Scholarly article on topic 'The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis'

The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis Academic research paper on "Biological sciences"

CC BY
0
0
Share paper
OECD Field of science
Keywords
{MicroRNA-155 / "lung fibrosis" / "liver X receptor" / fibroblasts / "alternatively activated macrophages"}

Abstract of research paper on Biological sciences, author of scientific article — Mariola Kurowska-Stolarska, Manhl K. Hasoo, David J. Welsh, Lynn Stewart, Donna McIntyre, et al.

Background Idiopathic pulmonary fibrosis (IPF) is progressive and rapidly fatal. Improved understanding of pathogenesis is required to prosper novel therapeutics. Epigenetic changes contribute to IPF; therefore, microRNAs may reveal novel pathogenic pathways. Objectives We sought to determine the regulatory role of microRNA (miR)-155 in the profibrotic function of murine lung macrophages and fibroblasts, IPF lung fibroblasts, and its contribution to experimental pulmonary fibrosis. Methods Bleomycin-induced lung fibrosis in wild-type and miR-155−/− mice was analyzed by histology, collagen, and profibrotic gene expression. Mechanisms were identified by in silico and molecular approaches and validated in mouse lung fibroblasts and macrophages, and in IPF lung fibroblasts, using loss-and-gain of function assays, and in vivo using specific inhibitors. Results miR-155−/− mice developed exacerbated lung fibrosis, increased collagen deposition, collagen 1 and 3 mRNA expression, TGF-β production, and activation of alternatively activated macrophages, contributed by deregulation of the miR-155 target gene the liver X receptor (LXR)α in lung fibroblasts and macrophages. Inhibition of LXRα in experimental lung fibrosis and in IPF lung fibroblasts reduced the exacerbated fibrotic response. Similarly, enforced expression of miR-155 reduced the profibrotic phenotype of IPF and miR-155−/− fibroblasts. Conclusions We describe herein a molecular pathway comprising miR-155 and its epigenetic LXRα target that when deregulated enables pathogenic pulmonary fibrosis. Manipulation of the miR-155/LXR pathway may have therapeutic potential for IPF.

Academic research paper on topic "The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis"

Accepted Manuscript

The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis

Mariola Kurowska-Stolarska, PhD, Manhl K. Hasoo, PhD, David J. Welsh, PhD, Lynn Stewart, BSc, Donna McIntyre, PhD, Brian E. Morton, BSc, Steven Johnstone, PhD, Ashley M. Miller, PhD, Darren L. Asquith, PhD, Neal L. Millar, MD, PhD, Ann B. Millar, MD, Carol A. Feghali-Bostwick, PhD, Nikhil Hirani, MBChB, PhD, Peter J. Crick, PhD, Yuqin Wang, PhD, William J. Griffiths, PhD, Iain B. McInnes, MD, PhD, Charles McSharry, PhD

TBI JOL'RNAI, OF

Allergy Clinical

PII: S0091-6749(16)31132-0

DOI: 10.1016/j.jaci.2016.09.021

Reference: YMAI 12396

To appear in: Journal of Allergy and Clinical Immunology

Received Date: 17 December 2015 Revised Date: 10 August 2016 Accepted Date: 6 September 2016

Please cite this article as: Kurowska-Stolarska M, Hasoo MK, Welsh DJ, Stewart L, McIntyre D, Morton BE, Johnstone S, Miller AM, Asquith DL, Millar NL, Millar AB, Feghali-Bostwick CA, Hirani N, Crick PJ, Wang Y, Griffiths WJ, McInnes IB, McSharry C, The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis, Journal of Allergy and Clinical Immunology (2016), doi: 10.1016/ j.jaci.2016.09.021.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Normal lung miR155-/- and IPF lung

Bleomycin

LXR = Liver X Receptor

The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis

Mariola Kurowska-Stolarska, PhD,a* Manhl K. Hasoo, PhD,a David J. Welsh, PhD,b Lynn Stewart, BSc,a Donna McIntyre, PhD,a Brian E. Morton, BSc,a Steven Johnstone, PhD,a Ashley M. Miller, PhD,c Darren L. Asquith, PhD,a Neal L. Millar, MD, PhD,a Ann B. Millar, MD,d Carol A. Feghali-Bostwick, PhD,e Nikhil Hirani, MBChB, PhD,f Peter J. Crick, PhD,g Yuqin Wang, PhD,g William J. Griffiths, PhD,g Iain B. Mclnnes, MD, PhD,a Charles McSharry, PhDa, h*

a Institute of Infection, Immunity and Inflammation, University of Glasgow, Glasgow G12 8TA, United Kingdom. b Scottish Pulmonary Vascular Unit, University of Glasgow, Glasgow G11 6NT, United Kingdom. c Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow G12 8TA, United Kingdom. d Academic Respiratory Unit, Second Floor, Learning and Research, University of Bristol, Bristol BS10 5NB, United Kingdom. e Division of Rheumatology & Immunology, Medical University of South Carolina, Charleston, USA. f The University of Edinburgh/MRC Centre for Inflammation Research, The Queen's Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4TJ, United Kingdom. g College of Medicine, Grove Building, Swansea University, Singleton Park, Swansea SA2 8PP, United Kingdom. h Greater Glasgow and Clyde Clinical Research and Development, Yorkhill Hospital, Glasgow G3 8SG, United Kingdom.

*Corresponding authors: Dr Mariola Kurowska-Stolarska, Institute of Infection, Immunity and Inflammation, University of Glasgow, 120 University Place, G12 8TA, Glasgow, UK; phone: +44 141 330 6085; fax: +44 141 330 4297; email Mariola.Kurowska-Stolarska@glasgow.ac.uk or Dr Charles McSharry, Institute of Infection, Immunity and Inflammation, University of Glasgow, 120 University Place, G12 8TA, Glasgow, UK; phone: +44 141 330 2282; fax: +44 141 4297; email: Charles.Mc Sharry@glasgow .ac.uk.

Funding: M K S. was supported by Arthritis Research UK (#19213); A.M.M. was supported by British Heart Foundation FS/08/035/25309. M.K.H. was supported by the Ministry of Higher Education and Scientific Research, Republic of Iraq. B.E.M. was supported by a joint grant (#657-2012) from Medical Research Scotland and Lamellar Biomedical Ltd. Bellshill. U.K. C.A.F-B was supported by NIH/NIAMS P30 AR061271 and K24 AR060297. S.J. and L.J. were supported by British Pigeon Fanciers' Medical Research Trust. Oxysterol measurement was supported by Biotechnology & Biological Sciences Research Council (BB/I001735/1, BB/L001942/1).

Author contributions: Conception and design: MKS., D.J.W., IB M., and C M S. Contribution of materials and technology: D.L.A., A.B.M., C.A.F-B., N.H., P.J.C., Y.W., and W.J.G. Data abstraction and statistics: M.K.S., M.K.H., A.B.M., W.J.G., C.A.F.B., P.J.C., Y.W., B.E.M., S. J., I B M., and C.M.S. Drafting the manuscript for intellectual content: MKS., M.K.H., D.J.W., W.J.G., LS., ER., D.M., A.M.M., D.L.A., N.M., N.H., W.J.F., I.B.M., and C.M.S. Manuscript revision and final approval: all authors. Competing interests: The authors declare no competing interests.

Abbreviation used

LXRa: Liver X receptor alpha

miR-155: microRNA-155

IPF: Idiopathic Pulmonary Fibrosis

TGFß/TGFßR2: Transforming growth factor beta/receptor II

3'UTR: 3-prime untranslated region of mRNA

22(S)HC: 22(S)-hydroxycholesterol

Arg2: Arginase 2

Ym1: Chitinase

IL13Ra2: Interleukin-13 receptor alpha 2

Col1a1/2: Collagen 1a1/2

Col3a1: Collagen 3a1

MFI: median fluorescence intensity

58 Abstract

59 Background: Idiopathic Pulmonary Fibrosis (IPF) is progressive and rapidly fatal.

60 Improved understanding of pathogenesis is required to prosper novel therapeutics.

61 Epigenetic changes contribute to IPF therefore microRNAs may reveal novel pathogenic

62 pathways.

63 Objectives: To determine the regulatory role of microRNA(miR)-155 in the pro-fibrotic

64 function of murine lung macrophages and fibroblasts, IPF lung fibroblasts and its

65 contribution to experimental pulmonary fibrosis.

66 Methods: Bleomycin-induced lung fibrosis in wild-type and miR-155-- mice was analyzed

67 by histology, collagen and pro-fibrotic gene expression. Mechanisms were identified by in

68 silico and molecular approaches; validated in mouse lung fibroblasts and macrophages, and

69 in IPF lung fibroblasts, using loss-and-gain of function assays, and in vivo using specific

70 inhibitors.

71 Results: miR-155"" mice developed exacerbated lung fibrosis, increased collagen deposition,

72 collagen 1 and 3 mRNA expression, TGFp production, and activation of alternatively-

73 activated macrophages, contributed by deregulation of the microRNA-155 target gene the

74 liver X receptor (LXR)a in lung fibroblasts and macrophages. Inhibition of LXRa in

75 experimental lung fibrosis and in IPF lung fibroblasts reduced the exacerbated fibrotic

76 response. Similarly, enforced expression of miR-155 reduced the pro-fibrotic phenotype of

77 IPF and miR-155-- fibroblasts.

78 Conclusion: We describe herein a molecular pathway comprising miR-155 and its epigenetic

79 LXRa target that when deregulated enables pathogenic pulmonary fibrosis. Manipulation of

80 the miR-155/LXR pathway may have therapeutic potential for IPF.

82 Key words: microRNA-155, lung fibrosis, Liver XReceptor, fibroblasts, alternatively

83 activated macrophages.

86 Key messages

88 ■ MicroRNA(miR)-155 deficiency exacerbates bleomycin-induced experimental

89 pulmonary fibrosis.

90 ■ In the absence of miR-155 epigenetic control, LXRa activity is deregulated in mouse

91 primary lung fibroblasts facilitating increased collagen and TGFp production, and in

92 macrophages enhancing alternative activation; each inhibited by LXR antagonism,

93 LXRa gene silencing or exogenous miR-155 mimic.

94 ■ The exacerbated bleomycin-induced pulmonary fibrosis in miR-155"" mice was

95 mitigated in vivo by LXR antagonism.

96 ■ Primary IPF lung fibroblasts had constitutively raised LXRa, deregulated from miR-

97 155, and their pro-fibrotic phenotype was inhibited by LXR antagonism, LXRa gene

98 silencing or exogenous miR-155 mimic.

100 Capsule Summary

101 Idiopathic pulmonary fibrosis (IPF) is a devastating disease with increasing incidence.

102 Therapeutic options are limited by poor understanding of underlying pathogenesis. We

103 propose a regulatory role for miR-155/LXRa in aberrant lung remodeling that may offer a

104 therapeutic target pathway.

106 Introduction

107 Idiopathic Pulmonary Fibrosis (IPF) affects more than 5 million people worldwide and the

108 incidence is increasing 1. Histology of IPF includes interstitial fibroblastic foci and

109 deposition of collagen-rich extracellular matrix , and pirfenidone targeting tissue remodeling

110 has improved therapeutic options . However, mechanisms controlling IPF progression

111 remain poorly understood. IPF is associated with age, male gender and cigarette smoking 4,

112 suggesting an epigenetic contribution to pathogenesis.

113 MicroRNAs (miRs) are 22-nucleotide non-coding RNAs that regulate gene expression 5.

114 Single miRs bind 6-8 nucleotide complementary sequences, mainly in the 3'UTR of target

115 mRNAs causing degradation or translation inhibition 6 and can fine-tune diverse mRNA

116 often within the same biological pathway . Identifying disease-specific miRs can reveal

117 novel target-mRNA/pathways and provide insight into pathogenesis and identify therapeutic

118 targets.

119 MiR-155 is required for normal immune function , ; its over-expression is associated with

120 inflammation, autoimmunity 8 10 and cancer 11, whereas miR-155-deficient mice develop age-

121 related airway fibrosis . MiR-155 may therefore act as a homeostatic rheostat contributing

122 to the onset and duration of inflammation and remodeling. Our hypothesis was that miR-155

123 attenuates pathways that induce lung remodeling. We revealed exacerbated experimental

124 fibrosis in miR-155"" mice upon lung injury. A novel miR-155 regulated pathway identified

125 in this context was the liver X receptor alpha (LXR)a, which is an oxysterol-activated

126 transcription-factor (NR1H3) 13.

128 METHODS

130 Bleomycin-induced lung fibrosis was induced in miR-155-- and control mice as described 14

131 15. Mouse lung fibroblasts and macrophages were derived from WT and miR-155-- mice by

132 lung digestion followed by FACS sorting. Primary lung fibroblasts from IPF patients (n=7)

133 and normal controls (n=8) were obtained and cultured as described 16. Experimental

134 interventions included transfecting cells with miR-155 mimic or LXRa siRNA, or incubated

135 with LXR agonist/antagonist or various alarmins. Comprehensive details are provided in the

136 Methods section in this article's online repository.

RESULTS

Experimental pulmonary fibrosis is exacerbated by miR-155 deficiency

To evaluate miR-155 epigenetic control of lung fibrosis, we used the murine model of

bleomycin-induced inflammation and pulmonary fibrosis . Bleomycin or control PBS was given to miR-155 gene-deleted (miR-155--) mice and wild-type (WT) controls. Bleomycin-induced weight loss (Fig 1, A), lung collagen deposition (Fig 1, B), and biomarkers of inflammation (Table E1) were exacerbated in miR-155- - mice compared with WT mice on day 18. This was accompanied by increased lung tissue expression of mRNA for collagen (Col)1a (mainly Col1a1 isoform) and Col3a1 (Fig 1, C and Fig E1), TGFP (Tf) expression and lung collagen protein (Table E1). The increased bronchoalveolar lavage (BAL) cell counts in bleomycin-treated miR-155- - mice (Table E1) were predominantly macrophages with the repair-associated, alternatively-activated (M2) phenotype (Fig 1, D) confirmed by increased arginase 2 (Arg2), chitinase (Ym1) and IL-13 receptor a2 (IL13ra2) expression, whereas the expression of the classically-activated macrophage (M1) phenotype marker, inducible nitric oxide synthase (Nos2) remained unchanged. Together, these data demonstrate that miR-155 deficiency exacerbated the pulmonary fibrotic response to bleomycin.

We next investigated the kinetics of lung tissue miR-155 expression in wild-type (WT) mice given bleomycin (Fig 1, E). MiR-155 expression in mice given PBS remained constant, whereas in response to bleomycin, miR-155 expression decreased at day 1, increased at day 7 and normalized by day 18. To investigate the factors that might regulate these changes, we established that bleomycin incubated in vitro with WT murine primary lung fibroblasts was sufficient to dose-dependently down-regulate expression of precursor miR-155 at 8h (Fig 1, F) and mature miR-155 at 24h (Fig E2, A). To mimic the effect of exposure to cytokines

17, 18

generated in the damaged lung , , miR-155 expression was measured in WT murine

17 19 18

primary lung fibroblasts incubated with exogenous alarmins IL-33 , IL-25 , IL-1a or

HMGB-1 released in response to injury. There was no change in response to IL-33, IL-25 or HMGB-1 (Fig E2, B), but IL-1a increased miR-155 expression (Fig E2, C). Thus the dynamic expression of miR-155 in vivo may reflect a homeostatic role in inflammation and repair in response to tissue injury.

Prediction analysis identified LXRa as a miR-155 target in the lung

Identifying mRNA targets under the epigenetic control of miR-155 was our strategy to identify cryptic pathways involved in lung fibrosis. We performed in silico analysis of predicted and validated conserved mouse and human miR-155 targets (TargetScan and miRTarBase) along with targets expressed in lungs or described in respiratory or fibrotic

diseases (Ingenuity Pathway Analysis database). This integrated approach identified target

mRNAs (Table E2), including hypoxia and TGFp pathways , among which we validated increased expression of Hif1a, Tgffir2 and Smad1 mRNA in lung tissue of miR-155" " mice given bleomycin (Fig E3). In addition to these recognized pro-fibrotic pathways we identified the Liver X Receptor-a (LXRa), which has not hitherto been described in lung fibrosis. LXRa has a conserved 3'UTR seed-region sequence (GCAUUAA) complementary to miR-155 therefore we highlighted this as a potential novel pathway to pathogenic fibrosis and this provides the basis of our study.

Endogenous miR-155 targets human LXRa

We recently demonstrated using a reporter assay that synthetic miR-155 could bind mouse

Lxra mRNA . To confirm that endogenous miR-155 targets human LXRa mRNA we used

an MS2-TRAP RNA affinity purification assay . Expression constructs encoding luciferase genes tagged with the MS2 binding domain (MS2BD) motif with either intact LXRa, or LXRa

mutated in the 3'UTR microRNA recognition element (MRE) as a negative control, were transfected into HEK293 cells together with the MS2GFP-expressing plasmid. The empty vector and a construct containing a tandem of 9 miR-155 binding sites (i.e. a miR-155 'sponge') were used as negative and positive controls, respectively. MS2BD-containing transcripts were isolated from transfected cells by immuno-precipitation of GFP and the enrichment of miR-155 in precipitates was measured by qPCR (Fig 2, A). Transcripts containing WT LXRa 3'UTRs showed significant enrichment in miR-155 compared to the mutated sequence, which showed miR-155 levels similar to the empty vector control, thus endogenous miR-155 could bind to human LXRa mRNA. To confirm and extend this observation, we reintroduced miR-155 into miR-155"" murine lung fibroblasts by transfection with a synthetic miR-155 mimic. After 24h culture, cytosolic LXRa protein concentrations (Fig 2, B) were reduced 60% by miR-155. Together these findings support a functional interaction between miR-155 and LXRa mRNA.

LXRa expression and activity are increased in miR-155-/- mice with lung fibrosis

Compared to WT mice given PBS, the expression of Lxra mRNA in lung tissue of WT mice given bleomycin was up-regulated on day 1 and had normalized by day 7 (Fig 2, C). This increase was confirmed at the protein level in lung fibroblasts of WT mice given bleomycin; peaking at day 2 and 3 and normalizing to control PBS levels at day 7 (Fig 2, D-F and Fig E4). This in vivo expression pattern of Lxra was reciprocal to that of miR-155 (Fig 1, E) in WT mice. Consistent with the homeostatic molecular interaction between miR-155 and Lxra mRNA, miR-155- - mice given bleomycin maintained higher levels of lung Lxra expression compared with WT mice (Fig 2, G). This increased expression was associated with an increase in Lxra activity as measured by the expression of its specific functional reporter Abcal in lung tissue mRNA (Fig 2, H). Together these data demonstrate that the

lack of epigenetic homeostatic regulation in miR-155-- mice was associated with a sustained increase in Lxra expression and activity in response to bleomycin.

Serum concentrations of LXR oxysterol ligands are unchanged in experimental fibrosis

Oxidized derivatives of cholesterol; oxysterols e.g. 24(S) hydroxycholesterol and 27-hydroxycholesterol are natural ligands that stimulate the expression and activation of LXRa

24 -/-

. We showed previously that miR-155- - mice have higher serum cholesterol concentrations

while on a high fat diet 22, therefore to test whether different oxysterol concentrations in miR-155-/- mice treated with bleomycin were responsible for the Lxra activation and exacerbated lung fibrosis, we profiled serum oxysterols using mass-spectrometry (Table E3). We found no differences between any of the known LXRa ligands 25, 26 suggesting that the increased activation of the Lxra pathway in miR-155- - mice was due to normal activation of more available LXRa.

MiR-155-/" lung fibroblasts and macrophages have an LXR-dependent pro-fibrotic phenotype

We next investigated the role of LXR pathway activation in primary lung fibroblasts and alveolar macrophages. Compared with WT cells, miR-155- - fibroblasts and macrophages had greater and constitutive expression of the LXRa reporter gene, Abca1 (Fig 2, I) suggesting that the LXRa pathway itself was constitutively activated. In miR-155- - macrophages this was associated with an increased pro-fibrotic (M2) phenotype characterized by increased expression of Arginase 2 (Arg2); a key enzyme controlling the bio-availability of

hydroxyproline for collagen synthesis (Fig 2, J). We demonstrated that this increased Arg2 expression in miR-155- - macrophages was restored to normal by Lxra-siRNA (Fig 2, K and Fig E5, A), and by LXR antagonist 22(S)-hydroxycholesterol; (22(S)HC) 28 (Fig 2, L). To

extend this to human cells, we investigated the regulatory inter-relationship between LXRa and miR-155 in the expression of ARG2 in human macrophages. Healthy human monocyte-derived macrophages were transfected with control siRNA or LXRa siRNA, each with miR-155 inhibitor or control inhibitor (Fig E5, C). To induce LXRa and ARG2 expression, the cells were cultured with LXR agonist GW3965 or control DMSO. The LXR agonist-induced increase in ARG2 expression was further increased by inhibition of miR-155, and this increase was restored to normal by LXRa-specific siRNA (Fig E6). Together these data suggest that LXRa-dependent regulation of ARG2 was governed by miR-155 in human and mouse macrophages.

We next explored whether miR-155 influenced the pro-fibrotic function of fibroblasts in an LXRa-dependent manner. In vitro proliferation, migration and collagen production were compared in primary lines derived from mouse lung tissue. MiR-155-- fibroblasts displayed greater proliferation to serum supplementation than WT fibroblasts (Fig 3, A), which was restored to the normal proliferation observed in WT cells by the LXR antagonist 22(S)HC in a dose-dependent manner (Fig 3, B). MiR-155-- fibroblasts also displayed increased migration compared with WT fibroblasts into the scratch-space of an in vitro wound-healing assay, which was normalized by 22(S)HC (Fig 3, C and D). The increased fibroblast infiltration was not due to proliferation because the culture medium was supplemented with 0.3% FCS, a concentration that did not support fibroblast proliferation (Fig 3, A). MiR-155--fibroblasts produced approximately 40-fold increased concentration of soluble collagen in culture than WT fibroblasts in response to 3% FSC (Fig 3, E), which was normalized in a dose-dependent manner by the 22(S)HC to concentrations produced by WT fibroblasts (Fig 3, F).

TGFp is the principal cytokine driving collagen gene expression, and oxysterol agonists of LXR can induce TGFp production 29, 30. Therefore, to investigate the role of miR-155 in

LXRa-dependent collagen production we quantified TGFp in WT and miR-155- - fibroblast supernatants cultured for 48h in 3% FCS, with/without 22(S)HC. MiR-155-- fibroblasts produced higher concentrations of TGFp than WT fibroblasts and this increase was inhibited either by LXR antagonism (Fig 3, G), or by restoring miR-155 by transfection (Fig 3, H and Fig E5, B). To investigate if arginase was involved in this process, we measured the expression of Arg2 in fibroblasts that were transfected with Lxra siRNA or control siRNA (Fig E5, B). MiR-155- - fibroblasts had higher expression levels of Arg2 than WT fibroblasts (Fig 3, I), and specific inhibition of Lxra by siRNA restored Arg2 expression in miR-155- -fibroblasts to the normal levels of WT fibroblasts. These observations indicate that excessive production of soluble collagen by miR-155- - fibroblasts may be due to an LXR-dependent increase in TGFp and increased arginase driven production of hydroxyproline.

The exacerbated bleomycin-induced lung fibrosis in miR-155-/- mice is LXR-dependent

To test the involvement of LXR in experimental lung fibrosis, miR-155- - and WT mice were given bleomycin or control PBS, and treated with the LXR antagonist 22(S)HC or control cyclodextrin excipient. The subsequent loss of body weights for miR-155- - and WT mice are shown on different panels for clarity in Fig 4, A. The exacerbated bleomycin-induced weight loss in miR-155-- mice was mitigated by treatment with 22(S)HC to the weight loss seen in WT mice given bleomycin, as was the exacerbated lung tissue collagen deposition (Fig 4, B), and the inflammatory BAL cytology (Fig E7). The miR-155---associated increased lung tissue Col1a1, Col3a1 and Arg2, and the BAL cell Arg2 mRNA expression were also attenuated by 22(S)HC (Fig 4, C and D). 22(S)HC had no significant effect on weight-loss in bleomycin-treated WT mice (Fig 4, B). These data demonstrate that the exacerbated inflammatory and fibrotic response to bleomycin in miR-155- - mice is at least partly dependent on LXRa and tractable in vivo by LXR antagonism.

The exacerbated pro-fibrotic behavior of IPF fibroblasts is normalized by neutralization of the LXR pathway

To investigate the contribution of LXR pathway activation to the exacerbated lung tissue-remodeling characteristic of Idiopathic Pulmonary Fibrosis (IPF), we obtained primary lung fibroblast lines from IPF patients and control subjects (details on Table E4 and E5). The constitutive cytosolic LXRa protein concentration was greater in IPF than normal lung fibroblasts (Fig 5, A). IPF lung fibroblasts showed increased collagen synthesis in vitro compared with control lung fibroblasts, which could be either, reduced in a dose-dependent manner by LXR antagonist (Fig 5, B), or further increased by the LXR agonist GW3965 (Fig 5, C). The contribution of LXR activation to the excess collagen production by IPF fibroblasts was further confirmed by transfecting IPF lung fibroblasts with LXRa siRNA (Fig E5, D), which attenuated the collagen production (Fig 5, D).

Control normal and IPF fibroblasts produced TGFp in culture supernatants, however only IPF fibroblasts increased TGFp production in response to 1% FCS supplementation; and this increased production was inhibited by LXR antagonist (Fig 5, E). Normal and IPF fibroblasts constitutively expressed similar levels of ARG2 mRNA, however, only IPF fibroblasts showed higher expression of ARG2 after stimulation with LXR agonist GW3965 (Fig 5, F); and this increased ARG2 expression was attenuated by transfection with LXRa siRNA (Fig 5, G). These data indicate that TGFp and ARG2 are regulated in an LXRa-dependent manner in IPF fibroblasts. In addition, compared with control lung fibroblasts, IPF lung fibroblasts showed greater in vitro proliferation in response to 1% FCS supplementation, which was reduced by LXR antagonist (Fig E8, A). IPF lung fibroblasts had increased migratory capacity into the scratch-space of an in vitro wound-healing assay, compared with normal lung fibroblasts, and this increased migration was reduced to levels of

normal fibroblasts by LXR antagonism (Fig E8, B and C). Thus, activation of the LXR pathway may drive the excessive pro-fibrotic phenotypic characteristics of IPF fibroblasts.

LXRa is deregulated from miR-155 in IPF lung fibroblasts

To test whether the LXRa-dependent collagen production by IPF fibroblasts was regulated by miR-155, control and IPF fibroblasts were transfected with miR-155 and stimulated by synthetic LXR agonist GW3965 in vitro. The increased collagen production by IPF fibroblasts was decreased (Fig E9, A), suggesting that collagen synthesis, as the prime exemplar of the LXRa-dependent pro-fibrotic function of IPF fibroblasts, can be regulated by miR-155.

Since constitutively increased LXRa expression (Fig 5, A) and activity contributed to IPF fibroblast phenotype we investigated if this was caused by altered serum concentrations of LXRa oxysterol ligands in IPF patients or by altered miR-155 expression. Comparing serum oxysterol concentrations in IPF and control subjects showed no differences in any of the LXRa ligands tested (Table E6). The constitutive miR-155 expression in IPF fibroblasts was similar to that of control lung fibroblasts (Fig E9, B) therefore we investigated whether the increased LXRa expression and activation in IPF fibroblasts was due to a deregulated interaction between miR-155 and LXRa. Since the consequence of LXRa deregulation resulting in exacerbated lung fibrosis became apparent in miR-155- - mice only when stressed with bleomycin, we compared the dynamic interaction between miR-155 and LXRa, in control and IPF fibroblasts cultured under the hypoxic stress (1% O2) which mimics the lung

environment in IPF . Compared with normoxia, miR-155 expression was increased by hypoxia in both healthy and IPF fibroblasts (Fig 9, C); however, LXRa and ABCA1 expression was increased by hypoxia only in IPF fibroblasts (Fig 9, D) suggesting selective deregulation of LXRa function. To explore the dynamics of the interaction between miR-155

338 and LXRa, we correlated the ratio of their relative expressions in normal and IPF lung

339 fibroblasts. The relative expression levels of LXRa and miR-155 in normal and IPF lung

340 fibroblasts cultured under normoxic conditions showed no significant correlation

341 (Spearman's rho and 95% C.I.): normal fibroblasts r = 0.263 (-0.310, 0.69) and IPF

342 fibroblasts r = 0.439 (-0.072, 0.767). However, under hypoxic conditions there was a negative

343 correlation in normal fibroblasts r = -0.655 (-0.868, -0.236) which was not apparent in IPF

344 fibroblasts r = -0.152 (-0.602, 0.375), (Fig E9, E). This suggested that there was tight post-

345 transcriptional control of LXRa expression by homeostatic miR-155 in response to a stressor

346 such as hypoxia in normal fibroblasts that was lost in IPF fibroblasts, potentially contributing

347 to the deregulated LXRa activity.

348 The mechanism of this deregulation may be due to increased competitive miR-155 binding

349 by other mRNA targets that contain multiple miR-155 seed-region binding sites . To test

350 this hypothesis we evaluated the expression of a validated miR-155 target ZNF652 that

351 contains 7 miR-155 binding sites (HumanTargetScan v7.0) in normal and IPF fibroblasts

352 cultured in normoxia and hypoxia. ZNF652 was up-regulated by hypoxia in IPF but not in

353 normal lung fibroblasts (Fig E9, F); and in contrast to LXRa, the expression of ZNF652

354 correlated negatively with miR-155 expression (Fig E9, G) suggesting that under hypoxic

355 stress, miR-155 may be preferentially bound by the increased ZNF652 leading to de-

356 repression of LXRa in IPF fibroblasts.

DISCUSSION

Characteristic IPF fibrosis is refractory to anti-inflammatory therapy 4 and anti-fibrotic

drugs underline the primacy of aberrant wound healing to pathogenesis 33. We provide new understanding of this process. Mouse models and IPF lung fibroblasts had constitutively increased LXRa transcription when deregulated from homeostatic miR-155; associated with LXR-dependent excessive fibrotic phenotype mediated by increased TGFp, arginase, and collagen production that could be mitigated by LXR antagonist (Summary Fig 6).

MiR-155 expression is rapidly and transiently reduced in wild-type mice after bleomycin, associated with transient reciprocally increased Lxra expression and protein; and the remodeling is self-limiting 15. In contrast, miR-155- - mice have constitutively increased Lxra and an exacerbated lung fibrosis; and this difference may provide novel insight into mechanisms of relentless lung remodeling. IPF lung fibroblasts also have constitutively more LXRa protein (and up-regulated LXRa and ABCA1: IPF data repositories GSE2052 34,

GEOD-24206 ), and greater LXR-dependent pro-fibrotic activation that was normalized by miR-155 over-expression, LXRa gene-silencing or by metabolic antagonism of LXRa activity using 22(S)HC.

LXR may exert pro-fibrotic effects by inducing Arginase 2 (Arg2) and Tgffi expression. The Arg2 promoter contains an LXR response element and is activated by LXR agonism in macrophages 36, and we extend this finding to mouse and human fibroblasts. Arginase 2 is the mitochondrial form involved in hydroxyproline production and is essential for collagen biosynthesis. Up-regulated Arg2 in miR-155- - macrophages and fibroblasts is normalized by inhibition of Lxra by siRNA, or its activity by metabolic antagonism. LXR may additionally exert pro-fibrotic effects by similarly regulating TGFp expression, and the excessively high concentrations of TGFp produced in vitro by miR-155- - and IPF fibroblasts were normalized by LXR antagonism.

Our pro-fibrotic LXR function in lung conflicts with the anti-fibrotic function of

T0901317-LXR activation in skin during experimental systemic sclerosis model . This can be reconciled; synthetic ligand T0901317 locks LXR into the conformation that recruits co-activators whereas natural oxysterol ligands and GW3965 induce the flexible conformation

which bind co-activators and co-repressors 38, and there are tissue-specific epigenetic changes

in chromatin that determine LXR-driven gene expression . Furthermore the multiple-dose bleomycin-induced skin fibrosis is driven by IL-6 from inflammatory macrophages which are

inhibited by LXR activation 37 whereas, in contrast, our single bolus bleomycin-induced lung

fibrosis is associated with repair M2 macrophage activation (Fig 1D ) which is enhanced by LXR activation 36, 40. Alveolar macrophages are uniquely enriched in genes of lipid metabolism that are cross-regulated by LXR supporting their role in lung homeostasis 41.

The cryptic involvement of LXRa in fibrosis became apparent when deregulated in miR-155-- mice plus the stressor of bleomycin. The mechanism of LXRa deregulation in IPF fibroblasts may be due to ineffective regulation by miR-155, which becomes apparent under

hypoxic stress equivalent to the IPF lung environment . IPF and control lung fibroblasts had similar miR-155 expression when cultured under normal oxygen tensions. Under hypoxic conditions the expression levels of miR-155 correlated negatively with LXRa in control lung fibroblasts, implying tight epigenetic control, whereas there was no equivalent engagement between miR-155 and LXRa in IPF fibroblasts thus enabling continued LXRa auto-activation 42 and pro-fibrotic behavior. This deregulation might be mediated by several mechanisms 39, including competition for available miR-155 by other targets with the AGCAUUAA seed-

region 7 as validated in cancer cells One strong miR-155 candidate target mRNA is ZNF652 which has 7 seed-region binding sites. ZNF652 is induced by hypoxia in IPF but not normal fibroblasts. We identified that in contrast to LXRa, its increased expression negatively correlated with miR-155 in IPF fibroblasts suggesting that ZNF652 mRNA competitively

408 bound miR-155 leading to de-repression of LXRa.

409 MiR-155 expression has been identified as increased 44 or reduced 45, and serum miR-

410 155 levels were normal 46 in IPF. This may reflect the dynamism of miR-155 expression in

411 experimental IPF. In lung tissue it is transiently down-regulated by bleomycin (Fig 1F) and

412 TGFp , and induced by inflammatory mediators e.g IL-1a (Fig E2, C) or hypoxia as a

413 counter-balance mechanism regulating homeostatic lung tissue repair.

414 Fibrosis of the lung is a common co-morbidity of systemic sclerosis (SSc). The

415 pathogenesis and clinical features of the autoimmune and inflammation -driven lung

416 pathology of systemic sclerosis differs from IPF 49 and two recent studies describe a

417 pathogenic role for miR-155 in the experimental skin and lung fibrosis associated with SSc 50'

418 51. This reflects the dual role of miR-155 driving chronic inflammation associated pathologies

419 and resolving fibrosis that we found aberrant in IPF.

References

1. Navaratnam V, Fogarty AW, Glendening R, McKeever T, Hubbard RB. The increasing secondary care burden of idiopathic pulmonary fibrosis: hospital admission trends in England from 1998 to 2010. Chest 2013; 143:1078-84.

2. Wynn TA. Integrating mechanisms of pulmonary fibrosis. The Journal of experimental medicine 2011; 208:1339-50.

3. King TE, Jr., Bradford WZ, Castro-Bernardini S, Fagan EA, Glaspole I, Glassberg MK, et al. A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. N Engl J Med 2014; 370:2083-92.

4. King TE, Jr., Pardo A, Selman M. Idiopathic pulmonary fibrosis. Lancet 2011; 378:1949-61.

5. Ruvkun G. Molecular biology. Glimpses of a tiny RNA world. Science 2001; 294:797-9.

6. Lim LP, Lau NC, Garrett-Engele P, Grimson A, Schelter JM, Castle J, et al. Microarray analysis shows that some microRNAs downregulate large numbers of target mRNAs. Nature 2005; 433:769-73.

7. Pasquinelli AE. MicroRNAs and their targets: recognition, regulation and an emerging reciprocal relationship. Nature reviews. Genetics 2012; 13:271-82.

8. Kurowska-Stolarska M, Alivernini S, Ballantine LE, Asquith DL, Millar NL, Gilchrist DS, et al. MicroRNA-155 as a proinflammatory regulator in clinical and experimental arthritis. Proceedings of the National Academy of Sciences of the United States of America 2011; 108:11193-8.

9. Thai TH, Calado DP, Casola S, Ansel KM, Xiao C, Xue Y, et al. Regulation of the germinal center response by microRNA-155. Science 2007; 316:604-8.

10. O'Connell RM, Kahn D, Gibson WS, Round JL, Scholz RL, Chaudhuri AA, et al. MicroRNA-155 promotes autoimmune inflammation by enhancing inflammatory T cell development. Immunity; 33:607-19.

11. O'Connell RM, Rao DS, Chaudhuri AA, Boldin MP, Taganov KD, Nicoll J, et al. Sustained expression of microRNA-155 in hematopoietic stem cells causes a myeloproliferative disorder. J Exp Med 2008; 205:585-94.

12. Rodriguez A, Vigorito E, Clare S, Warren MV, Couttet P, Soond DR, et al. Requirement of bic/microRNA-155 for normal immune function. Science 2007; 316:608-11.

13. Jakobsson T, Treuter E, Gustafsson JA, Steffensen KR. Liver X receptor biology and pharmacology: new pathways, challenges and opportunities. Trends in pharmacological sciences 2012; 33:394-404.

14. Moeller A, Ask K, Warburton D, Gauldie J, Kolb M. The bleomycin animal model: a useful tool to investigate treatment options for idiopathic pulmonary fibrosis? The international journal of biochemistry & cell biology 2008; 40:362-82.

15. B BM, Lawson WE, Oury TD, Sisson TH, Raghavendran K, Hogaboam CM. Animal models of fibrotic lung disease. American journal of respiratory cell and molecular biology 2013; 49:167-79.

16. Pilewski JM, Liu L, Henry AC, Knauer AV, Feghali-Bostwick CA. Insulin-like growth factor binding proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular matrix deposition. The American journal of pathology 2005; 166:399-407.

17. Li D, Guabiraba R, Besnard AG, Komai-Koma M, Jabir MS, Zhang L, et al. IL-33 promotes ST2-dependent lung fibrosis by the induction of alternatively activated macrophages and innate lymphoid cells in mice. J Allergy Clin Immunol 2014; 134:1422-32 e11.

18. Suwara MI, Green NJ, Borthwick LA, Mann J, Mayer-Barber KD, Barron L, et al. IL-1alpha released from damaged epithelial cells is sufficient and essential to trigger inflammatory responses in human lung fibroblasts. Mucosal Immunol 2014; 7:684-93.

19. Hams E, Armstrong ME, Barlow JL, Saunders SP, Schwartz C, Cooke G, et al. IL-25 and type 2 innate lymphoid cells induce pulmonary fibrosis. Proceedings of the National Academy of Sciences of the United States of America 2014; 111:367-72.

20. Rowe SM, Jackson PL, Liu G, Hardison M, Livraghi A, Solomon GM, et al. Potential role of high-mobility group box 1 in cystic fibrosis airway disease. American journal of respiratory and critical care medicine 2008; 178:822-31.

21. Wynn TA, Ramalingam TR. Mechanisms of fibrosis: therapeutic translation for fibrotic disease. Nat Med 2012; 18:1028-40.

22. Miller AM, Gilchrist DS, Nijjar J, Araldi E, Ramirez CM, Lavery CA, et al. MiR-155 has a protective role in the development of non-alcoholic hepatosteatosis in mice. PloS one 2013; 8:e72324.

23. Yoon JH, Srikantan S, Gorospe M. MS2-TRAP (MS2-tagged RNA affinity purification): tagging RNA to identify associated miRNAs. Methods 2012; 58:81-7.

24. Janowski BA, Willy PJ, Devi TR, Falck JR, Mangelsdorf DJ. An oxysterol signalling pathway mediated by the nuclear receptor LXR alpha. Nature 1996; 383:728-31.

25. Song C, Liao S. Cholestenoic acid is a naturally occurring ligand for liver X receptor alpha. Endocrinology 2000; 141:4180-4.

480 26. Griffiths WJ, Crick PJ, Wang Y, Ogundare M, Tuschl K, Morris AA, et al. Analytical strategies for

481 characterization of oxysterol lipidomes: liver X receptor ligands in plasma. Free radical biology &

482 medicine 2013; 59:69-84.

483 27. Endo M, Oyadomari S, Terasaki Y, Takeya M, Suga M, Mori M, et al. Induction of arginase I and II in

484 bleomycin-induced fibrosis of mouse lung. American journal of physiology. Lung cellular and

485 molecular physiology 2003; 285:L313-21.

486 28. Tranheim Kase E, Nikolic N, Pettersen Hessvik N, Fjeldheim AK, Jensen J, Thoresen GH, et al.

487 Dietary supplementation with 22-S-hydroxycholesterol to rats reduces body weight gain and the

488 accumulation of liver triacylglycerol. Lipids 2012; 47:483-93.

489 29. Schwartz Y, Khoshchenko OM, Dushkin MI, Feofanova NA. Effects of cholesterol and nuclear

490 hormone receptor agonists on the production of transforming growth factor-beta in macrophages. Bull

491 Exp Biol Med 2009; 148:406-9.

492 30. Kikuchi T, Sugiura H, Koarai A, Ichikawa T, Minakata Y, Matsunaga K, et al. Increase of 27493 hydroxycholesterol in the airways of patients with COPD: possible role of 27-hydroxycholesterol in

494 tissue fibrosis. Chest 2012; 142:329-37.

495 31. Tzouvelekis A, Harokopos V, Paparountas T, Oikonomou N, Chatziioannou A, Vilaras G, et al.

496 Comparative expression profiling in pulmonary fibrosis suggests a role of hypoxia-inducible factor-

497 1alpha in disease pathogenesis. American journal of respiratory and critical care medicine 2007;

498 176:1108-19.

499 32. Neilsen PM, Noll JE, Mattiske S, Bracken CP, Gregory PA, Schulz RB, et al. Mutant p53 drives

500 invasion in breast tumors through up-regulation of miR-155. Oncogene 2013; 32:2992-3000.

501 33. King TE, Jr., Bradford WZ, Castro-Bernardini S, Fagan EA, Glaspole I, Glassberg MK, et al. A phase

502 3 trial of pirfenidone in patients with idiopathic pulmonary fibrosis. The New England journal of

503 medicine 2014; 370:2083-92.

504 34. Wang XM, Zhang Y, Kim HP, Zhou Z, Feghali-Bostwick CA, Liu F, et al. Caveolin-1: a critical

505 regulator of lung fibrosis in idiopathic pulmonary fibrosis. The Journal of experimental medicine 2006;

506 203:2895-906.

507 35. Meltzer EB, Barry WT, D'Amico TA, Davis RD, Lin SS, Onaitis MW, et al. Bayesian probit regression

508 model for the diagnosis of pulmonary fibrosis: proof-of-principle. BMC medical genomics 2011; 4:70.

509 36. Marathe C, Bradley MN, Hong C, Lopez F, Ruiz de Galarreta CM, Tontonoz P, et al. The arginase II

510 gene is an anti-inflammatory target of liver X receptor in macrophages. The Journal of biological

511 chemistry 2006; 281:32197-206.

512 37. Beyer C, Huang J, Beer J, Zhang Y, Palumbo-Zerr K, Zerr P, et al. Activation of liver X receptors

513 inhibits experimental fibrosis by interfering with interleukin-6 release from macrophages. Ann Rheum

514 Dis 2015; 74:1317-24.

515 38. Albers M, Blume B, Schlueter T, Wright MB, Kober I, Kremoser C, et al. A novel principle for partial

516 agonism of liver X receptor ligands. Competitive recruitment of activators and repressors. The Journal

517 of biological chemistry 2006; 281:4920-30.

518 39. Steffensen KR, Jakobsson T, Gustafsson JA. Targeting liver X receptors in inflammation. Expert

519 opinion on therapeutic targets 2013; 17:977-90.

520 40. Hasty AH, Yvan-Charvet L. Liver X receptor alpha-dependent iron handling in M2 macrophages: The

521 missing link between cholesterol and intraplaque hemorrhage? Circ Res 2013; 113:1182-5.

522 41. Gautier EL, Shay T, Miller J, Greter M, Jakubzick C, Ivanov S, et al. Gene-expression profiles and

523 transcriptional regulatory pathways that underlie the identity and diversity of mouse tissue

524 macrophages. Nat Immunol 2012; 13:1118-28.

525 42. Laffitte BA, Joseph SB, Walczak R, Pei L, Wilpitz DC, Collins JL, et al. Autoregulation of the human

526 liver X receptor alpha promoter. Mol Cell Biol 2001; 21:7558-68.

527 43. Kumar MS, Armenteros-Monterroso E, East P, Chakravorty P, Matthews N, Winslow MM, et al.

528 HMGA2 functions as a competing endogenous RNA to promote lung cancer progression. Nature 2014;

529 505:212-7.

530 44. Pandit KV, Corcoran D, Yousef H, Yarlagadda M, Tzouvelekis A, Gibson KF, et al. Inhibition and role

531 of let-7d in idiopathic pulmonary fibrosis. American journal of respiratory and critical care medicine

532 2010; 182:220-9.

533 45. Cho JH, Gelinas R, Wang K, Etheridge A, Piper MG, Batte K, et al. Systems biology of interstitial

534 lung diseases: integration of mRNA and microRNA expression changes. BMC medical genomics

535 2011; 4:8.

536 46. Li P, Zhao GQ, Chen TF, Chang JX, Wang HQ, Chen SS, et al. Serum miR-21 and miR-155

537 expression in idiopathic pulmonary fibrosis. J Asthma 2013; 50:960-4.

538 47. Pottier N, Maurin T, Chevalier B, Puissegur MP, Lebrigand K, Robbe-Sermesant K, et al.

539 Identification of keratinocyte growth factor as a target of microRNA-155 in lung fibroblasts:

540 implication in epithelial-mesenchymal interactions. PLoS One 2009; 4:e6718.

541 48. O'Connell RM, Rao DS, Chaudhuri AA, Baltimore D. Physiological and pathological roles for

542 microRNAs in the immune system. Nat Rev Immunol; 10:111-22.

543 49. Herzog EL, Mathur A, Tager AM, Feghali-Bostwick C, Schneider F, Varga J. Review: interstitial lung

544 disease associated with systemic sclerosis and idiopathic pulmonary fibrosis: how similar and distinct?

545 Arthritis Rheumatol 2014; 66:1967-78.

546 50. Christmann RB, Wooten A, Sampaio-Barros P, Borges CL, Carvalho CR, Kairalla RA, et al. miR-155

547 in the progression of lung fibrosis in systemic sclerosis. Arthritis Res Ther 2016; 18:155.

548 51. Yan Q, Chen J, Li W, Bao C, Fu Q. Targeting miR-155 to Treat Experimental Scleroderma. Sci Rep

549 2016; 6:20314.

550 52. Blackwell TS, Tager AM, Borok Z, Moore BB, Schwartz DA, Anstrom KJ, et al. Future directions in

551 idiopathic pulmonary fibrosis research. An NHLBI workshop report. American journal of respiratory

552 and critical care medicine 2014; 189:214-22.

554 Acknowledgements: We thank Dr. Tranheim-Kase (University of Oslo, Norway) for advice

555 on 22(S)HC, Ms. Lisa Jolly for cytokine assays, Dr. Li for providing some mouse lung tissue,

556 Dr. J. Aponte for oxysterol analysis, Dr. D. Gilchrist for LXRa-3'UTR-MS2 assays and Drs

557 Macleod and Hansell for protocols for lung digestion and FACS-gating.

558 Figure Legends

559 FIG 1. MiR-155 deficiency exacerbates experimental model of pulmonary fibrosis. (A)

560 Bleomycin (one dose-intranasal) exacerbated a decrease in body weight and (B) lung

561 collagen deposition (turquoise staining) in miR-155-- mice (n=8). (C) miR-155-- bleo mice

562 show an increase in lung Collal and Col3a1 mRNA; and (D) M2 macrophage polarization.

563 (E) miR-155 is dynamically regulated by bleomycin in WT. (F) Precursor (pre-)miR-155 is

564 decreased in lung fibroblasts of WT mice (pooled n=5) cultured with bleomycin. Data shown

565 as mean±SEM or median and inter-quartile range, *P<0.05.

567 FIG 2. LXRa is regulated by miR-155. (A) miR-155 binds to human LXRa. HEK293 were

568 transfected with either empty vector (pmiRGLO-MS2BD) or miR-155 sponge (pmiRGLO-

569 MS2BD-miR155Sp or 3'UTR-LXRa (pmiRGLO-MS2BD-LXRa WT) or MS2 mutated in

570 MRE 3'UTR-LXRa (pmiRGLO-MS2BD-Lxra-MT), and miR-155 captured in the immuno-

571 precipitate quantified by qPCR. Data presented as mean±SEM of 2 technical replicates;

572 representative of three experiments. (B) MiR-155"" fibroblasts show down-regulation of

573 LXRa protein after transfection with miR-155 mimic. (C) Time-course of Lxra mRNA

574 expression in lungs of WT mice after bleomycin (n=4-7 per group). (D) Lung fibroblast

575 gating strategy. (E) Representative histograms and (F) quantitative evaluation of an increase

576 in LXRa expression in lung fibroblasts during fibrosis. (G) Expression of Lxra and (H)

577 Abcal in lungs of WT and miR-155- - mice on day 18. (I) Constitutive expression of Abcal in

578 lung fibroblasts (n=4) and in alveolar macrophages (n=4); and (J) constitutive expression of

579 Arg2 in alveolar macrophages (n=5) and (K) after transfection with Lxra siRNA or (L)

580 treatment with 22(S)HC (30^M). Data presented as mean±SEM or median and IQR.

581 *P<0.05.

FIG 3. The phenotype of miR-155-/" lung fibroblasts is driven by LXR. (A) MiR-155--fibroblasts showed higher proliferation in response FCS (%S) than WT (pooled lungs of 4 mice). (B) LXR-antagonist 22(S)HC reduced the proliferation of miR-155-/". (C-D) miR-155-/- fibroblasts had greater migration capacity; partially inhibited by 22(S)HC; and (E) produced more collagen than WT. (F) Collagen and (G-H) TGFp production by miR-155"" fibroblasts was inhibited by 22(S)HC (G) and by miR-155 mimic (H). (I) Arg2 was higher in miR-155-- than WT fibroblasts; and this was reduced by LXRa siRNA. Data are presented as mean±SEM of 4 biological replicates. *P<0.05.

FIG 4. Inhibition of LXR ameliorates lung fibrosis in miR-155- - mice. From two days prior to administration of bleomycin (n=10), mice were treated with daily injections of 22(S)HC. (A) Weight loss and (B) collagen deposition (turquoise) in miR-155- - mice was mitigated by 22(S)HC. (C) The expression of Col1a1, Col3a1 and Arg2 in lung tissues and (D) Arg2 in BAL cells in miR-155-- mice were reduced by 22(S)HC. Data presented as mean±SEM or median (IQR). *P<0.05.

FIG 5. The pro-fibrotic phenotype of IPF fibroblasts can be normalised by neutralization of LXR. Synchronized normal (n=4) and IPF (n=6) primary lung fibroblasts were cultured with 1% FCS (%S). (A) IPF fibroblasts contained higher concentrations of LXRa protein. (B) Collagen production by IPF fibroblasts was: (B) reduced by 22(S)HC or (C) potentiated by GW3965. (D) LXRa siRNA reduced collagen production by IPF fibroblasts. (E) 22(S)HC inhibited TGFp production, and (F) GW3965 potentiated serum-induced ARG2 expression by IPF fibroblasts. (G) Lxra siRNA reduced ARG2 expression by IPF fibroblasts. Data are presented as mean±SEM of 4 biological replicates. *P<0.05.

607 FIG 6. Deregulation of the miR-155/LXRa axis contributes to exacerbated pulmonary

608 fibrosis. LXRa, Liver X Receptor a; ARG2, Arginase 2; TGFP, Transforming growth factor

609 P; (22(S)HC), 22(S)hydroxycholesterol (metabolic blocker of LXR).

FIG 1 A

■§ 8001 tn

S- 600-1

io 400-

miR-155-'-

o 1200-

ci 900-

to 600-

O 300-

PBS bleo PBs bleo

miR-155

ra + c i_

■ WT + PBS

EES miR-155-/- + PBS

■ WT + ble * o 1-1 * 1 miR-155-/- + bleo

* i i * i rff1

Ym1 IL-13Ra2 Arg2 Nos2

g 5? 100T E £ I £ £ oJ

□ PBS

m 5 u" - 1.

day 18

W to lL

pre-miR-155

□ PBS

bleo 1 fxg/ml bleo 10 fxg/ml

FIG 2 A

miR155m Control

LXRa P-actin

64 kDa 43 kDa

oo 0.9

i/t <D

js 0.3.

Lxra (WT lung)

ormiR155m Control

MS2 miR-155 MS2- MS2-sPonge 3'UTR 3'UTR LXR LXR-MT

days: d1

d7 d18

-> CD140a

: isotype day 1 --PBS ~ -bleo J>A day 3 A

J>~ ay 7 I

PBS bleo

F _ 300-i Ll_ LXRa

2 200- ot CO CO CL * » -J T *

, , £ ¿P

se° xeP . ,£> xeP ¿5

Lxra (I ung)

Abcal (lung)

1.2 0.9. 0.6. 0.3

fibroblasts

miR-155

I 5.5-l

jG 3.0 cL

cu 0.501

> 0.25—1

0.40.2.

bleo niR-155-7-

Abcal 2.0,

1.5 1.00.5-

WT miR-155

WT miR-155

WT miR-155-'

siRNA C

miR-155-'

DMSO DMSO 22(S)HC WT miR-155-7-

«IS 0.5

2 ■a 1.0-

US OT E 0.5-

C 22(S)HC C 22(S)HC

WT miR-155

30|M 22(S)HC +3%S 20|M 22(S)HC +3%S 10|M 22(S)HC + 3%S 3%S 0%S

miR-155

30% S »«3% S + 22(S)HC (20|M) 13% S ^3% S + 22(S)HC (30|M) 13% S + 22(S)HC (10|M)

2500-,

0 0.3 1 3 miR-155 -/-

520' E 270'

c 20' <D 20'

miR-155-/-F

30|M 22(S)HC +3%S| 20|M 22(S)HC +3%S 10|M 22(S)HC + 3%S DMSO + 3%S 3%S 0%S

0%S + DMSO 0.3%S + DMSO 0.3%S + 22(S)HC

m.iR-15>

% S 0 3 0 3

WT miR-155'

£ 25'

TGF-P (pg/ml)

co||'~gen (|ig/ml)

2.0 | 1.5

g- 1.0 t 0.5

Cm miR-155m

siRNA C LXRa C LXRa

WT miR-155'

miR-155-'-

-e-PBS + excipient

0.80-days 0

-i—i—i—i—i—r

8 10 12 14 16 18 20

1.00-1 0.95 -

■ bleo + 22(S)HC 0.90-

-B- bleo + excipient

0.80 days 0

bleo i

—i—

- bleo + 22(S)HC

- bleo + excipient

t—i—i—i—i—i—i

8 10 12 14 16 18 20

I miR-155-'- PBS □mjRe15i5i-enteo

■ miR-155-'- bleo + 22(S)HC

miR-155-'- PBS miR-155-'- bleo+excipient miR-155-'- bleo+22(S)HC

Col1a. 'lungs)

Col3a1 (lungs)

Arg2 (lungs)

Arg2 (BAL)

LXRa ß-actin

Normal

■ * m

= f 4- r'n

^ S 32. I

g Normal IPF

Sä 100-

■■Normal □ IPF

1%S - + + + 22(S)HC-^M - - 10 30

c 300-

Sä 200-

GW3965 (^M)

"&400-I

300-1 ^200-I

C siRNA LXRasiRNA IPF

E 2000

Normal IPF

1%S -22(S)HC (^M) -

^■Normal 09Hd|IPF

^ 0.7-

g 0.5-1

a) 0.3H

f 0.1-

GW3965 (^M)

+ + 0.5 4

J 75 ■

c\j 50 ■ a

CsiRNA LXRa siRNA IPF

m kurowska-stolarska sm 160912 links removed.docx

Mariola Kurowska-Stolarska et al

The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis

SUPPLEMENTARY MATERIALS AND METHODS Mice

C57BL/6 microRNA-155-negative (miR155 1 ) mice were obtained from Jackson Laboratories. Wild-type (WT) control littermates from miRl55-/- mice were backcrossed from C57BL/6 mice 1. Mice were maintained in a pathogen-free facility. Experiments were performed according to UK Home Office guidelines (Animals Scientific Procedures Act 1986).

Reagents and experimental solutions

The synthetic LXR agonist GW3965 was provided by Schering-Plough Corporation (UK) and dissolved in DMSO (Riedel de Haen) for in vitro studies. The LXR antagonist 22(S)-

hydroxycholesterol, 22(S)HC ( Sigma) is an enantiomer of the natural LXR ligand 22(R)HC

and acts as an antagonist of LXR pathway function . For in vitro use the antagonist was dissolved in DMSO, and for in vivo use, in a 40% solution of 2-hydroxypropyl-ß-cyclodextrin in water; and the cyclodextrin was used as excipient control. Bleomycin sulphate, from sireptomyces verticillus (Sigma) was dissolved in PBS at a stock 1mg/ml concentration.

Experimental pulmonary fibrosis

Bleomycin administration generates a well-established murine model of pulmonary fibrosis 4 One dose (0.06units/30pl/25g mouse) or control PBS was administered intranasally under light anesthesia 5. Daily control excipient 2-hydroxypropyl-P-cyclodextrin (40%) or therapeutic LXR antagonist 22(S)HC (in vivo ti/2 8h ) i.p. at 30 mg/kg were started two days prior. 7-10 mice per group were used per experiment.

Endpoints used to assess experimental fibrosis

The mice were monitored daily for wellbeing and changes in body weight. Serum, broncho-alveolar lavage (BAL) fluid, and lung tissues were harvested typically at day 18, and on other days as indicated, and analyzed as described previously 6 Lung tissue was harvested and one lobe was collected into RNAlater, one lobe stored f ozen for collagen content, and one lobe collected into 4% neutral buffered forma';n and processed within 24h for histology. The extent of lung tissue inflammation and fibrosis was assessed by histological analysis of tissue sections stained with haematoxylin and eos.n (H&E) and Masson's Trichrome respectively. Two observers blinded to the nature of the samples scored the sections independently. Each scored the inflammation and fibrosis in 10 different microscope low-power fields per slide. The assessment of inflammation in each field was scored as follows: 1=histologically normal; 2=mild, 3=moderate and 4=severe inflammatory infiltrate; 5=complete loss of lung architecture. The inflammation score for each tissue section was the mean of the scores for each field. To evaluate the extent of the collagen deposition, a similar arbitrary score for fibrosis was as follows: 1=histologically normal; and 2=mild, 3=moderate and 4=severe collagen deposition across the parenchyma. The fibrosis score for each tissue section consisted of the mean of the scores for each field. Cytokine concentrations in serum and BAL fluid were measured using a 20-plex multi-cytokine assay (Life Technology) using a

Luminex platform (Bio-Rad), and TGFpi was quantified using paired antibodies (R&D systems). Acid/pepsin soluble collagen, appropriate for lung soft-tissue or cell culture was measured by quantitative Sirius Red binding (Sircol assay, Biocolor).

Human lung fibroblasts

Primary fibroblast lines were cultured from explanted lung tissues from idiopathic pulmonary

fibrosis (IPF) and normal subjects obtained with informed consent under a protocol approved

by the University of Pittsburgh Institutional Review Board (details on Tables E4 and E5). Fibroblast cell cultures

Lung fibroblast lines (IPF n=7 and control normal n=8) were ^ultured as previously described . Fibroblasts from pooled lungs of WT (n=4-5) and :mR- 155- - (n=4-5) mice were isolated and cultured as described 10. For functional experiments, sub-confluent cells in culture were first synchronised in serum-free DMEM medium for 24h followed by incubation with medium containing different concentration of fetal calf serum (FCS; 0.3%-3%) for 48h. Different fibroblast functions were best demonstrated at different FCS concentrations; and pilot titration experiments howed that the optimum FCS (%) in WT and miR-155- - mouse lung fibroblasts for prJiferation was (1 to 3%), for collagen production (3%) and for cell migration (0.3%). Similarly, for IPF and control lung fibroblast functions: for proliferation (3%), collagen production (3%) and migration (0.3%). Experiments were performed in 24 well plates, at a cell confluency of 70-80%. In some experiments LXR agonist GW3965 (0.5 pM, 4 pM) or LXR antagonist 22(S)HC (10, 20, 30 pM) or IL-33, IL-1a, IL-25, HMGB1 (10-100 ng/ml, all from Peprotech) or bleomycin (1, 10 pg/ml, Sigma) or bleomycin excipient (DMSO) were added. For proliferation assays, cells were pulsed with ( H) thymidine 4 hours before cell harvest. Control and IPF fibroblasts were tested typically in-

group numbers of 2 each in several independent experiments. Experiments were performed in standard condition (normoxia, 21% oxygen) or in hypoxia (1% oxygen). Each variant of stimulation was typically done in 3 to 4 technical replicates.

Evaluation of LXRa protein expression in lung cells by flow cytometry

Wild type mice were treated with Bleomycin (n=24) with dose of 0.06 U per 25 g mouse in 30 pl pro-rata or PBS (n=7) by intranasal installation on day 0 and then lungs were isolated on days 1, 2, 3 7 and 10. Mouse lungs were digested in RPMI containing Dispase (3.2 mg/ml, Roche), Collagenase P (0.4 mg/ml, Roche) and DNAse i (0 2 mg/ml) in 370C with rotation for 1h. Digests were filtered through 100 pm strainer and stained with Fixable Viability Dye eFluor® 780, 65086514, eBioscience) followed by the incubation with the antibodies against CD45 (109822), EpCAM (1x8220), CD141a (135908) and F4/80 (123110, all from BioLegend) 11. After 30 minutes, cells were washed and fixed/permeabilised with Transcription Factor Staining Buffer set (00552300, eBioscience) followed by incubation with anti-LXRa antibody (ab176323, Abcam) and Brilliant Violet 421™ Donkey anti-rabbit IgG Anybody (406410; BioLegend). LXRa protein expression was then evaluated in different lung cell types (fibroblasts, macrophages and epithelial cells). The gating strategy is shown on Fig, E1. Median fluorescence intensity (MFI) of isotype control was subtracted from MFI of LXRa and presented as a % of MFI of PBS group.

Manipulation of expression of miR-155 and LXRa.

To evaluate the contribution of LXRa to activation of fibroblasts, cells were transfected with LXRa-specific siRNA or All Stars Negative Control siRNA unlabeled or labeled with Alexa Fluor 488 (50 nM, both from Qiagen) using DharmaFECT 3 transfection reagent (Thermo Scientific). 16h later, cells were stimulated with medium containing 1% FSC or synthetic

LXR agonist GW3965 for a further 24h. To investigate the effect of miR-155 on fibroblast function, cells were transfected with mouse or human miR-155 mimic or control nucleotide mimic (C. elegans miR-67) unlabeled or labeled with Dy-547 (40 nM, both from Dharmacon). 16h later, the cells were stimulated with medium containing 1% FCS or supplemented with LXR agonist as described above. In all experiments, the transection efficiency of siRNA and microRNA mimic was checked by flow cytometry; and only when more than 60% of cells were positive were they used for further analysis. Cells were collected for RNA and the culture supernatants collected for assay of soluble collagen and TGFp.

Cell migration measured by in vitro wound-healing scratch assay

To evaluate the cell migration consistent with the wound healing properties of fibroblasts, an

in vitro scratch assay was performed as described previously . Briefly, cells were allowed to grow into a confluent monolayer and a single uniform scratch was applied to the cell layer in each well using a 100-p.l pipette tip. The cells were allowed to settle for 3h and then treated with either: vehicle, 22(S)HC (.O^M) in medium for the next 24h. The scratch was visualized using phase-con^asi microscopy (*10 objective) and photographed, and the width of the wound was measured at 0 and 24h. A line was drawn along each side of the scratch to remove minor irregularities generated by the mechanical scratch from the pipette tip, and the distance between the lines measured at 4 separate points. This was quantified in arbitrary units describing the scratch closure; from 1 = completely open scratch to 2 = total closure. Each variant of stimulation was done in 3 to 4 technical replicates.

Human monocyte-derived macrophages

CD14 positive cells were isolated from buffy coats using microbeads (Militenyi Biotech). Cells were differentiated into macrophages by incubation with M-CSF (50ng/ml) in RPMI 1640/10%FCS. After 6 days, macrophages (0.25 x106/well in 24 well plates) were transfected (DharmaFECT3) with human miR-155 inhibitor or control inhibitor (40 nM, Qiagen). 16h later the culture medium was changed and cells were transfected (DharmaFECT3) with LXRa siRNA or control siRNA (50nM, Qiagen). Cells were harvested 24h later and RNA was purified with miRNeasy kit (Qiagen). Each variant of stimulation was done in technical replicates (n=3-4).

Mouse alveolar macrophage culture

Cells obtained by bronchoalveolar lavage (BAL) fro m WT (n= 11) and miR-155- - (n=11) mice were counted and seeded in 200 pl of RPMI1640/'10%FSC (0.1 x106/well in 96-well plates). After 24h incubation, non-adherent cells were washed away and adherent alveolar macrophages were harvested for RNA ;sola+ion (n=5) or transfected with LXRa siRNA or control siRNA, n=3 (50 nM, Qiagen). In some experiments, alveolar macrophages (n=3) were incubated with DMSO or °2SHC (30pM). Cells and supernatants were collected 24h later.

Total RNA, including the small RNA fraction was isolated using the miRNeasy kit (Qiagen). cDNA was prepared using either the TaqMan microRNA Reverse Transcription (RT) kit or the high capacity cDNA RT kit (both from Life Technologies). TaqMan mRNA or microRNA assays (both from Life Technologies); or pairs of specific primers (Integrated DNA Technologies) in conjunction with SYBRgreen method (Life Technologies) were used for semi-quantitative determination of the expression of human and mouse microRNAs and

mRNAs. Details of the sequence and name of the assays are provided in Table E7. Expression of RNU6 snRNA or 18S or ß-actin or TBP (TATA binding protein) was used as endogenous control. Gene expression was analysed on an Applied Biosystems ABI7900HT

machine with SDS 2.2 software. Data is presented as relative values (2- ) with the expression relative to endogenous control; RNU6 for miR-155; 18S or ß-actin foi human mRNAs and 18S or TBP for mouse mRNAs. In some cases data are presented as fold change compared to control sample (2-AACT).

Western blot for LXRa

Sub-confluent normal and IPF lung fibroblast primary cell l:nes in culture were synchronised for 24h in serum-free medium and then cultured for 24h in medium supplemented with 1% FCS. Similarly, miR-155-- murine lung fibroblasts were synchronised for 24h and then transfected with control (C. elegans miR-67) or mouse miR-155 mimic (40nM; both from Dharmacon) in medium containing 1% FCS. After 24 hours, whole cell lysates were prepared using M-PER (Thermo Scientific) lysis buffer. Total protein (10^g) was applied to and run on a 10% SDS Page gel followed by transfer to PVDF membrane, and incubation with rabbit anti-mouse/human LXRa (2^g/ml; Lifespan Biosciences) or anti-mouse/human ß-actin (1^g/ml, Santa Cru^ Biotechnology).

MS2 pull-down assay

MS2-TRAP RNA affinity purification was performed as described in Yoon et al . Briefly, pmiRGLO vector (Promega) was modified to incorporate an MS2 binding domain (MS2BD) to create pmiRGLO-MS2BD. The 3'UTR from the human LXRa gene containing the predicted miR-155 binding sites were cloned downstream of the luciferase open reading frame (ORF) and MS2BD, creating pmiRGLO-MS2BD-Lxra (WT). pmiRGLO-MS2BD-

Lxra (MT) in which the seed-region of the predicted miR-155 MRE was mutated using Quick Change Lightning site-directed mutagenesis kit (Agilent) was used as a negative control. pmiRGLO-MS2BD-155 sponge (MS2BD-155s), containing 9 miR-155 binding sites, was synthesised to order from Integrated DNA Technology and cloned downstream of luciferase ORF in pGLO-MS2BD. This was used as a positive control. 1 million ^TEK293 cells were transfected with 1pg pMS2GFP plus 1pg of the relevant pmiRGLO-MS2BD vector using Lipofectamine 2000 (Invitrogen). Cells were grown for 48 hours. MS2GFP complexes were immuno-precipitated using anti-GFP IP kit (Milenyi) supplemented with RNAase inhibitor (10U/ml) (Invitrogen). Precipitated RNA were purified using miRNEASY kit (Qiagen). cDNA was synthesised using miRSCRIPT II kit (Qiagen) with HiFlex buffer. miR-155 was measured by qPCR using miR 155 Primer assay kit (Qiagen). miR-155 expression was normalised to luciferase transcript levels and expressed as relative expression to pmiRGLO-MS2BD empty vector.

Oxysterol assay

The serum samples from IPF patients and healthy control subjects was obtained and approved by the Lothian Research Ethics Committee. Oxysterols in these and mouse serum samples were analysed ^s described in Griffiths et al. 14 Briefly, sterols and oxysterols were extracted with ethanol and separated according to hydrophobicity on a reversed phase Oasis HLB column. The oxysterol fraction was split into two sub-fractions. Fraction A was treated with cholesterol oxidase followed by Girard P (GP) reagent, while fraction B was derivatised with GP reagent directly in the absence of enzyme. This allows the separate detection of oxysterols with and without an oxo group, respectively. The GP derivatised oxysterols were then separated from excess derivatisation reagent on a C18 column and analysed by liquid chromatography - mass spectrometry (LC-MS) with multistage fragmentation (MSn).

Oxysterols were identified by retention time, exact mass (5 ppm) and MSn by reference to authentic standards. Quantification was by isotope dilution mass spectrometry.

Statistical analysis

Data was analyzed by GraphPad prism or Minitab. Between-category differences in biomarker concentrations and gene expression levels were compared by Mann-Whitney u-test or one-way Anova with Tukey's multiple comparison tests Results from related cell cultures were compared by paired-t-test. Data illustrated as mean ± SEM, or box-and-whisker and dot-plot plots with medians and means respectively. Mass-ipectrometry data presented as mean±SD. P<0.05 was considered statistically significant.

SUPPLEMENARY TABLES

PBS Bleomycin

WT miR-155-- P WT miR-155-- p

Lung histology fibrosis score inflammation score 1.20 (1.10, 1.30) 1.22 (1.13, 1.35) 1.20 (1.15, 1.30) 1.27 (1.21, 1.52) 0.698 0.406 2.79 (2.27, 2.99) 2.44 (2.21, 2.9 1) 3.0 (2.69, 3.59) 3.80 (3.26, 4.15) 0.037 0.001

Lung lavage cell count x106 macrophage x106 lymphocytes x104 0.10 (0.10, 0.20) 0.10 (0.09, 0.19) 0.07 (0.03, 0.10) 0.20 (0.10, 0.40) 0.10 (0.09, 0.30) 0.16 (0.04, 0.55) 0.178 0.443 0.096 1.33 (1.20, 1.44) 0.92 (0.90, 1.08) 2 6 (0 44, 4.4) 7.55 (6.42, 8.15) 4.52 (3.09, 5.97) 23.3 (18.9, 31.1) 0.001 0.001 0.001

Lung tissue TGFp mRNA (tgfb) collagen (^g/g) 0.98 (0.93, 1.07) 2.00 (1.87, 2.23) 0.98 (0.95, 1.06) 2.19 (1.93, 2.30) 1.000 0.523 3.96 (2.94, 5.34) 2.60 (2.43, 2.91) 6.25 (5.05, 7.99) 2.90 (2.73, 3.35) 0.024 0.016

Table E1. Comparison of biomarkers of lung fibrosis and inflammation be tween wild-type (WT) and miR-155-/- mice given control PBS or bleomycin. Lung tissue and lavage were harvested 18 days after PBS or bleomycin. Inflammation and fibrosis was scored histologically (Methods in this article online repository). Lavage cytology was determined by total and differential cell counting of H&E stained cyto-centrifuge slide preparations. Expression of TGFp mRNA in lung tissue was determined by qPCR. Lung collagen was quantified by Sircol assay. Median (IQR), with Mann-Whitney p-values.

The following cytokines were detectable in BAL fluid (TNf a, IL-ip, IL-4, IL-10, IL-12, CXCL10, CXCL1, CXCL9, CCL3, basic FGF, VEGF), and in serum (TNF-a, IL-la, IL-1p, IL-4, IL-5, IL- 0, L-12, CXCL10, CXCL1, CXCL9, CCL3, basic FGF, VEGF, TGFp, IFN-y, CCL2). The concentrations of VEGF and CXCL1 in serum and bFGF in BAL were changed by bleomycin compared with PBS treatment, however none were changed further in miR-155- - mice compared with WT mice both given bleomycin. The following cytokines were undetectable in BAL fluid (GM-CSF, IFN-y, IL-la, IL-2, IL-5, IL-6, IL-13, IL-17, CCL-2), and in serum (GM-CSF, IL-2, IL-6, IL-13, IL-17).

Target Name Reference

AGTR1 Angiotensin II receptor, type 1 Martin MM 15

SMAD1 SMAD* family member 1 Rai D 16

SMAD5 SMAD family member 5

SMAD2 SMAD family member 2 Louafi F 17

HIF-1a Hypoxia-inducible factor 1 -alpha Brüning U 18

Table E2. Conserved and validated miR-155 targets that are expressed in lungs. The stepwise prediction-target identification strategy identified LXRa among experimentally validated miR-155 targets such as AGTR1, HIF-1a, BMP and TCFp signalling molecules; SMAD2, SMAD1 and 5, as potentially targets involved in miR-155 fine-tuning of the remodeling process in the lung. Among them however only HIF-1a and SMAD5 are miR-155 conserved target in mouse and human. We confirmed that HIF-1a mRNA was a target by demonstrating that its expression was increased in lung tissue of miR-155--mice given bleomycin (Fig. E3) compared to WT or PBS controls, as well as confirming the integrity of the miR-155 gene deleted mouse. *SMAD is an acronym for protein homologs of Drosophila protein (mothers against decapentaplegic; MAD) and C. elegans protein SMA (gene sma; small size).

Mean concentration . g/mL

PBS treatment Bleomycin treatment

Wild type miR-155"" Wild type miR-155--

Sterol Systematic Name (Common name) Mean SD Mean SD Mean SD Mean SD

7a-Hydroxy-3 -oxocholest-4-en-26-oic acid6 38.18 19.41 51.84 27.68 20.86 1.95 31.35 3.45

7a-Hydroxycholest-4-en-3-one 14.93 6.18 16.87 12.73 12.46 3.11 12.36 3.28

3-Oxocholesta-4,6-dien-26-oic acid6 8.11 6.85 11.52 9.71 3.84 1.86 3.10 1.34

Cholest-5 -ene-3p,7a-diol (7a-Hydroxycholesterol)5 5.04 4.62 11.91 20.62 2.55 1.28 2.19 1.51

3p-Hydroxycholest-5-en-7-one (7-Oxocholesterol)5 4.09 3.83 1.40 1.74 3.52 1.70 2.91 1.61

Cholest-5-ene-3p,7p-diol4 (7p-Hydroxycholesterol) 3.81 3.35 2.90 0.80 2.66 0.50 1.44 0.66

3p-Hydroxycholest-5-en-26-oic acid3 2.16 0.35 1.93 0.24 3.12 0.49 3.02 0.42

Cholest-5-ene-3p,24S -diol3 (24S-hydroxycholesterol) 1.78 0.25 1.18 0.41 1.86 0.94 1.08 0.43

3p,7a-Dihydroxycholest-5-en-26-oic acid7 1.37 1.34 0.81 1.63 0.82 0.47 1.29 1.47

7a,26-Dihydroxycholest-4-en-3-one 1.30 0.41 ^ 1.91 0.96 0.76 0.17 0.96 0.17

7a,25 -Dihydroxycholest-4-en-3 -one 1.14 0.31 | 0.97 0.44 1.02 0.67 0.83 0.15

7a-Hydroxy-26-nor-cholest-4 -ene -3,24 -dione1,2 1.10 0.48 1 1.31 0.29 1.88 0.46 2.54 1.45

3p,7p-Dihydroxycholest-5-en-26-oic acid3 0.95 0 i8 0.88 0.20 0.87 0.32 1.13 0.17

Cholest-5-ene-3p,26-diol3 ((25R),26-Hydroxycholesterol) 0.83 0.42 0.56 0.45 0.87 0.90 0.48 0.47

Cholest-5-ene-3p,25-diol3 (25-hydroxycholesterol) 0.56 0.58 0.18 0.15 0.30 0.39 0.17 0.12

3-Oxocholest-4-en-26-oic acid 0.55 (.23 0.56 0.32 0.64 0.20 0.62 0.12

3p,22,25-Trihydroxycholest-5-en-24-one1 0.38 0.30 0.53 0.47 0.43 0.44 0.44 0.33

Cholest-5-ene-3p,7a,25-triol (7a,25-Dihydroxycholesterol) 0.04 0.04 0.15 0.12 0.07 0.08 0.02 0.04

Cholest-5-ene-3p,7a,26-triol (7a,26-Dihydroxycholesterol) r 0.01 0.02 0.03 0.05 0.03 0.04 0.00 0.00

3p-Hydroxycholesta-5,7-dien-26-oic acid7 0 00 0.00 0.00 0.00 0.07 0.14 0.09 0.16

Table E3. Oxysterols and cholestenoic acids in mouse sei^m after control PBS or bleomycin treatment. Oxysterols and cholestenoic acids identified by LC-ESI-MSn in serum following solid phase extraction and charge-tagging with GP-hydrazine. In the absence of authentic standards presumptive identifications based on exact mass, MSn spectre and retention time are given. Quantitation was by stable isotope dilution. Samples from four PBS-treated wild type and miR-155"" mice and four bleomycin-treated wild-type and miR-155"" mice were analysed. We have adopted the systematic sterol nomenclature recommended by the Lipid Maps consortium http://www.lipidmaps.org/. In this nomenclature hydroxylation of the terminal carbon of the sterol side-chain introducing R stereochemistry at C-25 is defined as C-26 hydroxylation. 1 Presumptive identification based on exact mass and MSn spectra. 2 26-Nor-sterol is

a likely decomposition product of a 24-oxo-26 acid. 3 LXR ligand. 4 May be formed by autoxidation. 5 Can be formed enzymatically and by autoxidation. 6 7a-Hydroxy-3-oxocholest-4-en-26-oic acid dehydrates to a minor degree to 3-oxocholesta-4,6-dien-26-oic acA Thus, the total 7a-hydroxy-3-oxocholest-4-en-26-oic acid corresponds to the sum of the two acids. 7 3p,7a-Dihydroxycholest-5-en-26-oic acid dehydrates to a minor degree to 3p-hydroxycholesta-5,7-dien-26-oic acid. Thus, the total 3p,7a-dihydroxycholest-5-en-26-oic acid corresponds to the sum of +he two acids. To be noted 27-hydroxycholesterol is (25R)26-hydroxycholesterol according to IUPAC nomenclature.

IPF- Code age sex race smoke onset FVC FEV1 TLC DLCO EF PA echo PA ath T PA mean PA WP 6-min walk

117 71 f h 0 n.a. 51 54 n.a. 29 55 39 27/9 17 6 610

124 64 m c 60 n.a. 67 80 n.a. 37 60 39 30/14 21 7 1430

131 63 f c 0 5 57 66 60 15 55 71 70/21 40 10 470

133 61 f a 0 10 20 23 n.a. n.a. 55 20 45/8 30 10 550

135 67 f c 30 < 5 60 64 67 26 65 46 44/17 27 10 60

136 75 f c 0 3 57 67 55 26 60 46 39/14 23 12 900

161 65 f c 50 n.a. 48 64 n.a. na. 55 46 48/22 31 20 280

Table E4. Clinical and demographic details of IPF patients. I°F cote I.D. number, Age (years), Gender (f=female, m=male), race (h=Hispanic, c=Caucasian, a=Arabic/middle-eastern), smoking history (pack years), onset of symptoms (years), lung function; % of predicted normal: Forced Vital Capacity (FVC), Forced Expiratory Volume in one second (FEV 1), Total Lung Capacity (TLC), Extinction Fraction (EF), Pulmonary Artery echo, pressures and WP, and 6-minute walk test (feet). No patient was taking therapy for pulmonary hypertension. All patients were taking immune-suppressive therapy (prednisolone plus mycophenolate mofetil).

NL-Code age sex race smoke Cause of Death Notes

51 76 f n.a. 0 CVA mild pulmonary edema on CXR

57 60 f n.a. 0 n.a. On vent 1+ days, patchy infiltrates sparing RLL, minimal secretions at Bronch.

59 50 f n.a. 0 CVA; donated after cardiac death Other lung was used for a . <ngle transplant; less than 1 hr of warm ischemia.

60 60* n.a. n.a. 0 CVA after neurosurgery basilar atelectasis

62 62 f n.a. 10 intracerebral bleed after initial CVA CXR clear; 10 yeais of diabetes; past history of resected squamous cell of nose; PO2 78 on 60% FiO2, likely related to atelectasis or transient arrest earlier today. On antibiotics so covered for aspiration at time of CVA.

67 25 m n.a. n.a. GSW to the head Mild basilar density c/w atelectasis or infiltrate

71 33 f b n.a. Unknown Autopsy

74 37 m c n.a. Lung resection; benign granuloma ^Surgery; not a transplant donor

Table E5. Clinical and demographic details of post-mortem donor s of control lung tissue. Normal lung code ID. number, Age (years, *60=early 60's), Gender (f=female, m=male), race (c=Caucasian, b=Black), smoking his+ory (pack years), details of cause of death and clinical notes. n.a. = data not available.

Serum concentration ng/mL

Control n=6 IPF n=9

Sterol Systematic Name (Common name) Mean SD Mean SD

3p-Hydroxycholest-5-en-26-oic acid3 90.68 31.98 79.11 26.58

7a-Hydroxy-3 -oxocholest-4-en-26-oic acid6 76.03 28.31 90.89 26.76

3p,7a-Dihydroxycholest-5-en-26-oic acid3,7 29.51 17.03 45.29 19.04

Cholest-5-ene-3p,26-diol3 ((25R),26-Hydroxycholesterol) 24.36 3.85 21.88 11.55

3p-Hydroxycholest-5-en-7-one5 (7-Oxocholesterol) 22.46 23.01 68.03 112.00

7a-Hydroxycholest-4-en-3-one 18.55 16.48 15.52 11.52

Cholest-5-ene-3p,7a-diol5 (7a-Hydroxycholesterol) 17.31 9.19 22.14 27.06

Cholest-5-ene-3p,7p-diol4 (7p-Hydroxycholesterol) 15.91 18.03 19.21 29.42

3-Oxocholesta-4,6-dien-26-oic acid6 15.11 7.41 A 19 02 7.55

Cholest-5-ene-3p,24S-diol3 (24S-hydroxycholesterol) 13.01 4.4] 1^70 3.29

3p,22,25-Trihydroxycholest-5-en-24-one1 6.60 1.r4 6.80 3.67

3p,7p-Dihydroxycholest-5-en-26-oic acid3 4.48 2.35 4.71 0.99

7a-Hydroxy-3 -oxochol-4-en-24-oic acid 4.35 2.12 4.55 1.73

3p-Hydroxychol-5-en-24-oic acid 3.24 1.09 3.04 1.02

3p,7a-Dihydroxychol-5-en-24-oic acid 3.11 2.14 3.17 1.94

7a,26-Dihydroxycholest-4-en-3-one 2.43 0.98 2.46 0.79

Cholest-5-ene-3 p,25 -diol3 (25 -hydroxycholesterol) 1.97 *>0.38 2.19 1.29

3-Oxocholest-4-en-26-oic acid t i no 0.53 2.51 0.91

3 p-Hydroxycholesta-5,7 -dien-26-oic acid7 1 12 1.53 1.81 1.55

7a,25-Dihydroxycholest-4-en-3-one 079 0.29 1.10 0.40

7a-Hydroxy-26-nor-cholest-4-ene-3,24-dione1,2 0.35 0.12 0.38 0.11

Cholest-5-ene-3p,7a,26-triol (7a,26-Dihydroxycholesterol) 0.31 0.26 0.27 0.24

Cholest-5-ene-3p,7a,25-triol (7a,25-Dihydroxycholestero'i 0.21 0.13 0.24 0.14

Table E6. Oxysterols and cholestenoic acids in human IPF and control sera. Oxysterols and cholestenoic acids identified by LC-ESI-MS in serum following solid phase extraction and charge-tagging with GP-hydrazine. In the absence of authentic standards presumptive identifications based on exact mass, MSn spectra and retention ^me are given. Quantitation was by stable isotope dilution. We have adopted the systematic sie~ol nomenclature recommended by the Lipid Maps consortium http://www.lipidmaps.org/. In this nomenclature hydroxylation of the terminal carbon of the sterol side-chain introducing R stereochemistry at C-25 is defined as C-26 hydroxylation.

1 Presumptive identification based on exact mass and MSn spectra. 2 26-Nor-sterol is a likely decomposition product of a 24-oxo-26 acid. 3 LXR ligand. 4 May be formed by autoxidation. 5 Can be formed enzymatica'ly and by autoxidation. 6 7a-Hydroxy-3-oxocholest-4-en-26-oic acid dehydrates to a minor degree to 3-oxocholesta-4,6-dien-26-oic acid. Thus, the total 7a-hydroxy-3-oxocholest-4-en-26-oic acid corresponds to the sum of the two acids. 7 3p,7a-Dihydroxycholest-5-en-26-oic acid dehydrates to a minor degree to 3p-hydroxycholesta-5,7-dien-26-oic acid. Thus, the total 3p,7a-dihydroxycholest-5-en-26-oic acid corresponds to the sum of the two acids. To be noted 27-hydroxycholesterol is (25R)26-hydroxycholesterol according to IUPAC nomenclature.

Forward 5'—3' Reverse 5'--3'/Company

TBP TGC TGT TGG TGA TTG TTG GT AAC TGG CTT GTG TGG GAA AG

18S 4310893E Life Technologies

Col1A1 AGC TTT GTG GAC CTC CGG CT ACA CAG CCG TGC CAT TGT GG

Col3A1 GTT CTA GAG GAT GGC TGT ACT A AA CAC A TTG CCT TGC GTG TTT GAT ATT C

Col1A1 Mm00801666 ml Life Technologies

Col1A2 Mm00483888 ml Life Techno' >gie s

Tgfßl ACC CCC CAT TGCT GTC CCGT CCT TGG TTC AGC CAC TGC CG

Abcal Mm00442646 ml Life Technologies

Rxral Mm00441185 ml Life Tech ologies

Arg2 ACC AGG AAC TGG CTG AAG TG TGA GC A TCA ACC CAG ATG AC

Lxrß (NR1H2) Mm00443451_m1 Life Technologies

Lxra (NR1H3) GCT CAG GAG CTG ATG ATC CA G CG CTT GAT CCT CGT GTA G

Il13ra2 TCT GGT ATG AGG GCT TGG AT GCT GGA GGT AAT CAG CAC ACT

Ym1 CAT GAG CAA GAC TTG CGT GAC GGT CCA AAC TTC CAT CCT CCA

Nos2 AGA CCT CAA CAG AGC CCT CA GCA GCC TCT TGT CTT TGA CC

miR-155 TM002623/MS00031486 Life Technologies/Qiagen

RNU6 TM001973/MS00033740 Life Technologies/Qiagen

Hifla Mm00468869 ml Life Technologies

ß-actin 4310881E Life Technologies

COL1A1 CAA TGC TGC CCT TTC TGC TCC TTT CAC TTG GGT GTT TGA GCA TTG CCT

COL3A1 TAT CGA ACA CGC AAG GTG AGA GGC CAA CGT CCA CAC CAA ATT CTT

LXRa (NR1H3) Hs00l72885_ml Life Technologies

ABCA1 Hs0l059ll8 ml Life Technologies

ARG2 Hs00982833 ml Life Technologies

miR-155 TM002623/MS°003l486 Life Technologies/Qiagen

ZNF652 Hs00977533 ml Life Technologies

Table E7. Primer sequences for TaqMan assays

References

1. Kurowska-Stolarska M, Alivernini S, Ballantine LE, Asquith DL, Millar NL, Gilchrist DS, et al. MicroRNA-155 as a proinflammatory regulator in clinical and experimental arthritis. Proc Natl Acad Sci U

5 A 2011; 108:11193-8.

2. Tranheim Kase E, Nikolic N, Pettersen Hessvik N, Fjeldheim AK, Jensen J, Thoresen GH, et al. Dietary supplementation with 22-S-hydroxycholesterol to rats reduces body weight gain and the accumulation of liver triacylglycerol. Lipids 2012; 47:483-93.

3. Moeller A, Ask K, Warburton D, Gauldie J, Kolb M. The bleomycin animal model: a useful tool to investigate treatment options for idiopathic pulmonary fibrosis? The international joui* al of biochemistry

6 cell biology 2008; 40:362-82.

4. B BM, Lawson WE, Oury TD, Sisson TH, Raghavendran K, Hogaboam CM. A^im^' models of fibrotic lung disease. American journal of respiratory cell and molecular biology 2013; 49:167-79.

5. Li M, Krishnaveni MS, Li C, Zhou B, Xing Y, Banfalvi A, et al. Epithelium-specific deletion of TGF-beta receptor type II protects mice from bleomycin-induced pulmonary fibrosis. The Journal of clinical investigation 2011; 121:277-87.

6. Kurowska-Stolarska M, Stolarski B, Kewin P, Murphy G, Corrigan CJ, Ying S, et al. IL-33 amplifies the polarization of alternatively activated macrophages that contribute to airway inflammation. J Immunol 2009; 183:6469-77.

7. Patel M, Xu D, Kewin P, Choo-Kang B, McSharry C, Thomson nc, et al. TLR2 agonist ameliorates established allergic airway inflammation by promoting Th1 response and not via regulatory T cells. Journal of immunology 2005; 174:7558-63.

8. Hsu E, Shi H, Jordan RM, Lyons-Weiler J, Pilewski JM, Feghali-Bostwick CA. Lung tissues in patients with systemic sclerosis have gene expression patterns unique to pulmonary fibrosis and pulmonary hypertension. Arthritis and rheumatism 2011; 63:783 -94.

9. Pilewski JM, Liu L, Henry AC, Knauer AV, Fe^hali-Bos. wick CA. Insulin-like growth factor binding proteins 3 and 5 are overexpressed in idiopathic pulmonary fibrosis and contribute to extracellular matrix deposition. The American journal of pathology 2005, 166:399-407.

10. Welsh DJ, Peacock AJ, MacLean M, Harnett M Chronic hypoxia induces constitutive p38 mitogen-activated protein kinase activity that correlates with enhanced cellular proliferation in fibroblasts from rat pulmonary but not systemic arteries America. journal of respiratory and critical care medicine 2001; 164:282-9.

11. Malhotra D, Fletcher AL, Astar-a J, Lui.acs-Kornek V, Tayalia P, Gonzalez SF, et al. Transcriptional profiling of stroma from inflamed ^d resting lymph nodes defines immunological hallmarks. Nat Immunol 2012; 13:499-510.

12. Costello CM, Howell K, C^toL E, McBryan J, Konigshoff M, Eickelberg O, et al. Lung-selective gene responses to alveolar hypoxia: potential role for the bone morphogenetic antagonist gremlin in pulmonary hypertension. American journal of physiology. Lung cellular and molecular physiology 2008; 295:L272-84.

13. Yoon JH, Srikartan S, Gorospe M. MS2-TRAP (MS2-tagged RNA affinity purification): tagging RNA to identify associated miRNAs. Methods 2012; 58:81-7.

14. Griffiths WJ, Crick u, Wang Y, Ogundare M, Tuschl K, Morris AA, et al. Analytical strategies for characterization of oxysterol lipidomes: liver X receptor ligands in plasma. Free radical biology & medicine 2013; ^9:69-84.

15. Martin mmm, L ~e EJ, Buckenberger JA, Schmittgen TD, Elton TS. MicroRNA-155 regulates human angiotensin II type 1 receptor expression in fibroblasts. The Journal of biological chemistry 2006; 281:18277-84.

16. Rai D, Kim SW, McKeller MR, Dahia PL, Aguiar RC. Targeting of SMAD5 links microRNA-155 to the TGF-beta pathway and lymphomagenesis. Proc Natl Acad Sci U S A; 107:3111-6.

17. Louafi F, Martinez-Nunez RT, Sanchez-Elsner T. MicroRNA-155 targets SMAD2 and modulates the response of macrophages to transforming growth factor-{beta}. The Journal of biological chemistry 2010; 285:41328-36.

18. Brüning U, Cerone L, Neufeld Z, Fitzpatrick SF, Cheong A, Scholz CC, et al. MicroRNA-155 promotes resolution of hypoxia-inducible factor lalpha activity during prolonged hypoxia. Molecular and cellular biology 2011; 31:4087-96.

m kurowska-stolarska sup figure legends 160912.docx

Mariola Kurowska-Stolarska et al

The role of microRNA-155/LXR pathway in experimental and Idiopathic Pulmonary Fibrosis

Supplementary Figure Legends

FIG E1. miR-155-- bleo mice show an increase in the expression of lung collagen 1a (Collal) but not collagen1a2 (Col1a2). Wild-type and miR-155-- mice were given bleomycin (bleo) or PBS control (n=8 per group) on day 1 and tissues harveste4 on day 18. QPCR data were normalised to endogenous control (18S); and presented as 2-deltadeltaCT (fold change relative to mean of WT PBS group). Data are shown as median and inter-quartile range from 2 independent experiments. *P<0.05.

FIG E2. Regulation of miR-155 in murine lung fibroblasts. (A to C) Primary fibroblasts from pooled lungs of WT mice (n=5) were synchronized in culture with serum-free medium and then cultured with bleomycin (1 and 10 pg/ml), or IL-33, IL-1a, IL-25 or HMGB1 (10 or 100 ng/ml) for 8 and 24h. Cells were collected and expression of miR-155 evaluated by qPCR. (A) Bleomycin inhibited the expression of miR-155 at 24h, (B) Exogenous IL-33, IL-25 and HMGB-1 did not affect miR 155 expression. (C) Exogenous IL-1a increased miR-155 expression. Data presented as mean ± SEM of 3 technical replicates, repeated in 2 independent experiments. *P<0.05.

FIG E3. Lung tissue Hifla, TGFP and Smadl mRNA expression was increased in miR-155--

mice given bleomycin. RNA was purified from lung tissue harvested on day 18 from wild-type (WT) and miR-155- - mice given bleomycin or control PBS (n=8 per group). The expression levels of Hifla, Tgffi and Smadl were evaluated by qPCR and presented as relative to 18S (2-deltaCT). All genes tested were up-regulated in bleomycin treated miR-155-/- compared to WT. Data are presented as box and whisker plots showing median and inter-quartile range. * P<0.05.

FIG E4. LXRa protein expression in lung cells during bleomycin iiduced fibrosis. (A-D) WT mice were given bleomycin (n=24) or PBS (n=7) on day 0 and ling: were harvested on days: 1, 2 ,3, 7 and 10. Lungs from 4-5 mice (bleo group) and 1-2 mice (PBS group) were harvested at each time point; and LXRa expression evaluated by flow cytometry. Expression of LXRa in PBS group remained stable through the time course of experiment. (A) Gating strategy for macrophages, fibroblasts and epithelial cell, in digested lungs. (B) Representative and quantitative expression of LXRa in small macrophage population (Mac-1) showing no difference between PBS and bleo group at any time points. (C) Representative and quantitative expression of LXRa in large macrophage population (Mac-2) showing no difference between PBS and bleo group at any tinm points. (D) Representative and quantitative expression of LXRa in epithelial cells showing no difference between PBS and bleo group at any time points. Data are presented as % of median fluorescence intensity (MFI) of LXRa in PBS groups after subtraction of isotype MFI ± SEM of biological replicates *P<0.05.

FIG E5. Validation of transfection efficiency of LXR siRNA. (A) Wild-type and miR-155-/- lung macrophages (n=3), (B) lung fibroblasts (pooled n=4), (C) human blood monocyte-derived macrophages (n=2) prior transfected with control inhibitor (Ci) or miR-155 inhibitor (miR155i)

and (D) human IPF primary lung fibroblasts (n=3) were transfected with control siRNA or siRNA for LXR (both 50 nM). All data were collected 48 h after transfection. Expression of LXRa quantified by qPCR are normalised to 18S (mouse) or /-actin (human). Data are presented as change in LXRa expression compared to cells transfected with control (C) ;;RNA.

FIG E6. Inhibition of miR-155 in human macrophages by gene-slencing leads to increased LXR-dependent ARG2 mRNA expression. Macrophages were cultured (as in Fig E5, C) for 16h after siRNA transfection. GW3965 (4pM) or excipient control DMSO was then added for a further 24h. Inhibition of miR-155 in human macrophag-s minced LXR agonist-induced ARG2 expression. Values quantified by qPCR were normalized to /3-actin and presented as 2-deltaCT Histograms show mean ± SEM of three technic^ replicates. *P<0.05.

FIG E7. The bleomycin-induced bronchoalveolar lavage leukocytosis in miR-155"" mice was reduced by 22(S) hydroxycholesterol treatment. Airway and alveolar leukocytes were harvested by bronchoalveolar lavage (BAL) from WT and miR-155- - mice on day 18 after bleomycin or PBS (as in Fig 4, A). The increased BAL leukocytosis in miR-155- - mice given bleomycin was reduced when treated with the LXR antagonist 22(S) hydroxy-cholesterol [22(S)HC] compared with control ex^ipien (40% 2-hydroxypropyl-/-cyclodextrin in water), and was then not different from wild-type mice given bleomycin and treated with excipient. Histograms show mean ± SEM of 6-10 mice. *P<0.05.

FIG E8. LXR antagonist reduces IPF fibroblast proliferation and migration. (A to C) Normal

and IPF primary lung fibroblasts were serum-starved for 24h to synchronize their growth and then medium containing 1% FCS with or without 22(S)HC was added for 48h. (A) Cells were pulsed with ( H) thymidine 4 hours before harvest. IPF fibroblasts had higher than normal proliferation and this was reduced to normal by 22(S)HC. (B to C) Fibroblast migration measured in a wound-healing assay. A uniform scratch was made through confluent layers of fibroblasts that were then cultured with zero-percent FCS or 0.3% FCS with 22(H)HC or excipient DMSO and after 24h the breadth of the scratch space was measured. (B) A typical image showing that IPF fibroblasts demonstrate higher migration than normal fibroblasts into the scratch space. (C) Migration was quantified and presented as mean ± SEM of 2-3 biological replicates. Each sample was done in technical replicates (~=4) *P<0.05.

FIG E9. Coordinated expression of miR-1^5 and LXRa is deregulated by hypoxia in IPF fibroblasts. (A) Synchronized normal (n=3, and IPF (n=3) fibroblasts were transfected with control miR or miR-155 (C = control miR mimic; miR-155 = miR-155 mimic) and cultured in 1% FCS for 16h then supplemented with GW3965 for 24h and collagen measured by colorimetric (Sircol) assay. (B) Primary lung fibroblast cell lines from control donors (n=8) and IPF (n=7) were expa nded to passage #4, then synchronized in serum-free medium and cultured in medium containing 1% FSC (%S) for 48h. The expression of miR-155 was quantified by qPCR. (C-D) Control (n=4) and IPF (n=4) lung fibroblast cultures were synchronized in serumfree medium for 24h which was then replaced with fresh serum free media or medium containing 1% FSC for a further 48h in 21% oxygen; normoxia (N) or 1% oxygen; hypoxia (H). The expression of (C) miR-155 and (D) mRNA for LXRa and ABCA1 were quantified by qPCR. (E)

The individual expression levels for miR-155 and LXRa from normal (n=4) and IPF (n=4) fibroblasts cultured in serum free media or media containing 1% FCS in two separated experiments under hypoxia were plotted as dot-plot graph. (F) Cells were cultured as in C and ZNF652 expression evaluated by qPCR. (G) The individual expression levels for miR-155 and ZNF652 from normal and IPF fibroblasts cultured under hypoxia were plotted as dot-plot graph. Values quantified by qPCR relative to /3-actin (mRNA) or snU6 (miR-155). Data are presented as dot plots with a mean bar, or mean ± SEM. *P<0.05; each indiv'dual was done in technical (n=2) and experimental (n=2) replicates; solid circles represent fbr^blasts in medium containing 1% FSC; solid triangles represent fibroblasts, in serum-free medium.

Collagen 1a1

miR-155-'

£ (/) <0 CO C^CL

tD o C 5

Collagen 1a2

miR-155-'

miR-155

□ PBS

bleo Vg/ml I bleo 10 |ag/ml

miR-155

4.Qx1Q-3i

W 3.0x10-3-

cd 2.Qx1Q-3-

cd 1.0* 10"3-1

(ng/ml) 0 (-)

6.0x10-3-5.0x10-3-4.0x10-3-3.0x10-3-2.0x10-3-1.0x10-3-

miR-155

100 10 100 IL-25 HMGB1

'10 (rt CD

Jo cd Qi

'io (rt

5x10 "5-4x10 "5-3x10 "5-

jo 2x10 "5-

1x10 "5-

8x10-4,

6x10-4

a 4x10-4'

15 2x10-4' Qi

Tgfbr2

miR-155-/- WT miR-155-/

Figure E4

Leukocyte gate

Stromal cell gate

Mac-1 Mac-2

-> FSC

Epithelial cells Fibroblasts

^ CD'l40a......

Leukocytes ^¿^j«^

Stromal cells

B Mac-1

C Mac-2

D Epithelial Cells

4? A* ^ ^ ^

<£> ¿S 6" i>N ^

? 100-

& & & P £ xeP xeP J

A Mouse macrophages

B Mouse fibroblasts

C siRNA WT LXRa siRNA

miR155 LXRa siRNA

Human macrophages

Ci+ C siRNA Ci+LXRa siRNA □ miR155i+C siRNA ЕПП miR155i+LXRa siRNA

IPF lung fibroblasts

C siRNA LXRa siRNA

C siRNA miR-155i LXRa siRNA

o 0.5-

miR-155-/-

miR-155-/-

miR-155-/-WT WT

+ PBS + Excipient

+ Bleo + Excipient

+ Bleo + 22(S)HC + Bleo + Excipient + Bleo + 22(S)HC

ç 1500. E

&1000.

o 500-

Normal

□ 0%S

□ 1°% S

■■ 1% S + DMSO HID 1% S + 22(S)HC (10^M) ^ 1% S + 22(S)HC (20^M) ^ 1% S + 22(S)HC (30^M)

ot u 1.0-'F 0

O E0.5J

I - 22(S)HC

'2(S)HC

q 0.20-

c 0.15-

£ 0.10-

° 3.25'

P 1.25'

^ 0.25'

£ 0.25-

S 0.15'

■ Normal -,-

□ ipf

miR-155

£ 0.1-1

■g 0.01-1

JÊ tu

C miR155 C miR155 gw3965

miR155 *

o _ o 00

Normal

0.650.550.450.350.250.150.05

Normal

Normal

ABCA 1

■ ■

Normal

Normal lung fibroblasts (Hypoxia)

n o (r = -0.655, p = 0.008) n o

V 12, is

s ■ s

^ 0.8 rp

UO 0.4.

Q£ 0 0 ■ QC

'S ■ 'S

_-0.4 g o

o -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0

ID|r lung fibroblasts (Hypoxia) (r = -0.152, p = 0.573)

log LXRa expression

-0.9 -0.6 -0.3 0.0 log LXRa expression

ZNF652

N H N H

Normal lung fibroblasts (Hypoxia) (r = -0.349, p = 0.202)

£ 12 s s

0 0.9 r

1 °.6

Lo 0.3 0.0

-3.0 -2.8 -2.6 -2.4 -2.2

° 1.2

to to a)

IPF lung fibroblasts (Hypoxia) (r = -0.533, p = 0.033)

1.0 0.8

<u 0.6

-2 log ZNF652 expression — log ZNF652 expression

No.mal IPF