If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Pulmonary hypertension (PH) is a fatal disease characterized by remodeling processes such as increased migration and proliferation of pulmonary arterial smooth muscle cells (PASMC), enhanced matrix deposition, and dysregulation of cytoskeletal proteins. However, the contribution of cytoskeletal proteins in PH is still not fully understood. In this study, we have used a yeast two-hybrid screen to identify novel binding partners of the cytoskeletal adaptor protein four-and-a-half LIM domains 1 (Fhl-1). This identified paxillin as a new Fhl-1 interacting partner, and consequently we assessed its contribution to vascular remodeling processes. Native protein–protein binding was confirmed by co-immunoprecipitation studies in murine and human PASMC. Both proteins co-localized in PASMC in vitro and in vivo. In lung samples from idiopathic pulmonary arterial hypertension patients, paxillin expression was increased on mRNA and protein levels. Laser-microdissection of murine intrapulmonary arteries revealed elevated paxillin expression in hypoxia-induced PH. Furthermore, hypoxia-dependent upregulation of paxillin was HIF-1α dependent. Silencing of paxillin expression led to decreased PASMC adhesion, proliferation, and increased apoptosis. Regulation of these processes occurred via Akt and Erk1/2 kinases. In addition, adhesion of PASMC to the extracellular matrix protein fibronectin was critically dependent on paxillin expression. To summarize, we identified paxillin as a new regulator protein of PASMC growth.
Pulmonary hypertension (PH) is characterized by pulmonary arterial remodeling, leading to increased pulmonary vascular resistance
The remodeling process can affect all three layers of the lung vasculature. Proliferation of pulmonary arterial smooth muscle cells (PASMC) and fibroblasts as well as deposition of extracellular matrix (ECM) proteins
Pulmonary vascular remodeling occurs in response to a wide variety of stimuli, including physical (eg, mechanical stretch, shear stress) and chemical (eg, hypoxia, vasoactive substances, growth factors)
that are important sites for signal transduction. Focal adhesions are dynamic structures that alter in their size and composition during cellular processes such as adhesion, proliferation, and apoptosis, which are hallmarks of pulmonary vascular remodeling. Integrin signaling causes phosphorylation/activation of several cytoskeletal proteins, eg, the scaffold protein focal adhesion kinase (FAK),
leading to downstream signaling and finally cytoskeletal rearrangements. Many cytoskeletal proteins contain LIM domains, cystein-rich double zinc finger motifs, which act as protein-binding interfaces of transcription factors, signaling, and cytoskeletal proteins.
LIM proteins are known to play important roles in a variety of fundamental biological processes including transcriptional regulation, cytoskeletal organization, and organ development.
Furthermore, LIM domain containing cytoskeletal proteins have been suggested to play a key role in biomechanical stress responses leading to muscle disease.
Accordingly, Fhl-1 was identified as part of a complex within the cardiomyocyte sarcomere that senses biomechanical stress–induced responses leading to cardiac hypertrophy.
Although our previous study demonstrated the functional significance of Fhl-1 in PH, the molecular mechanisms by which Fhl-1 regulates the organization of the actin cytoskeleton are still not known. Therefore, we hypothesized that the identification of new Fhl-1 binding partners may provide a deeper understanding of the importance of cytoskeletal proteins in pulmonary vascular remodeling underlying PH and that these binding partners of Fhl-1 have impact on PASMC growth properties such as proliferation, apoptosis, and adhesion.
Materials and Methods
Yeast Two-Hybrid Screen
To identify novel Fhl-1 interacting proteins, a yeast two-hybrid screen was performed as previously described.
The bait construct for yeast two-hybrid screening was made by subcloning the cDNA encoding Fhl-1 (NM_010211) N- and C-terminally in the LexA domain of the vector pLexA-DIR (Dualsystems Biotech AG, Schlieren, Switzerland). The bait construct was transformed into the strain NMY32 (MATa his3Δ200 trp1-901 leu2-3,112 (lexAop)8-ADE2 LYS2::(lexAop)4-HIS3 URA3::(lexAop)8-lacZ GAL4) using standard procedures.
Correct expression of the bait was verified by Western blotting of cell extracts using a mouse monoclonal antibody directed against the LexA domain (Dualsystems Biotech AG, # P06004). The absence of self-activation was verified by co-transformation of the bait together with control preys (p53 or lamin C) and selection on minimal medium (SD) lacking the amino acids tryptophan, leucine, and histidine (selective medium). For the yeast two-hybrid screen, the bait was co-transformed together with a mouse lung cDNA library into NMY32. 6.0 × 106 (LexA-Fhl-1) and 9.5 × 106 (Fhl-1-LexA) transformants were screened, yielding 22 and 91 transformants that grew on selective medium. Positive transformants were tested for β-galactosidase activity using a PXG β-galactosidase assay (Dualsystems Biotech AG). In all, 21 of the 22 and 47 of the 91 initial positives showed β-galactosidase activity and were considered to be true positives. Library plasmids were isolated from positive clones and assayed in a bait dependency test with a control bait encoding LexA-lamin C or LexA-p53 fusion protein using a mating strategy.
Twenty of the 21 and 47 of the 91 positives showed β-galactosidase activity when co-expressed with the bait but not when co-expressed with the control bait and were considered to be bait-dependent positive interactors. The identity of positive interactors was determined by sequencing.
Cell Culture
Human primary PASMC were purchased from Lonza (Cologne, Germany) and cultured in Smooth Muscle Cell Growth Medium 2 containing supplement-mix (PromoCell, Heidelberg, Germany). Cells were incubated at 37°C in a humidified atmosphere of 5% CO2. Human PASMC between passages 5 and 9 were used for the experiments. Murine microvascular PASMC were isolated from precapillary pulmonary arterial vessels, according to a previously reported protocol.
Microvascular PASMC in passage 1 were used for experiments. When required, cell culture plates were first coated with 5 μg/cm2 fibronectin (Sigma-Aldrich, Munich, Germany) for 1 hour at room temperature before seeding human PASMC (100,000 cells/10 cm2).
Hypoxia experiments were performed in a chamber equilibrated with a water-saturated gas mixture of 1% O2, 5% CO2, and 94% N2 at 37°C. Cells were incubated in hypoxia or normoxia chambers for indicated time points and lysed in NP-40 buffer.
Co-Immunoprecipitation
Native human PASMC or mouse microvascular PASMC (550,000 cells/55 cm2) were disrupted in lysis buffer containing 20 mmol/L Tris (pH 7.6), 100 mmol/L NaCl, 2 mmol/L EDTA, 10% Glycerol, and 1% NP-40, incubated for 30 minutes on ice, and then centrifuged at 14,000 × g for 10 minutes at 4°C. Protein concentration was measured using a commercial spectrophotometric assay (BCA assay, Pierce, Rockford, MD). Equal amounts of the proteins were incubated overnight at 4°C with 2 μg of mouse anti-paxillin (Abcam, Cambridge, UK; ab3127) or goat anti–Fhl-1 antibody (ab23937; Abcam) or IgG as an isotype control (for paxillin: R&D Systems, Minneapolis, MN; MAB002; for Fhl-1: R&D Systems; AB-108-C). Samples were transferred to tubes containing protein A-Sepharose CL-4B beads (GE Healthcare, Munich, Germany). After 2 hours incubation at 4°C, the immunoprecipitates were repetitively washed with lysis buffer, boiled in SDS sample buffer, separated by SDS-PAGE under reducing conditions, and transferred to a PVDF membrane (Immobilon-P, Millipore Corporation, Bedford, MA). Immunoblots were analyzed using rabbit anti-paxillin (ab32084; Abcam) or rabbit anti–Fhl-1 antibodies (ab49241; Abcam).
Human Tissue
Tissue samples from human lungs were obtained from eight donors (mean age, 48 ± 17 years; 4 female, 4 male) and 8 IPAH patients (mean age, 34 ± 13 years; 5 female, 3 male), as described previously.
Samples were either snap-frozen in liquid nitrogen or placed in 4% (m/v) paraformaldehyde within 30 minutes after lung explantation. Only areas containing remodeled vessels were used for investigations. This selection process was performed by hemalaun and eosin staining. The study protocol was approved by the Ethics Committee of Justus-Liebig-University Giessen.
Chronic Hypoxia Exposure of Mice
C57BL/6J mice were exposed to normobaric normoxia [inspiratory O2 fraction (FiO2) 0.21] or normobaric hypoxia [FiO2 of 0.10] for 21 days. For determination of the occurrence of PH, right ventricular hypertrophy (RVH) measurements were performed as described previously.
Briefly, the right ventricle (RV) was separated from the left ventricle plus septum (LV+S). The ratio of RV to LV+S [RV/(LV+S)] was calculated as a measurement for RVH (see Supplemental Figure S1D at http://ajp.amjpathol.org). Isolation and preparation of mouse lungs was performed as described in detail previously.
All animal experiments were approved by the local authorities (Regierungspraesidium Giessen, Germany).
Laser-Assisted Microdissection
Cryosections (10 μm) from normoxic (21 days, 21% O2; 50 vessel profiles per animal; n = 5 animals) or hypoxic (21 days, 10% O2; 50 vessel profiles per animal; n = 5 animals) mouse lung tissue were mounted on glass slides and subsequently stained with hemalaun for 45 seconds (visualization of nuclei), following immersion in 70%, 96%, and 100% ethanol. Intrapulmonary arteries with a diameter of <100 μm were selected and microdissected under optical control using the Laser Microbeam System (P.A.L.M., Bernried, Germany; objective ×20 to ×40). Adjacent tissue was removed by UV-laser photolysis. Vessel profiles were isolated by a sterile syringe needle and transferred into a reaction tube containing 200 μL of RNA lysis buffer.
with an additional amplification step using the WT-Ovation Pico RNA Amplification System (Nugene, Bemmel, the Netherlands) according to the manufacturer's instructions. RNA was purified by using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. From mouse lung tissue (human and murine) and PASMC (100,000 cells/10 cm2) RNA was isolated with the RNeasy Miniprep Kit (Qiagen).
cDNA Synthesis and Real-Time PCR
For cDNA synthesis reagents and incubation steps were applied as described previously.
Real-time PCR was performed in an ABI 7900HT Sequence Detection System (Applied Biosystems, Freiburg, Germany). The PCR reactions were set up using PlatinumSYBRGreen qPCR SuperMix UDG (Invitrogen, Karlsruhe, Germany). Cycling conditions were as follow: 6 minutes at 95°C, [5 seconds at 96°C, 5 seconds at 59°C, and 10 seconds at 72°C] × 45. Because of the nonselective double strand DNA binding of the SYBRGreen I dye, melting curve analysis and gel electrophoresis were performed to confirm the exclusive amplification of the expected PCR product. The ΔCt values for each target gene were calculated by ΔCt = Cthousekeeping gene − Cttarget gene. Primer sequences are as follows: human PBGD (forward primer: 5′-CCCACGCGAATCACTCTCAT-3′; reverse primer: 5′-TGTCTGGTAACGGCAATGCG-3′); human paxillin (forward primer: 5′-TGGACAGCCCTACTGTGAAA-3′; reverse primer: 5′-AGAAGTGTTCAGGGTGCCA-3′); human fibronectin (forward primer: 5′-CCGACCAGAAGTTTGGGTTCT-3′; reverse primer: 5′-CAATGCGGTACATGACCCCT-3′); human ki67 (forward primer: 5′-GCAAGCACTTTGGAGAGC-3′; reverse primer: 5′-TCTTGACACACACATTGT-3′); mouse PBGD (forward primer: 5′-GGTACAAGGCTTTCAGCATCGC-3′; reverse primer: 5′-ATGTCCGGTAACGGCGGC-3′); mouse B2M (forward primer: 5′-AGCCCAAGACCGTCTACTGG-3′; reverse primer: 5′-TTCTTTCTGCGTGCATAAATT-3′) and mouse paxillin (forward primer: 5′-AATTCCAGTGCCTCCAACAC-3′; reverse primer: 5′-GAGCTCATGACGGTAGGTGA-3′). The primers were intron-spanning.
Transfection of Human Primary PASMC
Human PASMC (100,000 cells/10 cm2) were transfected in serum-free medium with 100 nmol/L small-interfering RNA (siRNA) using 1 μL/cm2 of X-tremeGENE siRNA Transfection Reagent (Roche, Mannheim, Germany) diluted in Opti-MEM I medium (Gibco, Darmstadt, Germany). The target sequences of siRNA were localized within the coding sequence of human paxillin (ON-TARGETplus, Thermo Scientific, Bonn, Germany), Fhl-1 (Eurogentec, Seraing, Belgium), FAK, or HIF-1α (both from BioSpring, Frankfurt am Main, Germany) sequence. Control siRNA (siR) was purchased from Ambion (Foster City, CA). The following siRNA sequence against Fhl-1: 5′-CCAGUAUUACUGCGUGGAUdTdT-3′, FAK: 5′-CUUAAAGCUCAGCUCAGCA55XUGCUGAGCUGAGCUUUAAG55-3′ or HIF-1α: 5′- UGUGAGUUCGCAUCUUGAUdTdT-3′ was applied.
Adhesion, Proliferation, and Apoptosis Assays
For determination of adhesion, human PASMC were transfected with siRNA against paxillin, Fhl-1 or a random sequence and then trypsinized and seeded 72 hours post transfection on 96-well plates coated with 2 μg/mL of bovine serum albumin (BSA, control; Sigma-Aldrich) or 2 μg/mL of fibronectin (Sigma-Aldrich). After incubation for 30 to 40 minutes at 37°C, the medium was removed and the cells were washed twice with PBS to remove nonadherent PASMC. Images were derived from adherent viable cells. Finally, absorbance was quantified after Crystal Violet staining at 550 nm in a VERSAmax lunable microplate reader (Molecular Devices, Ismaning, Germany). Proliferation of human PASMC was assessed after 72 hours post siRNA-transfection against paxillin or a random sequence by cell counting in a hemocytometer. For assessment of apoptosis, the colorimetric CaspACE Assay System (Promega, Mannheim, Germany), according to the manufacturer's instruction was applied 72 hours after siRNA-transfection against paxillin or a random sequence.
Protein Isolation and Western Blot Analysis
For protein extraction, human (donor and IPAH) and murine (normoxia and hypoxia) lung samples were disrupted by grinding in liquid nitrogen. PASMC (100,000 cells/10 cm2) were lysed with a cell scraper. A 150-μL quantity of NP-40 lysis buffer containing 2 mmol/L Na3VO4 (pH 10) and 1 × Complete (Roche) were added to PASMC or ground tissue. After incubation on ice for 30 minutes, samples were centrifuged (20,000 × g, 15 minutes, 4°C). The protein concentration in the supernatant was determined by a spectrophotometric assay (BCA assay, Pierce). A 20-μg/μL quantity of protein containing supernatant was used for Western blotting. Protein samples were run on a 12% SDS polyacrylamide gel, followed by electrotransfer to a 0.45-μm-pore PVDF membrane (Immobilon-P, Millipore Corporation). After blocking with 5% nonfat dry milk in TBS-T buffer (Tris-buffered saline with 0.1% Tween20), the membrane was incubated overnight at 4°C with one of the following antibodies: anti-paxillin (Abcam; ab32084); anti–phospho-paxillin (Tyrosine 118; Cell Signaling, Boston, MA; #2541); anti-phospho-FAK (Tyrosine 397; Tyrosine 576 both from Epitomics, Burlingame, CA; #2211 and #2103); anti-FAK (#3285); anti–phospho-Akt (Serine 473; #9272); anti-Akt (#9272); anti–phospho-Erk1/2 (Threonine 202/Tyrosine 204; #9101); anti-Erk1/2 (#9102; all raised in rabbits and 1:1000 diluted; all from Cell Signaling); and mouse anti–β-actin (A2228; Sigma-Aldrich) dilution 1:3000. After washing 3 × for 10 minutes with TBS-T buffer, the membrane was incubated for 1 hour with horseradish-peroxidase–labeled secondary antibodies (anti-rabbit W4021 and anti-mouse W4011; Promega, or anti-goat sc-2020; Santa Cruz, Heidelberg, Germany) dilution 1:5000. Afterward, the membranes were washed 3 × 15 minutes. Final detection of proteins was performed using the Amersham ECL Plus Western Blotting Detection System (GE Healthcare). Antibodies were removed by incubating the membrane for 30 minutes with stripping buffer containing 10 mL of H2O, 5 mL of 1mol/L glycine, and 750 μL of 25% HCl.
Immunohistochemistry and Immunofluorescence
Paxillin and α–smooth-muscle actin (α-SMA) staining was assessed on 3-μm paraffin-embedded lung sections. Sections were deparaffinized in xylene for 3 × 10 minutes and rehydrated in 100% ethanol for 2 × 5 minutes, 95% ethanol for 1 × 5 minutes, 70% ethanol for 1 × 5 minutes, and PBS for 2 × 5 minutes. Antigen retrieval was performed with 0.05% trypsin (Zymed Laboratories, Invitrogen) for 10 minutes at 37°C. Tissue was blocked for 1 hour with 10% BSA before incubation with rabbit anti-paxillin (1:100; Abcam; ab32084), and mouse-α-SMA (1:1000; A2547; Sigma-Aldrich) on human sections. Rabbit anti-paxillin (1:150; ab32084; Abcam) and rabbit–α-SMA (1:350; RB-9010-PO; NeoMarkers, Fremont, CA) were applied on mouse sections followed by ImmPRESS α-Rabbit Ig (peroxidase) Polymer Detection Kit or VECTOR VIP Peroxidase Substrate Kit (Vector Laboratories, Burlingame, CA). Negative control experiments were performed with the isotype control (matched host species and isotype; mouse IgG2a isotype control antibody (ab18415; Abcam) was used for mouse–α-SMA; rabbit IgG control antibody (MAB002; R&D Systems) was used for rabbit anti-paxillin and rabbit–α-SMA).
For immunocytochemistry, cells (25,000/cm2) were grown on eight-well chamber slides (BD Falcon, Heidelberg, Germany), fixed for 20 minutes in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 (Sigma-Aldrich), and rinsed 3 × in PBS. Afterward, cells were blocked for 1 hour with 5% BSA (Sigma-Aldrich) in PBS and then incubated overnight at 4°C with rabbit anti-paxillin (ab32084), goat anti–Fhl-1 (ab23937), rabbit anti–Ki-67 (1:100; ab833-500; all from Abcam), and rabbit anti-FAK (1:100; #3285; Cell Signaling) antibodies. For staining of F-actin, Alexa Fluor 568 phalloidin (1 U/mL; A12380; Invitrogen) was used. Finally, slides were washed 5 × with 0.1% BSA in PBS, incubated with corresponding fluorescent-labeled secondary antibodies (Alexa Fluor Dye 555—red, Alexa Fluor Dye 488 – green; Invitrogen; 488 rabbit A21441, 488 mouse A21200, 488 goat A21457; 555 rabbit A21429, 555 mouse A21427, 555 goat A21431) and mounted with fluorescence Vectashield mounting medium including DAPI (Vector Laboratories). Negative control experiments were performed with the isotype control (rabbit IgG control antibody; R&D Systems; MAB002) was used for rabbit anti-paxillin, rabbit anti–Ki-67, and rabbit anti-FAK antibodies; normal goat IgG (R&D Systems; AB-108-C) was used for goat anti–Fhl-1).
For immunofluorescence analysis of human or murine lung tissue, paraffin-embedded sections were deparaffinized and rehydrated as previously mentioned for immunohistochemistry. Antigen retrieval was performed with 0.05% trypsin (Zymed Laboratories, Invitrogen) for 10 minutes at 37°C. Tissue was blocked for 1 hour with 10% BSA before incubation with rabbit anti-paxillin (1:100; ab32084), goat anti–Fhl-1 (1:100; ab23937), rabbit anti-fibronectin (1:50; ab23750; all from Abcam;), or mouse α-SMA (1:100; A2547; Sigma-Aldrich) antibody for 1 hour. After washing 4 × 5 minutes each with 0.1% BSA in PBS, sections were incubated with Alexa Fluor Dye 488 and Alexa Fluor Dye 555 (488 rabbit A21441, 488 mouse A21200, 488 goat A21457; 555 rabbit A21429, 555 mouse A21427, 555 goat A21431; Invitrogen) antibodies for 1 hour. Tissue was fixed for 10 minutes with 4% paraformaldehyde and mounted with fluorescence Vectashield mounting medium including DAPI. Staining specificity was assessed via simultaneous staining of control sections with an isotype control (rabbit IgG control antibody; MAB002; R&D Systems) was used for rabbit anti-paxillin and rabbit anti-fibronectin; mouse IgG2a isotype control antibody (ab18415; Abcam) was used for mouse α-SMA; normal goat IgG (R&D Systems; AB-108-C) was used for goat anti–Fhl-1). For microscopic inspection, a Leica DMR microscope (Leica Microsystems, Wetzlar, Germany) equipped with the following filters was used: AL 380, 420 nm (DAPI); I3 420, 512 nm; and Alexa Fluor Dye 488 and N 2.1., 500 to 590 nm (Alexa Fluor Dye 555) (all from Leica).
Promoter Analysis
Sense and antisense strands of the human paxillin promoter were screened upstream and downstram of the coding sequence of the paxillin gene (NM_002859) for potential hypoxia response elements (HRE:gcgtg). The paxillin promoter sequence was obtained online (http://www.ncbi.nlm.nih.gov/mapview).
Electrophoretic Mobility Shift Assay
Electrophoretic mobility shift assays (EMSA) were performed with nuclear extracts (NE-PER Nuclear and Cytoplasmic Extraction Kit, Thermo Scientific, Rockford, IL) from human PASMC (550,000 cells/55 cm2), which had been grown for 2 hours at 1% or 21% O2, respectively. The following sense 3′-biotin–labeled or unlabeled and antisense oligonucleotide probes corresponding to the HRE consensus sequences in the human Paxillin promoter were designed: HRE1: 5′-GCTGGAAAACCACGCAATAGAGTGT-3′; HRE2: 5′-GGCGGGACCAGCGTGCGCAGGGGGC-3′; HRE3: 5′- GGCGGGGCGCGCGTGCACAGGGGGC-3′). For EMSA analysis, the LightShift Chemiluminescent EMSA Kit (Thermo Scientific), according to the manufacturer's instructions, was used.
Chromatin Immunoprecipitation
A chromatin immunoprecipitation (ChIP) Assay was performed using the Chromatin Immunoprecipitation (ChIP) Assay kit (Millipore Corporation) according to the manufacturer's protocol. Briefly, human PASMC (550,000 cells/55 cm2) grown for 2 hours at 1% O2 were cross-linked with 1% formaldehyde for 10 minutes. Cells were washed, harvested by scraping, lysed, and sonicated. Lysates were centrifuged, and an aliquot of supernatant (20 μL) was saved as input DNA. Supernatants were then immunoprecipitated with an HIF-1α (Novus Biologicals, Littleton, CO; NB100-449) or an IgG (Millipore Corporation; negative control; PP64) antibody. Immunoprecipitates were recovered by addition of Salmon Sperm DNA/Protein A Agarose-50% Slurry. After washing, elution, and reverse cross-linking, DNA was purified by phenol-chloroform extraction. Purified DNA was analyzed by conventional PCR with primers designed to amplify one single HRE. Due to the high GC content in the promoter region, HRE2 and HRE3 were amplified with the same primer pair (HRE1: Forward primer: 5′-AATAGAGGTATGTGCAAGAT-3′; Reverse primer: 5′-TTCCAATGCTCCCACCATGTTCA-3′; HRE2 and 3: Forward primer: 5′-TGCACAGACACATGGCTGAC-3′; Reverse primer: 5′-CGTCGAGGTCGTCCATCC-3′).
Statistical Analysis
Data are presented as mean ± SEM. For comparison of two groups, Student's t-test was performed. Differences between more than two groups were assessed by analysis of variance followed by Dunnett's multiple-to-one comparison post hoc tests. Adhesion values were calculated from normalized data to account for experiment-wise differences in cell numbers. Normalization was done by (xi,treated-xi,control)/(xi,control), where xi is measured values from experiment i. A P value less than 0.05 was considered significant for all analyzes.
Results
Screening for Novel Interaction Partners of Fhl-1
To identify novel binding partners of Fhl-1, a murine lung cDNA library was screened for Fhl-1 interactors using a LexA-based yeast two-hybrid system. Three independent clones of paxillin, corresponding to the LIM domain region, were recovered from the screen using both N- and C-terminal Fhl-1 baits (Figure 1A). Two control bait proteins, transcription factor p53 and lamin C, revealed no interaction with Fhl-1, indicating true binding between Fhl-1 and paxillin (Figure 1A). The binding between Fhl-1 and paxillin was confirmed by co-immunoprecipitation analysis in mouse microvascular PASMC using endogenous paxillin (Figure 1B). Species independent interaction between Fhl-1 and paxillin was demonstrated by co-immunoprecipitation in human PASMC (Figure 1C). To further examine the interaction between paxillin and Fhl-1, we analyzed their expression and localization in vitro and in vivo. Immunofluorescence analysis of mouse microvascular and human PASMC revealed strong cytoplasmic co-localization of both proteins (Figure 1D). In addition, paxillin and Fhl-1 were strongly expressed in the mouse and human pulmonary vasculature (Figure 1E). Because of the localization of paxillin in the pulmonary vasculature and the known involvement of Fhl-1 in PH, we next investigated whether paxillin was differentially regulated in human IPAH compared to their respective healthy controls.
Figure 1Paxillin, a novel interaction partner of Fhl-1 in murine and human lungs. A: Identification of paxillin by yeast two-hybrid analysis using N-terminal and C-terminal Fhl-1 clones (first panel, left and right, respectively). Both p53 or lamin C served as a specificity control. Schematic representation of paxillin (Prey) and Fhl-1 (Bait) proteins (second panel). Ade = adenine; aa =amino acids; His = histidine; Leu =leucine; SD = selective medium; Trp = tryptophan. B and C: Immunoprecipitation (IP) of endogenous paxillin (anti-Pxn) from (B) murine microvascular PASMC and (C) human PASMC. Western blots (WB) were probed with an anti–Fhl-1 or anti-paxillin (Pxn) antibody. IgG was used to exclude nonspecific interaction. Arrow indicates Fhl-1 (30 kDa) and paxillin (68 kDa); The asterisk indicates a heavy chain. D: Co-localization of paxillin and Fhl-1 in murine microvascular (left column) and human (right column) PASMC. E: Co-localization of paxillin and Fhl-1 in murine (left column) and human (right column) lungs. NC = isotype control staining.
Paxillin Expression in IPAH Patients and Influence of Fibronectin on Paxillin Signaling
Analysis of lung tissue samples from IPAH patients revealed significantly enhanced paxillin mRNA and protein expression compared to donor controls (Figure 2, A–C). Immunohistochemistry detected paxillin mainly in PASMC (Figure 2D). The increased expression of paxillin in IPAH samples and its localization in PASMC might imply a potential role for paxillin in the pathogenesis of PH.
Figure 2Paxillin expression is increased in lungs from IPAH patients and influenced by fibronectin. A: Real-time PCR analysis of paxillin expression in donor (control) and IPAH lung homogenates (n = 5 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. B: Western blot analysis of paxillin expression in donor and IPAH lung homogenates. C: Densitometric analysis of the Western blot (n = 4 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. D: Immunohistochemical staining of paxillin (in red) and α–smooth muscle actin (α-SMA; in dark red) in donor and IPAH lungs. NC = isotype control staining. E: Real-time PCR analysis of fibronectin expression in donor (control) and IPAH lung homogenates (n = 9 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. F: Co-localization of fibronectin and α–smooth muscle actin (α-SMA) in donor and IPAH lungs. Phase-contrast (PC) micrographs are shown in addition. NC = isotype control staining. G: Analysis of paxillin tyrosine (Y) 118 phosphorylation and total paxillin levels in human PASMC after seeding on fibronectin-coated or uncoated plates (with or without FN) for 40 minutes (min; n = 4). H: Quantification of Western blot analysis by densitometry (n = 4 each); protein levels were normalized to β-actin. Control was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM. I: Co-immunoprecipitation of paxillin by Fhl-1 (anti–Fhl-1) from cell lysates of human PASMC seeded on fibronectin-coated or uncoated (with or without FN) plates. Immunoprecipitates were probed with an anti-paxillin (Pxn) or anti–Fhl-1 antibody. IgG was used to exclude nonspecific interaction. Arrow indicates paxillin (68 kDa) and Fhl-1 (30 kDa). IP = immunoprecipitation; WB = Western blot.
Persistence, re-expression, and induction of pulmonary arterial fibronectin, tropoelastin, and type I procollagen mRNA expression in neonatal hypoxic pulmonary hypertension.
we next investigated the expression of fibronectin in human lungs and its impact on paxillin signaling in human PASMC. Increased fibronectin expression was detected in lung tissue from IPAH patients compared to donor controls, as assessed by real-time PCR and immunofluorescence staining (Figure 2, E and F). Culturing human PASMC on fibronectin-coated plates resulted in elevated paxillin phosphorylation (Tyr118) (Figure 2, G and H). Interestingly, fibronectin enhanced the interaction between paxillin and Fhl-1 as demonstrated by co-immunoprecipitation experiments (Figure 2I). The enhanced fibronectin expression in IPAH and its influence on paxillin phosphorylation links fibronectin to paxillin signaling pathways.
Effects of Paxillin on PASMC Adhesion
Because changes in the ECM induce alterations in the signaling and interaction of cytoskeletal proteins and thereby potentially modify cellular processes, we next examined the effect of fibronectin on human PASMC adhesion. Immunofluorescence staining of paxillin or Fhl-1 and actin filaments in human PASMC suggested that fibronectin increased cell adhesion and promoted the localization of paxillin to focal adhesion complexes (Figure 3, A, B, and H upper panel). PASMC adhesion to fibronectin was dependent on expression of both paxillin and its interaction partner Fhl-1, as demonstrated by siRNA knockdown experiments (Figure 3, C–F). Paxillin silencing disturbed the integrity of the actin cytoskeleton as shown by phalloidin staining and consequently impaired cell adhesion (Figure 3, G, and H right panel). These data demonstrated that ECM and cytoskeletal proteins are essential for cell adhesion as the absence of fibronectin, paxillin, or Fhl-1 adversely affected this process.
Figure 3PASMC adhesion is critically dependent on fibronectin, paxillin, and Fhl-1. A and B: Adhesion of human PASMC on fibronectin-coated or uncoated plates (with or without FN) assessed by immunofluorescence staining against F-actin (phalloidin) and (A) paxillin or (B) Fhl-1. C: siRNA transfection against paxillin (siPxn) or a random sequence (siR) assessed by Western blotting after 48 or 72 hours (h). D: Densitometric analysis of the Western blot analysis (n = 5). Control (siR) was set to 100%. Data were analyzed by analysis of variance followed by Dunnett's multiple-to-one comparison post hoc tests and are presented as mean ± SEM. Significant differences between control (siR) and siPxn are noted with an asterisk. E and F: Effects of paxillin (siPxn) or Fhl-1 knockdown (siFhl-1) on human PASMC adhesion determined by (E) light microscopy and (F) Crystal Violet staining (n = 6 performed in quatriplicate). Control (siR) was set to 1. Data were normalized to control and analyzed by analysis of variance followed by Dunnett's multiple-to-one comparison post hoc tests. Data are presented as mean ± SEM. Significant differences between control (siR) and siPxn or siFhl-1, respectively, are noted with an asterisk. G: Immunofluorescence staining of paxillin and phalloidin after paxillin (siPxn) or random sequence (siR) knockdown. NC = isotype control staining. H: Example of changes in paxillin distribution in human PASMC after siRNA transfection against paxillin (siPxn) or a random sequence (siR) and seeding on fibronectin-coated or uncoated plates (with or without FN) assessed by immunofluorescence staining. NC = isotype control staining.
Immunofluorescence staining of human PASMC revealed that paxillin and FAK were abundant in focal adhesion complexes (Figure 4A). Silencing of FAK in PASMC was accompanied by decreased paxillin phosphorylation (Y118) and expression (Figure 4, B and C). Consequently, silencing of paxillin led to decreased levels of FAK phosphorylation (Y397 and Y576), but did not affect its expression (Figure 4, D and E). These results indicate that both proteins are linked with each other in human PASMC.
Figure 4Paxillin and focal adhesion kinase (FAK) phosphorylation is dependent on their reciprocal expression. A: Co-localization of paxillin and FAK in focal adhesion complexes of human PASMC assessed by immunofluorescence staining. B: Representative Western blot analysis for FAK, paxillin tyrosine (Y) 118 phosphorylation and paxillin after silencing of FAK (siFAK) or a control sequence (siR) for the indicated time points. C: Densitometric quantification of the Western blot analysis at the 72-hour (h) timepoint (n = 3–7). Control (siR) was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM. D: Western blot for FAK tyrosine (Y) 397 or tyrosine (Y) 576, FAK and paxillin 72 hours after paxillin (siPxn) or a random sequence (siR) knockdown by siRNA transfection. E: Densitometric analysis of the Western blot analysis (n = 6–10). Control (siR) was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM.
Involvement of Paxillin in Proliferation and Apoptosis Processes of Human PASMC and Akt/Erk1/2 Signaling Pathways
Because of the important role of paxillin in regulating human PASMC adhesion, we next assessed the effect of paxillin knockdown on proliferation and apoptosis, critical processes that underlie the development of PH.
Paxillin silencing was associated with decreased proliferation of human PASMC as assessed by real-time PCR for Ki-67 mRNA expression, Ki-67 immunofluorescence staining and cell counting (Figure 5, A–D). In addition, knockdown of paxillin resulted in increased apoptosis as measured by a caspase-3 assay (Figure 5E).
Figure 5Paxillin knockdown hampers proliferation and enhances apoptosis. A: Real-time PCR analysis of paxillin expression after siRNA transfection against paxillin (siPxn) in comparison to control (siR) in human PASMC (n = 5 each). Control (siR) was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM. B: Real-time PCR analysis of Ki-67 expression after siRNA transfection against paxillin (siPxn) in comparison to control (siR) in human PASMC (n = 5 each). Control (siR) was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM. C: Immunofluorescence staining of Ki-67 after paxillin silencing (siPxn) in human PASMC in comparison to control (siR). As an example, arrows indicate Ki-67–positive cells. NC = isotype control staining. D: Proliferation of human PASMC assessed by cell counting after knockdown of paxillin expression (siPxn) in comparison to control (siR, n = 8 each). Control (siR) was set to 1. Data were analyzed by Student's t-test and are presented as mean ± SEM. E: Assessment of apoptosis by caspase-3 activation after paxillin knockdown (siPxn) in comparison to control (siR, n = 10 each). Control (siR) was set to 1. Data analyzed by Student′s t-test and are presented as mean ± SEM. F: Western blot for paxillin tyrosine (Y) 118 phosphorylation, Akt serine (S) 473 phosphorylation, and Erk1/2 threonine/tyrosine (T/Y) 202/204 phosphorylation after paxillin (siPxn) or random sequence (siR) silencing. G: Phosphorylation state of Akt (left panel;n = 6) and Erk 1/2 (right panel;n = 5) quantified by densitometry; protein levels were normalized to β-actin. Control (siR) was set to 100%. Data were analyzed by Student's t-test and are presented as mean ± SEM.
Protein kinases such as the phosphatidylinositol 3-kinase-Akt and the extracellular signal-regulated kinase (Erk1/2) have key roles in processes such as proliferation and apoptosis.
Thus, we investigated whether paxillin would affect the activation status of Akt and Erk1/2. Indeed, silencing of paxillin in human PASMC reduced the phosphorylation of Akt and Erk1/2 (Figure 5, F and G). These data demonstrated that changes in paxillin expression altered human PASMC homeostasis and cellular processes possibly via the Erk1/2 and Akt signaling pathways.
Paxillin Expression in the Chronic Hypoxic Mouse Model of PH
We next studied the role of paxillin in an experimental murine model of hypoxia-induced PH. Examination of lung homogenate from mice exposed to normoxia (21 days, 21% O2) or chronic hypoxia (21 days, 10% O2) did not reveal any significant changes in paxillin expression on either the mRNA or protein level (see Supplemental Figure S1, A–C at http://ajp.amjpathol.org). Because of the strong immunohistochemical staining of paxillin in SMC of the vascular wall (Figure 6A), we performed laser microdissection to specifically analyze expression in this compartment. Real-time PCR analysis of laser-microdissected intrapulmonary vessels in contrast to the investigations in lung homogenate revealed a significantly higher paxillin mRNA expression in pulmonary vessels of chronically hypoxic mice compared to the normoxic control (Figure 6B). Moreover, in contrast to the pulmonary vasculature, paxillin expression was unchanged in the systemic vasculature [aorta: Figure 6, C (aorta) and D (Arteria carotis)].
Figure 6Elevated paxillin expression in laser-microdissected intrapulmonary arteries from chronically hypoxic mice with PH. A: Localization of paxillin (in red) and α–smooth muscle actin (α-SMA; in dark red) in normoxic (N) and hypoxic (H) murine lung tissue. NC = isotype control staining. B: Paxillin mRNA expression in laser-microdissected normoxic (N) and hypoxic (H) murine lung vessels (outer diameter <100 μm, n = 5 each, 50 vessel profiles per animal) assessed by real-time PCR. Data were analyzed by Student's t-test and are presented as mean ± SEM. C and D: Paxillin mRNA expression assessed by real-time PCR in isolated (C) aortas or (D) Arteriae carotis from mice exposed to 1 day of normoxia (1dN) or to 1, 7, and 21 days of chronic hypoxia (1dH, 7dH, and 21dH; n = 6 each). Data were analyzed by analysis of variance followed by Dunnett's multiple-to-one comparison post hoc tests and are presented as mean ± SEM; ns = not significant.
To elucidate the molecular mechanisms of the transcriptional regulation of paxillin, we subjected human PASMC to hypoxia (1% O2) and normoxia (21% O2) for 24 hours. Hypoxia led to significantly elevated paxillin mRNA as well as protein expression in comparison to control (Figure 7, A–D). Of note, under hypoxic conditions, paxillin intensity increased in focal adhesions compared to control, as assessed by immunofluorescence staining. Identification of three potential hypoxia response elements (HRE) in the paxillin promoter region (Figure 7E) suggested a HIF-dependent regulation of paxillin. Indeed, silencing of the conserved transcription factor HIF-1α
diminished paxillin mRNA expression under hypoxic conditions (Figure 7F). Electrophoretic mobility shift assays (EMSA; Figure 7G, lines 2 and 4) confirmed functionality of all three HREs. Furthermore, binding to the promoter was augmented in hypoxic nuclear extracts (Figure 7G, line 4). Hypoxia-induced binding of HIF-1α to the HRE in the paxillin promoter was verified by chromatin immunoprecipitation (ChIP) (Figure 7H, line 3).
Figure 7Hypoxia-enhanced paxillin expression in PASMC is HIF-1α dependent. A and B: Paxillin expression in human PASMC exposed to normoxic (N; 21% O2) or hypoxic (H; 1% O2) conditions for 24 hours. A: Real-time PCR (n = 6 each). Data were analyzed by Student′s t-test and are presented as mean ± SEM. B: Western blot analysis. C: Densitometric analysis of the Western blot (n = 4 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. D: Paxillin expression in human PASMC after normoxic (N; 21% O2) or hypoxic (H; 1% O2) exposure assessed by immunofluorescence staining. NC = isotype control staining. E: Paxillin promoter analysis. Potential hypoxia response elements (HRE) in promoter (1500-bp) region are shown. The coding sequence of the paxillin gene is marked with + 1. F: Paxillin mRNA expression in human PASMC after siRNA transfection against HIF-1α (siHIF-1α) on hypoxic exposure (24 hours 1% O2, 24 hours H) in comparison to control (siR, n = 5 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. G: Representative electrophoretic mobility shift assays (EMSA) for each hypoxia response element (HRE) using normoxic (2 hours, 21% O2) and hypoxic (H; 2 hours, 1% O2) nuclear extracts (NE) from human PASMC and biotin-labeled or unlabeled (Cc = cold competitor) probes; n = 3 for each HRE. Arrow indicates shift complex. H: Chromatin immunoprecipitation was performed from hypoxic (2 hours, 1% O2) human PASMC using an HIF-1α (anti–HIF-1α) or an IgG antibody for precipitation. Representative agarose gel electrophoresis of PCR products amplified using specific primers for hypoxia response elements (HREs) in the paxillin promoter region are shown (n = 2–3 for each HRE). bp = Base pairs; input = input control. Arrow indicates size of product.
However, the contribution of cytoskeletal proteins to the pathogenesis of PH is poorly understood.
Previously, we identified the cytoskeletal protein Fhl-1 as one of the most highly upregulated proteins in hypoxia-induced PH in mice and human IPAH and as a key regulator of human PASMC proliferation and migration.
Moreover, Fhl-1 has been implicated in disorders such as muscle and left heart hypertrophy, by influencing cell spreading and activating of the Erk signaling cascade.
To decipher the mechanisms by which Fhl-1 affects vascular remodeling in PH, a murine lung cDNA library was screened for novel Fhl-1 binding partners using the yeast two-hybrid system. This analysis and subsequent co-immunoprecipitation studies identified the focal adhesion protein paxillin as a new binding partner of Fhl-1. Interestingly, in a previous study, we identified the cytoskeletal protein talin1, a known interaction partner of paxillin,
Similar to Fhl-1 and talin, paxillin belongs to the LIM protein family, containing four LIM domains and five leucine-rich regions termed LD motifs, which are essential for protein assembly.
Src and FAK kinases cooperate to phosphorylate paxillin kinase linker, stimulate its focal adhesion localization, and regulate cell spreading and protrusiveness.
In addition, paxillin is critically important for ontogenesis. Paxillin knockout mice exhibit embryonic lethality at E9.5 with defects in the amnion and allantois (an important site of vasculogenesis) as well as impaired growth, abnormal heart, and somite development.
As there are no paxillin inhibitors currently available for in vivo studies, investigations only on the cellular level are possible to elucidate the functional role of paxillin. In our study, silencing of paxillin expression resulted in decreased cell adhesion and proliferation, as well as increased apoptosis of human PASMC. Regulation of these processes may occur via Akt and Erk1/2 signaling, as paxillin silencing diminished phosphorylation of these two kinases. Both Akt and Erk1/2 are the converging points for several pathways leading to cell proliferation and have been implicated in progression of hyperplastic diseases such as pulmonary fibrosis and cancer,
Several studies have reported that targeting abnormal PASMC proliferation in the vascular wall attenuates development of PH in both rodents and humans.
The association of PH with the increased deposition of ECM proteins suggests that changes in the ECM could potentiate cellular signaling and thereby affect processes such as adhesion, proliferation, and apoptosis.
Persistence, re-expression, and induction of pulmonary arterial fibronectin, tropoelastin, and type I procollagen mRNA expression in neonatal hypoxic pulmonary hypertension.
This is of particular importance as cell adhesion to the ECM induces integrin clustering, formation of focal adhesion complexes, intracellular signaling, reorganization of cytoskeleton, and ultimately regulation of cellular processes.
Our in vitro experiments revealed the importance of fibronectin in enhancing all stages of this signaling cascade, resulting in increased focal adhesion complexes, paxillin-Fhl-1 interaction, and cell adhesion. In addition, we observed elevated phosphorylation of paxillin at tyrosine 118 in human PASMC cultured on fibronectin-coated plates. This may result in recruitment of other signaling molecules to focal adhesion sites that can further enhance downstream responses.
Although we have specifically examined the influence of fibronectin on paxillin signaling, it cannot be excluded that other ECM proteins, such as laminin or collagen, can also affect paxillin expression and PASMC homeostasis. The cytoskeletal disruption observed after paxillin knockdown can be attributed to the key role of paxillin in facilitating communication between integrins and actin filaments. In this regard, it was shown that such cytoskeletal alterations can be caused by loss of its interaction with proteins regulating the cytoskeletal architecture, such as the focal adhesion kinase (FAK).
Accordingly, in our study, silencing of paxillin impaired FAK phosphorylation. In addition, we observed reduced paxillin phosphorylation after FAK silencing, indicating that both proteins might regulate each other bidirectionally.
Besides the functional aspects, the possible relevance of paxillin in PH is further demonstrated by the following: its elevated expression in IPAH lungs; its presence in laser-microdissected intrapulmonary arteries from chronically hypoxic mice with PH; its predominant expression in the media of intrapulmonary vessels, major sites of vascular remodeling in PH
; and its specific regulation in the pulmonary and not in the systemic vasculature. Moreover, our detailed analysis demonstrated that paxillin was expressed in PASMC, a key site of vascular remodeling in PH, and that hypoxia-driven expression of paxillin in PASMC is controlled by HIF-1α. In this regard, it is important to mention that elevated levels of HIF-1 have been detected in arterial lesions in IPAH lungs.
The role of paxillin in PH is further supported by its interaction with other molecules, such as RhoA, which has recently been demonstrated to be hyperactive in PH.
being a further link to vascular remodeling in PH.
To summarize, we identified paxillin as a new regulator protein promoting human PASMC growth, as silencing of paxillin decreased adhesion and proliferation but increased apoptosis. Its upregulation in both human IPAH and hypoxia-induced PH in mice and its functional aspects suggest an important role in stimulating the vascular remodeling processes. The involvement of paxillin in PH pathogenesis is further supported by its interaction with Fhl-1, which was previously shown to be more abundant in PH.
Moreover, our data substantiate a critical role of cytoskeletal proteins in regulating the development of PH, with paxillin being linked with FAK and talin signaling (see Supplemental Figure S2 at http://ajp.amjpathol.org).
Acknowledgments
We thank Lisa Froehlich and Karin Quanz for technical support.
Paxillin expression in the chronic hypoxic murine model of pulmonary hypertension (PH). A and B: Paxillin expression in normoxic (N) and chronically hypoxic (H, 21 days, 10% O2) mouse lung homogenates; (A) real-time PCR (n = 5 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. B: Representative Western blot analysis. C: Densitometric analysis of the Western blots (n = 9 each). Data were analyzed by Student's t-test and are presented as mean ± SEM. D: Right to left ventricular plus septum weight ratio [RV/(LV+S)] of normoxic (N) and hypoxic (H) mice (n = 14 each). Data were analyzed by Student's t-test and are presented as mean ± SEM.
Model of paxillin signaling in PASMC, leading to vascular remodeling. The cytoskeletal signaling protein paxillin is expressed at a higher level in hypoxic human PASMC, and is positively regulated by HIF-1α. Furthermore, paxillin is phosphorylated on cell adhesion to fibronectin, enhancing the interaction between Fhl-1 and paxillin. Paxillin and FAK bidirectionally influence their phosphorylation. Activation of paxillin in turn affects cellular processes such as adhesion, proliferation, and apoptosis via the Akt and Erk1/2 signaling cascades, leading to vascular remodeling.
References
Jeffery T.K.
Wanstall J.C.
Pulmonary vascular remodeling: a target for therapeutic intervention in pulmonary hypertension.
Persistence, re-expression, and induction of pulmonary arterial fibronectin, tropoelastin, and type I procollagen mRNA expression in neonatal hypoxic pulmonary hypertension.
Src and FAK kinases cooperate to phosphorylate paxillin kinase linker, stimulate its focal adhesion localization, and regulate cell spreading and protrusiveness.
Supported by the Deutsche Forschungsgemeinschaft (WE 1978/4-1), Excellence Cluster Cardio-Pulmonary System EU FP6 “PULMOTENSION” (LSHM-CT-2005-018725) and Graduiertenstipendium of the Justus-Liebig-University Giessen, Germany.
Disclosures: Portions of the doctoral thesis of Christine Veith are integrated into this report.
00596-2&id=gr1.jpg){kind=link}
00596-2&id=gr2.jpg){kind=link}
00596-2&id=gr3.jpg){kind=link}
00596-2&id=gr4.jpg){kind=link}
00596-2&id=gr5.jpg){kind=link}
00596-2&id=gr6.jpg){kind=link}
00596-2&id=gr7.jpg){kind=link}