*3.4. Small RNA Cargo Is Altered in Vesicles from Asthmatic Subjects*

We additionally checked whether EVs derived from epithelial cells of healthy and asthmatic subjects differed in their miRNA expression profiles regardless of the secretion side, and found overall 12 miRNAs differentially represented in the two groups, with 6 being up- and 6 down-regulated in the vesicles of asthmatics (Supplementary Figure S2). A more complex picture was obtained when additionally taking the EV secretion side into account (Figure 7A). In apical EVs, 32 miRNAs showed a significant difference in abundances in the two groups, 29 of which were up- and 3 of which were down-regulated in the vesicles of asthmatics (Figure 6A, left). On the basolateral side, 23 miRNAs with a divergent expression profile were detected, 9 being significantly upregulated and 14 being downregulated in asthmatics (Figure 7A, right), with 5 out of 12 being family members of the let-7 family and with 3 out of 8 being members of the miR-10 family (Figure 6C). Specifically, the miR-9 family showed significant differences between the healthy and the asthmatic phenotype origin in both EV secretion compartments (Figure 7B,C). The KEGG pathway terms and biological functions associated with differentially abundant miRNAs in apical and basolateral EVs from asthmatic versus healthy subject's bronchial epithelial cell cultures are shown in Supplementary Figure S3A,B, respectively. Enriched associated diseases included a variety of inflammatory conditions, among them asthma, especially when the analysis was based on the signals from the basolateral side (Supplementary Figure S3C). Significantly enriched target regulatory proteins in asthmatic subjects were HIF1A and NFKB1 (Supplementary Figure S3D). The target genes of the EV-derived miRNAs differentially expressed between both conditions are shown in Supplementary Figure S4A,B, associated with some of the most significant pathways.

**Figure 7.** Differential miRNA expression analysis per disease condition (asthma versus healthy) and

compartment (apical and basolateral). (**A**) Volcano plot and heat map of differentially expressed miRNAs of asthma versus healthy comparison in EVs from the apical (left) and basolateral (right) cell culture compartments at *p*adj < 0.05 and FC ≥ 2, and (**B**,**C**) donut charts showing the distribution of the mainly represented miRNA families in the asthma versus healthy comparison in both compartments. Digits in brackets depict the number of enriched miRNA family members out of total number of miRNAs belonging to the respective family, \*\* *p* < 0.01. miRNA—microRNA; EV—extracellular vesicle; FC—fold change.

### **4. Discussion**

In the past decade, the role of EVs as communication structures between neighboring or remote cells has been increasingly recognized. The presence of EV-associated RNAs has been attested by next-generation sequencing in numerous body fluids, including blood plasma and sputum [35]. Specific miRNA signatures hold the potential for being used as fingerprints, helping to identify phenotypes or states of diseases and gain more insights into their underlying pathological mechanisms. Plenty of studies have analyzed miRNA profiles in patients with asthma compared to subjects not affected by this disease, some with a specific focus on the role of airway epithelial cells in EV-related miRNA generation [36]. However, to date, no investigations have explicitly distinguished between EV RNA profiles in polarized airway epithelial cells depending on their direction of secretion. Hence, no distinction has been made between their supposed location of action being either the outer epithelial environment, e.g., sputum, or compartments within the body, such as lung interstitium, tissue, or even blood plasma.

According to the results of the NanoFCM particle characterization, EV populations isolated from the apical cell side were mainly composed of vesicles with diameters matching the size range of the exosomes. Contrarily, on the basolateral cell side, median vesicle diameters were noticeably larger, consistent rather with the size of the microvesicles than with the exosomes [37,38]. FACS analysis for exosomal marker proteins revealed the presence of exosomes in both apical and basolateral samples, although apical EVs seemed to express exosomal marker proteins to a greater extent than the basolateral vesicles.

In our study, we were able to observe distinct differences in EV-associated miRNA patterns secreted by bronchial epithelial cells from healthy and asthmatic subjects, depending on the side of EV secretion. Interestingly, secretion patterns of EV miRNAs varied more distinctly based on the side of secretion than the pathophysiological condition. This knowledge might be essential for future investigations into potential biomarkers analyzed in different compartments such as sputum and plasma. Bartel et al. recently published a PCR-based study comparing the expression of specific miRNAs in EVs secreted by normal human bronchial epithelial cells to the basolateral and apical cell side. Interestingly, there were some notable overlaps in miRNAs, including miR-34b and miR-21 preferentially identified on the apical side, while other differentially expressed miRNAs in this study could not be observed in our analysis [39]. Notably, serum levels of miR-21 have been previously reported to be an efficient biomarker for asthma patients [40]. Further, treatment with a miR-21-specific antagomir was demonstrated to reduce airway hyperresponsiveness and restore steroid sensitivity in mice with ovalbumin-induced allergic airway inflammation [41]. The differentially expressed miR-10 on the apical side was found to regulate the proliferation of airway smooth muscle cells by suppressing the phosphoinositide 3-kinase (PI3K) pathway [42]. Moreover, MAPK and mTOR signaling pathways were enriched as potential targets for differentially expressed miRNAs on the apical side. Activation of the mTOR pathway has been shown to lead to tight junction susceptibility and epithelial–mesenchymal transition (EMT), which can in turn play an essential role in airway remolding in asthma pathogenesis [43,44]. Accordingly, inhibition of MAPK signaling pathway led to a significant reduction in the allergic inflammation of the airways [45]. In contrast, miR-221, which was downregulated on the basolateral side, was shown to play a unique role in controlling the differentiation of Th17 and regulatory T (Treg) cells through targeting SOCS-1 (suppressor of cytokine signaling 1) [46]. Likewise, T and B cell

receptor signaling pathways have been associated with miRNAs dominantly secreted to the basolateral side. These pathways have been intensively investigated as key regulators of the antigen recognition in the adaptive immune response and have been utilized as key therapy targets in asthmatic patients [47,48]. Taken together, our observations strongly indicate a compartmentalized packaging and side-specific release of EVs by bronchial epithelial cells, pointing to site-specific functions of these structures at least partially mediated by their miRNA cargo.

While differences in miRNA profiles from apical versus basolateral sides were more pronounced, a number of significant alterations in the levels of certain EV-associated miR-NAs (or their families) were also observed between cells of an asthmatic or healthy origin, either at one or at both sides. For example, EV miRNAs belonging to the miR-9 family were upregulated in both apical and basolateral vesicles of asthmatic patients. Increased levels of miR-9 have already been linked to steroid-resistant and neutrophilic, but not eosinophilic asthma [49]. Interestingly, Bazzoni et al. observed a miR-9-dependent inhibition of NFκB1 transcription in human neutrophils and monocytes exposed to proinflammatory signals, suggesting that the rapid induction of miR-9 operates as a feedback control of NFκB1 dependent cellular response [50]. Accordingly, NFκB1 was found among the predicted targets of miRNAs, specifically present in asthmatic samples in our investigation. Several independent studies further identified enhanced NFκB-pathway activation in asthmatic tissue [51].

Interestingly enough, miR-34b and miR-34c were downregulated when associated with EVs released by epithelial cells from asthmatic subjects at both compartments, even though they were generated from opposite DNA strands, dependent on the compartment. The levels of miR-34b/c have been found to be significantly lowered in murine mouse models of ovalbumin-induced allergic airway inflammation and have been suggested to play a regulatory role in the activation of the Nrf2-/ARE pathway [52]. Disruption of Nrf2 expression augmented airway inflammation and hyperresponsiveness [53–55]. Moreover, in a clinical study by Solberg et al., treatment with corticosteroids resulted in increased levels of miR-34b/c in BALF of patients suffering from asthma, while the administration of IL-13 was able to repress its expression in an air−liquid interface bronchial epithelial cell culture model [56]. This suggests a protective role of miR-34b/c against allergic and asthmatic cellular responses, and hints toward a diagnostic potential of the miR-34 family.

Generally, miRNA of the let-7 family showed a higher expression in apical compared to basolateral EVs. However, in basolateral EVs, the let-7 family was the miRNA subgroup with the largest divergence between asthmatic and healthy subjects, with 5 out of 12 total family members showing significant alterations, while on the apical side, no differences in expression were found. In contrast to this finding, Levänen et al. described significant variations in 16 miRNA, including the let-7 and miR-200 families and miR-99 (as seen in our study) in BALF [24]. Interestingly, while let-7c and let-7d were downregulated, let-7a showed an upregulation in EVs released by cells from asthmatics. miRNAs of the let-7 family are among the most intensely studied miRNAs, with two studies proposing the exosome-mediated transfer of let-7 miRNAs to various immune cells as a suppressive mechanism used by Treg cells (let-7d), and reporting the inhibition of Treg cell generation and function by these miRNAs (let-7i) [57,58].

There are some limitations to this study that should be mentioned. One limitation is the small sample size, yet the major goal of the study was a general overview of EV fingerprints in different cellular compartments, rather than assessing the function of singular EV miRNAs. For a more detailed analysis, further studies investigating the role of single miRNA are required to give the data a clinical significance. Another limitation is that there was no information available about the clinical details and/or the asthmatic phenotype of the patients whose cells were involved in the study. Lastly, as miRNAs can be secreted by almost every cell type, the overall in vivo situation can largely dissociate from in vitro observed conditions. Dissecting the contribution of individual cell types in

the production of specific miRNA is, on the other hand, essential to understand their role in disease pathogenesis.

In summary, in this study, we provide a general overview of miRNA-composition in EVs secreted by airway epithelial cells. We were able to reveal distinctly differing miRNA expression profiles depending on the vesicle side of secretion and disease condition. This emphasizes the importance of taking the vesicle site of action into consideration for further research to which the data presented in this study provide a sound basis.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biomedicines10030622/s1. Supplementary Figure S1: Comparison of EV total RNA composition between basolateral and apical compartments regardless of disease expression in percentage of reads mapped to indicated RNA species. EV—extracellular vesicle; rRNA—ribosomal RNA; mRNA—messenger RNA; lncRNAs—long-non-coding RNAs; piRNA— PIWI-associated RNAs; miRNA—microRNA. Supplementary Figure S2: Differential miRNA expression analysis per disease condition (asthma versus healthy) regardless of compartment (apical and basolateral). (A) Volcano plot and (B) heat map exhibiting the differently expressed miRNAs of the asthma versus healthy comparison at *p*adj < 0.05 and FC ≥ 2. miRNA—microRNA; FC—fold change. Supplementary Figure S3: Functional analysis of the differentially expressed miRNAs in the asthma versus healthy comparison in each apical and basolateral compartment. (A) KEGG pathway and (B) biological function analyses of differentially expressed vesicular miRNAs of asthma versus healthy in each apical and basolateral compartment, (C) enriched diseases, and (D) target proteins potentially regulated by the differentially expressed miRNAs. miRNA—microRNA. Supplementary Figure S4: mRNA−miRNA network analysis involving differentially regulated miRNAs of the asthma versus healthy comparison in (A) the apical and (B) the basolateral compartments associated with biological pathways of potential target genes. miRNA—microRNA.

**Author Contributions:** H.G., D.P.P., V.E.M.S. and F.A.—designed the experiments. V.E.M.S.—isolated the material and performed EV and RNA isolation. F.A.—prepared the material for sequencing. A.N. and T.S.—implemented sequencing analysis. C.P., B.A.A. and E.P.v.S.—implemented NanoFCM analysis. V.E.M.S. and L.H.—performed FACS analysis. C.T. and F.A.—performed the bioinformatics analysis. V.E.M.S., H.G. and F.A.—drafted the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the German Center for Lung Research (DZL); the German Academic Exchange Service (DAAD; F.A., personal reference number: 91726294); the HessenFonds, World University Service (WUS; F.A.); the Hessen State Ministry for Higher Education, Research and the Arts (HMWK; F.A.); and the German Research Foundation (DFG; E.P.v.S., B.A.A. and T.S., Grant 512416910386–GRK 2573/1).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study and underlying raw data are available on reasonable request from the corresponding author.

**Acknowledgments:** We would like to thank our funding institutions.

**Conflicts of Interest:** Authors declared no conflict of interest.
