Reduced expression of miR-146a in human bronchial epithelial cells alters neutrophil migration
The role of miRNAs in the pathogenesis and determining the phenotypes of asthma is not fully elucidated. miR-146a has been previously shown to suppress inflammatory responses in different cells. In this study, we investigated the functions of miR-146a in human bronchial epithelial cells (HBECs) in association with neutrophilic, eosinophilic, and paucigranulocytic phenotypes of asthma.
Bronchial brushing specimens and brochial mucosal biopsy samples were collected from adult patients with asthma and from age- and gender-matched non-asthmatic individuals. The expression of miR-146a in bronchial brushing specimens, bronchial biopsy tissue sections or cultured primary bronchial epithelial cells was analyzed by RT-qPCR or by in situ hybridization. The expression of direct and indirect miR-146a target genes was determined by RT-qPCR or ELISA. The migration of neutrophils was studied by neutrophil chemotaxis assay and flow cytometry. For statistical analysis, unpaired two-way Student’s t test, one-way ANOVA or linear regression analysis were used.
Reduced expression of miR-146a was found in bronchial brushing specimens from asthma patients as compared to non-asthmatics and irrespective of the phenotype of asthma. In the same samples, the neutrophil attracting chemokines IL-8 and CXCL1 showed increased expression in patients with neutrophilic asthma and increased IL-33 expression was found in patients with eosinophilic asthma. Linear regression analysis revealed a significant negative association between the expression of miR-146a in bronchial brushings and neutrophil cell counts in bronchoalveolar lavage fluid of patients with asthma. In bronchial biopsy specimens, the level of miR-146a was highest in the epithelium as determined with in situ hybridization. In primary conventional HBEC culture, the expression of miR-146a was induced in response to the stimulation with IL-17A, TNF-α, and IL-4. The mRNA expression and secretion of IL-8 and CXCL1 was inhibited in both stimulated and unstimulated HBECs transfected with miR-146a mimics. Supernatants from HBECs transfected with miR-146a had reduced capability of supporting neutrophil migration in neutrophil chemotaxis assay.
Our results suggest that decreased level of miR-146a in HBECs from patients with asthma may contribute to the development of neutrophilic phenotype of asthma.
KeywordsAirway epithelium HBEC Endotype microRNA miR-146a
Asthma affects up to 300 million people worldwide . Patients with asthma suffer from bronchoconstriction and increased mucus production in the airways that results in symptoms like coughing, wheeze and chest tightness [2, 3]. As asthma is a highly heterogeneous disease, attempts have been made to define asthma phenotypes and/or endotypes on the basis of involved immune cells or molecular processes, respectively [4, 5, 6]. As one option, this allows categorization of asthmatics into eosinophilic, neutrophilic, mixed granulocytic and paucigranulocytic asthma phenotypes [4, 5]. The eosinophilic phenotype is characterized by the presence of eosinophils in the lung, is frequently associated with increased Th2 responses, involvement of type 2 innate lymphoid cells (ILC2) and increased cytokines IL-4, IL-5 and IL-13 [7, 8]. Patients with paucigranulocytic asthma tend to have fewer overall immune cells in the airways, as measured in the bronchoalveolar lavage fluid (BALF) [4, 5]. In neutrophilic asthma, neutrophil chemoattractants are secreted by bronchial epithelial cells and fibroblasts, for example in response to IL-17 secretion from Th17 cells or ILC3 [9, 10]. Independent from phenotype, airways of all asthma patients are influenced by chronic inflammation, which also leads to the changes in genes affecting cell proliferation, such as HBEGF and FGF2 and eventually, airway remodeling may take place [1, 2, 3].
Inhaled glucocorticosteroids represent hallmark of asthma management, often together with other controller medicines that include β2-agonists, theophylline, antileukotrienes and anticholinergics [11, 12, 13]. In addition, several biologics have become available recently for severe asthma and more will be introduced in the future [6, 14]. However, because these treatment options are not always applicable and occasionally fail to prevent asthma exacerbations, there is still a need for better understanding of the molecular background of asthma to achieve more effective treatment [15, 16].
miRNAs are short endogenous RNAs capable of regulating gene expression through binding their target mRNA via partial complementarity, which leads to inhibition of the translation or mRNA degradation [17, 18, 19]. The majority of protein coding genes in human cells and most biological processes, including innate and adaptive immune responses influencing the development of asthma, are regulated by miRNAs [20, 21]. Previous studies have shown that miR-146a has important role in the suppression of inflammatory responses in different cell types [22, 23, 24] and therefore is involved in modulation of immune responses. In line with that, we have previously shown that miR-146a inhibits inflammatory responses in human primary keratinocytes and in a mouse model of atopic dermatitis [25, 26]. One of the main target genes of miR-146a is IL-1 receptor-associated kinase 1 (IRAK1) , which suppression leads to reduced activity of the nuclear factor (NF)-κB pathway and inhibition of numerous pro-inflammatory chemokines, including C–X–C Motif Chemokine Ligand (CXCL)1 and IL-8 [26, 28]. A recent genome-wide association study of a broad allergic disease phenotype (GWAS; n = 360,838) identified MIR3142–MIR146A as a shared risk loci of asthma, hay fever and atopic dermatitis . The capacity of miR-146a to inhibit the expression of pro-inflammatory chemokines IL-8 and CCL5 , to augment the anti-inflammatory effect of glucocorticosteroids  in human lung alveolar epithelial cell line A549, and to inhibit neutrophil elastase-induced gene MUC5AC in immortalized human bronchial epithelial cell line 16HBE  has been formerly shown. Very recently, the increased expression of miR-146a in cultured commercial human airway or bronchial epithelial cells (HBECs) was reported . However, no data about miR-146a expression in airway epithelial cells of asthma patients is available, as well as the potential influence of miR-146a on the development or contribution to the heterogeneity of asthma has not been studied before.
Here we analyzed the expression of miR-146a and selected genes either known to be influenced by miR-146a or associated with Th2 or Th1 responses in brush biopsy specimens from patients with asthma. In addition, the expression regulation and function of miR-146a in the modulation of molecular processes associated with asthma was studied in primary conventional HBEC cultures. Our results reveal that reduced expression of miR-146a in HBECs may lead to cellular processes that skew immune responses towards neutrophilic one and thus contribute to the development of neutrophilic phenotype of asthma.
Materials and methods
Clinical characteristics of studied patients
Age (y), mean ± SD
45.6 ± 13.1
39.8 ± 9.5
48.1 ± 18.7
49.3 ± 11.3
52.2 ± 16.6
Duration of asthma (y), mean ± SD
15.0 ± 16.3
8.3 ± 8.0
13.9 ± 7.9
21.8 ± 22.7
FEV1 % predicted
81.8 ± 24.0
86.6 ± 22.7
74.2 ± 32.9
81.9 ± 20.1
111.2 ± 9.6
FEV1 % reversibility
7.2 ± 7.2
6.5 ± 5.4
6.2 ± 5.7
8.4 ± 9.5
3.0 ± 3.5
Positive skin prick test with common allergens n, (%)
Severity of asthma, median (IQR)d
Inhaled glucocorticosteroids (%, average dose μg, [range])
96.6, 969 (0–4000)
100, 940 (250–2000)
100, 929 (500–2000)
Systemic glucocorticosteroids n, (%)
Cell culture, stimulation and transfection of HBECs
HBECs were isolated from bronchial biopsies by initial short-term pronase (Roche, Basel, Switzerland) and DNase (Sigma-Aldrich, USA) digestion. In total, cell lines were isolated from 1 asthma and 4 control subjects. Primary HBECs were cultured as monolayer in supplemented BEGM™ (Bronchial Epithelial Cell Growth Medium, Lonza, Basel, Switzerland) medium. For stimulation, IFN-γ (final concentration 20 ng/ml, eBiosciences, USA), TNF-α (20 ng/ml, Biolegend, USA), IL-17A (10 ng/ml, Peprotech, UK), IL-22 (20 ng/ml, Peprotech, UK) and IL-4 (40 ng/ml, Peprotech, UK) were used. For transfections, miRIDIAN microRNA Mimic Negative Control #1 (Dharmacon™, USA) or miRIDIAN microRNA hsa-miR-146a-5p mimic (Dharmacon™, USA) were transfected using MIRFECT (RNAexact, Estonia) according to manufacturer’s protocol. Additional information can be found in Additional file 1.
RNA isolation, cDNA synthesis and RT-qPCR
For RNA isolation Qiazol (Qiagen, Germany) and Total RNAzole out Mini kit (A&A Biotechnology, Poland) were used according to the manufacturer’s instructions. To analyze miRNA expression, either TaqMan® MicroRNA Assays (Life Technologies, California, USA) and 5× HOT FIREPol® Probe qPCR Mix Plus (ROX) (Solis BioDyne, Tartu, Estonia) or miScript II RT Kit, miScript SYBR Green PCR Kit and Hs_miR-146a_1 miScript Primer Assay (cat. MS00003535) by Qiagen were used according to the manufacturer’s protocols. For normalization let-7a and ΔΔCt calculation were used. For mRNA RT-qPCR, cDNA was synthesized using oligo-dT (TAG Copenhagen, Denmark), RevertAid Reverse Transcriptase (Thermo Scientific) followed by qPCR with 5× HOT FIREPol EvaGreen qPCR Supermix (Solis BioDyne, Estonia). Target gene expression was normalized to EEF1A1 expression using ΔΔCt calculation. Additional information is provided in Additional file 1.
ELISA and neutrophil chemotaxis assay
Supernatants from HBECs transfected with miRNA mimics and stimulated with cytokines were used for ELISA and neutrophil chemotaxis assay. ELISA MAX™ Deluxe Set (BioLegend, 431504) for IL-8 and human CXCL1/GRO alpha DuoSet ELISA (R&D Systems, DY275-05) were used according to the manufacturers’ instructions. For neutrophil chemotaxis assay, 4 × 105 of primary human neutrophils were seeded on ThinCert cell culture inserts (3-μm pore size) (Greiner Bio-One, Kremsmünster, Austria) placed in 24 well plate containing the supernatants from HBECs. 60 min after incubation at 37 °C, the number of migrated neutrophils was analyzed by using BD LSRFortessa (BD Biosciences, USA) cell analyzer. More detailed information is provided in Additional file 1.
In situ hybridization (ISH)
ISH was performed using 10 μm sections of frozen bronchial mucosal biopsy specimens. Using miRCURY LNA miRNA ISH Buffer Set (FFPE), hsa-miR-146a-5p miRCURY LNA miRNA Detection probe (cat. YD00619856) and control probe miRCURY scrambled ISH 49 °C (cat. YCD0074470-BCG) were used according to the manufacturer’s (Qiagen) protocol. Densitometry analysis of staining in in situ hybridization images were performed with ImageJ software. Additional information is provided in Additional file 1.
For visualization and statistical analysis, GraphPad Prism 5 (GraphPad Software Inc, USA) and unpaired two-way Student’s t-test, One-way ANOVA or linear regression analysis were used. The results were considered significant at *P < 0.05; **P < 0.01; ***P < 0.001. For heatmapping, online software Morpheus (https://software.broadinstitute.org/morpheus/) was used.
The expression of miR-146a is reduced in bronchial brush specimens and is in negative association with the number of neutrophils in the airways of patients with asthma
The expression of miR-146a-influenced chemokines is increased in the airway epithelial cells from patients with neutrophilic asthma phenotype
Pro-inflammatory cytokines induce the expression of miR-146a in HBECs
Overexpression of miR-146a leads to decreased expression of pro-inflammatory chemokines, diminished expression of airway remodeling associated growth factors and moderate increase in interferon response genes
miR-146a inhibits the expression of secretable factors needed for neutrophil migration
Asthma is a heterogeneous disorder, which is not always well controlled. Hence, there is an unmet need for better understanding of the mechanisms underlying the phenotypic diversity of asthma. In this study, we report significantly lower miR-146a expression in brushing biopsy airway epithelial cells from patients with asthma and a negative association between the reduced miR-146a expression and the number of neutrophils present in the airways. In the bronchial biopsy specimens, miR-146a was most highly expressed in the epithelium layer as determined by ISH. Further analyses showed expected excessive expression of neutrophil attracting chemokines IL-8 and CXCL1 in HBECs from the patients with neutrophilic asthma, while eosinophilic subgroup showed increased IL-33 expression. In cultured HBECs, the expression of miR-146a was induced in response to stimulation with pro-inflammatory cytokines characteristic of Th17 and Th2 driven inflammation. Overexpression of miR-146a in HBECs inhibited the expression of neutrophil attracting chemokines IL-8 and CXCL1 and reduced the number of migrating neutrophils in neutrophil chemotaxis assay. Together, our results indicate that reduced expression of miR-146a in airway epithelial cells from asthma patients may contribute to the development of neutrophilic phenotype of asthma.
Several studies have reported dysregulation of miRNAs in airway epithelial cells from asthmatics and suggest that these changes may contribute to the regulation of inflammatory responses and airway remodeling [37, 38, 39, 40]. However, to our knowledge, our study is the first describing the downregulated expression of miR-146a in airway epithelial cells of asthma patients. This finding is consistent with a previous study showing that patients with severe asthma have lower expression of miR-146a in CD8+ and CD4+ T cells . However, our observation is quite unusual as the expression of miR-146a has been more often reported to be increased in site of inflammation, for example in keratinocytes and in the skin of patients with atopic dermatitis and psoriasis [25, 26]. Using in situ hybridization, we observed that miR-146a is expressed in the lung epithelium and in bronchial, bronchiolar, and alveolar epithelium with no difference in the relative signal intensity of miR-146a between asthmatic patients and controls as analyzed by densitometry analysis. We propose that ISH using only a few samples is not sufficiently quantitative approach, however it is a suitable method to determine the expression location of miR-146a in lung epithelium. Interestingly, although miR-146a was downregulated in all asthma phenotypes, the linear regression analysis revealed a negative association between miR-146a levels in bronchial brushing samples and the number of neutrophils in BALF fluid, which prompted us to further address the question about the possible influence of miR-146a on the development of neutrophilic asthma phenotype.
To assess the relevance of the finding of the lower miR-146a expression in airway epithelial cells, we also tested the expression of selected protein coding genes. Increased expression of IL-8, CXCL1, IL-4R and IRF1 in patients with neutrophilic asthma and enhanced IL-33 in samples from eosinophilic patients confirmed phenotypic and endotypic similarity of these subgroups. As previous studies have shown miR-146a capacity to suppress IL-8 and CXCL1 in other cell types [28, 42], our results suggest that the lower expression of miR-146a may be linked to enhanced production of these chemokines by the airway epithelial cells in asthma. Although there was a negative association between the miR-146a levels and the number of neutrophils in BALF fluid, additional linear regression analysis did not reveal any association between relative miR-146a expression and mRNA levels of studied indirect and direct genes (data not shown). Also, we did not observe significant changes in IRAK1 levels in the samples from the patients with asthma regardless of the phenotype. We propose that miR-146a may act partially independently from IRAK1 in primary bronchial epithelial cells of patients with asthma, as previously demonstrated in primary human airway smooth muscle cells  and human alveolar A549 epithelial cell line . Further studies would be needed to confirm this hypothesis and to delineate additional miR-146a direct targets in HBECs. Taken together, these data indicate that miR-146a may act through more complex network of genes not fully described by the current study.
To better understand the expression regulation of miR-146a, we also analyzed its level in HBECs stimulated with inflammatory cytokines. Similarly to keratinocytes , the expression of miR-146a was induced by TNF-α or IL-17A and additionally with IL-4 in HBECs. Even higher level of miR-146a was detected following treatment with combination of TNF-α and IL-17A, suggesting that these cytokines have a synergistic effect as in airway smooth muscle cells . In line with previous publications [46, 47, 48], in our current study, HBECs from healthy donors expressed high amounts of IL-8 and CXCL1 in response to stimulation with TNF-α and IL-17A, as well as was the expression of interferon-regulated genes IFITM1 and IRF1 increased in response to IFN-γ. However, none of the used cytokines suppressed the expression of miR-146a leaving the question about downregulation mechanism in the patient samples unanswered. We propose that results from cell culture experiments using high concentrations of cytokines rather mimic processes of an acute inflammation, whereas patients with asthma had suffered from the disease on average for 15 years and therefore many changes in cells from bronchial brushing samples may represent characteristics of chronic inflammation. Furthermore, it is possible that the expression of miR-146a in HBECs of patients with asthma is decreased because of specific gene expression regulators, epigenetic and/or genetic differences. More specifically, microRNA expression relies on correct and functional microRNA biogenesis machinery, which can be disturbed in case of disease. For example, in lung cancer patients, a reduced expression of Dicer has been shown, which potentially might be responsible for decreased levels of different microRNAs in cancer cells . Secondly, microRNA expression could be dysregulated due to epigenetic changes occuring during development of asthma as it is demonstrated for airway smooth muscle cells from asthmatic subjects . Third possible explanation for downregulated miRNA expression could be genetic alterations. For example, in several types of lymphomas, the cause for dysregulated miR-15a and miR-16-1 is a deletion in 13q14 . Interestingly, it has also been shown that in airways of healthy individuals, genes regulating cell cycle and immune system are less expressed with age and among them miR-146 family expression was significantly decreased, further indicating that miR-146a is one of the tightly regulated genes in the airways .
Several studies have demonstrated that miRNAs, such as miR-155 [7, 53, 54], miR-21 , miR-181 , and miR-223 , are involved in the regulation of airway inflammation. Although anti-inflammatory function of miR-146a has been demonstrated in various cell types, including bronchial epithelial cell lines [28, 30, 31], the function of miR-146a in primary HBECs had not been studied before. Our study shows that the expression of IL-8 and CXCL1 were decreased both on mRNA and protein level in HBECs transfected with miR-146a. We also demonstrate a strong reduction of IRAK1 and a moderate upregulation of IRF1 and IFITM1 mRNA expression in miR-146a-transfected HBECs. It should be noted that both the suppression of IL-8 and CXCL1 and increase of IRF1 and IFITM1 are most probably indirect effects of miR-146a. As no influence of miR-146a transfection on the expression of IL-33 and IL-4R was detected (data not shown), our results suggest that miR-146a does not influence Th2-type inflammatory responses in HBECs. However, our results revealed that miR-146a expression affects HBEGF expression in the presence of IFN-γ, TNF-α, IL-17A or IL-4 and FGF2 in the presence of IL-17A. Previous publications have shown upregulated expression of HBEGF in asthmatic tissues and its involvement in airway remodeling [58, 59, 60], indicating that reduced expression of miR-146a in airways of asthmatic patients may also be associated with increased airway remodeling, however, further investigation would be necessary to confirm this hypothesis.
Previous publications have demonstrated that HBECs express high amounts of IL-8 and CXCL1 in response to TNF-α and IL-17A in order to recruit neutrophils into the airways [46, 47, 48, 57]. Several other studies have shown that insufficient regulation of NF-κB signaling could be the origin of abnormally high expressions of IL-8 [61, 62] and CXCL1 , which in turn may be the potential cause of neutrophilic infiltration in the lower airways of asthma patients. As we observed a negative association between the relative miR-146a expression and the number of neutrophils present in the airways and reduced neutrophil migration towards the supernatants of HBECs transfected with miR-146a and stimulated with TNF-α or IL-17A, our data suggest that miR-146a is capable of inhibiting the production of neutrophil chemoattractants also in the airways of patients with asthma. These results together suggest that there is a strong link between reduced expression of miR-146a in airway epithelial cells in asthmatic patients and the development of neutrophilic phenotype of asthma.
In addition to data indicating that reduced expression of miR-146a may be linked to the recruitment of neutrophils in asthmatic airways, our study reveals other interesting asthma-related findings. In particular, we observed a tendency toward increased expression of IL-4R in airway epithelial cells from patients with neutrophilic asthma. As IL-4 binding with its receptor IL-4R mediates signaling during allergic inflammation and could rather be associated with airway eosinophilia, this result conveys certain controversies . Nevertheless, Lavoie-Lamoureux et al. have demonstrated that IL-4 may also activate neutrophils by increasing the secretion of IL-8 and TNF-α and thereby providing a link between allergic, eosinophilic non-allergic and neutrophilic inflammation .
In summary, we show that miR-146a is expressed in the bronchial epithelium of the airways, whereas airway epithelial cells from patients with asthma express less miR-146a as compared to controls, whereas relative levels of miR-146a were detected to be in negative association with the number of neutrophils in BALF fluid. The cause of the reduced miR-146a expression in asthmatic airways is not known, however, it may lead to the increased neutrophil-attracting chemokines IL-8 and CXCL1 and thereby contribute to the development of neutrophilic phenotype of asthma. Our results provide a deeper understanding of inflammatory processes in asthma, and in long-term perspective, they might contribute to new potential options for therapeutic interventions.
The current study indicates that reduced level of miR-146a in the airway epithelial cells of the patients with asthma may lead to the development of neutrophilic phenotype of asthma. miR-146a may have potential as a novel therapeutic molecule in the modulation of immune responses in asthma.
AK performed the experiments, analyzed the data, contributed with design of the study and wrote the manuscript. JS, KS, MS, AAl, BJ, GB contributed with design of the study, were responsible for the clinical part of the study, and participated in writing the manuscript. AAab contributed with design of the study and data analysis. AR designed the study, supervised the project and writing the manuscript. All authors have read and approved the final manuscript.
This work was supported by personal research Grant PUT1669 from the Estonian Research Council, European Union through the European Regional Development Fund (Project No. 2014-2020.4.01.15-0012; Centre of Excellence for Genomics and Translational Medicine) and by Polish National Science Center Grants: UMO-2014/15/B/NZ5/01539 and UMO-2014/13/B/NZ3/02393.
Ethics approval and consent to participate
This study was approved by the Ethics Committee of the Jagiellonian University Medical College and by the Research Ethics Committee of the University of Tartu. Each participant provided written informed consent.
Consent for publication
The authors declare that they have no competing interests.
- 26.Rebane A, Runnel T, Aab A, Maslovskaja J, Ruckert B, Zimmermann M, et al. MicroRNA-146a alleviates chronic skin inflammation in atopic dermatitis through suppression of innate immune responses in keratinocytes. Allergy. 2014;69:100.Google Scholar
- 30.Tsai MJ, Tsai YC, Chang WA, Lin YS, Tsai PH, Sheu CC, et al. Deducting MicroRNA-mediated changes common in bronchial epithelial cells of asthma and chronic obstructive pulmonary disease-a next-generation sequencing-guided bioinformatic approach. Int J Mol Sci. 2019;20(3):E553.PubMedCrossRefPubMedCentralGoogle Scholar
- 42.Urgard E, Lorents A, Klaas M, Padari K, Viil J, Runnel T, et al. Pre-administration of PepFect6-microRNA-146a nanocomplexes inhibits inflammatory responses in keratinocytes and in a mouse model of irritant contact dermatitis. J Control Release. 2016;235:195–204.PubMedCrossRefPubMedCentralGoogle Scholar
- 47.Akdis M, Burgler S, Crameri R, Eiwegger T, Fujita H, Gomez E, et al. Reviews and feature article: interleukins, from 1 to 37, and interferon-γ: Receptors, functions, and roles in diseases. J Allergy Clin Immunol. 2011;127(701–21):e70.Google Scholar
- 55.Kim RY, Horvat JC, Pinkerton JW, Starkey MR, Essilfie AT, Mayall JR, et al. MicroRNA-21 drives severe, steroid-insensitive experimental asthma by amplifying phosphoinositide 3-kinase–mediated suppression of histone deacetylase 2. J Allergy Clin Immunol. 2017;139(2):519–32.PubMedCrossRefGoogle Scholar
- 62.Mori N, Oishi K, Sar B, Mukaida N, Nagatake T, Matsushima K, et al. Essential role of transcription factor nuclear factor-kappa B in regulation of interleukin-8 gene expression by nitrite reductase from Pseudomonas aeruginosa in respiratory epithelial cells (Retracted article. See vol. 79, pg. 3473, 2011). Infect Immun. 1999;67(8):3872–8.PubMedPubMedCentralGoogle Scholar
- 63.Burke SJ, Lu DH, Sparer TE, Masi T, Goff MR, Karlstad MD, et al. NF-kappa B and STAT1 control CXCL1 and CXCL2 gene transcription. Am J Physiol-Endoc M. 2014;306(2):E131–49.Google Scholar
- 65.Lavoie-Lamoureux A, Moran K, Beauchamp G, Mauel S, Steinbach F, Lefebvre-Lavoie J, et al. IL-4 activates equine neutrophils and induces a mixed inflammatory cytokine expression profile with enhanced neutrophil chemotactic mediator release ex vivo. Am J Physiol-Lung C. 2010;299(4):L472–82.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.