Acute severe lung injury caused by H1N1 virus infection leads to elevated IL-18 expression
To evaluate the therapeutic effect of primed hUC-MSCs on severe lung injury, it is necessary to induce a severe mouse model with weight loss of more than 20% (reaching the endpoint of the experiment). Four viral infection doses (TCID50: 1×106/ml, 1×107/ml, 5×107/ml, and 1×108/ml) were used inH1N1 (PR8) virus-induced acute lung injuries in vivo. The results demonstrated that the body weight of mice gradually decreased after viral infections, while only the highest infection dose (TCID50: 1×108/ml) could reduce the body weight to 80% lower than their original level 6 days post-infection (6 dpi) (Figure 1A), and the survival rate at 5 dpi and 6 dpi decreased to 66.67% and 33.33%, respectively (Figure 1B). As such, a severe disease mouse model was successfully established.
Various diseases induce unique cytokine expression profiles. In this study, we investigated changes in several important cytokines in a mouse model of severe lung injuries. We found that five key genes (TNF-α, IFN-γ, IL-1β, IL-6, and IL-10) first increased and then decreased, peaking at 5 dpi (Figure 1C). This phenomenon is consistent with typical inflammatory responses to viral infections.
Interestingly, we confirmed that the gene expression of proinflammatory factor IL-18 in the injured lung was higher than that of the other five cytokines, and its changes were similar to that of these genes (Figure 1C). Additionally, the IL-18 protein concentration in bronchoalveolar lavage fluid (BALF) was higher (approximately 2000-3500 pg/ml) than in the others, which were lower than 650 pg/ml (Figure 1D). Together, the expression of proinflammatory factor IL-18 is elevated in the severe H1N1 virus-induced acute lung injury model, suggesting that it could stimulate transplanted hUC-MSCs in vivo.
IL-18-primed hUC-MSCs exhibit robust immunosuppressive ability
We first investigated the expression of the IL-18 receptor (IL-18R) in hUC-MSCs to ensure activation of the downstream signaling pathway of IL-18-IL-18R. RT-qPCR analysis indicated that IL-18R expression is approximately 1/500th that of GAPDH (Supplementary figure 1A). To better understand IL-18R expression levels, we compared the expression of several common factor receptors, including IFN-γ receptors (IFNGR1 and IFNGR2), TNF receptors (TNFR1 and TNFR2), TGF-β receptors (TGFbR1 and TGFbR2), IL-1 receptors (IL1R1 and IL1R2), and the IL-17 receptor (IL17RA). We found that the mRNA expression of IL-18R was higher than that of IL1R2 and lower than that of the other eight receptors (Supplementary figure 1B). These results indicate that hUC-MSCs express IL-18R at relatively lower levels. We then investigated the characteristics and functions of IL-18-primed hUC-MSCs, as illustrated in the scheme (Figure 2A). In detail, hUC-MSCs were first cultured in a complete medium from passage 0 (P 0) to P 3, with passaging every 3-4 days. P 3 hUC-MSCs were re-seeded in a culture dish in a complete medium by the 5000 cells/cm2. Recombinant human IL-18 protein was added into a fresh medium after 48 hours of culture, and the P 4 IL-18-primed hUC-MSCs (IL18-hUCMSC) or control hUC-MSCs (hUCMSCcon) were obtained with or without 24 hours of priming. Then, the surface markers, tri-lineage differentiation potential, proliferation ability, migration ability, paracrine secretion, and immunosuppression ability were analyzed in the following experiments. First, we performed immunophenotyping of IL18-hUCMSC and hUCMSCcon. At P4, more than 95% of these two hUCMSCs were positive for typical mesenchymal cell surface markers (CD73, CD90, and CD105), while hematopoietic cell markers (CD34 and CD45) and HLA-DR were almost completely absent (Supplementary figure 2A). We also assessed the ability of IL18-hUCMSC and hUCMSCcon to differentiate into osteocytes, adipocytes, and chondrocytes on day 21 of culture in the conditioned medium. The results indicated that IL18-hUCMSC and hUCMSCcon had similar tri-differentiation abilities (Supplementary figure 2B). Second, the results of cell proliferation showed that IL18-hUCMSC expanded faster than hUCMSCcon during the seven-day culture, especially on days 2 and 3 (Figure 2B). The population doubling time (DT) was significantly lower for IL18-hUCMSC compared with hUCMSCcon (22.06 ± 0.63 h versus 29.65± 1.47 h, Figure 2C). We found no significant difference in cell migration between IL18-hUCMSC and hUCMSCcon in a scratch wound assay, with a similar healing ratio from 4h to 24h (Figure 2D and Supplementary figure 3A). Third, the qPCR analysis demonstrated that IL-18 priming could increase the mRNA expression of vascular cell adhesion molecule-1 (VCAM-1) and matrix metalloproteinase-1 (MMP-1), but not intercellular cell adhesion molecule-1 (ICAM-1) and MMP-2 in the IL18-hUCMSC group, compared with that of the hUCMSCcon group (Figure 2E, and Supplementary figure 3B). Adhesion and matrix degradation are two prerequisites for MSCs to move into injured tissues. Compared to the hUCMSCcon group, many chemokines have increased expression in the IL18-hUCMSC group, including CCL2, CCL7, CXCL1, CXCL2, CXCL8, and CXCL12 (while CCL5 and CXCL5 have no obvious change) (Figure 2E, and Supplementary figure 3B), suggesting that IL18-hUCMSC can recruit a variety of immune cells. Transforming growth factor-beta 1 (TGF-β1), an immunomodulatory factor [43], significantly increased after IL-18 priming in the IL18-hUCMSC group (Figure 2E), but other IDO, PGE-2, TSG-6, and PD-L1 expressions did not obviously increase compared to hUCMSCcon (Supplementary figure 3C). Additionally, many growth factors were analyzed by qPCR. The expression of nerve growth factor (NGF) in IL18-hUCMSC exceeded that of hUCMSCcon, but many other IGF-1, EGF, FGF-2, and HGF did not increase after IL-18 priming (Figure 2E, and Supplementary figure 3D). According to qPCR data, the most important immunosuppressive capacity of hUC-MSCs was evaluated by an in vitro coculture experiment.
The flow cytometric data in Figure 2F demonstrates that the proliferation percentage of T-cells not cocultured with hUC-MSCs was 76.10 ± 0.94%. After four days of coculture, hUCMSCcon could significantly suppress the proliferation of T-cells, from 76.10 ± 0.94% to 45.03 ± 2.63%. Importantly, compared with hUCMSCcon, IL18-hUCMSC significantly reduced the inhibition of T-cells (21.43 ± 1.46% versus 45.03 ± 2.63%) (Figure 2G). Together, the induction of hUC-MSCs by IL-18 in vitro promotes MSC proliferation, secreting some adhesion/matrix degradation/chemokine/growth paracrine factors and enhancing the immunosuppressive ability of T-cells.
IL18-hUCMSC enhances therapeutic effects by attenuating acute lung injuries in PR8-infected mice
The schematic of protocols used for establishing a severe lung injury model at day 0, included hUC-MSCs injection (i.v.) at 3 dpi, and analysis of weight loss, survival rate, serum, BALF, and lung tissue at 7 or 14 dpi (Figure 3A). The body weight of model mice significantly reduced after PR8 infection from 0 to 8 dpi. hUCMSCcon transplantation could increase the body weight from 6 dpi compared with the saline treatment group, but there was no significant difference between these two groups. Importantly, body weight increased in the IL18-hUCMSC group from 5 dpi, and there are significant differences at 7 dpi, 8 dpi, and 9 dpi, compared with the saline treatment group (Figure 3 B). The survival rates significantly decreased in the Model + Saline group compared with the Mock group and Model + IL18-hUCMSC group (25.0% versus 100.0%, and 60.0%, respectively; Figure 3C). Importantly, model mice with IL18-hUCMSC treatment had higher survival rates than those with hUCMSCcon treatment (60.0% versus 37.5%; Figure 3C). There was no change in the general appearance of the Mock group mice. In the Model + Saline group, flu-like symptoms began to appear at 4 dpi, such as reduced activity, ruffled fur, hunched back, and weight loss. The symptoms of the Model + hUCMSCcon group were slightly better than those of the Model + Saline group, while IL18-hUCMSC treatment could restore milder clinical symptoms than the Model + hUCMSCcon group (Figure 3D); the morphological scores in these four groups also had lower scores in the IL18-hUCMSC treatment group, similar with the Mock group (Figure 3E). This demonstrated that IL18-hUCMSC had enhanced therapeutic effects after assessing clinical symptoms. The results of general lung tissue analysis showed that in PR8-infected mice, the lungs exhibited different degrees of damage, and the color of the injured parts changed from pink to dark red with the presence of edema. The extent of the lung injury in the Model + Saline group was significantly more severe than in the Model + hUCMSCcon group and the Model + IL18-hUCMSC group; the lung color was darker, and the lesion area was larger. Interesting, the degree of lung injury in the Model + IL18-hUCMSC group was significantly less severe than in the Model + hUCMSCcon group (Figure 3F). The results of the lung index showed that the lung index of the Model + Saline group significantly increased compared with the Mock group, from 0.682 ± 0.059 % to 2.384 ± 0.297 %. hUCMSCcon and IL18-hUCMSC treatment reduced the lung index, from 2.384 ± 0.297 % to 1.885 ± 0.273 % or 1.413 ± 0.086 %, respectively (Figure 3G). The area of lung injury in different groups displayed a similar change trend regarding the lung index; the area was 0.620 ± 0.117, 0.420 ± 0.075, and 0.220 ± 0.075 in the Model + Saline group, Model + hUCMSCcon group, and Model + IL18-hUCMSC group, respectively (Figure 3H). Figure 3F-H demonstrates that IL-18 priming on hUC-MSCs could significantly decrease lung damage and promote lung repair. Altogether, IL18-hUCMSC showed enhanced therapeutic efficacy in PR8-infected mice.
IL18-hUCMSC attenuated acute lung injuries by reducing inflammation, fibrosis, and cell apoptosis
Histological examinations of lung tissues by HE staining showed the occurrence of alveolar edema, inflammation, bleeding, and interstitial tissue. PR8 infection induced severe alveolar edema, large infiltration of inflammatory cells, slight bleeding, hyperplasia of interstitial tissue in the Model + Saline group; and hUC-MSCs administration could suppress the occurrence of these symptoms at 7 dpi and 14 dpi. Compared with hUCMSCcon treatment in the Model + hUCMSCcon group, IL18-hUCMSC significantly attenuated these four aspects of acute lung injuries in the Model + IL18-hUCMSC group (Figure 4A). The histopathological scores significantly decreased by IL18-hUCMSC treatment compared to hUCMSCcon treatment; the suppression rate ranged from 10.00 ± 1.79 to 6.20 ± 1.60 at 7 dpi and from 8.80 ± 1.94 to 5.20 ± 2.04 at 14 dpi, respectively (Figure 4B). To assess whether hUC-MSCs regulate viral replication in damaged lungs, qPCR was used to detect changes in the viral matrix protein 1 (M1) expression in the lungs of PR8-infected mice, which could indirectly reflect the viral load. The viral load in the lungs of the Model + Saline group greatly increased after PR8 infection at 7 dpi, and hUC-MSCs treatment significantly reduced M1 expression. The M1 gene was barely expressed in the Model + IL18-hUCMSC group, which demonstrated that IL18-hUCMSCs have an antiviral function (Figure 4C). In addition, we could not find M1 expression in any group at 14 dpi (Figure 4C). Moreover, collagen deposition was analyzed in the lung tissue at 14 dpi using Masson’s Trichrome staining; we found that PR8 infection induced much lung fibrosis in the Model + Saline group, as indicated by the blue area in the pulmonary interstitium (Figure 4D). hUC-MSCs injection significantly reduced fibrosis, in particular, IL18-hUCMSC showed enhanced performance. The percentages of collagen area were 15.34 ± 2.24 %, 5.64 ± 1.56 %, and 2.44 ± 0.80 % in the Model + Saline group, Model + hUCMSCcon group, and Model + IL18-hUCMSC group, respectively (Figure 4E). Figure 3F shows lung necrosis after PR8 infection. We next analyzed the cell apoptosis in lung tissue at 7 dpi and 14 dpi using 7AAD staining. The flow cytometric analysis was used to identify the percentage of 7AAD positive cells in all groups (Figure 4F) and demonstrated that hUC-MSCs injection significantly decreased cell apoptosis at 7 dpi compared to the Saline treatment, but there was no obvious difference between hUCMSCcon and IL18-hUCMSC at 7 dpi and 14 dpi (Figure 4G). Collectively, IL18-hUCMSC protected the lungs by reducing inflammation, fibrosis, and cell apoptosis at the cellular level.
IL18-hUCMSC had better immunosuppression on T-cells in BALF
Next, the change of T-cells and their subpopulations in BALF were analyzed by flow cytometry after PR8 infection and hUC-MSCs treatment. The percentages of CD3+, CD4+, and CD8+ T-cells in BALF at 7 dpi were shown (Figure 5A). Compared with Saline treatment in the Model + Saline group, IL18-hUCMSC largely reduced the number of total cells in BALF (Figure 5B); further, we found that IL18-hUCMSC treatment could significantly decrease the infiltration of CD3+, CD4+, and CD8+ T-cells into BALF (Figure 5C-5E). In contrast, hUCMSCcon therapy effectively reduced CD3+ and CD4+ T-cell in BALF at 7 dpi (Figure 5C, 5D). Meanwhile, the protein levels of four proinflammatory cytokines were evaluated in BALF at 7 dpi. The results demonstrated that IL18-hUCMSC treatment largely suppressed IFN-γ, TNF-α, IL-1β, and IL-6 expression in the Model + IL18-hUCMSC group compared with the Model + Saline group; but there was no statistical difference in IL-1β and IL-6 expression between hUCMSCcon and IL18-hUCMSC (Figure 5F). Altogether, IL18-hUCMSC was a more effective immunosuppressant in BALF.
IL18-hUCMSC has no enhanced performance in suppressing proinflammatory cytokine expression in serum and lung tissue
PR8 infections in the lung typically induce systemic inflammation, while proinflammatory cytokines are also overexpressed in serum. Our results showed that hUC-MSCs therapy could significantly reduce proinflammatory cytokine expression in serum at 7 dpi compared with Saline treatment, but no difference was observed between hUCMSCcon and IL18-hUCMSC (Figure 6A). Then, we assessed proinflammatory cytokine expression in the lung tissue. Compared to Saline treatment in the Model + Saline group, IL18-hUCMSC could significantly reduce mRNA expression of proinflammatory cytokines, especially IFN-γ, TNF-α, and IL-1β; meanwhile, IL-10 (anti-inflammatory cytokine) was more highly expressed in the Model + IL18-hUCMSC group (Figure 6B). However, there was no obvious statistical difference between hUCMSCcon and IL18-hUCMSC (Figure 6B). We also observed similar trends in the protein levels of the above proinflammatory cytokines in lung tissue homogenate, and there was no change between the hUCMSCcon and IL18-hUCMSC group (Figure 6C). While IL18-hUCMSC treatment did not enhance performance in serum and lung tissue compared with hUCMSCcon, IL18-hUCMSC still had a therapeutic effect on N1N1 virus-induced lung damage.