DPP4 is involved in the regulation of hASC proliferation
To determine the DPP4 status of hASCs, we first analysed cell populations defined by the cell surface proteins CD90 and CD34. The results revealed that approximately 98% of the hASCs were positive for CD90 and negative for CD34 (Figure S1, Fig. 1A), which is consistent with our previous studies (19, 30). Then, hASCs from three different donors were subjected to FACS analysis of the DPP4+ and DPP4− cell populations. Relatively substantial donor-dependent variations in the percentage of DPP4+ hASCs from different donors were detected (Figure S1C). Approximately 33.3% of the cells were CD90+/DPP4+, and 68% were CD90+/DPP4− from one donor, as presented in Fig. 1B. To further confirm DPP4 expression, the sorted hASCs were analysed via flow cytometry (Fig. 1C). The majority of both DPP4+ hASCs and DPP4− hASCs exhibited the typical uniform spindle-shaped appearance of morphogenic fibroblasts (Fig. 1D).
To evaluate the proliferation properties of DPP4+ hASCs and DPP4− hASCs, immunofluorescence analysis was performed. The results revealed that the proportion of Ki67 positive cells among DPP4+ hASCs was significantly lower than that among DPP4− hASCs (Fig. 2A). To further investigate whether the reduced proliferative capacity of DPP4+ hASCs is responsible for cellular senescence, senescence-associated beta-galactosidase activity (SA-β-gal) staining was performed. The results revealed that the proportion of SA-β-Gal positive DPP4+ hASCs was significantly greater than that of DPP4− hASCs (Fig. 2B). Quantitative RT-PCR indicated that the expression of the galactosidase beta 1 (GLB1) gene was upregulated in DPP4+ hASCs (Fig. 2C).
Additionally, the mRNA levels of the cell cycle inhibitor, cyclin-dependent kinase inhibitor 1A (CDKN1A), CDKN2A, and CDKN2B, but not tumour protein p53 (TP53) were significantly greater in DPP4+ hASCs than in DPP4− hASCs (Fig. 2D). The expression levels of SASP-related genes, including insulin like growth factor binding protein 3 (IGFBP3), interleukin 1 receptor type 1 (IL1R1), and interleukin 6 (IL6), in DPP4+ hASCs were also significantly greater than those in DPP4− hASCs. These data indicated that the proliferative capacity of DPP4+ hASCs was lower than that of DPP4− hASCs.
To investigate the effect of DPP4 on the proliferative capacity of hASCs, DPP4 expression was knocked down in DPP4+ hASCs using DPP4 siRNA (Figure S2), and the properties of DPP4+ hASCs were evaluated. Immunofluorescence staining revealed that Ki67 expression was significantly greater in DPP4 siRNA-treated cells than in control siRNA-treated cells (Figure S3A). Moreover, the percentage of positive areas for SA-β-gal staining was also significantly lower in DPP4 siRNA-treated cells than in control siRNA-treated cells (Figure S3B). The results of quantitative RT-PCR indicated that the mRNA levels of IGFBP3 but not the GLB1, TGFB1 and CDKN2B were also significantly lower in DPP4 siRNA-treated DPP4+ hASCs than in the control siRNA-treated cells (Figure S3C, S3D). These data indicated that DPP4 may be directly involved in regulating the proliferation of hASCs.
DPP4+ hASCs have a greater capacity than DPP4− hASCs for differentiation into hepatocytes
Previously, we demonstrated that the activation of Wnt/β-catenin signalling induces definitive endoderm specification, which may mediate hASC differentiation into functional hepatocytes (19, 39). To compare the ability of DPP4− hASCs and DPP4+ hASCs to differentiate into hepatocyte-like cells (HLCs), hASCs from two groups (DPP4− hASCs and DPP4+ hASCs or control siRNA-treated DPP4+ hASCs and DPP4 siRNA-treated DPP4+ hASCs) were differentiated using a three-stage differentiation protocol, as previously described (19). The properties of the differentiated cells in the two groups were analysed at the following stages: definitive endodermal progenitor cells (EPCs), hepatic progenitor cells (HPCs) and HLCs during the process of hepatic differentiation of hASCs.
The results of quantitative RT-PCR revealed that the mRNA levels of GATA binding protein 4 (GATA4), GATA6, sex determining region Y (SRY)-related high-mobility group (HMG) box (SOX) protein 17 (SOX17), and the fork head domain protein FOXA2, which is a marker of definitive endoderm cells, were also significantly greater in DPP4+ hEPCs than in DPP4− hEPCs (Figure S4A). The mRNA levels of definitive endoderm specific transcription factors were lower in DPP4 siRNA-treated DPP4+ hEPCs than in the control siRNA-treated cells (Figure S4B). Immunofluorescence staining data verified that the relative intensity of FOXA2 staining in DPP4+ hEPCs was greater than that in DPP4− hEPCs (Fig. 3A). Moreover, the protein level of FOXA2 was lower in DPP4 siRNA-treated DPP4+ hEPCs than in control siRNA-treated cells (Fig. 3B). These results indicated that DPP4+ hASCs presented increased EPC differentiation efficiency.
On day 10 of the differentiation process, the expression of α-fetoprotein (AFP), which is a marker of HPCs, in the cells was determined. The results revealed that the protein level of AFP was greater in DPP4+ hHPCs than in DPP4− hHPCs (Fig. 3C). Compared with that in the control siRNA-treated cells, AFP expression was decreased in DPP4 siRNA-treated hHPCs (Fig. 3D).
The expression of albumin (ALB) and glutathione S-transferase alpha2 (GSTA2), which are markers of hepatocytes, was subsequently analysed. The results revealed that the protein levels of ALB and GSTA2 were greater in DPP4+ hHLCs than in DPP4− hHLCs (Fig. 3E, F). The expression of GSTA2 was lower in DPP4 siRNA-treated DPP4+ hHLCs than in control siRNA-treated cells (Figure S4C). These data indicated that DPP4+ hASCs can be efficiently differentiated into hHLCs by mimicking liver embryogenesis and maturation induction.
DPP4 controls hASC secretion of inflammatory factors
The well-characterized self-renewal properties of hASCs are coupled with their immune modulating function. The immunomodulatory effect of hASCs relies on paracrine effects and the release of various soluble factors (2). To evaluate the immunomodulatory effects of these cells, 6 growth factors, 33 cytokines, 18 chemokines, and 8 soluble receptors were assessed via the Luminex xMAP approach using the supernatant and cell lysate from hASCs. Radar map profiling revealed different patterns of immunomodulatory factors between the supernatant (Fig. 4A) and the cell lysate (Fig. 4B). The factors with concentrations above 500 pg/mL in the supernatant of hASCs were VEGF-A (3098.025 pg/mL), IL-6 (773.815 pg/mL), MCP-1 (703.915 pg/mL), MIP-3alpha (604.465 pg/mL), and MCP-3 (575.595 pg/mL) (Table S1). The factors presented at concentrations between 100 pg/mL and 500 pg/mL in the supernatant of hASCs included MMP-1, ENA-78, SDF-1alpha, IL-17A, IL-2, IL-27, IL-4, IL-8, LIF, M-CSF, APRIL, IL-2R, TRAIL, and TWEAK (Table S1).
The differentially expressed factors in the supernatant and cell lysate of hASCs were further analysed. The results revealed that the expression of 3 growth factors (MMP-1, VEGF-A, and NGF-beta), 7 chemokines (ENA-78, eotaxin, MCP-1, MCP-2, MCP-3, MIP-3alpha, and SDF-1alpha), 13 cytokines (IFN-alpha, IL-17A, IL-2, IL-20, IL-27, IL-6, IL-7, IL-8, IL-9, LIF, TNF-alpha, TNF-beta, and TSLP), and 2 soluble receptors (CD30, and CD40L) was significantly upregulated in the supernatants of hASCs compared with the cell lysates of hASCs (Fig. 4C). The factors significantly increased in the cell lysates of hASCs compared with those in the supernatant of hASCs are shown in Fig. 4D. The protein levels of 10 factors did not differ between the supernatant and cell lysates of hASCs (Fig. 4E).
To elucidate whether DPP4 controls hASC secretion of inflammatory factors, the differences in the levels of these factors in the supernatants and cell lysates of DPP4+ hASCs and DPP4- hASCs were further studied. The results revealed that the levels of matrix metalloproteinase (MMP) 1, Eotaxin-3, Fractalkine (FKN, CX3CL1), growth-related oncogene-alpha (GRO-alpha, CXCL1), monokines induced by interferon-gamma (MIG), macrophage inflammatory protein (MIP)-1beta, and macrophage colony-stimulating factor (M-CSF) were significantly greater in the supernatant of DPP4+ hASCs than in that of DPP4- hASCs. The protein level of HGF in the supernatant of DPP4- hASCs was greater than that in the supernatant of DPP4+ hASCs (Fig. 5A, Table S2a).
To investigate the differences in the secretory profiles of hASCs, DPP4- hASCs and DPP4- hASCs, the differentially expressed factors in the supernatant of three groups of hASCs were further analysed. The results revealed that the protein level of Eotaxin-2 and BLC in the supernatant of DPP4- hASCs and DPP4- hASCs was higher than in hASCs. The protein level of HGF, VEGF-A, Eotaxin, Fractalkine, MIG, MIP-3alpha, IL-22, M-CSF, TSLP, and BAFF in the supernatant of DPP4- hASCs and DPP4- hASCs was lower than in hASCs (Table S2b).
The factors with upregulated expression (P < 0.05, fold change > 1.2) in cell lysates from DPP4+ hASCs and DPP4- hASCs included SCF, VEGF, BLC, ENA-78, eotaxin-3, GRO-alpha, I-TAC, MCP-3, MIG, MIP-1beta, MIP-3alpha, G-CSF, GM-CSF, IFN-alpha, IL-1beta, IL-21, IL-22, IL-6, IL-7, IL-8, IL-9, M-CSF, MIF, APRIL, CD40L, and TRAIL (Fig. 5B and Table S3). The protein level of HGF in both the supernatant and cell lysate of DPP4- hASCs was lower than that in DPP4+ hASCs.
To investigate the effect of DPP4 on inflammatory factor expression in hASCs, the supernatants and cell lysates from control siRNA-treated DPP4+ hASCs and DPP4 siRNA-treated DPP4+ hASCs were studied. The results revealed that the protein levels of 17 factors in the supernatant and 2 factors in the cell lysates of DPP4 siRNA-treated DPP4+ hASCs were significantly greater than those in the supernatant and lysate of control siRNA-treated DPP4+ hASCs (Fig. 5C, Table S4a). Compared with those in control siRNA-treated DPP4+ hASCs, the protein level of HGF in both the supernatant and the cell lysate was greater in DPP4 siRNA-treated DPP4+ hASCs (Fig. 5C, Table S4b).
These data confirmed that hASCs can secrete many important factors that may be involved in distinct immunomodulatory properties, including vascular development and angiogenesis, maintenance of homeostasis, mediation of innate immunity and tissue inflammation. Moreover, the results indicated that DPP4 regulates HGF expression in hASCs.
DPP4 expression in hASCs is involved in the regulation of T-cell polyfunctionality
T-cell immunity plays a critical role in the body’s immune defend against infectious diseases, sterile inflammation, tumours, and autoimmune diseases. Once appropriately activated, T cells infiltrate the microenvironment and recognize and defend against antigens. The range of T‑cell functions includes the ability to proliferate or induce the proliferation of other cells (through the secretion of growth factors), organize immune responses (by secreting chemoattractants) and carry out effector functions by directly killing infected cells through cytolytic mechanisms or secretion of the cytokines (40).
To determine whether hASCs could regulate the functional properties of T cells, especially their polyfunctionality, which is the ability to secrete multiple (> 2) cytokines per cell, a multiplexed antibody-coated chip that allows the analysis of thousands of T cells at the single-cell level to determine the frequency and intensity of secretion of 32 cytokines was used, as described in previous studies(37, 38). Single-cell functional profiles based on secreted proteins can be categorized into effector (granzyme B, IFN-γ, MIP-1α, perforin, TNF- α, and TNF-β), stimulatory (GM-CSF, IL-2, IL-5, IL-7, IL-8, IL-9, IL-12, IL-15, and IL-21), regulatory (IL-4, IL-10, IL-13, IL-22, TGF-β1, sCD137, and sCD40L), chemoattractive (CCL-11, IP-10, MIP-1β, and RANTES), and inflammatory (IL-1B, IL-6, IL-17A, IL-17F, MCP-1, and MCP-4) groups (IsoCode Chip, Figure S5). A prespecified T-cell functionality strength index (FSI) and polyfunctionality strength index (PSI) were combined with cytokine secretion profiles to evaluate T cell functions in activated PBMCs with anti-CD3/CD28 antibodies from health donors cultured with the supernatant of hASCs or hASCs (Figure S6). The exposure of hASCs to activated PBMCs from healthy donors significantly increased T-cell functionality, including FSI and PSI values (Figure S7). Therefore, activated PBMCs were used in subsequent studies.
The immunomodulatory effect of hASCs relies on paracrine effects and the release of various soluble factors (2). To evaluate the effect of hASCs on T-cell functionality, activated PBMCs were cultured with hASCs or the medium of hASCs in vitro for 4 h, after which the FSI, PSI and cytokines secreted by CD4+ T cells or CD8+ T cells were determined. The results revealed that the functional profiles of CD4+ T cells and CD8+ T cells changed when the cells were cultured with hASCs. CD4+ T-cell functions, including effector, stimulatory, chemoattractive, and inflammatory functions, were increased, and the regulatory FSI value was decreased when the cells were cultured with hASCs. CD8+ T-cell functions, including stimulatory and chemoattractive functions, were increased, and the effector FSI and inflammatory FSI values were decreased when the cells were cultured with hASCs (Fig. 6A). Further analysis of the polyfunctionality of T cells revealed increased PSI values in CD4+ T cells, including effector and stimulatory T cells, and decreased PSI values in CD8+ T cells, including effector and inflammatory T cells (Fig. 6B). These results indicate that there is a significant difference in the immunomodulatory effect of hASCs on CD4+ T cells and CD8+ T cells. The reduced PSI in CD8+ T cells cultured with hASCs indicates that hASCs might suppress CD8+ T cell activation or cytokine secretion.
The major cytokines and chemokines induced in CD4+ T cells and CD8+ T cells upon treatment with hASCs were IL-8 and MIP-1b (Fig. 6C, 6D). The number of TNF-α-producing functional or polyfunctional CD4+ T cells increased upon treatment with hASCs, whereas the number of CD8+ T cells that produced TNF-α decreased upon treatment with hASCs (Fig. 6C, 6D). The levels of Granzyme B, RANTES, IL-9, and MPC-4 produced by functional or polyfunctional CD4+ T cells and CD8+ T cells decreased upon treatment with hASCs (Fig. 6C, 6D).
To evaluate the effect of hASC supernatants on T-cell functionality, activated PBMCs were cultured with or without hASC supernatants in vitro for 4 h, after which the FSI value, PSI value and cytokines secreted by CD4+ T cells and CD8+ T cells were determined. The results revealed that the functional profiles of CD4+ T cells and CD8+ T cells also changed when the cells were cultured with supernatants from hASCs (Fig. 6E). The effector, stimulatory and regulatory function PSI values of CD4+ T cells decreased when the cells were cultured with the medium from hASCs (Fig. 6F). The cytokines and chemokines produced by CD4+ T cells whose levels were most strongly decreased were Granzyme B, IL-9, TGF-β1, IL-17F, and TNF-β after culture with medium from hASCs (Fig. 6G). The effector PSI and chemoattractive PSI values of CD8+ T cells were increased, but the inflammatory PSI value was increased when these cells were cultured with medium from hASCs (Fig. 6F). Polyfunctionality analysis revealed that the number of CD8+ T cells that secreted multiple factors (IFN-γ, MIP-1b and TNFα) was increased, whereas the number of cells that secreted both Granzyme B and MCP-4 was decreased after culture with supernatants from hASCs (Fig. 6H). Single-cell PAT principal component analysis (PCA) revealed the polyfunctionality landscape of T cells, as presented in Figure S8.
To further assess the impact of DPP4 expression in hASCs on the regulation of T-cell functionality, activated PBMCs were cultured with or without DPP4−hASCs and DPP4+hASCs supernatants or cells. The results revealed that CD8+ T cell PSI values, including effector, stimulatory, chemoattractive and inflammatory values, were significantly greater in the DPP4+ hASC-treated group than in the DPP4− hASC-treated group (Fig. 7A, 7B). Compared with those observed after treatment with DPP4− hASCs, the number of CD4+ T cells that secreted single cytokines, including MIP-1b, IL-8 and TNF-α, was greater after treatment with the DPP4+ hASCs (Fig. 7C). Compared with DPP4− hASCs, DPP4+ hASCs promoted the secretion of multiple factors, including IL-8, MIP-1a, MIP-1b, and TNF-α; Granzyme B and MIP-1b; IL-6 and IL-8, and secretion of single cytokines, including MIP-1b and Granzyme B, by CD8+ T cell (Fig. 7D). These findings indicated that the regulation of CD8+ T-cell functionality by DPP4+hASCs was stronger than that by DPP4−hASCs.
An analysis of the effects of supernatants derived from DPP4−hASCs and DPP4+hASCs on the functional properties of T cells revealed that the FSI and PSI values of CD4+ T cells and CD8+ T cells were greater in those cultured with supernatants of DPP4+ hASCs than in those cultured with the supernatants of DPP4− hASCs (Fig. 7E, F). Compared with those of cells cultured with the supernatant of DPP4− hASCs, the effector PSI values of CD4+ T cells and the effector PSI values and stimulatory PSI values of CD8+ T cells were greater when the cells were cultured with the supernatant of DPP4+ hASCs (Fig. 7F). Compared with those in cultures with the supernatant of DPP4− hASCs, the regulatory and inflammatory functionas of CD8+ T cells were lower in cultures with the supernatant of DPP4+ hASCs (Fig. 7F).
Compared with those cultured with supernatants derived from DPP4− hASCs, the percentages of CD4+ T cells that secreted MIP-1b and TNF-α were increased, whereas the percentages of cells that secreted Granzyme B and perforin were decreased upon treatment with supernatants derived from DPP4+ hASCs (Fig. 7G). The number of CD8+ T cells that secreted TNF-α, IL-8, and MIP-1a was greater, the number of cells that secreted Granzyme B and MIP-1b, and the number of cells that secreted IL-17A, TNF-α, and TNF-β was lower among cells cultured with supernatants derived from DPP4+ hASCs than among those cultured with supernatants derived from DPP4−hASCs (Fig. 7H).
To further assess the impact of DPP4 expression in hASCs on the regulation of T-cell functionality, activated PBMCs were cultured with control siRNA-treated DPP4+ hASCs and DPP4 siRNA-treated DPP4+ hASCs. The results revealed that the effector, inflammatory and stimulatory functions were decreased in both CD4+ T cells and CD8+ T cells upon treatment with DPP4 siRNA-treated DPP4+ hASCs compared with those in control siRNA-treated DPP4+ hASCs (Fig. 8A, 8B). Compared with that of control siRNA-treated DPP4+ hASCs, the polyfunctionality of CD4+ T cells was decreased upon treatment with DPP4 siRNA-treated DPP4+ hASCs (Fig. 8B). The regulatory PSI value of CD8+ T cells was greater in DPP4 siRNA-treated DPP4+ hASCs than in control siRNA-treated DPP4+ hASCs (Fig. 8B). The analysis of polyfunctionality revealed that most of CD4+ T cells and CD8+ T cells secreted factors such as GM-CSF, Granzyme B, IL-8, MIP-1a, MIP-1b and TNF-α; however, the IL-6, IL-8, MIP-1a, MIP-1b, and TNF-α levels were lower after culture with DPP4 siRNA-treated DPP4+ hASCs than after culture with control siRNA-treated DPP4+ hASCs (Fig. 8C, 8D). The secretion of factors, such as IL-22, IP-10, IL-4 and IL-9 by CD8+ T cells was increased upon treatment with DPP4 siRNA-treated DPP4+ hASCs (Fig. 8D). These data were similar to those of the cells cultured with DPP4− hASCs and DPP4+ hASCs (Fig. 7C,7D).
Analysis of the polyfunctionality of T cells cultured with medium derived from control siRNA-treated DPP4+ hASCs and DPP4 siRNA-treated DPP4+ hASCs revealed that the numbers of inflammatory, chemoattractive, stimulatory, and effector CD8+ T cells were decreased (Fig. 8E, 8F). Compared with those of cells cultured with supernatants derived from control siRNA-treated hASCs, the percentages of CD4+ T cells that secreted cytokines such as GM-CSF, Granzyme B, IL-8, MIP-1a, MIP-1b and TNF-α or IL-6, IL-8, MIP-1a, MIP-1b, and TNF-α were lower, whereas the percentages of cells that secreted factors such as IL-6, IL-8, IP-10, MIP-1a, MIP-1b, TNF-α, and TNF-β were greater among cells treated with supernatants derived from DPP4 siRNA-treated hASCs (Fig. 8G). The percentages of CD8+ T cells that secreted polyfunctional factors, such as IL-6, MIP-1a, MIP-1b, and TNF-α, IL-8, perforin, and TNF-α, decreased upon treatment with supernatants derived from DPP4 siRNA-treated hASCs (Fig. 8H).
These data confirmed that hASCs have stronger immunomodulatory effects on T-cell polyfunctionality than do their secreted factors. DPP4 expression in hASCs is involved in the regulation of CD4+ T-cell and CD8+ T-cell polyfunctionality. CD8+ T cells are more sensitive to the cellular regulation of hASCs, especially the expression of DPP4 in hASCs.