The homing ability and secretory function of mesenchymal stem cells (MSCs) are key factors that influence cell involvement in wound repair. These factors are controlled by multilayer regulatory circuitry, including adhesion molecules, core transcription factors (TFs) and certain other regulators. However, the role of adhesion molecules in this regulatory circuitry and their underlying mechanism remain undefined. In the present paper, we demonstrate that an adhesion molecule, junction adhesion molecule A (JAM-A), may function as a key promoter molecule to regulate skin wound healing by MSCs. In in vivo experiments, we show that JAM-A up-regulation promoted both MSC homing to full-thickness skin wounds and wound healing-related cytokine secretion by MSCs. In vitro experiments also showed that JAM-A promoted MSC proliferation and migration by activating T-cell lymphoma invasion and metastasis 1 (Tiam1). We suggest that JAM-A up-regulation can increase the proliferation, cytokine secretion and wound-homing ability of MSCs, thus accelerating the repair rate of full-thickness skin defects. These results may provide insights into a novel and potentially effective approach to improve the efficacy of MSC treatment.

CLINICAL PERSPECTIVES

  • Patients with large skin wounds show a long wound-healing process which may exaggerate this population's complications. Experimental evidence suggests that migration and paracrine abilities of MSCs can improve the healing rate and quality of wound healing.

  • The results of the present study show that JAM-Aov MSCs can improve the skin wound-healing rate, and promote skin appendages, e.g. hair follicle reformation.

  • These findings suggest that, in wound healing, JAM-Aov MSCs could represent a new gene-modified stem-cell therapeutic strategy. The application in a clinical setting may not be that far away, with the positive effects of JAM-Aov MSCs being visible after a few days of treatment.

INTRODUCTION

Mesenchymal stem cells (MSCs) are a group of adult stem cells with the ability for self-renewal and potential multidirectional differentiation. These cells can be induced to differentiate into various cell types and to participate in the repair and reconstruction of damaged tissues [1,2].

Studies have demonstrated that the homing ability and secretory function of MSCs are key factors controlling cell participation in wound repair [3]. Xie et al. [4] combined tumour necrosis factor α (TNF-α) and TNF-β and interferon (IFN-β) β and IFN-γ, and found that the in vitro migratory ability of MSCs was increased, but that the effect of migrating to the affected site was ‘not sufficient to enhance wound repair’. In another study, Cheng et al. [5] reported that MSCs with high expression of chemokine receptor 4 (CXCR4) administered via the tail vein after transfection homed to the ischaemic myocardium better than control cells did. However, with an increase in CXCR4 expression, the risk of developing atherosclerosis also increases [6]. Although gene transfection to increase the expression level of integrins on MSCs can promote MSC migration to the bone marrow, it may also induce the transformation of MSCs into epidermal cells and increase the potential risk of tumorigenesis [7]. The use of granulocyte–colony-stimulating factor (G-CSF) and stem cell factor (SCF) can also induce homing of MSCs, but it does not seem to promote wound healing effectively [8,9]. This limitation is most probably because the wound repair ability of MSCs is not only related to the number of homing MSCs in the local wound but also closely related to the cells’ physiological activity and secretory function, e.g. certain researchers injected MSCs directly into a wound to increase their number and found that the outcome of wound healing was unsatisfactory, most probably because the MSCs were unable to exert their physiological secretory function [10]. Thus, we wondered whether there is a factor or method that can both promote homing and enhance the wound-healing function of MSCs.

Junction adhesion molecule A (JAM-A), also known as JAM-1, JAM1 or F11R, represents a group of cell adhesion molecules independent of calcium ions that belong to the cell immunoglobulin (Ig) superfamily. One other study [11] has found that JAM-A participates in angiogenesis by mediating the movement and migration of CD34+ haematopoietic progenitor cells to damaged vascular walls, inducing them to differentiate into vascular endothelial cells. Our previous study [12] also demonstrated that over-expression of JAM-A could not only promote MSC migration to diseased follicular tissue but also enhance the ability of the cells to differentiate into follicular tissue. In the present study, we also found and confirmed that the proliferative activity and secretory function of MSCs at the surface of a wound increased significantly when the cells were transfected with constructs that up-regulated JAM-A. In particular, JAM-A promoted the chemotaxis of MSCs, helping them to repair full-thickness skin defects.

MATERIALS AND METHODS

Reagents and antibodies

The reagents and antibodies used in this study are as follows: rabbit anti p-T-cell lymphoma invasion and metastasis 1 (Tiam1), rabbit anti-epidermal growth factor (EGF), goat anti-Tiam1, rabbit anti-p-Rac1, mouse anti-stromal cell-derived factor (SDF)-1a and rabbit anti-Rac1 polyclonal antibodies (Abcam); rabbit anti-JAM-A, rabbit anti-keratin, goat anti-CD31, mouse anti-CD11b and rabbit anti-fibroblast growth factor (FGF)-2 polyclonal antibodies (Santa Cruz); mouse anti-β-actin monoclonal antibody and secondary antibodies (Invitrogen); an immunohistochemistry staining kit and a diaminobenzidine horseradish peroxidase (HRP) Color Development Kit (Boster Biological Technology); transwell chamber and cell culture materials (Corning Costar); media, sera and antibiotics (Gibco, Invitrogen); SDF-1a [R&D Systems (350-NS-050)]; Tiam1 siRNA [Santa Cruz Biotechnology (C36670)]; the Rac inhibitor NSC23766 (Tocris Bioscience); and the YF-Tiam1 expression plasmid (Addgene) [13].

Isolation and culture of MSCs

Human MSCs were derived from a single 5-week embryo limb bud and somite. The MSCs were of low immunogenicity, had multiple differentiation and were engrafted [12]. The MSCs were grown as described previously and stored at the Department of Histology and Embryology of the Second Military Medical University [12].

Lentivirus-mediated transduction of MSCs

JAM-A was obtained by PCR amplification of wild-type cDNA (Shanghai GeneChem), using the following set of primers for JAM-A (both 5′ and 3′): sense: CCTGAAGCTTATGGGGACAAAGGC; antisense: ACCAGGATCCAACACCAGGAATGACGAGG. The cDNA was then subcloned into the pGC-FU lentiviral vector using the AgeI restriction enzyme (Shanghai GeneChem), and the recombinant virus was packaged using Lentivector Expression Systems (Shanghai GeneChem).

The pFU-GW-iRNA plasmid was used to generate a JAM-A shRNA-expressing lentivirus. The sequence was as follows: 5′-AAAGATGGGATAGTGATGCCT-3′ [14]. The negative-control scrambled sequence was 5′-TTCTCCGAACGTGTCACG-3′ [14].

MSCs in their fifth passage were transfected with pGC-FU-JAM-A-GFP (JAM-Aov MSCs) or pFU-GW-JAM-A-RNAi-GFP (JAM-Akd MSCs), or with pGC-FU-GFP {VEC [VEC-MSC injection mice group (vector control for JAM-A over-expression)]-MSCs} or pFU-GW-scrambled-shRNA-GFP {SCR [SCR-MSC injection mice group (scrambled siRNA control for JAM-A knockdown)]-MSCs} lentivirus alone, using enhanced infection solution, and were cultured in Dulbecco's modified Eagle's medium, low glucose (DMEM-LG or Gibco, Invitrogen) supplemented with 10% FBS (Invitrogen). At passage 7, MSCs were analysed for purity and epitope expression using fluorescence-activated cell sorting (FACS) analysis [12]. MSC proliferation was examined using a Cell Counting Kit (CCK)-8, assay.

Scratch wound assays

When confluent, an average of 6 units of MSCs was added to a six-well plate in triplicate. Vertical scratches were then made using a 200-μl plastic filter tip to create a ‘wound’ of approximate diameter 200 μm. To eliminate dislodged cells, the culture medium was removed, and the wells were washed with PBS. The wells were examined every 48 h and photomicrographs taken with a Nikon Eclipse Ti. The migrated cell was measured on photomicrographs by the Image-Pro Plus software, using the same area of the well for each measurement [15].

MSCs and epidermal cells in co-culture experiment

JAM-Aov MSCs, VEC-MSCs, JAM-Akd MSCs and SCR-MSCs were seeded on to filters on a 12-well transwell plate that was coated with Matrigel (BD Biosciences); an average of 12 units of epidermal cells was added to a 12-well plate in triplicate, when confluent, and then a 200-μl plastic filter tip was used to create a ‘wound’ of approximate diameter 200 μm. Dislodged cells were eliminated with PBS and examined 48 h later. Wound width was measured on photomicrographs; invasion of the MSCs through the Matrigel to the underside of the filter was assessed 48 h later by staining with Crystal Violet and counting under a bright-field microscope. For the scratch wound assay, epidermal cells were labelled with green fluorescent protein (GFP), and counted with Image-Pro Plus software. These epidermal cells were isolated by digestion of human foreskin and cultured in vitro.

Invasion assay

JAM-Aov MSCs, VEC-MSCs, JAM-Akd MSCs and SCR-MSCs were seeded on to filters in a 24-well transwell plate that was coated with Matrigel. Invasion of the cells was assessed 48 h later. Recombinant human SDF-1a at 100 ng/ml was added to the lower compartment [16].

Co-immunoprecipitation

The co-immunoprecipitation experiments were performed as described by Čajánek et al. [17]. Briefly, for the immunoprecipitation of JAM-A-containing complexes, cell extracts were prepared in a specific lysis buffer. Cell extracts for the immunoprecipitation of talin-containing complexes were prepared in a distinct lysis buffer. After centrifugation (10 000 g, 10 min at 4°C) and Bradford titration, 500 μg of protein was incubated for 4 h at 4°C on an orbital agitator in the presence of anti-JAM-A, anti-Tiam1, anti-Rac1 or non-specific IgGs.

Experimental wound preparation and analysis

All animal experiments were performed according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. The study was approved by the Institutional Animal Care and Use Committee of the Second Military Medical University, Shanghai, China [certificate SYXK2007-0003 (Shanghai)]. Male BALB/c-nu/nu mice, aged 6 weeks and weighing 16–20 g, were randomly assigned to 5 groups, with 20 animals in each group. Full-thickness skin wounds were created in the dorsal skin under sterile conditions, as described by Tian et al. [10]. MSCs or saline was injected via the tail vein 24 h after the wounding.

The groups included a saline injection group [saline (saline injection mice group) group], a JAM-Aov MSC injection group {OV [JAM-Aov MSC injection mice group (JAM-A over-expression)] group}, a VEC-MSC injection group (VEC group), a JAM-Akd MSC injection group {KD [JAM-Akd MSC injection mice group (JAM-A knockdown)] group} and an SCR-MSC injection group (SCR group). In total, 5×106 cells in a volume of 150 μl were injected.

The investigators measuring samples were blinded to group and treatment. Digital photographs of wounds were taken at days 3 and 7. Wound closure time was defined as the time at which the wound bed was completely re-epithelialized and filled with new tissue. Skin wound area was measured by tracing the wound margin and calculated using an image analysis program (Image-Pro Plus). The percentage of wound closure was calculated as follows:

Area of original wound − Area of actual wound)/Area of original wound × 100.

Mononuclear cell isolation and flow cytometric analysis

The skin wound and 4 mm of the surrounding skin were harvested. Mononuclear cells were isolated from wound tissue homogenates, bone marrow, spleen or kidney, as described in the literature [18,19]. For quantification of GFP+ MSC engraftment in wounded skin, four to six wounds and 4 mm of surrounding skin were excised at day 7 post-wounding. Tissue was minced and incubated in 0.2% collagenase for 1 h at 37°C. The wounds of saline injection mice were used as controls. The resulting single-cell suspensions were counted by FACS Calibur flow cytometer (BD Biosciences), and the data obtained were analysed using CellQuest Pro software (BD Biosciences). To validate the MSC counts obtained by FACS analysis, several calibration steps were employed to distinctly gate the GFP-labelled human MSCs at FL1 mode and count them. We gated out first ‘Lin-neg’ cells in the single-cell suspensions of the saline group, and then from ‘Lin-neg’ we hierarchically gated all other populations. Dead/dying cells collect as a band along the bottom of a forward-scatter (FSC-A) versus a side-scatter (SSC-A) two-parameter plot. Propidium iodide-based FACS was also performed as described by Misra et al. [19].

Histopathological analysis

Mice were sacrificed at 3, 7, 14 and 28 days post-wounding, at which time skin samples, including the wound and 4 mm of the surrounding skin, were harvested. For each wound specimen, one part was subjected to Western blot analysis and the other was used for GFP-positive cell detection by flow cytometric analysis. The remainder was fixed in 4% formaldehyde buffered with PBS and then embedded in paraffin to prepare sections.

The sections were stained with haematoxylin and eosin (H&E) for histopathological analysis or were used for immunohistochemical analysis to evaluate neovascularization, as described by Tian et al. [10]. Epidermal migration at day 7 post-wound shows the distance of epidermal creeping. The length of the new epidermal migration was started from the wound periphery and finished in the newborn skin-piece end. Immunofluorescence analysis was also conducted to identify localization of keratin [12].

CD31+ blood vessel assay

Quantification of blood vessels was performed by determining the number of CD31-positive (+) blood vessels per unit area [10]. Only CD31+ blood vessels that were 2–10 μm in diameter were counted as one vessel. The number of blood vessels was quantified by Image Pro-Plus.

Histological characterization and quantification of inflammation

The presence of inflammatory cells, which were recruited into wound skin at day 7 post-wounding, was quantified in H&E-stained serial sections of paraffin-embedded sections fixed in 4% formaldehyde after enucleation. In addition, for further characterization of inflammatory cells recruited to the skin wound, immunohistochemistry was performed on paraffin sections with the macrophage marker CD11b as described by Alexander et al. [20].

Western blot analysis

Whole-cell lysates were prepared in radioimmunoprecipitation assay lysis buffer. Blots were probed with antibodies against β-actin, JAM-A, Tiam1, p-Tiam1, Rac1, p-Rac1, EGF, SDF-1a and FGF-2. β-Actin was used as the internal reference. HRP-conjugated anti-rabbit, anti-mouse and anti-goat secondary antibodies were then used, after which signals were detected using ECL Plus (GE Healthcare). The band densities were then analysed using Image-Pro Plus software.

Statistical analysis

The data are expressed as the means±S.E.M.s. Differences between two groups or among multiple groups were analysed using a two-tailed Student's t-test and one-way ANOVA, respectively. A P value <0.05 was considered significant [18].

RESULTS

JAM-A up-regulation increases the number of MSCs homing to a skin wound

Human MSCs labelled with GFP (GFP+ cells) were detected in all groups 3 days post-wounding or 2 days after injection into the tail vein (Figure 1B). The number of GFP+ MSCs was the largest in the OV group (564±16), although the number decreased at day 7 (Figure 1B).

JAM-A up-regulation promotes the homing of MSCs to local wounds

Figure 1
JAM-A up-regulation promotes the homing of MSCs to local wounds

(A) The number of MSCs (GFP+) homing to the wound surface in the OV group 7 days post-wounding is significantly greater than in the KD group; there is no significant difference between the VEC and the SCR groups. The GFP+ MSCs were counted by FACS (n=5). (B) At days 3 and 7 post-wounding, GFP+ MSCs were observed at the wound surface in all groups (green, as indicated by the arrows). The number is greatest in the OV group and smallest in the KD group. At 28 days, a small number of GFP- and keratin-positive cells (arrows) were observed, indicating that MSCs are able to transform into epidermal cells. A typical cystic structure is also observed in the OV group (scale bars=100 μm).

Figure 1
JAM-A up-regulation promotes the homing of MSCs to local wounds

(A) The number of MSCs (GFP+) homing to the wound surface in the OV group 7 days post-wounding is significantly greater than in the KD group; there is no significant difference between the VEC and the SCR groups. The GFP+ MSCs were counted by FACS (n=5). (B) At days 3 and 7 post-wounding, GFP+ MSCs were observed at the wound surface in all groups (green, as indicated by the arrows). The number is greatest in the OV group and smallest in the KD group. At 28 days, a small number of GFP- and keratin-positive cells (arrows) were observed, indicating that MSCs are able to transform into epidermal cells. A typical cystic structure is also observed in the OV group (scale bars=100 μm).

MSCs were mainly distributed in the local wound 7 days post-wounding, and were undetectable in the liver, lung, spleen, kidney and bone marrow. In the OV group, 130±16 human cells (GFP+ cells) were detected per 10 000 cells in the local wound, compared with 51±7, 52±8 and 31±4 GFP+ cells in the VEC, SCR and KD groups, respectively. The number of GFP+ MSCs in the OV group was significantly greater than in the VEC group (n=5, P<0.01), whereas there was no significant difference between the VEC and the SCR groups. In addition, the number of GFP+ MSCs in the KD group in the wound was significantly smaller than in the SCR group (Figure 1A).

Preliminary formation of skin-like adnexal structures was observed 4 weeks post-wounding, at the wound surface of the OV group, and double immunofluorescence staining showed that certain cells were positive for both keratin and GFP (Figure 1B). Immature hair follicles have been found in the VEC and SCR groups, indicated by the white arrows in Figure 1(B).

JAM-Aov MSCs accelerate epidermal migration and promote vascularization and wound healing

The wound healing rates at days 3 and 7 post-wounding were 55±3.83% and 82.11±3.53%, respectively, in the OV group, compared with: 49±2.5% and 74.75±2.1%, respectively, in the VEC group; 41±4.365% and 70±3.2%, respectively, in the KD group; and 50±5.84% and 76.75±3.538%, respectively, in the SCR group. These results indicated that the wound-healing rate in the OV group was significantly higher than in the VEC group (n=5, P<0.01). Moreover, the wound-healing rate in the KD group was significantly lower than that in the SCR group (n=5, P<0.01), and there was no significant difference between the VEC and the SCR groups (Figure 2).

JAM-Aov MSCs apparently promote wound healing

Figure 2
JAM-Aov MSCs apparently promote wound healing

(A) Gross observation shows that wound healing is the quickest in animals in the OV group, in which the wound surface becomes smaller 3 and 7 days post-wounding, compared with the other groups. (B) Statistical data show that, 3 and 7 days post-wounding, the wound-healing rate in the OV group is significantly higher than that in the VEC group, the rate in the KD group is lower than that in the SCR group, and there is no significant difference between the VEC and the SCR groups (n=5, *P<0.01, OV vs VEC; #P<0.01, KD vs SCR; P<0.01, KD vs saline).

Figure 2
JAM-Aov MSCs apparently promote wound healing

(A) Gross observation shows that wound healing is the quickest in animals in the OV group, in which the wound surface becomes smaller 3 and 7 days post-wounding, compared with the other groups. (B) Statistical data show that, 3 and 7 days post-wounding, the wound-healing rate in the OV group is significantly higher than that in the VEC group, the rate in the KD group is lower than that in the SCR group, and there is no significant difference between the VEC and the SCR groups (n=5, *P<0.01, OV vs VEC; #P<0.01, KD vs SCR; P<0.01, KD vs saline).

At day 7, the distance of epidermal migration in the OV group was significantly further than that in the VEC group (1134.5±35.35 μm vs 905.5±74.79 μm; n=5, P<0.01) (Figure 3A). The distance of epidermal migration in the KD group was significantly less than that in the SCR group (704±16.26 μm vs 884.5±74.79 μm; n=5, P<0.01) (Figure 3). In contrast, there was no significant difference between the VEC and the SCR groups.

JAM-Aov promotes MSC migration to a local wound surface

Figure 3
JAM-Aov promotes MSC migration to a local wound surface

(A) Epidermal migration is the fastest in the OV group at day 7 post-wounding (arrows); at day 14, the wound surface has basically healed; and at day 28 more cystic structures are observed in the JAM-Aov MSC group (arrow) (scale bars=200 μm). (B) Epidermal migration at day 7 shows that the distance of epidermal creeping in the OV group is significantly longer than in the VEC group, the distance in the KD group is significantly shorter than in the SCR group, and there is no significant difference between the VEC and SCR groups (n=5, &P<0.01).

Figure 3
JAM-Aov promotes MSC migration to a local wound surface

(A) Epidermal migration is the fastest in the OV group at day 7 post-wounding (arrows); at day 14, the wound surface has basically healed; and at day 28 more cystic structures are observed in the JAM-Aov MSC group (arrow) (scale bars=200 μm). (B) Epidermal migration at day 7 shows that the distance of epidermal creeping in the OV group is significantly longer than in the VEC group, the distance in the KD group is significantly shorter than in the SCR group, and there is no significant difference between the VEC and SCR groups (n=5, &P<0.01).

The wounds had healed well 14 days post-wounding, in all groups except the saline group. On day 28 post-wounding, follicular structures were observed in the OV group, and peg-like structures in the VEC and SCR groups (Figure 3A). The neoepidermis in the OV group was the thickest (52.3±5.2 μm), followed by that in the VEC group (43±1.7 μm) and the SCR group (41±1.2 μm). These findings showed that the neoepidermis in each of these groups was thicker than the neoepidermis (17±1.5 μm) in the KD group (n=5, P<0.01).

CD31 immunohistochemical staining showed that the number of blood vessels in the OV group was 23.4±1.5/0.24 mm2 (×400) at day 7 post-wounding, which was higher than the numbers in the VEC group (20±1.58), the SCR group (18.2±3.14) and the KD group (16.2±1.48). The number was the smallest in the saline group (12.4±1.14) (n=5, P<0.01) (Figures 4A and 4C).

Transplantation of JAM-Aov MSCs attenuates the inflammatory response and promotes angiogenesis in local wounds

Figure 4
Transplantation of JAM-Aov MSCs attenuates the inflammatory response and promotes angiogenesis in local wounds

All examinations were performed at 7 days post-wounding. (A, C) The number of the newly formed vessels in the OV group (arrow) is significantly greater than the numbers in the other three groups (n=5, P<0.01). (B, D) The fewest inflammatory cells are in the OV group, and there are more inflammatory cells in the saline group (triangles) (n=5, *P<0.05, **P<0.05). (E, F) In the OV group, EGF, FGF-2 and SDF-1a expression is highest. The next highest expression is in the VEC, SCR, KD and saline groups, and there is no significant difference between the VEC and the SCR groups (n=5, **P<0.01, OV vs VEC; #P<0.05, SCR vs KD; &P<0.05, KD vs saline). The β-actin and objective to protein come from the same sample, and the same volume of sample was used.

Figure 4
Transplantation of JAM-Aov MSCs attenuates the inflammatory response and promotes angiogenesis in local wounds

All examinations were performed at 7 days post-wounding. (A, C) The number of the newly formed vessels in the OV group (arrow) is significantly greater than the numbers in the other three groups (n=5, P<0.01). (B, D) The fewest inflammatory cells are in the OV group, and there are more inflammatory cells in the saline group (triangles) (n=5, *P<0.05, **P<0.05). (E, F) In the OV group, EGF, FGF-2 and SDF-1a expression is highest. The next highest expression is in the VEC, SCR, KD and saline groups, and there is no significant difference between the VEC and the SCR groups (n=5, **P<0.01, OV vs VEC; #P<0.05, SCR vs KD; &P<0.05, KD vs saline). The β-actin and objective to protein come from the same sample, and the same volume of sample was used.

The number of CD11b+ cells in the OV group was 50.25±4.7/0.24 mm2 (×400) at day 7 post-wounding, which was significantly lower than in the VEC group (95.75±4.9). Meanwhile, the number of CD11b+ cells in the KD group was significantly greater than in the SCR group (149±19 vs 99.75±5.7). However, the number of CD11b+ cells in the saline group (170±21) was significantly higher than in the KD group (n=5, P<0.01) (Figures 4B and 4D). CD11b is expressed on myeloid cells and natural killer cells.

The expression of SDF-1a, FGF-2 and EGF in the local wound area was the highest in the OV group 7 days post-wounding compared with expression in the VEC, SCR, KD and saline groups (Figure 4E). There was no significant difference between the VEC and the SCR groups, whereas the expression of SDF-1a, FGF-2 and EGF was significantly lower in the KD group than in the SCR group, and was the lowest in the saline group (Figures 4E and 4F). The housekeeping gene, β-actin, was tested as an internal reference for Western blot in the same sample panel, using Image-Pro Plus software.

JAM-A promotes chemotaxis and migration of MSCs by activating the Tiam1/Rac1 pathway

An MSC scratch test showed that the number of JAM-Aov MSCs undergoing migration was significantly greater than the number of migrating VEC-MSCs (194.3±5.85/3.80 mm2 vs 151±9.84/3.80 mm2; n=3, P<0.01). In addition, the number of JAM-Akd MSCs undergoing migration was significantly lower than the number of migrating SCR-MSCs (140.6±5.5 vs 153.33±6.0; n=3, P<0.05). However, there was no significant difference between VEC-MSC and SCR-MSC migration (Figures 5A and 5B). For the cell scratch experiment, all the cells were labelled with GFP. A photograph was taken 24 or 48 h later, and the migrated cell numbers counted using Image-Pro Plus software.

JAM-A promotes MSC migration by activating the Tiam1/Rac1 pathway

Figure 5
JAM-A promotes MSC migration by activating the Tiam1/Rac1 pathway

(A, B) The migratory ability is the strongest in JAM-Aov MSCs, followed by VEC-MSCs, SCR-MSCs and JAM-Akd MSCs (n=3, **P<0.01, *P<0.05) (scale bars=100 μm). (A, C) The migratory ability of JAM-Aov MSCs decreases after Tiam1 interference or the addition of NSC23766 (n=3, **P<0.01), and Tiam1 over-expression partially reverses the decreased migratory ability of JAM-Akd MSCs (n=3, **P<0.01). (D) The expression of Tiam1 and p-Tiam1 increases with JAM-A up-regulation. (E) Tiam1 over-expression partially restores the expression of Tiam1, p-Tiam1, Rac1 and p-Rac1. (F) Tiam1 interference down-regulates the expression of Tiam1, p-Tiam1, Rac1 and p-Rac1, and NSC23766 down-regulates the expression of Rac1 and p-Rac1. The β-actin and objective to protein come from the same sample, and the same volume of sample was used. (G) Co-immunoprecipitation confirms that JAM-A and Tiam1 can be precipitated together.

Figure 5
JAM-A promotes MSC migration by activating the Tiam1/Rac1 pathway

(A, B) The migratory ability is the strongest in JAM-Aov MSCs, followed by VEC-MSCs, SCR-MSCs and JAM-Akd MSCs (n=3, **P<0.01, *P<0.05) (scale bars=100 μm). (A, C) The migratory ability of JAM-Aov MSCs decreases after Tiam1 interference or the addition of NSC23766 (n=3, **P<0.01), and Tiam1 over-expression partially reverses the decreased migratory ability of JAM-Akd MSCs (n=3, **P<0.01). (D) The expression of Tiam1 and p-Tiam1 increases with JAM-A up-regulation. (E) Tiam1 over-expression partially restores the expression of Tiam1, p-Tiam1, Rac1 and p-Rac1. (F) Tiam1 interference down-regulates the expression of Tiam1, p-Tiam1, Rac1 and p-Rac1, and NSC23766 down-regulates the expression of Rac1 and p-Rac1. The β-actin and objective to protein come from the same sample, and the same volume of sample was used. (G) Co-immunoprecipitation confirms that JAM-A and Tiam1 can be precipitated together.

With the up-regulation of JAM-A expression, the migratory ability of JAM-Aov MSCs increased, and the expression of Tiam1 and p-Tiam1 also increased significantly (Figures 5A, 5B and 5D). However, after Tiam1 siRNA interference, the number of migrating JAM-Aov MSCs decreased 1.42-fold (n=5, P<0.01) and, at the same time, the expression levels of Tiam1, p-Tiam1, total RacGTPase-activating protein 1 (Rac1) and p-Rac1 decreased to a certain extent (Figures 5A, 5C and 5F). After Rac1 blockage, the number of migrating JAM-Aov MSCs decreased 1.39-fold, and the overall expression levels of Rac1 and p-Rac1 in JAM-Aov MSCs also decreased (Figures 5A, 5C and 5F).

After the over-expression of Tiam1 in JAM-Akd MSCs, the expression levels of Tiam1, p-Tiam1, total Rac1 and p-Rac1 increased to varying degrees (Figure 5E). In addition, the number of migrating JAM-Akd MSCs increased 1.37-fold, to 193.6±14.5, compared with the number before over-expression (n=3, P<0.01) (Figures 5A and 5C).

A transwell assay confirmed that the number of JAM-Aov MSCs migrating to the lower compartment of the culture dish was 2.23-fold higher than the number of VEC-MSCs and 2.8-fold higher than the number of JAM-Akd MSCs (47±3.24/0.95 mm2 vs 21±1.58 mm2 and 16.6±2.07 mm2; n=3, P<0.01). The number of migrating JAM-Akd MSCs was also smaller than the number of migrating SCR-MSCs (20.6±2.4) (n=3, P<0.05). In contrast, there was no significant difference between VEC-MSC and SCR-MSC migration. After Tiam1 siRNA interference and Rac1 blockage, the number of migrating JAM-Aov MSCs was significantly down-regulated. After Tiam1 over-expression, the number of migrating JAM-Akd MSCs was increased significantly (results not shown).

The result of co-precipitation showed that the protein complex precipitated by an anti-JAM-A antibody contained Tiam1 and that the complex precipitated by an anti-Tiam1 antibody contained JAM-A. In addition, the band appearing on a Western blot was consistent with the band for total MSC protein (Figure 5G). These results indicate that JAM-A mediates the migration and chemotaxis of MSCs by activating the Tiam1–Rac1 pathway.

To find out the role of JAM-A or Tiam1 in cell migration, we added three test groups to the cell scratch test: JAM-Aov MSCs vs JAM-Aov + Tiam1ov MSCs; JAM-Akd MSCs vs JAM-Akd + Tiam1kd MSCs; and control MSCs (VEC-MSCs) vs Tiam1ov MSCs vs Tiam1kd MSCs. The cell scratch test results showed that, with the up-regulation of JAM-A and Tiam1 expression, the migratory ability of JAM-Aov + Tiam1ov MSCs was better than JAM-Aov MSCs. With the down-regulation of JAM-A and Tiam1 expression, the migratory number of JAM-Akd + Tiam1kd MSCs was lower than that of JAM-Akd MSCs. With Tiam1 over- or under-expression, the ability for cell migration was also up- or down-regulated (see Supplementary Figure S1).

JAM-Aov MSCs enhance the proliferation and migration of epidermal cells

Figure 6(A) is the experimental model for MSCs promoting epidermal cell migration. Under serum-free culture conditions, non-contact co-culture of scratched epidermal cells and MSCs showed that the number of migrating epidermal cells in the JAM-Aov MSC group was significantly greater than that in the VEC-MSC group (560.8±8.78/3.80 mm2 vs 386.4±11.14 mm2; n=3, P<0.01) and the number of migrating epidermal cells in the JAM-Akd MSC group was significantly smaller than that in the SCR-MSC group (327.6±23.09 vs 394.6±16.43; n=3, P<0.01) (Figures 6B and 6D). The number of MSCs in the JAM-Aov MSC group recruited by the scratched epidermal cells which migrated to the lower compartment of the culture dish was 2.2-fold higher than in the VEC-MSC group and 2.5-fold higher than in the JAM-Akd MSC group (35.05±2.19/0.95 mm2 vs 14.03±1.96/0.95 mm2 and 11.36.05±2.07/0.95 mm2; n=3, P<0.01). Moreover, the number of JAM-Akd MSCs was significantly down-regulated compared with the number of SCR-MSCs (14.11±2.04; n=3, P<0.05) (Figures 6B and 6C). JAM-Aov MSCs could promote the migration of epidermal cells and exhibited the strongest ability to migrate after being recruited by scratched epidermal cells. Meanwhile, there was no significant difference between VEC-MSCs and SCR-MSCs.

JAM-Aov MSCs promote the proliferation and migration of epidermal cells

Figure 6
JAM-Aov MSCs promote the proliferation and migration of epidermal cells

(A) The experimental model of the present study. (B, C) The number of JAM-Aov MSCs attracted by the scratched epidermal cells is significantly greater than the number of VEC-MSCs, and the lowest number is observed in the JAM-Akd MSCs (n=3, **P<0.01, *P<0.05) scale bars=100 μm). (B, D) JAM-Aov MSCs promote the migration of epidermal cells more markedly than SCR-MSCs or VEC-MSCs. The promoting effect is the weakest in the JAM-Akd MSCs (n=3, **P<0.01, *P<0.05). (E) Of the epidermal cells co-cultured with JAM-Aov MSCs, 46.13% of cells entered the S phase, which is significantly more than in the other groups. (F) The CCK-8 assay shows that JAM-Aov MSCs promote the proliferation of epidermal cells (n=3, *P<0.05).

Figure 6
JAM-Aov MSCs promote the proliferation and migration of epidermal cells

(A) The experimental model of the present study. (B, C) The number of JAM-Aov MSCs attracted by the scratched epidermal cells is significantly greater than the number of VEC-MSCs, and the lowest number is observed in the JAM-Akd MSCs (n=3, **P<0.01, *P<0.05) scale bars=100 μm). (B, D) JAM-Aov MSCs promote the migration of epidermal cells more markedly than SCR-MSCs or VEC-MSCs. The promoting effect is the weakest in the JAM-Akd MSCs (n=3, **P<0.01, *P<0.05). (E) Of the epidermal cells co-cultured with JAM-Aov MSCs, 46.13% of cells entered the S phase, which is significantly more than in the other groups. (F) The CCK-8 assay shows that JAM-Aov MSCs promote the proliferation of epidermal cells (n=3, *P<0.05).

After 3 days of co-culturing the epidermal cells with MSCs, the number of S-phase cells (46.13±1.4%) was significantly greater in the JAM-Aov MSC group than in the VEC-MSC group (37.35±4.08%), the SCR-MSC group (39.64±3.5%) and the JAM-Akd MSC group (25.35±3.9%) (n=3, P<0.01) (Figure 6E). These results showed that co-culture with JAM-Aov MSCs accelerated the rate of epidermal cell proliferation. The result of a CCK-8 assay further demonstrated that JAM-Aov MSCs promoted the proliferation of epidermal cells (Figure 6F).

JAM-A over-expression enhances proliferative activity and improves the secretory function of MSCs

MSCs highly expressed the MSC-specific antigens CD29, CD44, CD90 and CD106, but did not express haematopoietic lineage antigens CD34 and CD45 before and after adjusting for the JAM-A expression level. At 72 h after viral infection, the green fluorescence in JAM-Aov MSCs was mainly located in the cell membrane and intracellular vesicle system. The green fluorescence was seen in cell bodies in the VEC-MSC, SCR-MSC and JAM-Akd MSC groups (Figures 7A and 7D).

JAM-A up-regulation promotes the proliferation, migration and secretion of MSCs

Figure 7
JAM-A up-regulation promotes the proliferation, migration and secretion of MSCs

(A) High expression of MSC-specific antigen, without expression of haematopoietic linkage-specific antigen. The GFP-positive rate in all MSC groups was >95%. (B, C) FGF-2, EGF and SDF-1a expression in MSCs increases with the increase of JAM-A (n=5, *P<0.01, JAM-Aov MSCs vs VEC-MSCs; #P<0.01, JAM-Akd MSCs vs SCR-MSCs). (D) MSCs are spindle shaped. The fluorescence in the JAM-Aov MSC group is spotty and mainly distributed in the cell membrane. (E, G) JAM-Aov MSCs show the strongest migration ability, followed by SCR-MSCs and VEC-MSCs, and JAM-Akd MSCs are weakest. (F) The proliferative ability of JAM-Aov MSCs is increased markedly, whereas that of JAM-Akd MSCs is decreased (n=3, **P<0.01).

Figure 7
JAM-A up-regulation promotes the proliferation, migration and secretion of MSCs

(A) High expression of MSC-specific antigen, without expression of haematopoietic linkage-specific antigen. The GFP-positive rate in all MSC groups was >95%. (B, C) FGF-2, EGF and SDF-1a expression in MSCs increases with the increase of JAM-A (n=5, *P<0.01, JAM-Aov MSCs vs VEC-MSCs; #P<0.01, JAM-Akd MSCs vs SCR-MSCs). (D) MSCs are spindle shaped. The fluorescence in the JAM-Aov MSC group is spotty and mainly distributed in the cell membrane. (E, G) JAM-Aov MSCs show the strongest migration ability, followed by SCR-MSCs and VEC-MSCs, and JAM-Akd MSCs are weakest. (F) The proliferative ability of JAM-Aov MSCs is increased markedly, whereas that of JAM-Akd MSCs is decreased (n=3, **P<0.01).

Western blot showed that JAM-A expression was increased significantly in the JAM-Aov MSC group, whereas it was reduced markedly in the JAM-Akd MSC group (Figures 7B and 7C). There was no significant difference in the expression of these molecules between VEC-MSCs and SCR-MSCs.

CCK-8 assay showed that the proliferative activity of JAM-Aov MSCs increased markedly compared with VEC-MSCs (n=3, P<0.01), whereas the proliferative activity of JAM-Akd MSCs decreased markedly compared with SCR-MSCs (n=3, P<0.01). There was no significant difference between the VEC-MSC and SCR-MSC groups (Figure 7F).

The transwell test showed that, with up-regulation of JAM-A expression, the migratory ability of MSCs also increased (Figures 7E and 7G).

DISCUSSION

One of the key factors in the process of wound healing is whether MSCs can be mobilized and induced to directionally migrate to the wound surface to play their role. This study was the first to observe that JAM-A can promote the chemotaxis and migration of MSCs by activating the Tiam1/Rac1 pathway, and that up-regulation of JAM-A expression not only increases the number of MSCs homing to a wound surface but also enhances their proliferative activity and secretory ability. These cells subsequently promote the proliferation and migration of epidermal cells into the periphery of the wound, thus accelerating the rate of wound healing.

Up-regulation of JAM-A increases the migration and homing ability of MSCs by activating the Tiam1/Rac1 pathway

Transplantation of stem cells has shown promise in curing multiple difficult diseases and disorders, such as myocardial infarction and liver impairment, among others [21]. However, one of the factors affecting the therapeutic efficacy of stem cells is that transplanted stem cells tend to be retained in non-target organs and are often difficult to directionally migrate to the locality of the affected tissue [22].

In this study, up-regulation of JAM-A improved the ability of MSCs to directionally migrate and home to a wound surface via the peripheral circulation, which significantly improved the ability of the MSCs to promote wound healing. The mechanism underlying JAM-A-mediated homing of MSCs remains unclear. It has been found that cis-dimerization of JAM-A can form a spatial structure that can bind to the PDZ structural and functional domain [23,24], and that Tiam1 has the PDZ functional domain [25]. However, whether JAM-A affects MSC migration via the Tiam1/Rac1 pathway remains unknown. Our results suggest that the up-regulation of JAM-A is closely associated with increased MSC homing and activation of the Tiam1/Rac1 pathway.

JAM-A up-regulation enhances the proliferative activity and secretory abilities of MSCs

In the present study, it was observed that JAM-A over-expression not only improved the directional chemotaxis of MSCs but also enhanced their proliferative activity and secretory ability (see Figure 7). Various active factors endogenously secreted by MSCs play important roles in helping MSCs to repair affected tissues and organs by regulating the extracellular microenvironment, such as IFN-β and vascular endothelial growth factor (VEGF) [26]. However, the above genes could act as vectors to transport the target gene to the wound surface without enhancing the secretory function of MSCs [27].

In vivo animal studies demonstrated that JAM-Aov MSCs retained their strong secretory ability after homing to a wound location and that the EGF, SDF-1a and FGF-2 detected in the local wound microenvironment were mainly secreted by the MSCs that homed to the wound surface (see Supplementary Figure S2). The above cytokines secreted by MSCs may all participate in the process of wound repair [28,29]. However, the pathways that JAM-A uses to promote the secretory ability of MSCs need to be investigated further.

JAM-A up-regulation promotes wound repair by enhancing both the homing and the secretory abilities of MSCs

The action of MSCs in regulating a wound surface is associated with multiple factors, including directional chemotaxis, the inherent proliferative activity and secretory function of the MSCs, and the local wound microenvironment [21,22]. Recently, most researchers have focused their efforts on regulating a single acting factor to improve the tissue-repairing ability of MSCs [26]. However, no study has reported a regulatory factor with multiple functions.

It would be interesting to study the role of MSCs in skin wound healing. In the present study, there was a difference between the groups observed, suggesting that JAM-A is directly implicated in this process of MSC-induced wound healing. Intravenous injection of MSCs or gene-modified MSCs is an acceptable method for clinical application. As MSCs were expanded by demonstrating their capacity for self-renewal and multilineage differentiation, and gene modification can promote migratory ability, directional differentiation and secretion ability, etc., these characteristics are good for various tissue injuries [26,27]. Of course, MSC or JAM-A/MSC treatment depends on the type of wound, and the health status of the patient. In the near future, intravenous injection of transgenic MSCs will be an acceptable method for clinical application.

In summary, we demonstrate that up-regulation of JAM-A not only markedly increases the number of MSCs homing to the wound surface, but also improves the secretory function of MSCs. These effects promote angiogenesis at the wound surface and accelerated epithelialization, as represented by the increased thickness of the neoepidermis after wound healing and the formation of more skin appendage-like structures, thus improving the quality of wound healing. This double effect of the JAM-A gene, enhancing both the homing and the secretory abilities of MSCs, provides a new research foundation for investigating the therapeutic efficacy of MSCs.

AUTHOR CONTRIBUTION

Minjuan Wu, Shizhao Ji and Shichu Xiao conceived and designed the study, and contributed equally to this work. Zhengdong Kong, He Fang, Yunqing Zhang, Kaihong Ji and Yongjun Zheng analysed and interpreted the data. Houqi Liu and Zhaofan Xia led the design and drafted the paper.

FUNDING

This work was supported by China Postdoctoral Science Foundation (No. 2013T60956), National Natural Science Foundation of China (No. 81301649) and Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry (No. 46).

COMPETING INTERESTS

The authors declare that they have no competing interests.

Abbreviations

     
  • CCK

    Cell Counting Kit

  •  
  • CXCR4

    chemokine receptor 4

  •  
  • EGF

    epidermal growth factor

  •  
  • FACS

    fluorescence-activated cell sorting

  •  
  • FGF

    fibroblast growth factor

  •  
  • G-CSF

    granulocyte–colony-stimulating factor

  •  
  • GFP

    green fluorescent protein

  •  
  • H&E

    haematoxylin and eosin

  •  
  • HRP

    horseradish peroxidase

  •  
  • Ig

    immunoglobulin

  •  
  • INF

    interferon

  •  
  • JAM-A

    junction adhesion molecule A

  •  
  • KD

    JAM-Akd MSC injection mice group (JAM-A knockdown)

  •  
  • MSC

    mesenchymal stem cell

  •  
  • OV

    JAM-Aov MSC injection mice group (JAM-A over-expression)

  •  
  • Saline

    saline injection mice group

  •  
  • SCF

    stem cell factor

  •  
  • SCR

    SCR-MSC injection mice group (scrambled siRNA control for JAM-A knockdown)

  •  
  • SDF

    stromal cell-derived factor

  •  
  • Tiam

    T-cell lymphoma invasion and metastasis

  •  
  • TF

    transcription factor

  •  
  • VEC

    VEC-MSC injection mice group (vector control for JAM-A over-expression)

References

References
1
Park
 
D.
Spencer
 
J.A.
Koh
 
B.I.
Kobayashi
 
T.
Fujisaki
 
J.
Clemens
 
T.L.
Lin
 
C.P.
Kronenberg
 
H.M.
Scadden
 
D.T.
 
Endogenous bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and regeneration
Cell Stem Cell
2012
, vol. 
10
 (pg. 
259
-
272
)
[PubMed]
2
Lee
 
C.H.
Shah
 
B.
Moioli
 
E.K.
Mao
 
J.J.
 
CTGF directs fibroblast differentiation from human mesenchymal stem/stromal cells and defines connective tissue healing in a rodent injury model
J. Clin. Invest.
2010
, vol. 
120
 (pg. 
3340
-
3349
)
[PubMed]
3
Guan
 
M.
Yao
 
W.
Liu
 
R.
Lam
 
K.S.
Nolta
 
J.
Jia
 
J.
Panganiban
 
B.
Meng
 
L.
Zhou
 
P.
Shahnazari
 
M.
, et al 
Directing mesenchymal stem cells to bone to augment bone formation and increase bone mass
Nat. Med.
2012
, vol. 
18
 (pg. 
456
-
462
)
[PubMed]
4
Xie
 
C.
Xie
 
D.Y.
Lin
 
B.L.
Zhang
 
G.L.
Wang
 
P.P.
Peng
 
L.
Gao
 
Z.L.
 
Interferon-β gene-modified human bone marrow mesenchymal stem cells attenuate hepatocellular carcinoma through inhibiting AKT/FOXO3a pathway
Br. J. Cancer
2013
, vol. 
109
 (pg. 
1198
-
1205
)
[PubMed]
5
Cheng
 
Z.
Ou
 
L.
Zhou
 
X.
Li
 
F.
Jia
 
X.
Zhang
 
Y.
Liu
 
X.
Li
 
Y.
Ward
 
C.A.
Melo
 
L.G.
Kong
 
D.
 
Targeted migration of mesenchymal stem cells modified with CXCR4 gene to infarcted myocardium improves cardiac performance
Mol. Ther.
2008
, vol. 
16
 (pg. 
571
-
579
)
[PubMed]
6
Burger
 
J.A.
Kipps
 
T.J.
 
CXCR4: a key receptor in the crosstalk between tumor cells and their microenvironment
Blood
2006
, vol. 
107
 (pg. 
1761
-
1767
)
[PubMed]
7
Kumar
 
S.
Ponnazhagan
 
S.
 
Bone homing of mesenchymal stem cells by ectopic alpha 4 integrin expression
FASEB J.
2007
, vol. 
21
 (pg. 
3917
-
3127
)
[PubMed]
8
Kastrup
 
J.
Ripa
 
R.S.
Wang
 
Y.
Jørgensen
 
E.
 
Myocardial regeneration induced by granulocyte–colony-stimulating factor mobilization of stem cells in patients with acute or chronic ischaemic heart disease: a non-invasive alternative for clinical stem cell therapy?
Eur. Heart. J.
2006
, vol. 
27
 (pg. 
2748
-
2754
)
[PubMed]
9
Wuchter
 
P.
Leinweber
 
C.
Saffrich
 
R.
Hanke
 
M.
Eckstein
 
V.
Ho
 
A.D.
Grunze
 
M.
Rosenhahn
 
A.
 
Plerixa for induces the rapid and transient release of stromal cell-derived factor-1 alpha from human mesenchymal stromal cells and influences the migration behavior of human hematopoietic progenitor cells
Cell Tissue. Res.
2014
, vol. 
355
 (pg. 
315
-
326
)
[PubMed]
10
Tian
 
H.
Lu
 
Y.
Shah
 
S.P.
Hong
 
S.
 
14S, 21R-Dihydroxydocosahexaenoic acid remedies impaired healing and mesenchymal stem cell functions indiabetic wounds
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
4443
-
4453
)
[PubMed]
11
Peddibhotla
 
S.S.
Brinkmann
 
B.F.
Kummer
 
D.
Tuncay
 
H.
Nakayama
 
M.
Adams
 
R.H.
Gerke
 
V.
Ebnet
 
K.
 
Tetraspanin CD9 links junctional adhesion molecule-A to αvβ3 integrin to mediate basic fibroblast growth factor-specific angiogenic signaling
Mol. Biol. Cell
2013
, vol. 
24
 (pg. 
933
-
944
)
[PubMed]
12
Minjuan
 
Wu.
Guo
 
X.
Yang
 
L.
Wang
 
Y.
Tang
 
Y.
Yang
 
Y.
Liu
 
H.
 
Mesenchymal stem cells with modification of JAM-A induce hair formation
Stem Cells Transl. Med.
2014
, vol. 
3
 (pg. 
481
-
488
)
[PubMed]
13
Inoue
 
T.
Heo
 
W.D.
Grimley
 
J.S.
Wandless
 
T.J.
Meyer
 
T.
 
An inducible translocation strategy to rapidly activate and inhibit small GTPase signaling pathways
Nat. Methods
2005
, vol. 
2
 (pg. 
415
-
418
)
[PubMed]
14
Mandell
 
K.J.
Babbin
 
B.A.
Nusrat
 
A.
Parkos
 
C.A.
 
Junctional adhesion molecule 1 regulates epithelial cell morphology through effects on beta1 integrins and Rap1 activity
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
11665
-
11674
)
[PubMed]
15
Crosby
 
L.M.
Luellen
 
C.
Zhang
 
Z.
Tague
 
L.L.
Sinclair
 
S.E.
Waters
 
C.M.
 
Balance of life and death in alveolar epithelial type II cells: proliferation, apoptosis, and the effects of cyclic stretch on wound healing
Am. J. Physiol. Lung Cell. Mol. Physiol.
2011
, vol. 
301
 (pg. 
L536
-
L546
)
[PubMed]
16
Aiuti
 
A.
Webb
 
I.J.
Bleul
 
C.
Springer
 
T.
Gutierrez-Ramos
 
J.C.
 
The chemokine SDF-1 is a chemoattractant for human CD34 +hematopoietic progenitor cells and provides a new mechanism to explain the mobilization of CD34+ progenitors to peripheral blood
J. Exp. Med.
1997
, vol. 
185
 (pg. 
111
-
120
)
[PubMed]
17
Čajánek
 
L.
Ganji
 
R.S.
Henriques-Oliveira
 
C.
Theofilopoulos
 
S.
Koník
 
P.
Bryja
 
V.
Arenas
 
E.
 
Tiam1 regulates the Wnt/Dvl/Rac1 signaling pathway and the differentiation of midbrain dopaminergic neurons
Mol. Cell. Biol.
2013
, vol. 
33
 (pg. 
59
-
70
)
[PubMed]
18
Rustad
 
K.C.
Wong
 
V.W.
Sorkin
 
M.
Glotzbach
 
J.P.
Major
 
M.R.
Rajadas
 
J.
Longaker
 
M.T.
Gurtner
 
G.C.
 
Enhancement of mesenchymal stem cell angiogenic capacity and stemness by a biomimetic hydrogel scaffold
Biomaterials
2012
, vol. 
33
 (pg. 
80
-
90
)
[PubMed]
19
Misra
 
S.K.
Naz
 
S.
Kondaiah
 
P.
Bhattacharya
 
S.
 
A cationic cholesterol based nanocarrier forthe delivery of p53-EGFP-C3 plasmid to cancer cells
Biomaterials
2014
, vol. 
35
 (pg. 
1334
-
13346
)
[PubMed]
20
Alexander
 
K.A.
Flynn
 
R.
Lineburg
 
K.E.
Kuns
 
R.D.
Teal
 
B.E.
Olver
 
S.D.
Lor
 
M.
Raffelt
 
N.C.
Koyama
 
M.
Leveque
 
L.
, et al 
CSF-1-dependant donor-derived macrophages mediate chronic graft-versus-host disease
J. Clin. Invest.
2014
, vol. 
124
 (pg. 
4266
-
4280
)
21
Gnecchi
 
M.
He
 
H.
Liang
 
O.D.
Melo
 
L.G.
Morello
 
F.
Mu
 
H.
Noiseux
 
N.
Zhang
 
L.
Pratt
 
R.E.
Ingwall
 
J.S.
, et al 
Paracrine action accounts for marked protection of ischemic heart by Akt-modified mesenchymal stem cells
Nat. Med.
2005
, vol. 
11
 (pg. 
367
-
368
)
[PubMed]
22
Karp
 
J.M.
Leng Teo
 
G.S.
 
Mesenchymal stem cell homing: the devil is in the details
Cell Stem Cell
2009
, vol. 
4
 (pg. 
206
-
216
)
[PubMed]
23
Severson
 
E.A.
Jiang
 
L.
Ivanov
 
A.I.
Mandell
 
K.J.
Nusrat
 
A.
Parkos
 
C.A.
 
Cis-dimerization mediates function of junctional adhesion moleculea
Mol. Biol. Cell
2008
, vol. 
19
 (pg. 
1862
-
1872
)
[PubMed]
24
Severson
 
E.A.
Lee
 
W.Y.
Capaldo
 
C.T.
Nusrat
 
A.
Parkos
 
C.A.
 
Junctional adhesion molecule A interacts with Afadin and PDZ-GEF2 to activate Rap1A, regulate beta1 integrinlevels, and enhance cell migration
Mol. Biol. Cell
2009
, vol. 
20
 (pg. 
1916
-
1925
)
[PubMed]
25
Songyang
 
Z.
Fanning
 
A.S.
Fu
 
C.
Xu
 
J.
Marfatia
 
S.M.
Chishti
 
A.H.
Crompton
 
A.
Chan
 
A.C.
Anderson
 
J.M.
Cantley
 
L.C.
 
Recognition of unique carboxyl-terminal motifs by distinct PDZ domains
Science
1997
, vol. 
275
 (pg. 
73
-
77
)
[PubMed]
26
Moon
 
H.H.
Joo
 
M.K.
Mok
 
H.
Lee
 
M.
Hwang
 
K.C.
Kim
 
S.W.
Jeong
 
J.H.
Choi
 
D.
Kim
 
S.H.
 
MSC-based VEGF gene therapy in rat myocardial infarction model using facial amphipathic bile acid-conjugated polyethylenei-mine
Biomaterials
2014
, vol. 
35
 (pg. 
1744
-
1754
)
[PubMed]
27
Mundra
 
V.
Wu
 
H.
Mahato
 
R.I.
 
Genetically modified human bone marrow derived mesenchymal stem cells for improving the outcome of human islet transplantation
PLoS One
2013
, vol. 
8
 pg. 
e77591
 
[PubMed]
28
Song
 
M.
Jang
 
H.
Lee
 
J.
Kim
 
J.H.
Kim
 
S.H.
Sun
 
K.
Park
 
Y.
 
Regeneration of chronic myocardial infarction by injectable hydrogels containing stem cell homing factor SDF-1 and angiogenic peptide Ac-SDKP
Biomaterials
2014
, vol. 
35
 (pg. 
2436
-
2445
)
[PubMed]
29
Kottakis
 
F.
Polytarchou
 
C.
Foltopoulou
 
P.
Sanidas
 
I.
Kampranis
 
S.C.
Tsichlis
 
P.N.
 
FGF-2 regulates cell proliferation, migration, and angiogenesis through an NDY1/KDM2B-miR-101-EZH2 pathway
Mol. Cell
2011
, vol. 
43
 (pg. 
285
-
298
)
[PubMed]

Supplementary data