Contraction Agonist Induced TGF-β Activation by HASM Cells
A. HASM cells in co-culture with TMLCs were stimulated with 0, 0.1, 1, 10 and 100μM LPA. Data is expressed as mean TGF-β activity ± SEM. n= 3 independent primary HASM cell lines derived from non-asthmatic donors. P<0.05 for 10μM LPA and P<0.05 for 100μM LPA.
B. HASM cells in co-culture with TMLCs were stimulated with 0, 0.01, 0.1, 1, and 10μM methacholine. Data is expressed as mean TGF-β activity ± SEM. N=2 independent experiments using non-asthmatic HASM cells. P<0.05 for 1μM methacholine.
C. HASM cells were stimulated with 20μM LPA in the presence (white bars), or absence (black bars), of 10μg/ml pan-TGF-β neutralising antibody and relative expression of PAI1 mRNA levels were measured.LPA increased PAI1 expression (P<0.05), which was abrogated by pan-TGF-β neutralising antibody (P<0.05) Data was normalised to housekeeping gene β2-M and expressed as mean fold increase in PAI1 mRNA ± SEM. Data replicated three times and representative example shown.
D. HASM cells were stimulated with 1μM methacholine in the presence (white bars), or absence (black bars), of 10μg/ml pan-TGF-β neutralising antibody and relative expression of PAI1 mRNA levels were measured. Methacholine increased PAI1 expression (P<0.005), which was abrogated by pan-TGF-β neutralising antibody (P<0.005) Data was normalised to housekeeping gene β2-M and expressed as mean fold increase in PAI1 mRNA ± SEM. Data replicated three times and representative example shown.
E. HASM cells were stimulated with 20μM LPA for 0, 60, 120, 180 and 240 minutes. Cytoplasmic (lanes 1-6) and nuclear (lanes 7-12) proteins were separated by SDS-PAGE. Loading controls of GAPDH for cytoplasmic proteins and lamin AC for nuclear proteins are shown. Data replicated three times and representative example shown.
F. HASM cells were stimulated with 1μM methacholine for 0, 60, 120, 180 and 240 minutes. Cytoplasmic (lanes 1-6) and nuclear (lanes 7-12) proteins were separated by SDS-PAGE. Loading controls of GAPDH for cytoplasmic proteins and lamin AC for nuclear proteins are shown. Data was replicated three times and representative example shown

LPA-Induced TGF-β Activation is Mediated via αVβ5 Integrin and Involves the Cytoskeleton
A. HASMs were stimulated in duplicate with 20μM with or without an αVβ5 neutralising antibody (10μg/ml) and PAI1 mRNA levels measured. Addition of αVβ5 neutralising antibody inhibited LPA-induced PAI1 (P<0.01). Data is expressed as mean fold increase PAI1 mRNA ± SEM. Data replicated three times and representative example shown.
B. HASM were stimulated with 1μM methacholine with or without an αVβ5 neutralising antibody (10μg/ml) and PAI1 mRNA levels measured. Addition of an αVβ5 neutralising antibody inhibited methacholine-induced TGF-β activation. Data is expressed as mean fold increased PAI1 mRNA ± SEM. Data replicated three times and a representative experiment shown.
C. HASM were stimulated with 20μM LPA with or without formoterol (1μM) and PAI1 mRNA levels measured. Formoterol inhibited LPA-induced PAI1 expression (P<0.05). Data is expressed as mean fold increase PAI1 mRNA ± SEM. Data replicated three times and representative example shown.
D. HASM were stimulated with 20μM LPA with or without cytochalasin D (1μM) and PAI1 mRNA levels measured. Cytochalasin D inhibited LPA-induced PAI1 expression (P<0.01). Data is expressed as mean fold increase PAI1 mRNA ± SEM. Data replicated three times and representative example shown.

A Polymorphism in the β5 Cytoplasmic Tail Abrogates αVβ5-Mediated TGF-β Activation
A. The cell surface expression of αVβ5 by CS-1 cells transfected with full length β5 construct (FNKFNK) was assessed by flow cytometry. Figure shows the mean PE fluorescence intensity.
B. The cell surface expression of αVβ5 by CS-1 cells transfected with polymorphic β5 construct (FNK) was assessed by flow cytometry. Figure shows the mean PE fluorescence intensity
C. FNKFNK transfected CS-1 cells were co-cultured with TMLC with or without a pan TGF-β neutralising antibody. Pan TGF-β neutralising antibody inhibited basal luciferase activity (P<0.05). Data is expressed as mean RLU ± SEM. Data replicated four times and a representative experiment shown.
D. FNK transfected CS-1 cells were co-cultured with TMLC with or without a pan TGF-β neutralising antibody. Data expressed as mean RLU ± SEM. Data replicated four times and a representative experiment shown.

Talin Interacts with Both Full-Length and Polymorphic β5 Subunits
A. The interaction between talin and the β5 subunit was assessed by co-immunoprecipitation. β5 was immunoprecipitated from 500μg total cell protein and any associated talin detected by western blot. Talin migrates at 220kDa and 190kDa. 10μg input protein (prior to immunprecipitation) was used as a loading control. Data replicated three times and representative example shown.
B. The interaction between talin and the β5 subunit was assessed by co-immunoprecipitation. Talin was immunoprecipitated from 500μg protein and any associated β5 detected by western blot. β5 migrates at approximately 95kDa. 10μg input protein (prior to immunprecipitation) was used as a loading control. Data replicated three times and representative example shown..

Asthmatic HASM Activate More TGF-β than Non-asthmatic HASM Cells
A. A co-culture of either non-asthmatic or asthmatic HASM cells with TMLCs was stimulated with increasing concentrations of LPA and TGF-β activity determined. Asthmatic cells activate more TGF-β in response to LPA than non-asthmatic cells (P<0.005). Data is expressed as mean TGF-β activity ± SEM. N=3 non-asthmatic cell lines and n=3 asthmatic cell lines
B. Non-asthmatic and asthmatic HASM cells were stimulated with 20μM LPA and PAI1 mRNA levels measured by QPCR. Asthmatic HASM cells expressed higher levels of PAI1 in response to LPA compared with non-asthmatic cells (P<0.05). Data was normalised to housekeeping gene β2-M and expressed as fold increased in PAI1 over time 0 ± SEM. N=3 non-asthmatic and n=3 asthmatic cell lines.
C. A co-culture of either non-asthmatic or asthmatic HASM cells with TMLCs was stimulated with increasing concentrations of methacholine and TGF-β activity was determined. Asthmatic cells activate more TGF-β in response to methacholine than non-asthmatic cells (P<0.005). Data is expressed as mean TGF-β activity ± SEM. N=3 non-asthmatic cell lines and n=3 asthmatic cell lines
D. Non-asthmatic and asthmatic HASM cells were stimulated with 1μM methacholine for 6 hours and PAI1 mRNA levels measured by QPCR. Asthmatic cells expressed more PAI1 in response to methacholine than non-asthmatic cells (P<0.05). Data was normalised to housekeeping gene β2-M and expressed as fold increased in PAI1 over time 0 ± SEM. N=3 non-asthmatic and n=3 asthmatic cell lines.
E. Non-asthmatic and asthmatic HASM cells were stimulated with 20μM LPA fibronectin mRNA levels measured. Asthmatic cells expressed more fibronectin following LPA stimulation than non-asthmatic cells (P<0.005). Data was normalised to housekeeping gene β2-M and expressed as fold increased in fibronectin over time 0 ± SEM. N=3 non-asthmatic and n=3 asthmatic cell lines.
F. Non-asthmatic and asthmatic HASM cells were stimulated with 20μM LPA and collagen mRNA levels measured. There was no significant difference in response between asthmatic and non-asthmatic cells. Data was normalised to housekeeping gene β2-M and expressed as fold increased in fibronectin over time 0 ± SEM. N=3 non-asthmatic and n=3 asthmatic cell lines.
G. The amount of total TGF-β sequestered extracellularly basally by non-asthmatic (n=3) and asthmatic (n=3) HASM was determined by acid-treatment of cells to activate all TGF-β present. TGF-β activity was determined using TMLCs. Asthmatic HASM expressed more total TGF-β than non-asthmatic HASM (p<0.005). Data is expressed as mean TGF-β concentration per 10⁴ cells ± SEM.

The αVβ5 integrin Mediates Increased TGF-β activity in Asthmatic Cells but this is not due to Increased Cell Surface αVβ5 integrin Expression
A. Co-culture of non-asthmatic HASM cells and TMLCs was stimulated with 20μM LPA with or without 10μg/ml anti-αVβ5. Anti-αVβ5 inhibited LPA-induced TGF-β activation (P<0.05). Data is expressed as mean RLU ± SEM. Data replicated three times and representative example shown.
B. Co-culture of asthmatic HASM cells and TMLCs was stimulated with 20μM LPA with or without 10μg/ml anti-αVβ5. Anti-αVβ5 inhibited TGF-β activation (P<0.01). Data is expressed as mean RLU ± SEM. Data replicated three times and representative example shown.
C. Cell surface expression of integrin αVβ5 on non-asthmatic HASM cells was assessed by flow cytometry. Data is expressed as mean PE fluorescence intensity. Data replicated three times and representative example shown.
D. Cell surface expression of integrin αVβ5 on asthmatic HASM cells assessed by flow cytometry. Data is expressed as mean PE fluorescence intensity. Data replicated three times and representative example shown.

A. Haematoxylin and eosin stain of lung section from a saline treated animal. Magnification x20
B. Haematoxylin and eosin stain of lung section from an OVA treated animal. Magnification x20
C. Expression of αVβ5 by murine airway smooth muscle in saline treated mice was assessed using an α-smooth muscle actin (αSMA) antibody with a FITC labelled secondary and an αVβ5 antibody with a DyLight649 labelled secondary antibody. Nuclei were stained with Dapi. Images were analysed by con-focal microscopy, and imaged through a single plane to confirm the same cell co-expressing αvβ5 integrin and smooth muscle cell actin. αvβ5 integrins are localised at the end of actin fibres (arrowed).
D. Expression of αVβ5 by murine airway smooth muscle in OVA treated mice was assessed using a FITC labelled α-smooth muscle actin (αSMA) antibody and a DyLight649 labelled αVβ5 antibody. Nuclei were stained with Dapi.
E. Expression of αVβ5 by murine airway smooth muscle in OVA treated mice was assessed using a FITC labelled α-smooth muscle actin (αSMA) antibody and a DyLight649 labelled αVβ5 antibody. Cross sectional views of smooth muscle cell shows αvβ5 staining at the apices of the triangular actin skeleton (arrowed). Nuclei were stained with Dapi.
F. Activation of TGF-β in vivo in an OVA model of asthma was confirmed by staining histological sections using an αSMA antibody with a FITC labelled secondary antibody and a phospho-Smad2 antibody with a Dylight 594 labelled secondary antibody. Nuclei were stained with Dapi. Co-staining of a smooth muscle cell nuclei expressing phospho-Smad2 is arrowed. Images were analysed by con-focal microscopy, and imaged through a single plane to confirm the same cell co-expressing nuclear phospho-Smad2 and αSMA.
G. Expression of murine PAI1 protein in lung homogenates from saline (n=4) and OVA (n=4) challenged mice was determined by sandwich ELISA. PAI1 levels were higher in OVA challenged mice (P<0.05). Data is expressed as mean PAI1 concentration (ng/ml) ± SEM.

A. Histological sections from mice treated with saline and an isotype control antibody were stained with FITC labelled anti-αSMA antibody (green) and a Dapi nuclear stain.
B. Histological sections from mice treated with saline and an αVβ5 antibody were stained with FITC labelled anti-αSMA antibody (green) and a Dapi nuclear stain.
C. Histological sections from mice treated with OVA? and an isotype control antibody were stained with FITC labelled anti-αSMA antibody (green) and a Dapi nuclear stain.
D. Histological sections from mice treated with OVA and an αVβ5 antibody were stained with FITC labelled anti-αSMA antibody (green) and a Dapi nuclear stain.
E. Treatment with an anti-αVβ5 antibody reduces OVA-induced increases in ASM thickness. ASM area was quantified by measuring area of αSMA staining around airways using Nikon NIS Elements software. Data were expressed as a mean ratio of ASM area: airway radius ± SEM.
F. The number of inflammatory cells present in BALF cytospins was counted. Data shown as mean cell number ± SEM.
G. Histological sections from mice treated with OVA and an isotype control antibody were stained with haematoxylin and eosin. Magnification x20.
H. Histological sections from mice treated with OVA and an αVβ5 antibody were stained with haematoxylin and eosin. Magnification x20.

A. Histological sections from control mice treated with saline were stained with FITC labelled anti-αSMA. The sections were counterstained with a Dapi nuclear stain.
B. Histological sections from itgb5^-/- mice treated with saline were stained with FITC labelled anti-αSMA antibody. The sections were counterstained with a Dapi nuclear stain.
C. Histological sections from control mice treated with Asp.f were stained with FITC labelled anti-αSMA antibody. The sections were counterstained with a Dapi nuclear stain.
D. Histological sections from itgb5^-/- mice treated with Asp.f were stained with FITC labelled anti-αSMA antibody. The sections were counterstained with a Dapi nuclear stain.
E. ASM area was quantified by measuring area of αSMA staining around airways using Nikon NIS Elements software. Data were expressed as a mean ratio of ASM area: airway radius ± SEM. β5 knockout mice had reduced Asp.f-induced ASM compared with wild type mice (P<0.01).
F. The number of inflammatory cells present in BALF was quantified. Asp.f treatment increased the number of inflammatory cells present. BALF from β5 knockout mice had more inflammatory cells present compared with BALF from wild type mice (P<0.05). Data shown as mean cell number ± SEM.
G. Histological sections from control mice treated with Asp.f were stained with haematoxylin and eosin. Magnification x20.
H. Histological sections from itgb5^-/- mice treated with Asp.f were stained with haematoxylin and eosin. Magnification x20.