Introduction

Cystic fibrosis (CF) is a monogenic autosomal recessive disorder, caused by mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, thought to affect up to 100,000 people worldwide1. CF disease leads to multifaceted clinical manifestations, affecting multiple organ systems, in particular the respiratory, gastrointestinal, endocrine, and reproductive systems. Recent advances in CF treatment include the use of CFTR small molecule modulators, which include correctors, potentiators, and amplifiers, and act to improve the function of the mutated CFTR protein2. However, modulators do not work on all CFTR mutations and alone they cannot reverse the damage observed in established lung disease. Supportive therapies such as MSC administration, which attenuate pulmonary inflammation, are anti-microbial and have the capacity to support repair may prove beneficial to CF patients alongside current therapies.

MSCs are non-hematopoietic, multipotent adult stem cells that can differentiate into bone, cartilage, muscle, and fat cells3. They have been isolated from almost all human tissues and are thought to function through trans-differentiation, cell contact and trophic mechanisms4. Cell to cell contact allows MSCs to deliver mitochondria using tunnelling nano-tubules to alter bioenergetics of recipient cells5. In addition, evidence has emerged that the MSC secretome has therapeutic potential. MSCs can respond to their environment and release soluble factors, growth factors, cytokines, chemokines and extracellular vesicles which can carry cargo containing lipids, proteins, RNA and microRNA6. Importantly, MSCs possess immunomodulatory activity, are immune evasive and can mediate effects on both the innate and adaptive immune systems through the modulation of monocytes/macrophages, dendritic cells, neutrophils, innate lymphoid cells, T cells, B cells and natural killer cells7,8. Apoptotic MSCs are now recognized to function through a mechanism of immunosuppression, once administered they are efferocytosed by recipient phagocytes to reprogramme immunosuppressive cells and upregulate molecules such as indoleamine 2,3-dioxygenase (IDO) and prostaglandin E2 (PGE2) to alleviate pathogenic inflammation in vivo9,10,11.

Research into the therapeutic potential of MSCs for various acute and chronic lung diseases including CF is now underway. Plastic adherent bone marrow cells were found to result in CFTR expression in airway epithelium, improve bacterial clearance, and increase the survival of CFTR − / − recipient mice12. These findings are supported by other studies using verified MSCs13and warranted investigation for safety and tolerance as a therapy in the phase-I clinical trial CEASE-CF14.

In this study, we explored the therapeutic potential of MSCs in the Scnn1b-Tg mouse model of CF lung disease. We hypothesized that through their anti-inflammatory and regenerative properties, MSCs may halt the progression of lung damage in these mice. To assess this hypothesis, we intravenously delivered MSCs to Scnn1b-Tg mice and characterized disease markers including inflammation, emphysema, and lung function.

Results

Characterization of lung disease in C57BL/6 Scnn1b-Tg mice at 8-weeks of age, MSC potency assay and lung confirmation of the presence of MSCs

Scnn1b-Tg mice have been characterized on different genetic backgrounds including C3H x C57BL/6 hybrids, C57BL/6, and BALB/c15,16. As genetic background can affect the severity of disease, we characterized our colony of C57BL/6 background Scnn1b-Tg mice by performing BALF inflammatory cell counts, histological analysis and lung function at 8 weeks of age i.e., in adult mice (Figure S1). Consistent with the literature, our colony of Scnn1b-transgenic mice showed significant inflammatory cell infiltration in lung BALF (Fig. S1a), histological structural damage (Fig. S1b) and emphysema with reduced lung function parameters on the FlexiVent system (Fig S1c and S1d).

Before mice received MSCs, each batch of cultured cells was evaluated using an in vitro potency assay where the anti-inflammatory effects of MSCs were assessed using primary human macrophages (monocyte derived macrophages/MDM) and murine macrophages (MH-S) (Fig. S2a). In addition, MSC detection was confirmed in the lungs of mice 24 h after administration using Q-PCR and human specific probes (Fig. S2b).

MSC cells are anti-inflammatory in the chronically inflamed Scnn1b-Tg mouse lung

After MSC therapy in Scnn1b-Tg mice (Schematic in Fig. 1a), airway inflammation was significantly reduced, with ~ 44% fewer immune cells in BALF recovered from these mice (Fig. 1b). Upon further investigation of the immune cell populations present in the BALF of these mice, we noted a significantly reduced number of monocytes/macrophages (P = 0.002) and neutrophils (P = 0.0005) after MSC treatment (Fig. 1b). Although the number of eosinophils was reduced after MSC therapy, due to the small number of cells detected, the data did not reach statistical significance (Supplemental Fig. S3). Consistent with the reduced inflammatory cell numbers, we observed significantly decreased levels of pro-inflammatory markers including, KC (P = 0.0067), osteopontin (P = 0.0076) and IL-1β (P = 0.0012) in BALF after MSC therapy (Fig. 1c). These factors were investigated as KC and IL-1β are factors involved in the Scnn1b-Tg mouse phenotype, and osteopontin can regulate inflammatory responses17,18,19.

Fig. 1
figure 1

MSC therapy significantly reduces inflammation in chronically inflamed Scnn1b-Tg lungs. Mice were given 5 × 105 MSCs intravenously at 4 and 6 weeks of age before BALF/cell/tissue analysis at 8 weeks. (a) Schematic showing the MSC therapy study model. (b) Total and differential cell counts of leukocyte populations in BALF from 8-week-old Scnn1b-Tg mice after MSC or vehicle (PBS) administration, n = 12–20 mice per group. (c) Cytokine analysis of BALF using KC, Osteopontin and Il1-β ELISAs with n = 9–15 mice per group. Data represents mean ± SEM, statistical analysis was conducted using t-test or ANOVA with Sidak’s multiple comparison test where appropriate, with ** = P ≤ 0.01, *** = P ≤ 0.001 and **** = P ≤ 0.0001.

Full size image

PPARγ and EGR1 transcription factors are implicated in the anti-inflammatory nature of MSCs in Scnn1b-Tg mice

As a regulatory loop between osteopontin and PPARg has been identified in a murine emphysema model, we were interested to see if MSC therapy had any impact on this axis in Scnn1b-Tg mice20. In vitro studies showed MSC-conditioned medium (MSC-CM) regulated the expression of PPARγ target genes (Arg1, Retnla and Mmp12) in Scnn1b-Tg alveolar macrophages (Fig. S4). These findings supported follow up in vivo findings, where at 24 h post MSC administration (schematic in Fig. 2a), we observed a significant (P < 0.05) restoration of Pparγ expression in lung tissue (Fig. 2b). Concomitant with the transcript data, we also observed an increase (~ 1.51 fold) in nuclear PPARγ protein expression (Fig. 2c and Fig. S5). These findings demonstrate that in the chronically inflamed Scnn1b-Tg lung, MSC administration can alleviate inflammation, and this may be due, at least in part, to the restoration and signaling of PPARγ.

Fig. 2
figure 2

PPARγ and EGR1 transcription factors have roles in MSC anti-inflammatory signaling in Scnn1b-Tg mice. Mice were given 5 × 105 MSCs intravenously at 4 weeks of age before collection of tissue/cells for analysis at the 24 h time point. (a) Schematic of the 24 h MSC study. (b) Q-PCR of PPARγ expression in lung tissue 24 h post MSC installation, n = 7–12 mice per group. Data represents mean ± SEM, statistical analysis was conducted ANOVA with Sidak’s multiple comparison test, with * = P ≤ 0.05 and **** = P ≤ 0.0001. (c) Western blot for nuclear accumulation of PPARγ and normalized densitometry using Lamin B1 expression. (d) Q-PCR analysis of Egr1, Cxcl1, Il1β, Il6, and Spp1 in lung tissue and alveolar macrophages, n = 5–8 mice per group. Data shown are expressed as group mean ± SEM with statistical analysis performed using Mann–Whitney t-test, with * = P ≤ 0.05.

Full size image

PPARg activation has been shown to be linked to the suppression of the early growth response protein 1 (Egr-1) transcription factor21 and its pro-inflammatory targets which include Cxcl1 and Spp1 that encode KC and osteopontin, respectively. Having identified that MSCs modulate PPARg expression as early as 24h after the commencement of therapy (Fig. 2b and 2c), we also investigated the expression of Egr-1 and its related pro-inflammatory genes in lung tissue at this timepoint. MSC therapy significantly reduced the expression of Il1β (P = 0.036), Il6 (P = 0.0196) and Cxcl1 (KC) (P = 0.0293) in lung tissue (Fig. 2d). We also detected a trend towards reduced expression of Egr1 and a marginal reduction in Spp1 in tissue samples, though this was not statistically significant. When similar analyses were performed on alveolar macrophages isolated from Scnn1b-Tg mice at 24h post-MSCs, we observed a significant reduction in both Egr-1 (P = 0.0177) and its direct target Cxcl1 (P = 0.017) in mice treated with MSCs (Fig. 2d). Taken together, these findings suggest that the effects of MSCs on inflammation in the Scnn1b-Tg model may be due to the interplay between the upregulation of anti-inflammatory PPARγ signaling and the downregulation of pro-inflammatory Egr1 target genes.

MSC therapy results in reduced emphysema and improved lung function in Scnn1b-Tg mice

Having considered the effects of MSCs on inflammation in Scnn1b-Tg mice, we subsequently assessed the effects on lung structure and function. After MSC treatment, lung morphology was assessed for histological markers of lung damage and emphysema including Mean Linear Intercept and Destructive Index analysis22,23,24. Airway sections stained with hematoxylin and eosin highlighted a difference in vehicle and MSC treated mouse lungs (representative images in Fig. 3a). To assess this observation, we compared PBS and MSC treated Scnn1b-Tg mice and found significantly lower MLI (P = 0.0387) and DI (P = 0.0095) scores in the MSC treated mice (Fig. 3b). These data demonstrate that MSCs can reduce damage and development of emphysema in this model.

Fig. 3
figure 3

MSC therapy reduces histological markers of emphysema and improves lung function in Scnn1b-Tg mice. Mice were given 5 × 105 MSCs intravenously at 4 and 6 weeks of age before tissue analysis or lung function testing at 8 weeks. (a) Representative hematoxylin and eosin images demonstrating morphological characteristics of PBS and MSC treated Scnn1b-Tg mouse lungs, scale bar 100 µm. (b) Markers of emphysema mean linear intercept and destructive index were assessed using immunohistochemistry. Data represents average counts ± SEM from 6–9 images per mouse, with n ≥ 12 per group. (c) Analysis of lung mechanics: Lung function pressure volume (PV) loop curves generated by stepped inflation and deflation maneuvers for vehicle and MSC treated mice. (d) Compliance and elastance lung function analysis. Statistical analysis performed using an unpaired or Mann–Whitney t-test as appropriate, with * = P ≤ 0.05 and ** = P ≤ 0.01.

Full size image

To further assess the effects of MSCs on lung damage, lung function analysis was performed. All lung function parameter data are outlined in supplemental table 1. The averaged PV-loop data demonstrates a reduction in emphysema in the MSC treated Scnn1b-Tg mice when compared to their vehicle control treated counterparts (Fig. 3c). In addition, lung function analysis revealed a significant improvement in compliance (P = 0.0067) and elastance (P = 0.0132) of the entire respiratory system i.e., whole thorax including chest and lung wall (Fig. 3d) in mice that received MSCs. Interestingly, a significant (P = 0.0202) improvement in tissue elasticity was observed when the data was separated to only look at the alveolar tissue, which indicates improved lung function in the alveolar compartment after MSC therapy (Fig. 3d). Overall, these data suggest that in this model of chronic inflammation and muco-obstructive disease, MSCs can reduce emphysema, lung damage and improve lung mechanics.

PPARγ activation is anti-inflammatory but cannot recapitulate the effects of MSCs on emphysema and lung function in Scnn1b-Tg mice

Having identified that MSCs may mediate effects via PPARγ, we next sought to determine whether specific activation of PPARγ signaling could recapitulate the effects of MSCs in Scnn1b-Tg mice. To assess this, we administered the PPARγ activator rosiglitazone (RGZ) to mice for 14 days before examining BALF for inflammatory markers and performing lung function analysis (Study Schematic in Fig. 4a). Daily subcutaneous delivery of RGZ appeared to be well tolerated in the mice with no significant differences in vehicle and rosiglitazone group weights (Fig. 4b).

Fig. 4
figure 4

Rosiglitazone targeting of PPARγ in Scnn1b-Tg is anti-inflammatory but cannot recapitulate the effects of MSCs on lung function. Rosiglitazone (RGZ) or vehicle control was administered to Scnn1b-Tg animals subcutaneously at 10 mg/kg daily for 14 days from 4 weeks of age. (a) Schematic of the rosiglitazone study. (b) RGZ study animal weights/growth curves during treatment (day 0–14 = 4–6 weeks of age). (c) Total and differential cell counts of leukocyte populations in BALF from untreated wild-type and Scnn1b-Tg vehicle/RGZ treated mice. (d) Cytokine analysis of BALF using KC and Osteopontin ELISAs. (e) Analysis of lung mechanics: Lung function pressure volume (PV) loop curves generated by stepped inflation and deflation maneuvers for vehicle and RGZ treated mice. Data shown are expressed as group mean ± SEM, with n = 5–6 per group. Statistical analysis was performed using t-test or ANOVA as appropriate, with * = P ≤ 0.05, ** = P ≤ 0.01, and **** = P ≤ 0.0001.

Full size image

As expected, RGZ was anti-inflammatory in this model, resulting in a significant reduction in total airway inflammatory cells (P = 0.0027) and myeloid cells (P = 0.0028) (Fig. 4c). However, in contrast to MSC administration, no significant changes were observed for neutrophil cell counts when mice were given RGZ. In line with these findings, cytokine analysis of the BALF revealed no significant change in the levels of KC, a neutrophil chemoattractant, whereas osteopontin was significantly (P = 0.0217) reduced (Fig. 4d) in mice receiving RGZ.

We investigated the effect of RGZ on lung function but did not observe any significant changes in PV loop curves after RGZ therapy (Fig. 4e). In addition, lung function parameters of compliance (C) and elastance (E) did not show any significant changes after RGZ treatment (Supplemental Fig. S6). Together, these results indicate that the anti-inflammatory effects of MSC may be partially attributable to PPARγ upregulation and activation, but the role of PPARγ in improving emphysema and lung function appears marginal in this model of CF lung disease.

MSCs promote regeneration in the inflamed emphysematous lungs of Scnn1b-Tg mice

To delineate whether the MSCs were reducing damage through the promotion of repair in vivo, we looked for the presence of the marker of proliferation, Ki67, in lung tissue collected from mice after two doses of MSCs. Immunohistochemical staining revealed a significant (P = 0.02) enhancement of Ki67 staining in the lungs of mice receiving MSCs versus PBS controls (Fig. 5a and b). It was noted that staining was concentrated in alveolar compartments. To investigate whether it was ATII cells that were proliferating, immunofluorescence was performed on lung tissue from mice 24 h after MSC treatment, with co-staining for Ki67 and the ATII epithelial progenitor cell marker, surfactant protein C (SPC). We observed a significant (P = 0.045) increase in the percentage of SPC + Ki67 + cells in MSC treated mice (Fig. 5c and d), indicating that MSCs are promoting proliferation of epithelial progenitor ATII cells in the emphysematous Scnn1b-Tg lung.

Fig. 5
figure 5

MSC therapy promotes parenchymal repair in Scnn1b-Tg mice. Mice were given 5 × 105 MSCs intravenously at 4 and 6 weeks of age before tissue analysis at 8 weeks. (a) Analysis of proliferation in parenchymal tissue using Ki67 immunohistochemistry, representative images shown, black arrows indicate positively stained nuclei, 3LL-Lewis Carcinoma tumor xenograft used as a positive control for Ki-67 staining, image scale bars 50 µm. (b) Quantification of the % of Ki67 positive cells in MSC and vehicle treated Scnn1b-Tg lung parenchyma. Data represents average counts ± SEM from ≥ 10 fields of view per mouse, with n ≥ 7 mice per group. Statistical analysis was performed using unpaired or Mann–Whitney t-test as appropriate, with * = P ≤ 0.05. (c) Immunofluorescent staining of lung parenchymal tissue 24 h post MSC administration. Cryosections (8 µm) were co-stained with Dapi (blue), Prosurfactant Protein C (SPC) (red) and ki67 (green). White arrows on representative images represent SPC+Ki67+Dapi+ cells, scale bars 50 µm. (d) Quantification of IF staining 24 h post MSC administration, where data represents average counts ± SEM from 4–6 fields of view per mouse, with n = 6 mice per group. Statistical analysis was performed using Mann–Whitney t-test with * = P ≤ 0.05.

Full size image

Discussion

Our goal in this study was to delineate the anti-inflammatory and pro-reparative capabilities of mesenchymal stromal cells in the Scnn1b-Tg mouse model, which recapitulates several characteristics of muco-obstructive lung disease seen in cystic fibrosis. To our knowledge this is the first study testing MSC therapy in the Scnn1b-Tg mouse model. Our main findings demonstrate that; (i) MSCs are anti-inflammatory in the chronically inflamed lungs of Scnn1b-Tg mice, an effect which is partly mediated by PPARγ and EGR1 signaling, (ii) MSCs reduce histological markers of lung damage and emphysema, (iii) MSCs improve multiple lung function parameters and (iv) MSCs promote proliferation of airway progenitor cells in areas of damaged parenchymal tissue in this model. However, this study has some limitations, in that a single MSC donor (24-year-old, Caucasian, male) was used for the studies outlined in this manuscript, we cannot therefore comment on the reproducibility of the effects seen in Scnn1b-Tg mice with that of different donors.

Current mouse models with CFTR mutations do not readily recapitulate human disease characteristics of CF25. This has led to the development of the Scnn1b-transgenic mouse model which exhibits airway over-expression of the β-ENaC channel, driving Na + absorption and resulting in lung pathology similar to that observed in CF lung disease15,26. In adult Scnn1b-Tg mice, the chronic lung disease phenotype includes elevated mucus expression, goblet cell metaplasia, epithelial hypertrophy and necrosis, airway neutrophilic inflammation and parenchymal destruction/emphysema15,27. More recently, infection studies with Pseudomonas aeruginosa in Scnn1b-Tg mice have shown these mice exhibited an increased bacterial burden following instillation of this bacteria into the lungs as a result of impaired bacterial clearance rates compared to that of their wild-type littermates28. Additionally, the Scnn1b-Tg mice develop inflammation characterised by increased levels of activated macrophages and neutrophils with upregulated inflammatory cytokines and proteases including neutrophil elastase and matrix metalloproteinase 12 (MMP-12) which contribute to emphysema formation29. Consistent with the literature, in the characterization of our adult (8-week-old) Scnn1b-Tg mice we observed upregulated macrophages and neutrophils in BAL fluid in addition to the presence of emphysema and a deterioration in lung function parameters.

Inflammation in CF airways is thought to be mostly neutrophilic, which results in increased ROS production and inhibition of antimicrobial activity in the lung30. In addition, in CF macrophages, CFTR expression and dysfunction can lead to increased metabolism and the elevated generation of proinflammatory cytokines31. The inflammation observed in Scnn1b-Tg mice is characterised by the presence of activated macrophages and increased neutrophils in the BAL fluid, leading to increased inflammation with elevated Osteopontin, KC and IL1-β.

Osteopontin, a factor produced following lung injury and can be found elevated in airways of CF patients19. In Scnn1b-Tg mice, the elevated levels of osteopontin in BALF were reduced to almost wild type littermate levels after MSC therapy. Osteopontin regulation by MSCs correlated with the restoration of Pparγ expression and its nuclear localization, a finding that is consistent with the identification of a regulatory loop between these two proteins20. This led us to the hypothesis that PPARγ may modulate some of the anti-inflammatory effects of MSCs in Scnn1b-Tg.

PPARγ is a ligand activated nuclear receptor that heterodimerizes with the retinoid-X-receptor (RXR) to regulate transcription of target genes using PPAR response elements in their promoters. Studies have identified PPARγ expression to be prominent in the airway epithelium and alveolar macrophages of mouse lungs32,33. Indeed, overexpression of a double negative PPARγ variant in alveolar type 2 cells (ATII) cells of mice causes upregulation of pro-inflammatory mediators and increased emphysema34. In addition, PPARγ negatively regulates MMP-12, a protease which can contribute to emphysema formation in Scnn1b-Tg mice29,35. CFTR dysfunction itself is known reduce PPARγ signaling in a process that may involve reduced expression, altered cofactor recruitment, ligand availability, or its sequestration into aggresomes36.

To gain more insight into the anti-inflammatory effects of MSCs in our model, we performed gene expression analysis 24 h post infusion, on a panel of inflammatory genes in lung tissue samples and alveolar macrophages. We observed ameliorated reduced expression of IL-1β, IL-6 and Cxcl1 in lung tissue, and in addition, significantly reduced levels of Egr1 and Cxcl1transcripts in alveolar macrophages. 15d-PGJ2, a natural endogenous activating ligand of PPARγ, has been shown to be secreted by MSCs37. Activation of PPARγ with 15d-PGJ2 in RAW 264.7 macrophages result in reduced expression of the transcription factor early growth response-1 (EGR-1)38. Consistent with these findings, in Scnn1b-Tg mice we see a significant reduction in Egr-1 expression in Scnn1b-Tg alveolar macrophages after MSC therapy, which may be due to PPARγ signaling. EGR-1 is an immediate early response gene (IEG) that is a master regulator of pro-inflammatory signaling with roles in the regulation of cell survival, proliferation, and cell death and has been recently identified as altered in CF39. Target genes for EGR-1 relevant to the disease pathophysiology of Scnn1b-Tg mice include Pparγ, IL-1β, Cxcl1 and Spp1. EGR1 has been shown to bind directly to the Cxcl1 and Spp1 gene promoters40,41. Consistent with these findings, our data reveal that PPARγ restoration after MSC therapy correlates with reduced Egr1 expression, in addition to ameliorated reduced Cxcl1 expression.

IL-1β and EGR-1 can reciprocally mediate each other’s expression42. IL-1 signaling has been implicated in the inflammation and emphysema of Scnn1b-Tg mice18. Overexpression of IL-1β in transgenic mice resulted in infiltration by neutrophil and macrophages infiltrates, enhanced mucus production, increases in CXCL1 and MMP-12 expression and distal airspace enlargement43. Furthermore, protective effects on neutrophilic inflammation and emphysema were seen after administration of the IL-1R antagonist anakinra44. The PPARγ promoter contains an overlapping Egr-1/Sp1 site and treatment with IL-1 can induce Egr1 recruitment and reduced Sp1 occupancy45. Taken together, MSC mediated restoration of PPARγ expression/signaling may negate the effects of IL-1β and EGR1 pathways to limit airway inflammation in Scnn1b-Tg mice.

The therapeutic targeting of PPARγ was anti-inflammatory in the airways of Scnn1b-Tg mice, with reduced airway inflammation and reduced osteopontin detected in the BALF of treated mice. However, although the anti-inflammatory effects of rosiglitazone were demonstrated, it did not result in improved lung function and emphysema, unlike the improvements observed with MSCs, which did improve these structural changes. Importantly, this suggests the role of PPARγ signaling alone may not fully account for the anti-inflammatory mechanisms of MSCs and their contribution to the emphysematous repair may may involve separate pathways. The dose of rosiglitazone, route of administration and bioavailability of the drug in the lungs may explain why we see observe only anti-inflammatory effects and not an improvement in lung function. However, PPARγ signaling was evident in Scnn1b-Tg mice, which suggests that the MSCs anti-inflammatory effects may function, at least in part, through this pathway.

Studies have identified early onset and progressive emphysema in CF lung disease46. One such study detected emphysema in CF from early adolescence (~ 13 years) and this emphysema increased with age. As patients with CF can now access CFTR modulator therapy, the resultant reductions in morbidity and mortality may increase the cohort of patients with emphysema due to an increased life expectancy. Thus, further treatments will be required to treat this specific CF cohort. Few therapies for emphysematous lung disease have shown much potential for lung regeneration. In our Scnn1b-Tg model, in addition to the reduction in inflammation and an improvement in lung function, we also observe an increased proliferation in parenchymal cells after MSC administration which is indicative of stimulation of regenerative processes in the lungs of these mice.

In summary, our investigation has shown that when MSCs were systemically administered to Scnn1b-Tg mice, we observed significantly reduced airway inflammation, a marked improvement in lung function and promotion of repair. We identified PPARγ as a potential regulator of the immune dysregulation in these mice and propose that activation of PPARγ signaling may play a critical role in maintaining alveolar homeostasis by suppressing pulmonary inflammation. Our results indicate that targeting of the PPARγ-IL1β-EGR1 axis with MSCs may be beneficial for the treatment of pulmonary inflammation and emphysema associated with cystic fibrosis.

Methods

Cell culture

Human bone marrow-derived MSCs (ATCC/LGC Standards, UK) were cultured in MEMα medium (Gibco, UK), supplemented with 16.5% fetal bovine serum (FBS) (Gibco, UK) and 2 mM L-glutamine. MSC Donor (ATCC lot: 63,208,778, male, Caucasian, 24-year-old). These cells met the criteria for defining multipotent MSCs as outlined by the International Society of Cellular Therapy47. ATCC confirmed positive expression for CD29, CD44, CD73, CD90, CD105, and CD166 and negative expression for CD14, CD34, CD19, and CD45, alongside experimental confirmation of differentiation experiments for adipocytes, chondrocytes and osteocytes. Low passage MSCs (p3 and p4) were used for all experiments described, for each study the cells were expanded from cryostorage at 2.1 × 104 per T175 cm2 flask and cultured for 10–14 days before collection. Cell viability was assessed with trypan blue extrusion and cells were routinely 92 – 98% viable before administration to animals. MH-S cells (ATCC, CRL-2019) were cultured in RPMI-1640 media containing 10% FBS (Gibco, UK) and 2 mM L-glutamine. Primary human monocyte derived macrophages (MDMs) were generated from buffy coat donors obtained from the Northern Ireland Blood Transfusion Service (NIBTS). Ethical approval and all experiments were performed in accordance with relevant named guidelines and regulations outlined by the Queens University Belfast Research Ethics Committee. Informed consent was obtained from all participants who were all aged over 18 years of age. Monocytes were extracted using Ficoll-Paque Premium (GE Healthcare, UK) and gradient centrifugation, cells were washed and counted before seeding at 3 × 105 cells per well (24-well plate) and allowed to adhere. Cells were then differentiated using recombinant human GM-CSF (granulocyte macrophage colony stimulating factor) (PeproTech, UK) at 10 ng/ml in 10% FBS RPMI-1640 medium at 37˚C for 5–7 days before use.

MSC therapy in the Scnn1b Transgenic Mouse Model of Chronic Inflammation

Scnn1btransgenic mice15 and wild-type (WT) littermates (C57BL/6 background) were housed in specific pathogen-free conditions with access to water and food ad libitum. Mice were weaned at 21 days and genotyped for experiments using MyTaq PCR kit (Bioline, UK) with forward (5’ – ctt cca aga gtt caa cta ccg – 3’) and reverse (5’ – tct acc agc tca gcc aca gtg – 3’) primers (Eurofins Genomics, Germany) to amplify the DNA. Using these primers WT mice had a PCR product of 350 bp, whereas Scnn1b-Tg mice had products at 350 bp and 254 bp indicating the presence of the transgene (hemizygous mice are sufficient to produce disease phenotype). For in vivo studies MSCs were seeded in T175cm3 flasks at 2.1 × 104 cells per flask (~ 120 cells/cm2) and cultured for 14 days before collection for I.V. implantation at 5 × 105cells in 100 µl endotoxin free sterile PBS (Gibco). PBS was used as a vehicle control comparison in this study, a fibroblast cellular control was not used in this instance as these cells are known to possess anti-inflammatory, immune-modulatory, and regenerative properties akin to MSCs48,49. For the MSC therapy study, mice were given an I.V. injection of MSCs at 4 and 6 weeks of age before sample/tissue collection at 8 weeks. All experimentation was conducted in accordance with the Animal (Scientific Procedures) Act 1986 and current guidelines approved by the Queen’s University Ethical Review Committee and carried out in compliance with the ARRIVE guidelines.

Bronchoalveolar lavage fluid collection and Inflammatory cell differential staining

Mice were terminally anesthetized with sodium pentobarbital, the trachea opened, and a 21-gauge cannula inserted. Bronchoalveolar lavage fluid (BALF) was collected after the lungs were lavaged with a weight-adjusted volume of PBS (0.035 ml/g)50. BALF was then centrifuged, the supernatant removed and stored at −80˚C and the cell pellet used for cell analysis. Total cell counts were performed using a trypan blue exclusion assay and differential cell counts were analyzed from cytospin slides stained with May-Grünwald and Giemsa stains (VWR, UK). Cells were imaged on a Leica DM5500B microscope and ≥ 400 cells counted per mouse using Image J software.

Alveolar macrophage extraction

Mouse lungs were lavaged using a total of 3 ml PBS containing 0.5 mM EDTA (five 600 µl volumes were used to lavage the lungs twice before collection, the second wash of each lavage was left in the lungs for 3 min). Cells were then pelleted and resuspended in 0.5 ml of ACK buffer (150 mM NH4Cl, 10 mM KHCO3, 0.1 mM Na2EDTA in H2O pH 7.4) for 2 min, before addition of 5 ml PBS, then pelleted again before resuspension in media. Cells were left to attach for 2 h, before washing and were then ready for experimentation.

ELISA

Cell supernatants from MSC potency assays were used for TNFα ELISA (Mouse TNFα ELISA from R&D Systems, Human TNFα ELISA from PeproTech). Mouse osteopontin Duoset ELISA (R&D Systems) and mouse Cxcl1 (KC) Duoset ELISA (R&D Systems) were performed on BALF. Concentrations of cytokines and osteopontin were calculated using four parameter logistical (4-PL) curve analysis.

Western Blot

Lung tissue was homogenized in 10 mM HEPES pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.2% Igepal. The nuclear fraction was then extracted from nuclear/cell debris pellets using 20 mM HEPES pH 7.9, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA and 25% glycerol with protease inhibitors (Roche). Protein concentrations of nuclear lysates were assessed using a Bradford Assay. Samples were then prepared in 1% SDS, 10% Glycerol, 0.05% bromophenol blue and 2% β-mercaptoethanol in 0.1 M Tris pH 6.8. Denatured nuclear lysates (10 µg) were separated on 12% SDS-PAGE gels and transferred onto nitrocellulose membrane (GE Healthcare, Buckinghamshire, UK). Membranes were blocked and incubated with anti-PPARγ (E-8) (#sc-7273, Santa Cruz Biotechnology), and anti-Lamin B1 (#12,987–1-AP, Proteintech, UK) overnight at 4 °C. Membranes were washed and incubated with HRP-conjugated secondary antibodies for 1 h at RT. Membranes were developed using chemiluminescent substrate (Western Lightning, PerkinElmer, Coventry, UK) and viewed using Syngene G:Box and GeneSnap software (Syngene, Cambridge, UK).

Lung function measurements

Mice were placed under surgical anesthesia with an I.P. injection of xylazine—ketamine (10 μL/g). The neck was surgically opened, and trachea cannulated (18-gauge blunt needle), the mouse was then placed on default ventilation delivered at a frequency of 150 breaths/min and a tidal volume of 10 ml/kg, before analysis of lung mechanics on a FlexiVent™ (SciReq, Canada) system. A Deep Inflation perturbation was performed prior to all analysis to inflate the lungs, reach the total lung capacity (TLC) and to enable recruitment of closed lung areas. FlexiVent perturbation maneuvers were then conducted with each perturbation repeated four times for each animal. Perturbations included a deep inflation for inspiratory capacity (IC) measurements, pressure volume (PV) analysis for PV loop curves, static compliance (Cst) and hysteresis, single frequency forced oscillations (single compartment model) for the measurement of resistance (R), elastance (E), compliance (C) of the respiratory system and broadband forced oscillations (constant phase model-which separates conducting airways from the alveolar tissue) for the measurement of Newtonian resistance (Rn), tissue damping (G) and tissue elastance (H).

Rosiglitazone PPARγ activation study

Adult (4–6 week-old) Scnn1b-Tg mice and WT littermates were treated with 10 mg/kg body weight of the PPAR agonist rosiglitazone (RGZ) (Sigma) by subcutaneous injection, or an equal volume of vehicle (4% DMSO/30% PEG300/5% Tween 80/ddH2O) daily for a period of 14 days before sample collection.

Immunofluorescence

Mice were culled, perfused with PBS until the lungs turned white before opening the trachea, cannulating, and inflating the lungs with 50% OCT/4% PFA. Lungs were then placed in 4% PFA overnight before transferring to 30% sucrose/PBS solution for 3–5 days. Samples were then frozen in moulds using OCT (Optimal Cutting Temperature) solution in a 2-methylbutane/dry ice bath and stored at −80°C. Cryosections were cut at 8 µm and placed on Superfrost + slides using a Leica CM1950 cryostat with the chamber set to −24°C. Slides were stored at −80°C until required. Slides were thawed at room temperature for 30 min before washing in TBS-T (0.25% tween-20). Samples were then antigen retrieved using citrate buffer pH 6 at 80°C for 5 min, before washing and permeabilization in 1% triton-X-100/TBS for 30 min. Tissue sections were then circled using a hydrophobic wax pen and blocked in 10% donkey serum/0.1 M glycine/TBS-T for 1 h at room temperature, before incubation with primary antibodies in 1% donkey serum/TBS-T overnight in a humidified chamber at 4°C. Primary antibodies; Rabbit anti-prosurfactant protein C (proSP-C) Antibody serum (1:100 dilution, AB3786, Chemicon®), Rat anti-Ki-67-FITC (SolA15) (1:100, 11–5698-82, eBioscience™), Rabbit IgG (I-1000) and Rat IgG (I-4000) isotype antibodies (Vector Laboratories). After washing in TBS-T, samples were incubated with secondary antibody; donkey anti-rabbit IgG H&L (Alexa Fluor® 568), preadsorbed (1:200 ab175692, Abcam). Sections were then stained with DAPI in TBS (~ 166 ng/ml) for 5 min, before incubation for 30 min with 0.3% Sudan black B/70% ethanol. Slides were washed and mounted overnight in Prolong gold antifade (Invitrogen), coverslips were sealed, and images acquired at 40X on a Leica DM5500B microscope and analysed on QuPath software (DAPI counts), Leica LasX software (SPC+Ki67+ counts) and image J (SPC+ counts) using 4–6 fields of view per mouse. Immunofluorescence antibody validation and staining controls can be seen in supplemental Fig. S7.

Histology and Immunohistochemistry

Lungs were inflated and fixed in 10% neutral buffered formalin (Sigma-Aldrich, UK) at a pressure of 25 cmH2O, before tissue processing on a Leica TP1020 System and wax embedded into paraplast (Leica). The lungs were then sectioned at 5 µm, stained for hematoxylin and eosin and images analysed for the histological emphysema markers, mean linear intercept (MLI)/chord length (Lm) and destructive index (DI)22,23,51. The MLI describes the mean free distance in air spaces, whereas the DI allows calculation of alveolar septal destruction by counting the intact or normal (N) and destroyed (D) alveoli, where DI = 100 × D / (D + N). Bright field images were taken at 20X on a Leica DM5500B microscope and analysed on Image J software using ≥ 6 fields of view.

For proliferation analysis Ki67 was assessed in lung parenchyma using immunohistochemistry (IHC). In brief, 5 μm sections were baked, dewaxed and brought to water, antigen retrieval was achieved using microwave irradiation in 0.1 M citrate buffer pH 6.0, sections were blocked with 3% hydrogen peroxide, then in 3% BSA and 10% normal horse serum before being probed with Anti-Ki67 primary antibody (1.667 µg/ml, Ab15580, Abcam, UK) or rabbit IgG isotype antibody (Vector Laboratories) overnight at 4 °C. Biotinylated horse anti-rabbit secondary antibody (3.75 µg/ml dilution, Vector Laboratories, BA-1100,) was added and detected using an avidin–biotin complex (ABC) peroxidase kit (Vector Laboratories, 111–217) according to the manufacturer’s instructions. DAB (3,3’-diaminobenzidine tetrahydrochloride) substrate (Dako, USA) was added to the slides for immunohistochemical detection before counterstaining in hematoxylin for 30 secs. Bright field images were taken at 40X on a Leica DM5500B microscope and analysed on Image J software using ≥ 10 fields of view.

Gene Expression Analysis

RNA was extracted using a miRNeasy kit (Qiagen, UK). RNA was quantified using a NanoDrop™ One Spectrophotometer (Thermo Scientific) and reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Thermo Scientific). Q-PCR was performed using TaqMan™Gene Expression Assays (FAM) and the TaqMan™ Gene Expression Master Mix (Thermo Scientific) as per the manufacturer’s instructions. Human specific probes included GAPDH (Hs03929097_g1) and MT-CO1 (Hs02596864_g1). Mouse probes included Gapdh (Mm99999915_g1), Pparg (Mm00440940_m1), Il6 (Mm00446190_m1), Cxcl1 (Mm04207460_m1), Egr1 (Mm00656724_m1), Spp1 (Mm00436767_m1) and Il1β (Mm00434228_m1). Gene expression was assessed on the LightCycler® 480 instrument II system (Roche) using the Advanced Relative Quantification software for analysis.

Statistics

Statistical methods relevant to each figure are highlighted in each of the legends. Analysis was performed in Prism 9 software (GraphPad Software, La Jolla, CA). Normality testing of data was performed using the D’Agostino and Pearson Omnibus test. Comparison of parametric data was achieved using the students t-test or one-way ANOVA. For non-parametric data, the Mann–Whitney or ANOVA with Kruskal Wallis was used. Paired t-test and Two-way ANOVA were also used when appropriate. The statistical significance levels were as follows * = P ≤ 0.05, ** = P ≤ 0.01, *** = P ≤ 0.001 and **** = P ≤ 0.0001.