Articles in PresS. Am J Physiol Lung Cell Mol Physiol (April 22, 2016). doi:10.1152/ajplung.00065.2016
1 1
Surfactant dysfunction during over-expression of TGF-β1 precedes profibrotic
2
lung remodeling in vivo
3 4
Elena Lopez-Rodriguez1, Caroline Boden1, Mercedes Echaide3, Jesus Perez-Gil3, Martin Kolb4,
5
Jack Gauldie4, Ulrich A. Maus2,5, Matthias Ochs1,5,6 and Lars Knudsen1,5.
6 1
7
Institute of Functional and Applied Anatomy, Hannover Medical School, Hannover,
8 9 10
Germany 2
Department of Experimental Pneumology, Hannover Medical School, Hannover, Germany 3
Department of Biochemistry and Molecular Biology, Faculty of Biology, Universidad
11
Complutense de Madrid, Madrid, Spain 4
12 13
Firestone Institute of Respiratory Health, McMaster University, Hamilton, Canada
5
Biomedical Research in Endstage and Obstructive Lung Disease Hannover (BREATH),
14
Member of the German Center for Lung Research (DZL) 6
15
REBIRTH Cluster of Excellence
16 17 18 19 20 21 22 23 24 25 26
Corresponding Author: Lars Knudsen, Prof MD Institute of Functional and Applied Anatomy, Hannover Medical School, Carl-Neuberg-Str-1 30625 Hannover Germany Phone number: +49 511 532 2888 Fax number: +49 511 532 166741 e-mail:
[email protected]
27 28
Running title: In vivo effects of TGF-β1 in the lung
29 30 31
Keywords:
32
TGF-β1, TTF-1, AE2C polarity, surfactant, pulmonary fibrosis
33
Copyright © 2016 by the American Physiological Society.
2 34
ABSTRACT
35
Transforming growth factor beta1 (TGF-β1) is involved in regulation of cellular proliferation,
36
differentiation and fibrogenesis, inducing myofibroblast migration and increasing
37
extracellular matrix synthesis. Here, TGF-β1 effects on pulmonary structure and function
38
were analysed. Adenoviral mediated gene-transfer of TGF-β1 in mice lungs was performed
39
and evaluated by design-based stereology, invasive pulmonary function testing and detailed
40
analyses of the surfactant system one and two weeks after gene-transfer. After one week
41
decreased static compliance was linked with a dramatic alveolar derecruitment without
42
edema formation or increase in the volume of septal wall tissue or collagen fibrils.
43
Abnormally high surface tension correlated with down-regulation of surfactant proteins B
44
and C. TTF-1 expression was reduced and using PLA (proximity ligand assay) technology, we
45
found Smad3 and TTF-1 forming complexes in vivo, which are normally translocated into the
46
nucleus of the alveolar epithelial type II cells (AE2C), but in the presence of TGF-β1 remain in
47
the cytoplasm. AE2C show altered morphology, resulting in loss of total apical surface area
48
per lung and polarity. These changes of AE2C were progressive 2 weeks after gene-transfer
49
and correlated with lung compliance. While static lung compliance remained low, the
50
volume of septal wall tissue and collagen fibrils increased two weeks after gene-transfer. In
51
this animal model, the primary effect of TGF-β1 signalling in the lung is downregulation of
52
surfactant proteins, high surface tension, alveolar derecruitment and mechanical stress
53
which precede fibrotic tissue remodelling and progressive loss of AE2C polarity. Initial TTF-1
54
dysfunction is potentially linked to down-regulation of surfactant proteins.
55 56 57 58 59 60
3 61
INTRODUCTION
62
Transforming growth factor beta 1 (TGF-β1) is a cytokine with a pivotal role in development
63
of lung fibrosis (1, 47). TGF-β1 has been previously related to fibroblast migration,
64
extracellular matrix deposition and collagen accumulation (22). Molecular pathways of TGF-
65
β1 start with the activation of two receptors, followed by numerous signalling cascades, one
66
of which includes activation of members of the Smad family. Complexes of Smad2, Smad3
67
and Smad4 translocate to the nucleus and are responsible, together with other transcription
68
factors, of the expression of different genes in the cell (30, 31). Adenoviral transfection of
69
active TGF-β1 has already been studied in rats and mice, where development of a fibrotic
70
phenotype after a minor inflammatory phase showed a prolonged deposition of collagen
71
and recruitment of myofibroblast (2, 12, 40, 48). On the other hand, TGF-β1 effects have
72
also been deeply studied in vitro, especially concerning alveolar epithelial cells and
73
surfactant regulation. Surfactant proteins B (SP-B) and C (SP-C) are physiologically important
74
components of pulmonary surfactant, playing a key role on the maintenance of opened
75
alveoli, reducing surface tension and avoiding alveolar collapse (45, 62). The effect of TGF-β1
76
over-expression in the regulation of SP-B has been described in alveolar epithelial-like cells,
77
NCI H441 and A549 cells (24, 27, 32), but never before in an in vivo model. SP-B is essential
78
for surfactant activity, and SP-B conditional knock-out models showed increased surface
79
tension leading to abnormal mechanical stress in the alveolar epithelium (36). Moreover, SP-
80
B levels are reduced in patients suffering different lung diseases such as lung fibrosis (15). In
81
addition to surfactant impairment, TGF-β1 has also been directly related to mechanical
82
stress (7, 17) in the lung, and closely linked to epithelial-to-mesenchymal transition (EMT) (6,
83
21, 28, 35, 57, 59) leading to loss of epithelial integrity (with deficiencies on cadherins and
84
integrins to maintain proper cell-cell and cell-basal membrane adhesions) and remodelling of
85
the lung tissue components (such as increased deposition of collagen) (3, 11, 46). The order
86
of the events triggering and enhancing remodelling of the lung is still unclear, but there is
87
growing evidence that surfactant dysfunction and alveolar collapse may be events preceding
88
fibrotic remodelling in the lung (29). As a result alveolar epithelial cells show diverse
89
phenotypes, from cells suffering ER stress and apoptosis, through those one migrating and
90
transitioning to a wound-healing phenotype to those displaying a partial EMT-like process
91
(46). Moreover, since alveolar epithelial type II cells (AE2C) hyperplasia is a feature of
92
idiopathic pulmonary fibrosis (IPF) (9), of special interest would be to analyse both the
4 93
temporal development and the effects of TGF-β1 on AE2C in vivo. Studying early stages in
94
development of fibrotic phenotype we aim to help elucidating the mechanisms undergoing
95
changes in AE2C, surfactant function, mechanical stress and chronic lung injury, thought to
96
be one of the triggering events leading to lung fibrotic remodelling.
97
5 98
MATERIALS AND METHODS
99
Animals: 80 male C57BL/6 mice (8 weeks old; 20-25g) were obtained from Charles River
100
Laboratories GmbH (Sulzfeld, Germany). The study protocol was approved by the authorities
101
of Lower Saxony LAVES (Niedersächsisches Landesamt für Verbraucherschutz und
102
Lebensmittelsicherheit, which houses the German equivalent of an institutional animal care
103
and use committee) following European Animal Welfare Regulations (approval TVA
104
12/0799).
105 106
Adenoviral transfection and sample collection: 108 plaque forming units of AdTGFβ1223-225
107
(48) or empty control vector were instilled intra-tracheally. Mice were sacrificed 1 week and
108
2 weeks after adenoviral transfection.
109 110
Invasive pulmonary function testing: Animals were anesthetized with intraperitoneal (ip)
111
injection of 80mg/kg Ketamin (Anesketin®, Albrech GmbH, Aulendorf, Germany) and 5mg/kg
112
Xylazin (Rompun® 2%, Bayer, Leverkusen, Germany), diluted in physiological NaCl. Following
113
the cessation of pain reflexes, a cannula was surgically inserted in the trachea. The animal
114
was then connected to the FlexiVent and an additional ip injection of the same anaesthesia
115
solution was applied in order to maintain the animal deep in narcosis during the mechanical
116
ventilation. Additionally, 0.8mg/kg Pancuroniumbromid (Pancuronium-Actavis 2mg/ml,
117
Actavis, München, Germany) was ip injected to prevent spontaneous breathing. Following a
118
5-10 minutes stabilization period ventilating the animal at PEEP 3cmH2O, de-recruitability
119
tests were conducted as described before (29). First two recruitment manoeuvres (deep
120
inflation of the lung consisting of a 3 seconds increasing ramp to 30cmH2O followed by 3
121
seconds pressure hold to standardize the volume history) were applied. Then forced
122
oscillation technique (FOT), consisting of baseline ventilation for 5 minutes with 30 seconds
123
intervals of 8 seconds broadband, low-amplitude of 17 sinusoidal frequencies (0.5-19 Hz)
124
perturbations was conducted. Respiratory impedance was calculated for each frequency by
125
Fourier Transform and fit to the constant phase model to compute tissue elastance (H).
126
Tissue hysteresivity is calculated dividing tissue damping (G) (data not shown) by tissue
127
elastance (H). The Flexiware software calculates therefore tissue hysteresivity (G/H) values
128
within the 5 minutes of FOT, we calculated an average per animal and plotted the average
129
per group. Finally, 3-pressure controlled quasi-static pressure-volume (PV) loops were
6 130
performed and quasi-static compliance (Cst) was calculated. The reported H values are the
131
mean value of all the animals per group, while Cst values are mean values from 3
132
measurements in each animal. Inflation of the lung before fixation was performed by deep
133
inflation of the lung (up to 30cmH2O) and subsequently decreased to 10cmH2O, pressure at
134
which the trachea was closed. Since mechanical differences between healthy controls and
135
empty vector controls (EV control) are not recognizable, next assays were performed with
136
both healthy and EV controls and referred from now on as control group to reduce the
137
number of animal experiments.
138 139
Tissue preparation and fixation: vascular perfusion-fixation was performed right after
140
inflation of the lungs to 30 cm H2O during the deflation limb at a constant airway opening
141
pressure of 10 cm H2O using the FlexiVent. Perfusion of the lungs was carried out with 0.9%
142
NaCl and 0.25% Heparin at a constant pressure of 25cmH2O and followed by perfusion with
143
the fixation solution (1.5% glutaraldehyde, 1.5% paraformaldehyde in 0.15M Hepes buffer).
144
Then the whole lung-heart block was excised and further fixed by immersion in the same
145
fixing solution for 24h at 4°C. Afterwards all tissue not belonging to the lungs was eliminated
146
and lung volume was determined by fluid displacement. Sampling of the fixing lungs was
147
performed following Systematic Uniform Random Sampling as described before (52) and
148
lung sections were embedded in Technovite 8100 (Kulzer Hereaus, Wehrheim, Germany)
149
(43) and further stained with 0.1% toluidine or 0.2% orceine for stereological analysis at light
150
microscopy level. Lung cubes were also randomly sampled, embedded in Epon (Serva,
151
Heidelberg, Germany) and contrasted with 1% Osmium tetraoxyde (EMS, Hatfield, PA, USA)
152
and 4% uranyl acetate (Serva, Heidelberg, Germany) for stereological analysis under electron
153
microscopy (19, 34, 37). Following the same procedure, another set of lungs were
154
instillation-fixed with 10% formalin and sampled before embedding in Paraffin (Paraplast®,
155
Leica, Wetzlar, Germany) and used for immunohistochemistry and proximity ligand assay.
156 157
Histological sections and staining: Formalin-fixed paraffin-embedded tissue was stained
158
with a Masson-Goldner staining kit (Merck, Darmstadt, Germany) for visualization of
159
connective tissue (green).
160
7 161
Design-based stereology in light microscopy and electron microscopy: A list of
162
recommended and determined stereological parameter (34, 37) is given in table 1. At the
163
light microscopic level, a newCAST-system (Viopharm A/S, Horsholm, Denmark) was used to
164
perform systematic uniform random area sampling and to superimpose an appropriate test
165
system on the fields of view. Generally a point grid was used and points hitting the structure
166
of interest or the reference volume were counted so that the ratio of the counted points
167
resulted in the volume fraction of the given structure within the reference space. Since the
168
volume of the reference space (volume of the lung) is known, volume fractions were
169
converted to total volumes in all cases. At light microscopy level from 8 animals per group 6
170
to 8 sections each and at least 60 fields of view to obtain 100-200 counting events per lung
171
were analysed. At electron microscopy level a transmission electron microscope (Morgagni,
172
Eindhoven, Netherlands) equipped with a digital camera (Olympus Soft Imaging Systems,
173
Münster, Germany) was used to obtain representatives fields of view from 5 animals per
174
group. 5 to 7 sections per lung and at least 60 fields of view were analysed to obtain 100-200
175
counting events per lung, providing appropriate precision of the stereological parameters.
176
For superimposing a test system STEPanizer© stereology tool (53) was used.
177 178
Western-Blot: for BAL sample preparation (n=5 per group), equal volumes of BAL samples
179
were taken and mixed with loading buffer (2% SDS, 10% glycerol, 0.1% bromophenol blue,
180
10% β-mercaptoethanol in 12.5 µ Tris-HCl (pH 6.8)) and boiled for 10 min. 20µl of BAL
181
samples were loaded in the 16% polyacrylamide (PA) electrophoresis gel for detecting SP-B
182
and SP-C and 10% PA gels to detect proSP-B and proSP-C. Proteins were transferred into
183
Immun-Blot PVDF membrane (BioRad Laboratories, München, Germany) and following 2h of
184
blocking (TBS-0.1% Tween 20, 5% non-fat milk, pH 7.4) at room temperature (RT), primary
185
antibodies were incubated over night at 4°C. Table 2 shows the list of primary antibodies
186
used for western-blot from BAL samples.
187
After washing with TBS-T (TBS-0.1% Tween 20), secondary antibody (1:2000, polyclonal
188
swine anti-rabbit, DAKO, Glostrup, Denmark) was incubated for 1h at RT. Blot membranes
189
were developed with ECL Prime Western blotting detection reagent (Amersham Bioscience,
190
Buckinghamshire, UK) and band density of exposed films (Amersham Bioscience,
191
Buckinghamshire, UK) was analysed by densitometry using Image J (Rasband, W.S., ImageJ,
192
U. S. National Institutes of Health, Bethesda, Maryland, USA).
8 193
For lung homogenate (LH) samples (n=5 per group), little pieces of lung tissue (approx.
194
15mg) were mixed with lysis buffer (50mM Tris-HCl pH 7.5, 150mM NaCl, 1% Triton, 0.5%
195
Na-deoxycholate, 5mM EDTA and protease inhibitor cocktail (complete mini, Roche,
196
Mannheim, Germany). A TissueLyser II (Qiagen, Hilden, Germany) was used to disrupt the
197
tissue. Soluble proteins were isolated from the supernatant of the disrupted tissue after
198
centrifugation (10000 xg, 10min, 4°C). Total protein content was determined using BCA
199
protein assay (Thermo Scientific, Illinois, USA) and total 20 µg of proteins per well in loading
200
buffer were loaded in 12% PA gels. Proteins were transferred into Immun-Blot PVDF
201
membrane (BioRad Laboratories, München, Germany) and following 2h of blocking (TBS-
202
0.1% Tween 20, 5% bovine serum albumin (BSA), pH 7.4) at room temperature (RT), primary
203
antibodies were incubated over night at 4°C. A list of primary antibodies is shown in table 3.
204
After washing with TBS-T (TBS-0.1% Tween 20), secondary antibodies (1:2000, polyclonal
205
swine anti-rabbit-HRP or 1:1000 polyclonal rabbit anti-goat-HRP, DAKO, Glostrup, Denmark)
206
were incubated for 1h at RT. Blot membranes were developed with ECL Prime Western
207
blotting detection reagent (Amersham Bioscience, Buckinghamshire, UK) and band density
208
of exposed high performance ECL X-ray films (Amersham Bioscience, Buckinghamshire, UK)
209
was analysed by densitometry using Image J (Rasband, W.S., ImageJ, U. S. National Institutes
210
of Health, Bethesda, Maryland, USA).
211 212
RT-PCR: Total cell RNA was isolated from frozen lung tissue using ISOLATE RNA Mini Kit
213
(Bioline GmbH, Berlin, Germany) and quantified by spectrometry at 260 nm. Reverse
214
Transcription was performed using 1 µg of total RNA and the iScript™ cDNA Synthesis Kit
215
(BioRad Laboratories, München, Germany) in a thermocycler (MasterCycler, Eppendorf,
216
Hamburg, Germany). Real Time PCR was performed using iQ™ SYBR® Green Supermix
217
(BioRad Laboratories, München, Germany) at a final primer concentration of 100 nm and
218
50ng cDNA. Antibody-mediated hot-start iTaq™ DNA polymerase was activated after an
219
initial 3 minutes denaturation step at 95°C. Denaturation of DNA was performed by holding
220
further 95°C for 10 seconds. Annealing and extension temperature was set to 59°C (30
221
seconds) for 40 cycles in a BioRad CFX96 RT-PCR system (BioRad Laboratories, München,
222
Germany). Primer design was performed using the GeneFisher software from the University
223
Bielefeld (Bielefeld, Germany) (13) and primer sequences are shown in table 4. RT-PCR data
224
is represented as normalized expression (∆∆Cq) which corresponds to the relative quantity
9 225
of the target gene normalized to the quantity of the reference gene (GAPDH). Bars
226
correspond to the mean and standard error of n=6 animals per group.
227 228
Captive Bubble Surfactometer (CBS): surfactant membranes
229
samples and were assayed at 20mg/ml phospholipid concentration in a CBS as described
230
before (44).
were isolated from BAL
231 232
Proximity Ligand Assay (PLA) and quantification: allow us to detect and quantify protein
233
interactions using two specific primary antibodies for the proteins of interest and two
234
different secondary antibodies which contain unique DNA strands, which are ligated, when
235
closed enough (