hUC-MSCs were provided by Dr. Tokiko Nagamura (Tokyo University, Tokyo, Japan). A two-dimensional collagen gel (Type I-A, Nitta Gelatin Inc. Nitta Gelatin, Tokyo, Japan) culture system was prepared following the manufacturer’s protocol. The collagen gel solution containing hUC-MSCs was poured into a plastic Petri dish, followed by incubation at 37°C to allow the gel to polymerize. Subsequently, hUC-MSCs were seeded onto the polymerized gels.

In this study, umbilical cord MSCs were seeded after coating plastic culture dishes with fibronectin, collagen gel, collagen coat, Laminin 211, Laminin 411, and Laminin 511 ECM. The cells were characterized by seeding umbilical cord MSCs in different culture environments created using these techniques. Normal culture without ECM was described as non-coated.

In addition, plates coated with a mixture of collagen and Laminin 411 were described as collagen coat-L411, in which 1 mL of diluted collagen concentration was mixed with 9 μl of Laminin 411, added to the wells and allowed to stand at room temperature for 1 hour. The collagen was used as collagen coat-L411. Gel-L411 plates were also prepared and used by adding 9 μl of Laminin 411 to 2 mL of dilute collagen solution, mixing, adding to wells, and heating at 37°C for 30 minutes to gelatinize.

The cells from the Gel and Gel-L411 culture environment were lysed from the gel using collagenase and used for gene expression analysis and flow cytometry. Cells from other culture methods were exfoliated using trypsin.

C57BL/6 (B6) mice were purchased from CLEA Japan Inc. (Tokyo, Japan) and maintained under conventional conditions at the Nagasaki University Animal Center. Induction of emphysema was performed according to a previous protocol established as previously reported (19, 20). Briefly, the animals were treated with 3% isoflurane under oxygen anesthesia and challenged with intranasal instillation of 30 μg porcine pancreatic elastase (catalog number 058-05361; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) in 50 µl of 0.9% saline solution. The animals were administered the dose only once on Day 0. Control animals were administered 50 µl of 0.9% saline solution (vehicle). Lung samples collected from treated and control animals were analyzed (see online Data Supplement for details).

The mice were anesthetized (isoflurane: 3% induction and 1% maintenance), and saline solution or cultured hUC-MSCs (3.0 × 10⁶ cells, total volume 100 μl) were slowly injected via their tail veins. Lung samples were collected at different time points, and their histology, gene expression profiles, and cell surface proteins were analyzed (see the online Data Supplement for details).

All values are expressed as means ± SD of three independent experiments. Data were analyzed using analysis of variance (ANOVA) followed by the Tukey–Kramer test. The statistical significance of differences between concentrations was set at *P < 0.05 or **P < 0.01, as indicated.

After 24 hours of incubation, non-adherent cells were removed by washing with PBS. Cells that remained adherent to the well plate were considered as cells with adhesive ability, and the number of those cells was measured and evaluated using the cell counting kit-8 (CCK-8) assay. The results of CCK-8 assay showed that all types of MSCs exhibited greater attachment to plastic surfaces than human umbilical vein endothelial cells (HUVECs), especially at high cell seeding densities ( Figure 1A ). The ratio of VEGF expression to GAPDH gene expression in each MSCs was analyzed using HUVECs, which are known to promote cellular responses involved in angiogenesis, as a control. The mRNA expression of vascular endothelial growth factor (VEGF), was upregulated in each MSCs group in comparison to HUVEC ( Figure 1B ), indicating that the therapeutic potential of UC-MSCs is comparable to that of hBM-MSCs and hADSC-MSCs. MSCs can secrete the angiogenic factor VEGF, which promotes local angiogenesis, suggesting that MSC-based therapy after tissue injury may increase microvascular density and preserve organ function. For this VEGF gene expression analysis, HUVECs, hBM-MSCs, hADSC-MSCs, and hUC-MSCs were cells cultured on plastic surfaces to which no special coating such as ECM or stimulating factors were added.

The cells isolated from the Wharton’s jelly of the human umbilical cord exhibited a spindle-shaped fibroblast-like morphology and adhered well to the plastic ( Figure 1C ). These cells expressed stem cell-specific transcription factors such as OCT4, NANOG, and SOX2. Flow cytometry analysis revealed up- or downregulated expression of the following surface markers in the sorted cells: CD105-PE, 90.2%; CD73-PE-Cy7, 92.3%; CD146-FITC, 97.9%; CD90-PE, 97.3%, and 0.47% CD45-FITC. These results suggest that our cell cultures exhibited the typical MSC immunophenotype, i.e., CD105⁺/CD73⁺/CD146⁺/CD90⁺/CD45^- ( Figure 1C ).

We assessed the expression of neovascularization-related factors in hUC-MSCs cultured on different ECM proteins that constitute the basement membrane of alveolar epithelial cells. VEGF expression was higher in collagen I gel-coated wells than in laminin- or fibronectin-coated wells ( Figure 1D ). Coating with laminin 411, but not laminin 211 or laminin 511, promoted PDGFα expression. Since type I collagen gel cultures of hUC-MSCs exhibited vaso-inductive potential, cDNA microarray analysis was performed to evaluate hUC-MSCs’ effects on the regulation of vascular endothelium-related gene expression. A comparison of the microarray data obtained from cells cultured without collagen-1 coating (non-coated) and cells cultured on type I collagen gel revealed differences in the expression of VEGF induction-related genes ( Figure 1E ).

High levels of induction of vascular EC growth factors, IL-6, and BCL2 were found in cells cultured on type-I collagen gels. Additionally, the expression of genes such as endothelial cell-specific molecule 1 (ESM1) and meis homeobox 1 (MEIS1), which negatively regulate cell proliferation, was suppressed ( Figure 1E ). hUC-MSCs cultured on type I collagen gels also exhibited high expression of several genes involved in myofiber differentiation, indicating that these hUC-MSCs may heal injured areas of the airway smooth muscles ( Figure 1F ).

The additive effects of different combinations of ECMs on the mRNA, protein, surface marker, vasoinductive marker expression, and morphology of hUC-MSCs attached to these ECMs were evaluated. ECMs containing a collagen coating, collagen gel, and laminin 411 showed increased expression of vasoinductive markers ( Figure 1D ). Although cells cultured on collagen gels (Gel: type I collagen gel) and collagen gels containing mixed substrates (Gel-L411: type I collagen gel and laminin 411) showed greatly enhanced VEGF expression, laminin 411 did not exhibit an additive effect ( Figure 2A ).

Flow cytometry was used to characterize the differentiation and maturation processes. Changes in the expression of the cell surface markers, CD31 and CD34, were analyzed in the CD73⁺/CD90⁺ MSC populations (these are the most reliable vascular endothelial progenitor cell markers). High levels of CD31 were observed in cells cultured on Gel and Gel-L411 (approximately 61.4, 71.5, and 13.3% CD31⁺ cells grown in Gel, Gel-L411, and uncoated-cell culture conditions, respectively). The collagen-coated cultures did not show marked changes in the expression of these markers. The percentage of CD31⁺/CD34⁺ double-positive cells increased from 1.26% in uncoated cells to 5.64% in gel cells and 7.49% in Gel-L411 cells ( Figure 2B ), suggesting that vascular endothelial progenitor cell differentiation was induced in cultures grown in the presence of collagen gel, and cell differentiation was further enhanced by the combination of collagen gel and laminin 411. To observe the effects of ECM on cell morphology, hUC-MSCs were extracted from coated culture dishes, isolated by collagenase digestion, and seeded onto plastic slides. Although spindle-shaped fibroblast-like cells were observed in hUC-MSCs extracted from uncoated cells, hUC-MSCs extracted from Gel or Gel-L411 exhibited a heterogeneous morphology, including small round cells and spindle-shaped cells that showed spontaneous aggregation, suggesting that culturing on collagen gels induced cell-to-cell adhesion of hUC-MSCs. The aggregated cells comprised Nanog-positive undifferentiated MSCs and CD31⁺ vascular endothelial progenitor cells ( Figure 2C ).

Given these observations, we evaluated whether cell aggregation enhanced the angiogenic potential by inducing spheroid formation ( Figure 2D ). Spheroids formed from uncoated cultured cells exhibited regular, tightly packed aggregates, whereas cells cultured on Gel and Gel-L411 formed loose, irregularly shaped aggregates. Regardless of the culture conditions, none of the hUC-MSC spheroids were positive for Nanog or CD31, indicating that the aggregation of undifferentiated hUC-MSCs promoted differentiation and angiogenesis. Western blot analysis revealed that the expression of the cell adhesion molecule integrin was markedly upregulated in cells cultured on Gel and Gel-L411; however, there was no observable difference in integrin expression in hUC-MSCs cultured in the presence or absence of laminin 411 ( Figure 2E ).

Alveolar epithelial progenitor cells (AEpCs) were isolated from adult human lung tissues for in vitro analysis ( Figure 3A ). These progenitor cells express the MSC surface marker CD90 and factors associated with alveolar type II cells, such as the epithelial cell adhesion molecule (EpCAM) and pro-surfactant protein C (pro-SPC). The half-maximal inhibitory concentration (IC₅₀) value of elastase against these AEpCs was 65.063 μg/ml ( Figure 3B ). To evaluate the effect of UC-MSC signaling on the alveolar epithelium, hUC-MSCs and AEpCs were indirectly co-cultured on Transwell^® membranes ( Figure 3C ). qRT-PCR showed significant inhibition of expression of the inflammatory marker, tumor necrosis factor (TNF)-α, and upregulation of expression of the anti-inflammatory marker, interleukin (IL)-10, in hUC-MSCs that were co-cultured with AEpCs.

A preclinical COPD model was used to determine the in vivo effects of human MSC therapy. The COPD model was established by intranasal administration of elastase to mice over 4 weeks to induce emphysema-like changes in the lungs ( Figure 3D ).

Four weeks after elastase administration, hUC-MSCs were administered to the mice, and the lungs were harvested on day 12 for analysis ( Figure 4A ). Alveolar damage was assessed using hematoxylin and eosin (H&E) staining. The total surface area of the alveoli increased after administration of hUC-MSCs, with a significant tissue repair effect, especially in the Gel-L411 cultured cell group ( Figure 4B ). The localization of alpha-smooth muscle actin (α-SMA)-positive cells was examined by immunostaining to confirm the contribution of hUC-MSCs in the formation of vascular smooth muscle cells along the small airways (bronchioles) and alveolar walls of the lungs. Immunofluorescence analysis of the elastase-treated lungs showed that the expression of elastin, a major component of the connective tissue that organizes the alveoli, was suppressed, whereas that of α-SMA increased after hUC-MSC injection ( Figure 4C ). The expression of CD31, also known as platelet/endothelial cell adhesion molecule-1 (PECAM-1, an adhesion molecule that accumulates at adhesion sites between vascular endothelial cells and increases integrin binding activity), was markedly upregulated in the Gel and Gel-L411 cell groups. The expression of the hematopoietic progenitor cell marker, CD34, did not markedly increase ( Figure 4C ). The antioxidant protein heme oxygenase-1 (HO-1) was detected mainly in alveolar type II cells, indicating the co-localization of HO-1 with proSPC ( Figure 4C ).

Alveolar macrophages may play an important role in the pathogenesis of COPD because they express proteases such as matrix metalloproteinases (MMPs) and cathepsins (21). Therefore, we examined the polarity of macrophages and analyzed the mechanisms underlying alveolar regeneration. hUC-MSCs were co-immunolabeled with CD11b, a monocyte marker, and CD206, an M2 anti-inflammatory macrophage marker, to investigate the phenotypic changes that occur upon the administration of hUC-MSCs to macrophages present in lung tissues. Flow cytometric analysis of isolated lung monocytes revealed an overall increase in CD11b levels in all elastase-treated groups. The hUC injection, especially the injection of Gel-L411 cells, resulted in an increased number of M2 macrophages ( Figure 4D ).

To determine whether the injected hUC-MSCs interacted with macrophages, lungs collected on day 12 were immunostained with antibodies against F4/80 (a pan-macrophage marker) and CD206 ( Figure 4E ). In line with the fluorescence-activated cell sorting (FACS) analysis, we confirmed that the overall polarization of macrophages in lung tissues shifted toward an M2-predominant type after injection of Gel-L411 cells.

We examined whether Gel- and Gel-L411-cultured cells, when co-cultured with receptor activator of NF-κB ligand (RANKL)-induced osteoclast progenitors and hUC-MSCs, affected osteoclastogenesis in vitro. Bone marrow from wild-type mice was extracted, seeded in well plates, and cultured in the presence of M-CSF and RANKL to differentiate into mature multinucleated tartrate-resistant acid phosphatase (TRAP)-positive osteoclasts. The contact approach between osteoclasts and hUC-MSCs was evaluated using a non-contact co-culture model of osteoclasts and hUC-MSCs using Transwell (left side of Figure 5A ; insert group) and a direct co-culture model in which hUC-MSCs were added directly to osteoclasts (right side of Figure 5A ; direct group). The group of bone marrow cells to which only M-CSF was added was designated “macrophages” and the group to which M-CSF and RANKL were added was designated “control (no hUCs)”. The hUCs added to osteoclasts in both the insert and direct groups were hUC-MSCs cultured without coating and “non-coated”, “gel” on gel and “Gel-411” on gel with Laminin 411.

TRAP staining (red) showed that osteoclastogenesis was almost completely inhibited in the Direct assay, indicating that cell-to-cell contact stimulated osteoclastogenesis, which occurs in the presence of soluble mediators ( Figure 5A ).

We also evaluated the morphology and number of osteoclast progenitor cells by culturing bone marrow cells (obtained from the tibiae of the control and experimental mice) in the presence of RANKL and M-CSF for five days. Numerous TRAP-positive multinucleated osteoclasts were observed in the control group. The number of osteoclasts and expression of the mature osteoclast marker, TRAP, in cultures of COPD mouse models were significantly lower in the Gel and Gel-L411 groups than in the control group ( Figure 5B ). Expression of the M1 markers, IL-6 and inducible nitric oxide synthase (iNOS), in bone marrow cells (isolated from experimental mice) decreased in the hUC-MSC-treated group and was suppressed to a greater extent in the Gel group than in the Gel-L411 group ( Figures 5C, D ).

To investigate the function of hUC-MSCs in vivo, we analyzed their behavior using luciferase-expressing cells (Luc-hUCs). hUC-MSCs almost disappeared 24 h after intravenous administration of Luc-hUCs. Several Luc-hUCs cultured on the Gel or Gel-L411 were observed in the lung samples ( Figure 6A ). To determine whether hUC-MSCs interacted with macrophages in vivo after injection, lung samples were immunolabeled with antibodies against the macrophage marker F4/80. In the Gel and Gel-L411 groups, hUC-MSCs were distributed throughout the lung tissue, accumulated on the outer margins, and co-localized with macrophages. Injection of hUC-MSCs induced the generation of CD206⁺ M2-type macrophages, while CCR7⁺ M1-type macrophages were not formed, suggesting that the injection of hUC-MSCs induced an M2 polarization shift in macrophages, resulting in anti-inflammatory effects.

A noticeable change in TNFα expression was not observed in the lungs 24 h after cell administration ( Figure 6B ) but those of IL-1β and IL-6 were suppressed, indicating that cells cultured on the ECM induced anti-inflammatory effects in the lungs.

Integrin β1 co-localized with proSPC in the COPD mouse models injected with Gel or Gel-L411 cells ( Figure 6C ). Considering that attached hUC-MSCs are phagocytosed by macrophages, in vitro experiments ( Figures 2C–E ) suggested that integrin β1 may be the key factor that induced the adhesion of hUC-MSCs, leading to alveolus formation in the COPD model. This result is agreement with previously reports that human mesenchymal stem cells cultured on collagen gels differentiate into epithelial cells through the formation of cytokeratin-18 (22).