Female 6-8-week-old C57BL/6 mice were divided into three groups (n=5 per group). Mice were sacrificed after 2 (short-term) and 9 (long-term) days after intranasal (i.n.) administration of the adipose-tissue-derived MSCs ( Figure 1A ; for detailed method description please see Supplementary Materials – methods ). Biological material was collected and biobanked for further analyses.

The presence of inflammatory infiltration and mucus production in the lung was assessed by histochemical stainings. First, lungs were fixed in 4% paraformaldehyde and paraffin-embedded. Next, 4μm microtome sections were placed on the glass slides (Thermofisher Scientific) and stained with hematoxylin-eosin (H+E) and Periodic acid-Shiff (PAS) according to the standard protocols. The slides were visualized using a digital slide scanner Nanozoomer SQ (Hamamatsu). Both H+E and PAS staining’s were quantified in ImageJ software.

Inflammatory infiltration within the lung tissue was quantified using ImageJ software (NIH) in H+E-stained slides. The default thresholding method and the HSB model for color space were selected to perform the analysis. The lung tissue surface was measured using a threshold tool. The slider in the brightness panel was appropriately adjusted to cover all tissue areas. Additionally, to evaluate the inflammation surface the slider in the hue panel was acquired to imply the dark pink-purple colors. Three independent measurements were performed and the mean was calculated. The results were presented as inflammation area to tissue (slide) area ratio. Additionally, all the values were normalized to the mean of control group measurements. To evaluate the mucus production in PAS-stained slides, the quantification was restricted only to bronchi. To maximize the relevance of the results two different bronchioles within one tissue slide were taken into consideration. Similarly, to assess the mucus area, the slider was adjusted to incorporate pink-red colors. Additionally, the surface inside all bronchi was measured. To simplify the calculations, the bronchi shape was assumed as a circle and the perimeter was calculated. Three measurements for all bronchioles were performed and the mean was calculated. The results were presented as a ratio of mucus area and bronchioles perimeter score.

Lung tissue dissociation was performed using Lung Dissociation Kit (Miltenyi Biotec). Next, cells were stimulated with Leukocyte Activation Cocktail with Golgi Plug (BD Pharmingen) for 3 hours. Extracellular and intracellular staining was performed according to the standard protocol using a panel of fluorochrome-labeled monoclonal antibodies ( Supplementary Materials – Supplementary Table S2 ). Cells were acquired using the FACSAria system (BD Biosciences) and analyzed with the FlowJo v.10 software (BD Biosciences). A representative gating strategy has been presented in the Supplementary Materials – Supplementary Figure S2A .

Snap-frozen lung tissues were cryosectioned (Leica CM3050 S) at 8μm, and subsequently fixed with 4% paraformaldehyde. Cryosections were submerged in blocking solution (10% goat serum, 1% bovine serum albumin, and 0,2% TritionX100) prior to incubation with polyclonal rabbit anti- ZO- 1 antibody (Invitrogen), monoclonal mouse anti-occludin (Invitrogen) at 1:200, and polyclonal rabbit anti-claudin 3 (Invitrogen) at 1:100 in 1% BSA in PBS, followed by incubation with Alexa Fluor 488 Goat anti-Rabbit IgG (Invitrogen) or Goat anti-Mouse IgG (Invitrogen) at 1:1000. Specimens were analyzed using a Zeiss LSM780 microscope (Zeiss). The detailed information on used antibodies has been presented in the Supplementary Materials – Supplementary Table S2 .

Statistical analysis was performed using GraphPad Prism v.9. Statistical significance was evaluated by the U Mann-Whitney test; p<0.05 was considered significant.

Lung lobes were stored in RNA later solution (Invitrogen) for 48 hours to stabilize RNA. Next, tissues were disrupted using TissueRuptor II (Qiagen) in RNeasy Lysis Buffer (Qiagen). Total RNA was isolated by using the RNeasy Mini Kit according to the manufacturer’s protocol (Qiagen). 1 µg of total RNA with RNA integrity number (RIN) > 8, was subjected to the cDNA library preparation according to TruSeq Stranded Total RNA protocol (Illumina), followed by the quality confirmation by TapeStation 2200 (Agilent, CA, USA). Next-generation sequencing (RNAseq) was performed using the Illumina HiSeq 4000 platform generating 150 bp paired-end reads (2 x 75 bp). Subsequently, transcriptomic profiling and analysis were performed. The entire data set has been submitted to the NCBI GEO database: accession number GSE200028 (The datasets are currently private and available under access token: “ovivksgmxdgzhcz”, and will be released immediately after manuscript acceptance).

Sequencing quality was evaluated by FastQC version 0.11.5. Reads were mapped to the reference genome of Mus musculus (GRCm38) using STAR aligner version 2.5.3a. The obtained read counts were used to differential expression analysis (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/; https://qubeshub.org/resources/fastqc) (15). Differential gene expression analysis was performed using DESeq2 (16). To adjust the Wald test p-value, the procedure of Benjamini & Hochberg was applied. Additionally, the count matrix was transformed into Transcripts Per Milion (TPM) to normalize gene expression. Differentially expressed genes (DEGs) with matched HGNC symbols were identified based on adjusted p-values < 0.1, and absolute Log2FoldChange > 0.5. To visualize the expression of DEGs, Venn Diagram, Volcano Plots, and heatmap with complete linkage clustering was generated in “R”. Gene Set Enrichment Analysis was performed to reveal Gene Ontology terms present in the dataset. Subsequently, the top 20 terms according to the Benjamini & Hochberg adjusted p<0.05 were plotted on the graphs. Moreover, based on The Mouse Genome Database (MGD; http://www.informatics.jax.org; access date: 29^th June 2022), individual branches of Gene Ontology terms were selected for further analysis. Commonly regulated genes were analyzed using string. db and Ingenuity Pathway Analysis (IPA, QIAGEN Inc., https://digitalinsights.qiagen.com/IPA). The STRING gene networks were generated concerning all commonly regulated genes, including the predicted and RIKEN genes (n=104). Nodes without any connections were excluded from the network on set medium confidence levels. Using IPA generated pathways with altered z-score, commonly regulated genes for two investigated models were analyzed individually. A two-tailed Mann U Whitney test was applied to assess the difference in the expression of mentioned genes. The delta of the expression was presented using the R package ggplot2 (https://ggplot2.tidyverse.org). The expression datasets were analyzed in IPA with the cutoff points for absolute Log2FoldChange > 0.5 and adjusted p-value < 0.1 with additional lung tissue filters applied.

First, we aimed to confirm that adipose tissue-derived cells fulfill the criteria of mesenchymal stem cells established by the International Society for Cellular Therapy. We expanded plastic adherent cells and confirmed the surface expression of MSCs characteristic markers, namely CD73, CD90, and CD105, with simultaneous lack of lineage marker CD45 and human leukocyte antigen (HLA-DR) expression ( Supplementary Figure S1A ). Moreover, we successfully differentiated the cells in vitro into adipocytes, osteocytes, and chondrocytes ( Supplementary Figure S1B ).

Having confirmed that cells isolated from adipose tissue fulfill the criteria of MSCs, next, we wished to investigate the effects of their i.n. administration on the induction of inflammation in the lower airways ( Figure 1A ). We found no signs of increased cellular infiltration and mucus production within the lungs both directly (two days, short-term) and extendedly (nine days, long-term) after MSCs transfer ( Figure 1B , Supplementary Materials - Figure S2A ). However, we observed an increase in the frequency of INFγ producing, but not IL-4, IL-17, and IL-10 producing, CD3⁺CD4⁺ T cells in the lungs as an effect of the long-term MSCs administration ( Figure 1C ; Supplementary Materials - Figure S2B ). Moreover, occludin and claudin 3, but not ZO-1 were decreased in the long-term model ( Figure 1D ). To better understand the effects of MSCs administration on the non-inflamed lungs, we next aimed to investigate MSC-mediated effects on the transcriptomic profiles of the lungs. We observed dynamic changes in the lung gene expression profiles after MSCs administration ( Figure 1E ). We found 674 differentially regulated genes unique for the short-term model, while only 75 genes were unique for the prolonged observations ( Figure 1F ), suggesting waning of the low-grade inflammation and active resolution in the longer time point. In addition, a total of 104 genes were common for both analyzed time points.

Having found significant changes in the transcriptomic profiles, we next wished to elucidate whether differentially regulated genes may be functionally related, integrated, and referred to the specific genes clusters and interaction nodes. Thus, we evaluated the enrichment of DEGs in the gene ontologies, signaling pathways, and mapped predicted interactions using clusterProfiler v.4.0.0 (17). In the top 20 most significant gene ontology terms we found changes in immunological pathways. The normalized enrichment score (NES) analysis indicated activation of the innate and adaptive immune responses. We found upregulation in phagocytosis and engulfment processes and B cell receptor signaling in both analyzed time-points ( Figure 2 ). Moreover, we found an increase in the ribosome function, and biogenesis in MSC-treated mice, longitudinally ( Figure 2 ). More precisely, the analysis of canonical pathways activation revealed the increase in IL-7 signaling, T and B cell signaling in the short- and long-term models compared to controls. Moreover, we found changes reflecting redox imbalance, such as an increase in HIF1α signaling and superoxide radicals degradation in the short-term model ( Figure 3A ). In addition to this observation, we noted the gradual downregulation in HIF1α signaling, IL-17 signaling, and B cell receptor signaling in the longer time point ( Figure 3A ). Furthermore, the analysis of the expression of genes clustered to the terms and processes related to airway inflammation, namely Th1-, Th2-, Th17-driven immune responses development, tight junction molecules, and mucins, revealed a relatively low number of significantly changed genes ( Figure 3B ), which stay in line with our ex vivo observations of effector T cells, epithelial barrier integrity, and histochemical staining’s.

Next, we analyzed deeper the expression profiles of genes related to oxidative stress and immune responses, macrophage activation and phagocytosis. We noted the dysregulated expression of pattern recognition receptors (PRRs, Figure 4A ), macrophage activation ( Figure 4B ), oxidative stress ( Figure 4C ), phagocytosis ( Figure 4D ), and inflammation of the respiratory system ( Figure 4E ). Importantly, clustered genes present a trend to be downregulated in the latter time point towards the level observed in the untreated controls. However, in the short-term model, we noted an upregulated expression of most analyzed genes. However, a relatively low number of genes were significantly upregulated (FDR < 0.05) in analyzed models in the cluster reflecting inflammation within the respiratory system (9 upregulated genes among 50 defined in the cluster, Figure 4E ).

Having found longitudinal changes in the gene expression profiles among MSCs-treated groups, next, we aimed to focus on the common genes for both short-term and long-term models. First, we found a low number of interactions among analyzed genes ( Figure 5 ). The observed ones mainly reflect dysregulation of immune responses, namely dendritic cell, T cell, and B cell function (Cd7, Cd37, Cd72, Cd79a, Spib, and Il-21r, Cxcr5, Ccl5, Zbp1). In addition, we observed interactions for ribosome biogenesis, function, and cell cycle (Rrs1, Gpatch4, Trp35, Ncl, Rasl11a, Rbm38, Hist1h1b), shock proteins (Cirbp, Hspb6), and circadian rhythm (Dbp, Arntl, Npas2, Nfil3) ( Figure 5A ). Furthermore, we analyzed the most significant genes with altered z-scores at the investigated time-points according to the Ingenuity Pathway Analysis. We found Rap2b relative expression as the one of most changed compared to the other delta expression of genes ( Figure 5B ). According to the Pathcards and Reactome database (https://pathcards.genecards.org; https://reactome.org; access date: 4^th July 2022), Rap2b is predicted to be involved in the neutrophils degranulation pathway, which is linked with the reactive oxygen species production (18). Moreover, we also noted a relatively relevant change in Trp53 ( Figure 5B ), which is also recognized as an important contributor to oxidative stress-induced necrosis (19). Additionally, a significant change in Gpr132 relative expression was observed, which is highly specific to infiltrating macrophages (20). Finally, we also observed a trend in a longitudinal decrease in the expression of the analyzed common genes ( Figures 6A, B ), which may indicate lung homeostasis reestablishment.