjected to significant expression changes by time (Supporting Information Fig. S3). Amongst these, GDF5 (fold regulation: −2.6294) and WNT10B (−1.5928) were considerably decreased at d7 compared with d1 (cut-off level 1.5-fold), however, the downregulation was significant only for WNT10B (−1.6849, p = 0.012501) at d5 compared with d1. On the other hand, the expression of BMP2 (+3.6300), FGF2 (+3.1252), IGF1 (+3.2193), and TGFβ3 (+1.5233) was clearly increased at d7 compared with d1. Of note, even though TGFβ3 exhibited the lowest gene expression increase, it was exclusively found to be significant (p = 0.001913, Fig. S3). Nevertheless, the commonly discrete regulation revealed a distinctive expression pattern, that is, only BMP4, IL6, and MCSF tended to be upregulated, while the majority of genes followed the trend of being downregulated by time (Fig. S3). Together, these data indicate that hMSCs may putatively act as trophic mediators for bone cells. Human Mesenchymal Stem Cells Release Bone Cell Chemoattractants Following detection of the biomolecules with respect to their gene expression magnitude, we next assessed whether the genes showing a strong mRNA expression were released as proteins by hMSCs. For this purpose, the culture supernatant of serum-deprived hMSCs at 1, 2, 3, 5, and 7 days was screened for the presence of VEGF, OPG, PgE2, TGFβ1, MCSF, and additionally RANKL. The release levels were normalized to the total protein amount, which helped to evaluate whether differences in protein levels result from a time progression-related biomolecule enrichment or cell proliferation, respectively, or whether they actually represent variations in chemoattractant biomolecule release. In concordance with the mRNA expression data, it was not surprising that the release of VEGF protein was strong (Fig. 1B). Of note, although its gene expression was not significantly modulated, the release of VEGF was increased at 2, 3, 5, and 7 d (p < 0.0001 for all compared with d1, Fig. 1B). The second most prominently secreted biomolecule was OPG, although its release was roughly 4.5-fold smaller than the amount of VEGF (Fig. 1B). In compliance with mRNA expression, the OPG release was not modulated by time but showed substantial inter-individual differences, which became apparent with regard to a sizeable standard deviation (Fig. 1B). Nevertheless, for all donors and time points investigated, OPG levels were significantly higher than RANKL release (p = 0.0342 at d1, p = 0.0071 at d2, p = 0.0070 at d3, p = 0.0145 at d5, and p = 0.0147 at d7). In addition to its globally low levels, the release of RANKL was even decreased at d5 (p = 0.038) and d7 (p = 0.001) compared with baseline (Fig. 1B). In consequence, the RANKL/OPG ratio was clearly shifted to OPG, suggesting that any supposable OPG action may overtop RANKL signals emitted by hMSCs in the osteoblast context. Unexpectedly, for all of the other biomolecules showing a high mRNA expression, the effective release of respective bioactive molecules was quite low (Fig. 1B, compare y-axis scales for all biomolecules). For TGFβ1, a small but significant increase was found at d2 (p = 0.014); however, at 3, 5, and 7 d, the levels of TGFβ1 did not differ from baseline (Fig. 1B). Of note, MCSF was increasingly released at d2 (p = 0.037), and d3, d5, and d7 (p < 0.0001 for all), indicating that it may potentially be more relevant in hMSC communication than TGFβ1, despite its generally very low levels (Fig. 1B). Similarly, PgE2 was expressed at low levels, which were even decreased at d3 (p = 0.017), d5 (p = 0.003), and d7 (p = 0.001). Taken together, the biomolecule release data show that from all biomolecules with high mRNA expression, only VEGF and, although to a lower extent, OPG are present at remarkable quantities, suggesting that these two may play a functional role in intercellular communication between hMSCs and alveolar bone cells. VEGF But Not OPG Attracts Alveolar Osteoblasts After having identified VEGF and OPG as the two biomolecules that are mainly released by hMSCs, we next performed a functional analysis to test whether the two biomolecules are actually able to attract alveolar osteoblasts. For this purpose, a chemotaxis assay was conducted measuring the migration orientation of alveolar osteoblasts in response to VEGF or OPG supply, respectively (Fig. 2). The tracking of the alveolar osteoblast migration in negative controls (without VEGF or OPG), test groups with unilateral VEGF or OPG supply, and positive controls with bilateral VEGF or OPG supply, respectively, was analyzed for the following parameters: the FMI which measures the parallel (FMI∥) and perpendicular (FMI⊥) migration direction with chemotaxis effects characterized by elevated |FM∥| but |FMI⊥| values similar to controls. Interestingly, the osteoblasts were clearly attracted by VEGF in the test groups, while the cellular migration was not oriented to a specific side in negative and positive controls, respectively (Fig. 2A; Supporting Information Table S2A). This observation was verified by Rayleigh distribution tests which evaluate the uniformity of the cell endpoint distribution, that is, chemotaxis effects are disclosed by nonhomogeneous cell arrangements. In fact, the Rayleigh distribution was significantly nonhomogeneous for osteoblasts with unilateral VEGF supply, while the cell track endpoints were uniformly distributed in negative and positive controls (Table S2A). This finding is reflected by the centre of mass (CoM)-values measuring the displacement magnitude of the whole osteoblast collective, which were found to be elevated only for unilateral VEGF supply (Fig. 2A; Supporting Information Table S2A). In positive controls, osteoblasts were evenly attracted by VEGF from both sides, hence it is not surprising that the Rayleigh distribution was nonsignificant and the CoM-values were low. On the other hand, the inhibition of VEGF triggered the osteoblasts to completely lose their orientation towards the VEGF-supplied reservoir in unilateral test groups (Fig. 2A; Supporting Information Table S2B). This was confirmed by small |FMI∥| and |FMI⊥| values, and Rayleigh tests revealed that the cell endpoint distribution was random and did not display any orientation (Fig. 2A; Supporting Information Table S2B). These data show that VEGF acts as chemoattractant for alveolar bone cells. Open in new tabDownload slide Alveolar osteoblast chemotaxis in response to vascular endothelial growth factor (VEGF) and osteoprotegerin (OPG). A diagram of the chemotaxis chamber visualizes the experimental setup with alveolar osteoblasts in the focus area, which was connected to reservoirs either without VEGF or OPG supply (−/− VEGF/OPG, negative controls), respectively, with unilateral VEGF or OPG supply (+/− VEGF/OPG, test groups), and with bilateral VEGF or OPG supply (+/ + VEGF/OPG, positive controls). (A): Representative chemotaxis diagrams indicating the number of cells in the different sectors after 24 hours (x-/y-axis: µm). Test groups: 2D trajectory plots of the alveolar osteoblasts migration paths with all starting points set to x/y = 0 and end-points at y > 0 in black color and end-points with y < 0 in red color (x-/y-axis: µm, green cross: centre of mass [CoM]). With unilateral VEGF supply, the osteoblasts were clearly attracted towards VEGF, which becomes visible from the single-edge chemotaxis diagrams, the multitude of red-labeled cell migration paths in the trajectory plots, and CoM ≠ [x/y = 0]. With simultaneous VEGF receptor inhibition, the cells were no longer able to follow the VEGF-triggered chemoattraction. (B): The unilateral addition of OPG did not attract alveolar osteoblasts as obvious from uniform chemotaxis diagrams and trajectory plots in the test groups, which are similar to matched controls. In consequence, further OPG neutralization did not modify the osteoblast migration response to OPG supply. Chemotaxis Chamber diagram copyright: ibidi GmbH, Planegg, Germany. Abbreviations: OPG, osteoprotegerin; VEGF, vascular endothelial growth factor. Open in new tabDownload slide Alveolar osteoblast chemotaxis in response to vascular endothelial growth factor (VEGF) and osteoprotegerin (OPG). A diagram of the chemotaxis chamber visualizes the experimental setup with alveolar osteoblasts in the focus area, which was connected to reservoirs either without VEGF or OPG supply (−/− VEGF/OPG, negative controls), respectively, with unilateral VEGF or OPG supply (+/− VEGF/OPG, test groups), and with bilateral VEGF or OPG supply (+/ + VEGF/OPG, positive controls). (A): Representative chemotaxis diagrams indicating the number of cells in the different sectors after 24 hours (x-/y-axis: µm). Test groups: 2D trajectory plots of the alveolar osteoblasts migration paths with all starting points set to x/y = 0 and end-points at y > 0 in black color and end-points with y < 0 in red color (x-/y-axis: µm, green cross: centre of mass [CoM]). With unilateral VEGF supply, the osteoblasts were clearly attracted towards VEGF, which becomes visible from the single-edge chemotaxis diagrams, the multitude of red-labeled cell migration paths in the trajectory plots, and CoM ≠ [x/y = 0]. With simultaneous VEGF receptor inhibition, the cells were no longer able to follow the VEGF-triggered chemoattraction. (B): The unilateral addition of OPG did not attract alveolar osteoblasts as obvious from uniform chemotaxis diagrams and trajectory plots in the test groups, which are similar to matched controls. In consequence, further OPG neutralization did not modify the osteoblast migration response to OPG supply. Chemotaxis Chamber diagram copyright: ibidi GmbH, Planegg, Germany. Abbreviations: OPG, osteoprotegerin; VEGF, vascular endothelial growth factor. Regarding OPG, no measurable chemotaxis effect became evident in the test groups showing homogenous osteoblast migration pattern similar to controls (Fig. 2B). This observation was confirmed by a nonsignificant Rayleigh distribution, and FMI∥ and CoM-values similar to matched controls (Supporting Information Table S3A). The same was found for unilateral OPG supply together with simultaneous neutralization, which did not influence the migration direction of alveolar osteoblasts (Fig. 2B; Supporting Information Table S3B). Irrespective of VEGF or OPG supply, respectively, the small directness and high distance values indicate that alveolar osteoblasts migrated in a meandering manner. The migration speed was similar in test groups and matched controls. These findings clearly prove VEGF as causative for the chemoattraction of alveolar osteoblasts, while OPG was found to have no mechanistic relevance in terms of osteoblast chemotaxis. Human Mesenchymal Stem Cells Attract Alveolar Osteoblasts via VEGF After identifying VEGF as mainly released biomolecule with the ability of attracting alveolar osteoblasts, we next tested the hypothesis that hMSCs themselves are capable of attracting alveolar osteoblasts and further that VEGF is the key molecule of the hMSC-mediated chemoattraction. To this end, the chemotaxis assay was performed for interactive hMSC cocultures with and without simultaneous VEGF receptor inhibition. To approach the tissue-specific conditions as far as possible in a highly relevant way that is close to the clinical reality, the osteoblasts were incorporated in 3D collagen gels and received medium supply from two reservoirs that were equipped with medium only (negative controls), or hMSCs either at both sides (bidirectional cocultures, positive controls), or at one side only (unidirectional cocultures, test groups, Fig. 3). Open in new tabDownload slide Alveolar osteoblast chemotaxis triggered by human mesenchymal stem cell (hMSC) coculture. A diagram of the chemotaxis chamber visualizes the experimental setup with alveolar osteoblasts in the focus area, which was connected to reservoirs either without hMSCs (−/− hMSCs, negative controls), with unidirectional interactive hMSC coculture (+/− hMSCs, test groups), and with bidirectional hMSC coculture (+/ + hMSCs, positive controls). (A): Representative chemotaxis diagrams indicating the number of cells in the different sectors after 24 hours (x-/y-axis: µm). Test groups: 2D trajectory plots of the alveolar osteoblasts migration paths with all starting points set to x/y = 0 and end-points at y > 0 in black color and end-points with y < 0 in red colour (x-/y-axis: µm, green cross: centre of mass [CoM]). Bar graphs: forward migration index (FMI∥) (y-axis: index range), FMI⊥ (y-axis: index range), CoM (y-axis: |µm|) and directness values (y-axis: index range) from three independent experiments (see Table 1, blue: negative controls, striped: test groups, red: positive controls). In test groups, the osteoblasts were clearly attracted towards hMSCs, which becomes visible from the single-edge chemotaxis diagrams, elevated |FMI∥| together with stable |FMI⊥| values, and high CoM values compared with matched controls. (B): Representative chemotaxis diagrams, trajectory plot, and bar graphs of the osteoblast chemotaxis show that the cells were not attracted by hMSCs with simultaneous vascular endothelial growth factor (VEGF) receptor inhibition. This becomes obvious from uniform chemotaxis diagrams and trajectory plots, small FMI∥ and CoM values similar to controls. Irrespective of VEGF receptor inhibition, the osteoblasts migrated in a meandering manner, as obvious from directness values similar to controls. Chemotaxis Chamber diagram copyright: ibidi GmbH, Planegg, Germany. Abbreviations: CoM, centre of mass; FMI, forward migration index; hMSC, human mesenchymal stem cell; VEGF, vascular endothelial growth factor. Open in new tabDownload slide Alveolar osteoblast chemotaxis triggered by human mesenchymal stem cell (hMSC) coculture. A diagram of the chemotaxis chamber visualizes the experimental setup with alveolar osteoblasts in the focus area, which was connected to reservoirs either without hMSCs (−/− hMSCs, negative controls), with unidirectional interactive hMSC coculture (+/− hMSCs, test groups), and with bidirectional hMSC coculture (+/ + hMSCs, positive controls). (A): Representative chemotaxis diagrams indicating the number of cells in the different sectors after 24 hours (x-/y-axis: µm). Test groups: 2D trajectory plots of the alveolar osteoblasts migration paths with all starting points set to x/y = 0 and end-points at y > 0 in black color and end-points with y < 0 in red colour (x-/y-axis: µm, green cross: centre of mass [CoM]). Bar graphs: forward migration index (FMI∥) (y-axis: index range), FMI⊥ (y-axis: index range), CoM (y-axis: |µm|) and directness values (y-axis: index range) from three independent experiments (see Table 1, blue: negative controls, striped: test groups, red: positive controls). In test groups, the osteoblasts were clearly attracted towards hMSCs, which becomes visible from the single-edge chemotaxis diagrams, elevated |FMI∥| together with stable |FMI⊥| values, and high CoM values compared with matched controls. (B): Representative chemotaxis diagrams, trajectory plot, and bar graphs of the osteoblast chemotaxis show that the cells were not attracted by hMSCs with simultaneous vascular endothelial growth factor (VEGF) receptor inhibition. This becomes obvious from uniform chemotaxis diagrams and trajectory plots, small FMI∥ and CoM values similar to controls. Irrespective of VEGF receptor inhibition, the osteoblasts migrated in a meandering manner, as obvious from directness values similar to controls. Chemotaxis Chamber diagram copyright: ibidi GmbH, Planegg, Germany. Abbreviations: CoM, centre of mass; FMI, forward migration index; hMSC, human mesenchymal stem cell; VEGF, vascular endothelial growth factor. Alveolar osteoblast chemotaxis in response to hMSC coculture (A) without and (B) with simultaneous VEGF inhibition . . Rayleigh distr. . FMI⊥ . FMIǁ . Dir . CoM . Accumulated distance (µm) . Euclidean distance (µm) . Velocity (µm/min) . (A) Alveolar osteoblasts −/− hMSCs n.s. 0.0 −0.03 0.19 5.65 247.9 ± 137.5 47.9 ± 46.2 0.17 ± 0.1 +/− hMSCs p = 0.0002 0.01 −0.12 0.28 80.07 649.3 ± 158.3 189.9 ± 136.8 0.45 ± 0.11 +/+ hMSCs n.s. 0.04 −0.03 0.26 25.13 533.9 ± 126.6 139.2 ± 83.0 0.37 ± 0.09 Alveolar osteoblasts −/− hMSCs n.s. −0.02 −0.02 0.21 28.27 761.3 ± 200.0 168.6 ± 118.8 0.53 ± 0.14 +/− hMSCs p < 0.0001 0.06 −0.15 0.28 117.04 675.3 ± 140.1 198.6 ± 122.5 0.47 ± 0.10 +/+ hMSCs n.s. −0.03 0.0 0.23 25.66 771.8 ± 195.5 182.9 ± 134.6 0.54 ± 0.14 Alveolar osteoblasts −/− hMSCs n.s. 0.01 0.0 0.25 15.73 623.6 ± 176.4 168.4 ± 144.3 0.43 ± 0.12 +/− hMSCs p = 0.0097 −0.04 −0.06 0.23 53.47 657.6 ± 117.1 155.5 ± 99.9 0.46 ± 0.08 +/+ hMSCs n.s. −0.00 −0.06 0.23 61.52 702.2 ± 212.6 172.1 ± 149.6 0.49 ± 0.15 (B) Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.02 −0.02 0.20 14.66 479.59 ± 175.24 93.73 ± 69.45 0.33 ± 0.12 +/− hMSCs n.s. 0.06 0.01 0.42 17.84 456.92 ± 158.54 187.2 ± 98.59 0.32 ± 0.11 +/+ hMSCs n.s. −0.05 −0.05 0.40 21.25 464.02 ± 189.38 179.56 ± 99.04 0.32 ± 0.13 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.07 −0.01 0.31 42.65 727.89 ± 133.15 232.75 ± 140.48 0.51 ± 0.09 +/− hMSCs n.s. −0.02 0.01 0.25 23.23 757.23 ± 126.99 185.69 ± 112.15 0.53 ± 0.09 +/+ hMSCs n.s. 0.03 0.01 0.27 23.28 743.14 ± 139.32 204.66 ± 134.15 0.52 ± 0.10 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. −0.01 −0.02 0.25 21.37 907.26 ± 211.99 229.96 ± 142.06 0.63 ± 0.14 +/− hMSCs n.s. −0.01 −0.02 0.25 23.50 909.18 ± 208.00 233.32 ± 163.23 0.63 ± 0.15 +/+ hMSCs n.s. −0.04 −0.04 0.25 64.57 952.81 ± 180.8 248.29 ± 151.6 0.66 ± 0.13 . . Rayleigh distr. . FMI⊥ . FMIǁ . Dir . CoM . Accumulated distance (µm) . Euclidean distance (µm) . Velocity (µm/min) . (A) Alveolar osteoblasts −/− hMSCs n.s. 0.0 −0.03 0.19 5.65 247.9 ± 137.5 47.9 ± 46.2 0.17 ± 0.1 +/− hMSCs p = 0.0002 0.01 −0.12 0.28 80.07 649.3 ± 158.3 189.9 ± 136.8 0.45 ± 0.11 +/+ hMSCs n.s. 0.04 −0.03 0.26 25.13 533.9 ± 126.6 139.2 ± 83.0 0.37 ± 0.09 Alveolar osteoblasts −/− hMSCs n.s. −0.02 −0.02 0.21 28.27 761.3 ± 200.0 168.6 ± 118.8 0.53 ± 0.14 +/− hMSCs p < 0.0001 0.06 −0.15 0.28 117.04 675.3 ± 140.1 198.6 ± 122.5 0.47 ± 0.10 +/+ hMSCs n.s. −0.03 0.0 0.23 25.66 771.8 ± 195.5 182.9 ± 134.6 0.54 ± 0.14 Alveolar osteoblasts −/− hMSCs n.s. 0.01 0.0 0.25 15.73 623.6 ± 176.4 168.4 ± 144.3 0.43 ± 0.12 +/− hMSCs p = 0.0097 −0.04 −0.06 0.23 53.47 657.6 ± 117.1 155.5 ± 99.9 0.46 ± 0.08 +/+ hMSCs n.s. −0.00 −0.06 0.23 61.52 702.2 ± 212.6 172.1 ± 149.6 0.49 ± 0.15 (B) Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.02 −0.02 0.20 14.66 479.59 ± 175.24 93.73 ± 69.45 0.33 ± 0.12 +/− hMSCs n.s. 0.06 0.01 0.42 17.84 456.92 ± 158.54 187.2 ± 98.59 0.32 ± 0.11 +/+ hMSCs n.s. −0.05 −0.05 0.40 21.25 464.02 ± 189.38 179.56 ± 99.04 0.32 ± 0.13 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.07 −0.01 0.31 42.65 727.89 ± 133.15 232.75 ± 140.48 0.51 ± 0.09 +/− hMSCs n.s. −0.02 0.01 0.25 23.23 757.23 ± 126.99 185.69 ± 112.15 0.53 ± 0.09 +/+ hMSCs n.s. 0.03 0.01 0.27 23.28 743.14 ± 139.32 204.66 ± 134.15 0.52 ± 0.10 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. −0.01 −0.02 0.25 21.37 907.26 ± 211.99 229.96 ± 142.06 0.63 ± 0.14 +/− hMSCs n.s. −0.01 −0.02 0.25 23.50 909.18 ± 208.00 233.32 ± 163.23 0.63 ± 0.15 +/+ hMSCs n.s. −0.04 −0.04 0.25 64.57 952.81 ± 180.8 248.29 ± 151.6 0.66 ± 0.13 Individual replicates of osteoblast chemotaxis assessment. (A): Alveolar osteoblasts were clearly attracted by hMSCs, as revealed by a significant Rayleigh distribution, high FMI∥, and elevated CoM-values in the test groups (bold). Nevertheless, the small directness and distance values indicate that alveolar osteoblasts migrate in a meandering manner irrespective of hMSC coculture. The migration speed was similar in test groups and matched controls. (B): With simultaneous VEGF receptor inhibition, the osteoblasts were not able to follow the chemoattraction triggered by the unilateral hMSC coculture in the test groups (bold), which is revealed by a nonsignificant Rayleigh distribution with FMI∥ and CoM values that are similar to matched controls. Abbreviations: Accumulated distance, length of the total cell path; CoM, centre of mass; Dir, directness of osteoblast migration; Euclidean distance, length of a straight line from the cell starting point to the cell endpoint; Forward migration index (FMI), perpendicular (FMI⊥) and parallel (FMI∥) migration direction; hMSC, human mesenchymal stem cell;−/− hMSCs, without hMSC coculture, negative controls; + /− hMSCs, unilateral hMSC coculture, test groups; + / + hMSCs, bilateral hMSC coculture, positive controls; Rayleigh distribution (distr.), uniformity of the cell endpoint distribution; VEGF, vascular endothelial growth factor; velocity: osteoblasts migration speed. Open in new tab Alveolar osteoblast chemotaxis in response to hMSC coculture (A) without and (B) with simultaneous VEGF inhibition . . Rayleigh distr. . FMI⊥ . FMIǁ . Dir . CoM . Accumulated distance (µm) . Euclidean distance (µm) . Velocity (µm/min) . (A) Alveolar osteoblasts −/− hMSCs n.s. 0.0 −0.03 0.19 5.65 247.9 ± 137.5 47.9 ± 46.2 0.17 ± 0.1 +/− hMSCs p = 0.0002 0.01 −0.12 0.28 80.07 649.3 ± 158.3 189.9 ± 136.8 0.45 ± 0.11 +/+ hMSCs n.s. 0.04 −0.03 0.26 25.13 533.9 ± 126.6 139.2 ± 83.0 0.37 ± 0.09 Alveolar osteoblasts −/− hMSCs n.s. −0.02 −0.02 0.21 28.27 761.3 ± 200.0 168.6 ± 118.8 0.53 ± 0.14 +/− hMSCs p < 0.0001 0.06 −0.15 0.28 117.04 675.3 ± 140.1 198.6 ± 122.5 0.47 ± 0.10 +/+ hMSCs n.s. −0.03 0.0 0.23 25.66 771.8 ± 195.5 182.9 ± 134.6 0.54 ± 0.14 Alveolar osteoblasts −/− hMSCs n.s. 0.01 0.0 0.25 15.73 623.6 ± 176.4 168.4 ± 144.3 0.43 ± 0.12 +/− hMSCs p = 0.0097 −0.04 −0.06 0.23 53.47 657.6 ± 117.1 155.5 ± 99.9 0.46 ± 0.08 +/+ hMSCs n.s. −0.00 −0.06 0.23 61.52 702.2 ± 212.6 172.1 ± 149.6 0.49 ± 0.15 (B) Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.02 −0.02 0.20 14.66 479.59 ± 175.24 93.73 ± 69.45 0.33 ± 0.12 +/− hMSCs n.s. 0.06 0.01 0.42 17.84 456.92 ± 158.54 187.2 ± 98.59 0.32 ± 0.11 +/+ hMSCs n.s. −0.05 −0.05 0.40 21.25 464.02 ± 189.38 179.56 ± 99.04 0.32 ± 0.13 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.07 −0.01 0.31 42.65 727.89 ± 133.15 232.75 ± 140.48 0.51 ± 0.09 +/− hMSCs n.s. −0.02 0.01 0.25 23.23 757.23 ± 126.99 185.69 ± 112.15 0.53 ± 0.09 +/+ hMSCs n.s. 0.03 0.01 0.27 23.28 743.14 ± 139.32 204.66 ± 134.15 0.52 ± 0.10 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. −0.01 −0.02 0.25 21.37 907.26 ± 211.99 229.96 ± 142.06 0.63 ± 0.14 +/− hMSCs n.s. −0.01 −0.02 0.25 23.50 909.18 ± 208.00 233.32 ± 163.23 0.63 ± 0.15 +/+ hMSCs n.s. −0.04 −0.04 0.25 64.57 952.81 ± 180.8 248.29 ± 151.6 0.66 ± 0.13 . . Rayleigh distr. . FMI⊥ . FMIǁ . Dir . CoM . Accumulated distance (µm) . Euclidean distance (µm) . Velocity (µm/min) . (A) Alveolar osteoblasts −/− hMSCs n.s. 0.0 −0.03 0.19 5.65 247.9 ± 137.5 47.9 ± 46.2 0.17 ± 0.1 +/− hMSCs p = 0.0002 0.01 −0.12 0.28 80.07 649.3 ± 158.3 189.9 ± 136.8 0.45 ± 0.11 +/+ hMSCs n.s. 0.04 −0.03 0.26 25.13 533.9 ± 126.6 139.2 ± 83.0 0.37 ± 0.09 Alveolar osteoblasts −/− hMSCs n.s. −0.02 −0.02 0.21 28.27 761.3 ± 200.0 168.6 ± 118.8 0.53 ± 0.14 +/− hMSCs p < 0.0001 0.06 −0.15 0.28 117.04 675.3 ± 140.1 198.6 ± 122.5 0.47 ± 0.10 +/+ hMSCs n.s. −0.03 0.0 0.23 25.66 771.8 ± 195.5 182.9 ± 134.6 0.54 ± 0.14 Alveolar osteoblasts −/− hMSCs n.s. 0.01 0.0 0.25 15.73 623.6 ± 176.4 168.4 ± 144.3 0.43 ± 0.12 +/− hMSCs p = 0.0097 −0.04 −0.06 0.23 53.47 657.6 ± 117.1 155.5 ± 99.9 0.46 ± 0.08 +/+ hMSCs n.s. −0.00 −0.06 0.23 61.52 702.2 ± 212.6 172.1 ± 149.6 0.49 ± 0.15 (B) Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.02 −0.02 0.20 14.66 479.59 ± 175.24 93.73 ± 69.45 0.33 ± 0.12 +/− hMSCs n.s. 0.06 0.01 0.42 17.84 456.92 ± 158.54 187.2 ± 98.59 0.32 ± 0.11 +/+ hMSCs n.s. −0.05 −0.05 0.40 21.25 464.02 ± 189.38 179.56 ± 99.04 0.32 ± 0.13 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. 0.07 −0.01 0.31 42.65 727.89 ± 133.15 232.75 ± 140.48 0.51 ± 0.09 +/− hMSCs n.s. −0.02 0.01 0.25 23.23 757.23 ± 126.99 185.69 ± 112.15 0.53 ± 0.09 +/+ hMSCs n.s. 0.03 0.01 0.27 23.28 743.14 ± 139.32 204.66 ± 134.15 0.52 ± 0.10 Alveolar osteoblasts + VEGF inhibitor −/− hMSCs n.s. −0.01 −0.02 0.25 21.37 907.26 ± 211.99 229.96 ± 142.06 0.63 ± 0.14 +/− hMSCs n.s. −0.01 −0.02 0.25 23.50 909.18 ± 208.00 233.32 ± 163.23 0.63 ± 0.15 +/+ hMSCs n.s. −0.04 −0.04 0.25 64.57 952.81 ± 180.8 248.29 ± 151.6 0.66 ± 0.13 Individual replicates of osteoblast chemotaxis assessment. (A): Alveolar osteoblasts were clearly attracted by hMSCs, as revealed by a significant Rayleigh distribution, high FMI∥, and elevated CoM-values in the test groups (bold). Nevertheless, the small directness and distance values indicate that alveolar osteoblasts migrate in a meandering manner irrespective of hMSC coculture. The migration speed was similar in test groups and matched controls. (B): With simultaneous VEGF receptor inhibition, the osteoblasts were not able to follow the chemoattraction triggered by the unilateral hMSC coculture in the test groups (bold), which is revealed by a nonsignificant Rayleigh distribution with FMI∥ and CoM values that are similar to matched controls. Abbreviations: Accumulated distance, length of the total cell path; CoM, centre of mass; Dir, directness of osteoblast migration; Euclidean distance, length of a straight line from the cell starting point to the cell endpoint; Forward migration index (FMI), perpendicular (FMI⊥) and parallel (FMI∥) migration direction; hMSC, human mesenchymal stem cell;−/− hMSCs, without hMSC coculture, negative controls; + /− hMSCs, unilateral hMSC coculture, test groups; + / + hMSCs, bilateral hMSC coculture, positive controls; Rayleigh distribution (distr.), uniformity of the cell endpoint distribution; VEGF, vascular endothelial growth factor; velocity: osteoblasts migration speed. Open in new tab The osteoblast migration was analyzed according to the parameters tested for VEGF-only supply, that is, FMI∥, FMI⊥, Rayleigh distribution, CoM values, and migration directness and velocity. Of note, the alveolar osteoblasts migrated towards hMSCs in unidirectional cocultures, while their migration was not oriented to a specific side in negative and positive controls, respectively (Fig. 3A; Table 1). This observation was verified by the Rayleigh distribution which was significant for osteoblasts only in unidirectional hMSC cocultures, while the cells migrated in a nondirected manner in negative and positive controls (Fig. 3A; Table 1). In addition, the CoM-values were found to be elevated for osteoblasts in unidirectional hMSC cocultures compared with controls (Fig. 3A; Table 1). These data show for the first time that hMSCs act as chemoattractors for alveolar bone cells. The directness of osteoblast migration was calculated by comparing the Euclidean distance, that is, the length of a straight line from the cell starting point to the cell endpoint, to the accumulated distance, that is, the length of the total cell path. The globally small directness values indicated that osteoblasts migrated in a meandering manner, even though being attracted by hMSCs in interactive cocultures (Fig. 3A; Table 1). The velocity analysis showed that neither unidirectional nor bidirectional hMSC coculture enhanced the osteoblast migration speed. Thus, hMSCs do not further stimulate the per se good migration performance, but navigate the migration orientation and hence attract alveolar osteoblasts. As hMSCs were found to successfully attract osteoblasts, the final step was to prove whether VEGF, showing a strong mRNA expression and protein release, is causative for the observed alveolar osteoblast chemotaxis. For this purpose, VEGF communication was inhibited in hMSC cocultures and controls, and the chemotactic response of alveolar osteoblasts was recorded. In fact, inhibiting the VEGF receptor entirely reversed the chemotactic response of osteoblasts triggered by hMSCs, that is, the cells migrated readily while completely losing their orientation towards hMSCs in unidirectional cocultures (Fig. 3B; Table 1B). This was confirmed by all migration parameters tested, because both |FMIǁ| and |FMI⊥| were close to zero, and Rayleigh tests revealed that the cell endpoint distribution was nonsignificant, that is, cells were homogeneously distributed, while lacking any orientation preference (Fig. 3B; Table 1B). With respect to the accumulated distance and the velocity, it was found that VEGF inhibition did not accelerate the osteoblast migration (Table 1B). These findings together with the mRNA expression and biomolecule release data clearly prove VEGF release as causative mechanism in hMSC-induced hOA chemoattraction. Discussion Although cell-based therapies using hMSCs are promising for alveolar bone regeneration purposes [1, 35], the underlying cellular interactions as well as the molecular mechanisms still remain elusive. Finally, evidence is rising that the formation of new tissue may largely rely on the attraction and activation of host tissue cells [7, 10, 23]. In this regard, we found that hMSCs express mRNAs of a panel of biomolecules for which osteoblasts are known to be responsive [24, 25, 27]. Nevertheless, we showed that only a few of them, namely VEGF and OPG, are released as proteins at noteworthy amounts. More important, our results show for the first time that hMSCs do in fact attract alveolar osteoblasts, and finally that the osteoblast chemotaxis is mediated by hMSC-released VEGF. Intriguingly, the biomolecules that are supposed to enhance the migratory activity and induce chemotaxis of osteoblasts are established so far rather in the context of cell differentiation than migration [24, 25, 27, 36, 37]. The broad spectrum of biomolecules which were transcribed and released by hMSCs revealed two critical issues: first, the genes that were regulated by time showed only small expression levels, and secondly, the gene expression levels did not necessarily coincide with the respective protein release magnitudes. The finding that the release of VEGF, TGFβ1, and MCSF was increased by time progression while the respective gene expression remained constant suggests that the relative amounts of other biomolecules may have substantially decreased with respect to the total protein amount. This applies for example for PgE2 and RANKL, and possibly, this may also reflect the measured downregulation of GDF5 and Wnt10b mRNAs. By all means, the significant increase of VEGF and MCSF emphasizes their meaningfulness as hMSC signals, but regarding the total protein amounts with VEGF roughly at a 100-fold higher magnitude than MCSF, we conclude that VEGF plays a pivotal role for hMSC communication. Admittedly, the measured biomolecule levels do not necessarily exclude the possibility of some of the lower expressed factors including MCSF having chemotactic properties. Nevertheless, the assumption that VEGF is the key is substantiated by the finding that inhibition of the VEGF receptor abolishes the hMSC-triggered chemotaxis of alveolar osteoblasts. The concept that VEGF promotes cell migration is well established for endothelial cells [38], but is new for osteoblast chemotaxis in the context of hMSC interaction. Among the multiple mechanisms by which VEGF is reported to support endothelial cell migration [39-41], the best established mode is the VEGF receptor-mediated actin filament reorganization involving phosphatidylinositol-3 kinase [42] and p38 mitogen-activated protein kinase signal transduction [43] as well as Src-related focal adhesion turnover [44, 45]. Although evidence is rising that other cell types than endothelial cells are equipped with VEGF receptors, data concerning the responsiveness of osteoblasts to VEGF are still conflicting [46]. Our novel finding that VEGF receptor tyrosine kinase inhibition prevented the hMSC-induced chemoattraction of osteoblasts supports evidence that the latter do sense VEGF similar to long bone-derived osteoblasts [24, 47]. In this light, the observation that VEGF inhibition leads to a stochastic and nondirected mode of osteoblast migration may possibly be interpreted as a result of a modified actin fiber assembly and altered focal adhesion turn-over [43, 44]. Regarding OPG, its release differed with respect to individual hMSC donors [48]; however its levels constantly exceeded RANKL ratios, thereby suggesting that hMSCs may support bone formation [18]. Nevertheless, OPG signaling is not essential for hMSC-related chemoattraction of alveolar osteoblasts, although indirect OPG-related effects on cell migration may possibly exist while being attributed to RANKL-mediated NFкB inducement [49, 50]. With regard to all other detected biomolecules, especially GDF5 is remarkable because it was recently identified as a new therapeutic tool for periodontal regeneration purposes [51]. We found that hMSCs expressed only small amounts of GDF5 mRNA, and that its levels even decreased with prolonged culture time. In other words, although GDF5 is reported to promote bone formation and support periodontal regeneration [51], it seems to be of minor importance within the repertoire of hMSC-innate molecules which drive the chemoattraction of alveolar osteoblasts, observed in the present study. Conclusion The current findings do not only answer the question whether hMSCs are capable of attracting alveolar osteoblasts, but do further provide information regarding the precise inter-cellular communication signals. This issue is captivating because it shows for the first time that hMSCs act as chemoattractors for alveolar osteoblasts, which adds to the existing body of evidence that bone regeneration essentially involves the reactions of host tissue cells to prospectively administered hMSCs [7, 10]. In addition, our data emphasize that on the mechanistic level, the phenomenon of hMSC-related chemoattraction of osteoblasts is driven by VEGF communication. The current results substantially contribute to the understanding of cellular interactions during hMSC-based alveolar bone regeneration. This knowledge will be precious for anticipating or even navigating the clinical results of hMSC administration in prospective therapeutic alveolar bone preservation and regeneration strategies. Acknowledgments We thank Heike Jahnke for her excellent technical assistance and the staff of the Life Imaging Center (LIC) in the Center for Biological Systems Analysis (ZBSA) of the Albert-Ludwigs-University Freiburg for help with their microscopy resources and support in image recording. This work was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft DFG, PR-1481/1) and in part by the Medical Faculty of Freiburg University (ProTissueMat). Author contributions: S.P.: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing; G.B.: provision of study materials/biopsies; K.V.: data analysis; R.N.: data collection, final approval; P.T.: manuscript writing, final approval; E.H.: financial support, final approval. Disclosure of Potential Conflicts of Interest The authors indicate no potential conflicts of interest. References 1 Kawaguchi H , Hirachi A, Hasegawa N et al. Enhancement of periodontal tissue regeneration by transplantation of bone marrow mesenchymal stem cells . J Periodontol 2004 ; 75 : 1281 - 1287 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Ohazama A , Modino SA, Miletich I et al. Stem-cell-based tissue engineering of murine teeth . J Dent Res 2004 ; 83 : 518 - 522 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Lee T . Stem cell therapy independent of stemness . World J Stem Cells 2012 ; 4 : 120 - 124 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Jones E , Yang X. Mesenchymal stem cells and bone regeneration: Current status . Injury 2011 ; 42 : 562 - 568 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Choi JK , Hwang HI, Jang YJ. The efficiency of the in vitro osteo/dentinogenic differentiation of human dental pulp cells, periodontal ligament cells and gingival fibroblasts . Int J Mol Med 2015 ; 35 : 161 - 168 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Mastri M , Lin H, Lee T. Enhancing the efficacy of mesenchymal stem cell therapy . World J Stem Cells 2014 ; 6 : 82 - 93 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Prockop DJ . Repair of tissues by adult stem/progenitor cells (MSCs): Controversies, myths, and changing paradigms . Mol Ther 2009 ; 17 : 939 - 946 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Manuguerra-Gagne R , Boulos PR, Ammar A et al. Transplantation of mesenchymal stem cells promotes tissue regeneration in a glaucoma model through laser-induced paracrine factor secretion and progenitor cell recruitment . Stem Cells 2013 ; 31 : 1136 - 1148 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Hughes FJ . Mesenchymal stem cells and periodontal regeneration . Curr Oral Health Rep 2014 ; 1 : 1 - 8 . Google Scholar Crossref Search ADS WorldCat 10 Tolar J , Le Blanc K, Keating A et al. Concise review: Hitting the right spot with mesenchymal stromal cells . Stem Cells 2010 ; 28 : 1446 - 1455 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Hoch AI , Binder BY, Genetos DC et al. Differentiation-dependent secretion of proangiogenic factors by mesenchymal stem cells . PloS One 2012 ; 7 : e35579 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Wang Z , Wang Y, Wang Z et al. Engineered mesenchymal stem cells with enhanced tropism and paracrine secretion of cytokines and growth factors to treat traumatic brain injury . Stem Cells 2015 ; 33 : 456 - 467 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Ando Y , Matsubara K, Ishikawa J et al. Stem cell-conditioned medium accelerates distraction osteogenesis through multiple regenerative mechanisms . Bone 2014 ; 61 : 82 - 90 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Osugi M , Katagiri W, Yoshimi R et al. Conditioned media from mesenchymal stem cells enhanced bone regeneration in rat calvarial bone defects . Tissue Eng Part A 2012 ; 18 : 1479 - 1489 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Katagiri W , Osugi M, Kawai T et al. Novel cell-free regeneration of bone using stem cell-derived growth factors . Int J Oral Maxillofac Implants 2013 ; 28 : 1009 - 1016 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Chen L , Tredget EE, Wu PY et al. Paracrine factors of mesenchymal stem cells recruit macrophages and endothelial lineage cells and enhance wound healing . PloS One 2008 ; 3 : e1886 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Matsuo K , Irie N. Osteoclast-osteoblast communication . Arch Biochem Biophys 2008 ; 473 : 201 - 209 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Walsh MC , Choi Y. Biology of the RANKL-RANK-OPG system in immunity, bone, and beyond . Frontiers Immunol 2014 ; 5 : 511 . Google Scholar Crossref Search ADS WorldCat 19 Proksch S , Steinberg T, Vach K et al. Shaping oral cell plasticity to osteogenic differentiation by human mesenchymal stem cell coculture . Cell Tissue Res 2014 ; 356 : 159 - 170 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Mundy GR , Boyce B, Hughes D et al. The effects of cytokines and growth factors on osteoblastic cells . Bone 1995 ; 17 : 71s - 75s . Google Scholar Crossref Search ADS PubMed WorldCat 21 Lind M , Deleuran B, Thestrup-Pedersen K et al. Chemotaxis of human osteoblasts. Effects of osteotropic growth factors . APMIS 1995 ; 103 : 140 - 146 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Fiedler J , Roderer G, Gunther KP et al. BMP-2, BMP-4, and PDGF-bb stimulate chemotactic migration of primary human mesenchymal progenitor cells . J Cell Biochem 2002 ; 87 : 305 - 312 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Hughes FJ , Ghuman M, Talal A. Periodontal regeneration: A challenge for the tissue engineer? Proc Inst Mech Eng H, 2010 ; 224 : 1345 - 1358 . Google Scholar Crossref Search ADS WorldCat 24 Mayr-Wohlfart U , Waltenberger J, Hausser H et al. Vascular endothelial growth factor stimulates chemotactic migration of primary human osteoblasts . Bone 2002 ; 30 : 472 - 477 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Dirckx N , Van Hul M, Maes C. Osteoblast recruitment to sites of bone formation in skeletal development, homeostasis, and regeneration . Birth Def Res 2013 ; 99 : 170 - 191 . Google Scholar Crossref Search ADS WorldCat 26 Caplan AI . Adult mesenchymal stem cells for tissue engineering versus regenerative medicine . J Cell Physiol 2007 ; 213 : 341 - 347 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Sood S , Gupta S, Mahendra A. Gene therapy with growth factors for periodontal tissue engineering—A review . Med Oral Patol Oral Cir Bucal 2012 ; 17 : e301 - 310 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 28 Maes C , Goossens S, Bartunkova S et al. Increased skeletal VEGF enhances β-catenin activity and results in excessively ossified bones . EMBO J 2010 ; 29 : 424 - 441 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Asano YH , E. Instructions chemotaxis and migration tool 2.0 . ibidi GmbH 2011;Version 1.0:1-21. 30 Bianco P , Robey PG, Simmons PJ. Mesenchymal stem cells: Revisiting history, concepts, and assays . Cell Stem Cell 2008 ; 2 : 313 - 319 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Dominici M , Le Blanc K, Mueller I et al. Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement . Cytother 2006 ; 8 : 315 - 317 . Google Scholar Crossref Search ADS WorldCat 32 Ruschke K , Hiepen C, Becker J et al. BMPs are mediators in tissue crosstalk of the regenerating musculoskeletal system . Cell Tissue Res 2012 ; 347 : 521 - 544 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Moore YR , Dickinson DP, Wikesjo UM. Growth/differentiation factor-5: A candidate therapeutic agent for periodontal regeneration?. A review of pre-clinical data . J Clin Periodontol 2010 ; 37 : 288 - 298 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Devescovi V , Leonardi E, Ciapetti G et al. Growth factors in bone repair . Chir Organi Mov 2008 ; 92 : 161 - 168 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Hynes K , Menicanin D, Gronthos S et al. Clinical utility of stem cells for periodontal regeneration . Periodontol 2000 2012 ; 59 : 203 - 227 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Zhang W , Zhu C, Wu Y et al. VEGF and BMP-2 promote bone regeneration by facilitating bone marrow stem cell homing and differentiation . Eur Cells Mater 2014 ; 27 : 1-1 - 12 . Google Scholar OpenURL Placeholder Text WorldCat 37 Anitua E , Tejero R, Zalduendo MM et al. Plasma rich in growth factors promotes bone tissue regeneration by stimulating proliferation, migration, and autocrine secretion in primary human osteoblasts . J Periodontol 2013 ; 84 : 1180 - 1190 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Matsumoto T , Claesson-Welsh L. VEGF receptor signal transduction . Sci STKE 2001 ; 2001 : re21 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Oommen S , Gupta SK, Vlahakis NE. Vascular endothelial growth factor A (VEGF-A) induces endothelial and cancer cell migration through direct binding to integrin {alpha}9{beta}1: Identification of a specific {alpha}9{beta}1 binding site . J Biol Chem 2011 ; 286 : 1083 - 1092 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Wang Y , Zang QS, Liu Z et al. Regulation of VEGF-induced endothelial cell migration by mitochondrial reactive oxygen species . Am J Physiol Cell Physiol 2011 ; 301 : C695 - 704 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Wang S , Li X, Parra M et al. Control of endothelial cell proliferation and migration by VEGF signaling to histone deacetylase 7 . Proc Natl Acad Sci USA 2008 ; 105 : 7738 - 7743 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Morales-Ruiz M , Fulton D, Sowa G et al. Vascular endothelial growth factor-stimulated actin reorganization and migration of endothelial cells is regulated via the serine/threonine kinase Akt . Circ Res 2000 ; 86 : 892 - 896 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Olsson AK , Dimberg A, Kreuger J et al. VEGF receptor signalling—In control of vascular function . Nature Rev Mol Cell Biol 2006 ; 7 : 359 - 371 . Google Scholar Crossref Search ADS WorldCat 44 Rousseau S , Houle F, Kotanides H et al. Vascular endothelial growth factor (VEGF)-driven actin-based motility is mediated by VEGFR2 and requires concerted activation of stress-activated protein kinase 2 (SAPK2/p38) and geldanamycin-sensitive phosphorylation of focal adhesion kinase . J Biol Chem 2000 ; 275 : 10661 - 10672 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Zachary I , Gliki G. Signaling transduction mechanisms mediating biological actions of the vascular endothelial growth factor family . Cardiovasc Res 2001 ; 49 : 568 - 581 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Clarkin CE , Gerstenfeld LC. VEGF and bone cell signalling: An essential vessel for communication? Cell Biochem Funct 2013 ; 31 : 1 - 11 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Jacobsen KA , Al-Aql ZS, Wan C et al. Bone formation during distraction osteogenesis is dependent on both VEGFR1 and VEGFR2 signaling . J Bone Mineral Res 2008 ; 23 : 596 - 609 . Google Scholar Crossref Search ADS WorldCat 48 Lepperdinger G . Developmental programs are kept alive during adulthood by stem cells: The aging aspect . Exp Gerontol 2013 ; 48 : 644 - 646 . Google Scholar Crossref Search ADS PubMed WorldCat 49 Dougall WC . RANKL signaling in bone physiology and cancer . Curr Opin Support Palliat Care 2007 ; 1 : 317 - 322 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Ando K , Mori K, Redini F et al. RANKL/RANK/OPG: Key therapeutic target in bone oncology . Curr Drug Discov Technol 2008 ; 5 : 263 - 268 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Lee J , Wikesjo UM. Growth/differentiation factor-5: Pre-clinical and clinical evaluations of periodontal regeneration and alveolar augmentation—Review . J Clin Periodont 2014 ; 41 : 797 - 805 . Google Scholar Crossref Search ADS WorldCat © 2015 AlphaMed Press This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - hMSC-Derived VEGF Release Triggers the Chemoattraction of Alveolar Osteoblasts
JF - Stem Cells
DO - 10.1002/stem.2119
DA - 2015-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/hmsc-derived-vegf-release-triggers-the-chemoattraction-of-alveolar-BPXezTXBF2
SP - 3114
EP - 3124
VL - 33
IS - 10
DP - DeepDyve
ER -