O

Sec

XSa

b

c

d

e

f

g

h

i

j

a

ARRAA

KSTF

1

postd

M

h1

Prostaglandins & other Lipid Mediators 108 (2014) 23–30

Contents lists available at ScienceDirect

Prostaglandins and Other Lipid Mediators

riginal Research Article

phingosylphosphorylcholine induces �-smooth muscle actinxpression in human lung fibroblasts and fibroblast-mediated gelontraction via S1P2 receptor and Rho/Rho-kinase pathway

.Q. Wanga,b, L.J. Maoc, Q.H. Fangd, T. Kobayashie, H.J. Kimf, H. Sugiurag,. Kawasakih, S. Togoi, K. Kamioj, X. Liua, S.I. Rennarda,∗

Pulmonary, Critical Care, Sleep and Allergy, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE, United StatesDepartment of Respiratory Disease, Affiliated Hospital of Hebei United University, Hebei Province, ChinaResearch Center of Occupational Medicine, Peking University Third Hospital, Beijing 100191, ChinaDepartment of Pulmonary and Critical Care, Beijing Shijitan Hospital, Capital Medical University, Beijing 100038, ChinaDepartment of Pulmonary and Critical Care Medicine, Mie University School of Medicine, Tsu, JapanDepartment of Internal Medicine, SanBon Hospital, WonKuang University School of Medicine, Seoul, South KoreaDepartment of Respiratory Medicine, Tohoku University Graduate School of Medicine, 1-1 Seiryo-machi, Aoba-ku, Sendai 980-8574, JapanDepartment of Respiratory Medicine, University of Tokyo Graduate School of Medicine, Tokyo, JapanDivision of Respiratory Medicine, Juntendo University Faculty of Medicine & Graduate School of Medicine, Tokyo, JapanDepartment of Pulmonary Medicine/Infection and Oncology, Internal Medicine, Nippon Medical School, Tokyo, Japan

r t i c l e i n f o

rticle history:eceived 18 September 2013eceived in revised form 21 February 2014ccepted 25 February 2014vailable online 5 March 2014

eywords:PCissue repair

a b s t r a c t

Chronic airway diseases like COPD and asthma are usually accompanied with airway fibrosis. Myofibro-blasts, which are characterized by expression of smooth muscle actin (�-SMA), play an important role ina variety of developmental and pathological processes, including fibrosis and wound healing. Sphingo-sylphosphorylcholine (SPC), a sphingolipid metabolite, has been implicated in many physiological andpathological conditions. The current study tested the hypothesis that SPC may modulate tissue remodel-ing by affecting the expression of �-SMA in human fetal lung fibroblast (HFL-1) and fibroblast mediatedgel contraction. The results show that SPC stimulates �-SMA expression in HFL-1 and augments HFL-1mediated collagen gel contraction in a time- and concentration-dependent manner. The �-SMA pro-

ibroblast tein expression and fibroblast gel contraction induced by SPC was not blocked by TGF-�1 neutralizingantibody. However, it was significantly blocked by S1P2 receptor antagonist JTE-013, the Rho-specificinhibitor C3 exoenzyme, and a Rho-kinase inhibitor Y-27632. These findings suggest that SPC stimulates�-SMA protein expression and HFL-1 mediated collagen gel contraction via S1P2 receptor and Rho/Rhokinase pathway, and by which mechanism, SPC may be involved in lung tissue remodeling.

© 2014 Published by Elsevier Inc.

. Introduction

Chronic airway diseases, such as asthma and chronic obstructiveulmonary disease (COPD) are characterized by fibrotic scarringf airways [1,2]. Myofibroblasts, distinguished by the expres-ion of smooth muscle actin (�-SMA), have widely been thought

o contribute to these processes [3,4]. �-SMA has already beenemonstrated to be related to fibroblast contraction and tissue

∗ Corresponding author at: University of Nebraska Medical, 985910 Nebraskaedical Center, Omaha, NE 68198-5910, United States. Tel.: +1 4025597313.

E-mail address: srennard@unmc.edu (S.I. Rennard).

ttp://dx.doi.org/10.1016/j.prostaglandins.2014.02.002098-8823/© 2014 Published by Elsevier Inc.

remodeling which are involved in normal wound healing and fibro-sis [4–6].

TGF-�1 induces fibroblast transition via stimulation of �-SMAexpression in fibroblasts which is thought to be mediated by Smad3protein [5,6]. Many other reagents, like thrombin [7,8], IL-4, and IL-13 [9] also can stimulate �-SMA expression, which is mediated bya different mechanism.

Sphingosylphosphorylcholine (SPC) has been recognized as animportant signal molecule and is believed to play an important rolein physiological and pathophysiological conditions. It is involved

in many biological processes including smooth muscle contrac-tion, wound healing and angiogenesis [10,11]. It has also beendemonstrated that SPC accelerates the healing of cutaneous woundin healing-impaired diabetic mice [12] and stimulates dermal

24 X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30

Fig. 1. Effect of pretreatment of SPC on fibroblast-mediated collagen gel contraction and �-SMA expression. Panel A: Collagen gel contraction. HFL-1 cells were pretreated withvarying concentrations of SPC for 48 h. Cells were then trypsinized and cast into type I collagen gels. Gel size was measured with an image analyzer daily for 4 days. Verticalaxis: gel size expressed as percent of initial size (%); horizontal axis: time (day). *p < 0.05 compared to SF-DMEM (control) on each day. Panel B and C: SPC concentration-dependent effect on �-SMA expression. HFL-1 cells were serum-deprived for 24 h, and then treated with various concentrations of SPC for 48 h. Cell lysate (5 �g/lane) wereseparated by SDS–polyacrylamide gel electrophoresis and immunoblotted with anti-�-SMA and anti-ß-actin antibodies. Panel B: One representative of immunoblotting.Panel C: Densitometric quantification of 3 separate experiments. *p < 0.05 compared to control. Panel D and E: Time-dependent effect of SPC on �-SMA expression. HFL-1c ) for

r ate ex

fiamidssbAt

2

2

tcBMcGGA

ells were serum-deprived for 24 h, and then treated with or without SPC (10−5 Mepresentative of immunoblotting. Panel E: Densitometric quantification of 3 separ

broblast mediated gel contraction [13]. However, the mech-nism is not fully defined. Thus, we hypothesized that SPCight also be involved in airway tissue remodeling by affect-

ng the phenotype of fibroblasts through specific receptor andown stream signaling pathways. To demonstrate this hypothe-is, we have conducted serial experiments to determine if SPCtimulates �-SMA expression in HFL-1 and if SPC pretreated fibro-last can augment fibroblast-mediated collagen gel contraction.dditionally, this study seeks to explore its pathway of signal

ransduction.

. Materials and methods

.1. Cell culture

Human fetal lung fibroblasts (HFL-1) were purchased fromhe American Type Culture Collection (Rockville, Maryland). Theells were cultured in 100 mm tissue culture dishes (FALCON,ecton-Dickinson Labware, Lincoln Park, New Jersey) in Dulbecco’sodified Eagle’s Medium (DMEM) supplemented with 10% fetal

alf serum (FCS, Biofluid, Rockville, Maryland), 50 U/ml penicillin sodium, 50 �g/ml streptomycin sulfate (Penicillin-streptomycin,IBCO, Life Technologies, Grand Island, New York) and 1 �g/mlmphotercin B (Pharma-Tek, Hutington, New York). Cells were

8–48 h. Cell lysate was used for immunobloting �-SMA and ß-actin. Panel D: Oneperiments. *p < 0.05.

cultured at 37 ◦C in a humidified atmosphere of 5% CO2/95% air.Passages 14–19 were used in all experiments.

2.2. Materials

Sphingosylphosphorylcholine (SPC), pertussis toxin, mono-clonal �-SMA-antibody and monoclonal �-actin antibody werepurchased from Sigma (St. Louis, MO). Recombinant TGF-�1 andits neutralizing anti-TGF�1 antibody were from R&D Systems Inc.(Minneapolis, MN). JTE-013 and W146 were from Cayman chemi-cal (Ann Arbor, MI). C3 exoenzyme was provided by courtesy of Dr.Myron Toews. Y27632 were obtained from Biochem (La Jolla, CA).Anti-phospho-smad3 and anti-Smad3 antibodies were obtainedfrom Cell signaling Technology (Beverly, MA).

2.3. Collagen gel contraction assay

Collagen gels were prepared by mixing rat tail tendon collagen(RTTC), distilled water, 4×-concentrated DMEM and cell suspen-sion so that the final mixture resulted in 0.75 mg/ml collagen,

3 × 105 cells/ml and 1× DMEM. One-half milliliter of the colla-gen/cell mixture was plated into each well of a 24-well tissueculture plate. Gels were allowed to solidify at room temperature,usually within 15 min. Gels were then released into a 60-mm tissue

X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30 25

Fig. 2. Role of TGF-ß1 in mediating SPC effect on fibroblast-mediated collagen gel contraction and �-SMA expression. Panel A: effect of anti-TGF-ß1 antibody on SPC-augmented collagen gel contraction. HFL-1 cells were pretreated with or without TGF-1 (10−10 M) or SPC (10−6 M) in the presence or absence of anti-TGF-ß1 antibody(100 �g/ml) for 48 h. Cells were then trypsinized and cast into the type I collagen gel. After polymerization, gels were released into serum free DMEM. Gel size was measuredon day 1 with an image analyzer. Vertical axis: gel size expressed as percent of initial size (%). Horizontal axis: pre-treatment group. Data are shown as meansSEM. * p < 0.05.Panel B: Effect of SPC on Smad3 phosphorylation and nuclei translocation. HFL-1 cells were serum-deprived for 24 h, and then treated with TGF-ß1 (10−10 M) for 30 minor 10−5 M SPC as indicated. Nuclear proteins were separated by SDS–polyacrylamide gel electrophoresis and immunoblotted with anti-phospho-Smad3 and anti-Smad3antibodies. Data presented is one representative of 3 separate experiments. Panel C and D: effect of TGF-ß1 neutralizing antibody on SPC-induced �-SMA expression. HFL-1c esencew th antP

ct

2

Ffal1altCcrefbasAe

ells were serum-deprived for 24 h, and then treated with TGF-ß1 or SPC in the prere separated by SDS–polyacrylamide gel electrophoresis and immunoblotted wi

anel D: densitometric quantification of 3 separate experiments. *p < 0.05.

ulture dish containing 5 ml of serum-free DMEM and the variousreatment agents.

.4. Immunoblotting

HFL-1 cells were cultured in 100 or 60 mm dishes with 10%CS-DMEM until sub-confluent. Cells were then serum starvedor 24 h followed by treatment with or without various reagentss indicated. After washing with PBS, cells were harvested withysing buffer containing 35 mM Tris–HCl, PH 7.4, 0.4 mM EGTA,0 mM MgCl2, 0.1% Triton X-100 and protease inhibitors (3 �g/mLprotinin, 1 �M phenylmethanesulphonylfluoride and 10 �g/mLeupeptin). For Smad3 nuclear translocation assay, nuclear pro-eins were prepared with nuclear extraction kit (Active Motif,arlsbad, CA) following the manufacture’s instruction. Proteinoncentrations were determined by the Bradford method (Bio-ad laboratories Hercules, CA). Equal amounts of protein werelectrophoresed on 10% or 15% SDS-polyacrylamide gels and trans-erred onto polyvinylidene difluoride (PVDF) membranes. Afterlocking with 5% milk, the membrane was incubated with primary

ntibodies overnight at 4 ◦C. Horseradish peroxidase conjugatedecondary antibody was applied for 1 h at room temperature.fter extensive washes, immunoreactive bands were visualized bynhanced chemiluminescence reaction.

or absence of neutralizing anti-TGF-ß1 antibody for 48 h. Cell lysate (10 �g/lane)i-�-SMA and anti-ß-actin antibodies. Panel C: one representative immunoblotting.

2.5. GST-Rhotekin pull-down assay

HFL-1 cells were grown to 90% confluence on 100 mm dishes,starved for serum in DMEM for 24 h, and stimulated with SPC andother agents as indicated. Cells were then harvested with ice-coldlysis buffer provided by the manufacture. RhoA pull-down assaywas performed following the manufacture’s instruction (Millipore,Billerica, MA). Briefly, cell lysates were incubated with GST fusion ofthe RhoA effector rhotekin beads for 45 min at 4◦C. The GST beadswere washed three times with lysis buffer, and then loaded into10% SDS–PAGE gels for electrophoresis. After protein transfered toPVDF membrane, bound RhoA was detected with anti-RhoA mono-clonal antibody. Following incubation with HRP-conjugated 2ndantibody, GTP bound RhoA as well as total RhoA were visualized byenhanced chemilunescence reaction.

3. Results

3.1. SPC stimulates collagen gel contraction and smooth muscleactin (˛-SMA) synthesis by HFL-1 cells

In the current study, we have examined the effect of sphin-gomyelin metabolites (sphingosylphosphorylcholine, ceramide,sphingosine, and sphingosine 1-phosphate) on fibroblast-mediatedcollagen gel contraction. Both sphingosylphosphorylcholine (SPC)

26 X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30

Fig. 3. Role of S1P2 receptor in mediating SPC effect on fibroblasts. Panel A: effect of W-146 (S1P1 specific antagonist), JTE013 (S1P2 specific antagonist), and Suramin (S1P3

specific antagonist) on fibroblast-mediated gel contraction. HFL-1 cells were pre-treated with SPC (10−6 M) or TGF- (10−10 M) in the presence or absence of W-146 (10−6 M),JTE013 (10−6 M) and Suramin (10−6 M) for 48 h. Cells were then cast into type I collagen gels. After polymeriztion, gels were released and transferred to 60 mm tissue culturedishes containing 5 ml of serum free DMEM. Gel size was measured with an image analyzer. Vertical axis: gel size on day 2 expressed as % of initial size. Horizontal axis:pre-treatment. Data are shown as meansSEM. **p < 0.01. Data presented is one representative of 3 separate experiments. Panel B and C: effect of JTE013 (S1P2 specificantagonist) on �-SMA expression. HFL-1 cells were serum-deprived for 24 h, pre-treated with or without SPC 10−6 M in the presence or absence of JTE-013 (10−6 M) for4 lectropr arate

atcvam

lc(icotpmotU�ct

3c

bS

8 h. Cell lysate proteins (5 �g/lane) were separated by SDS–polyacrylamide gel eepresentative immunoblotting data. Panel C: densitometric quantification of 3 sep

nd sphingosine 1-phosphate (S1P) augmented collagen gel con-raction, while ceramide and sphingosine had minimal effect onollagen gel contraction (data not shown). While we have pre-iously reported that S1P augmented lung fibroblast migrationnd collagen gel contraction [14], the effect of SPC on fibroblast-ediated collagen gel contraction has not been reported.To further evaluate effects of SPC on fibroblast-mediated col-

agen gel contraction, HFL-1 cells were pretreated with differentoncentration (10−7–10−5 M) of SPC for 48 h in serum-free mediumSF-DMEM). Cells were then cast into type I collagen gels and floatedn SF-DMEM. During the 4 days of contraction, SPC at 10−5 M signifi-antly augmented collagen gel contraction from day1, while 10−6 Mn day 2 and 10−7 M on day3 significantly augmented contrac-ion (Fig. 1A, p < 0.05 compared to control at corresponding timeoint). Since �-smooth muscle actin (�-SMA) is involved in aug-enting fibroblast-mediated collagen gel contraction [6], effects

f SPC on �-SMA synthesis was examined. Under the normal cul-ure condition, �-SMA expression by HFL-1 cells was detectable.pon stimulation with varying concentrations of SPC, however,-SMA synthesis was significantly up-regulated by SPC in a con-entration dependent manner (10−8–10−5 M, Fig. 1B and C) andime-dependent manner (8–48 h, Fig. 1D and E).

.2. TGF-ß1 does not mediate SPC-augmented collagen gelontraction and ˛-SMA synthesis

TGF-ß1 is known to stimulate not only collagen gel contraction,ut also �-SMA expression by fibroblasts. Thus, it is plausible thatPC stimulates fibroblast-mediated collagen gel contraction and

horesis and immunoblotted with anti-SMA or anti-actin antibodies. Panel B: oneexperiments.* p < 0.05.

�-SMA expression indirectly through stimulating TGF-ß1. There-fore, the role of TGF-ß1 was evaluated by the following differentmethods. First, TGF-ß1 production by HFL-1 cells stimulated withSPC was quantified by ELISA. However, SPC had no effect on TGF-ß1 synthesis by HFL-1 cells (data not shown). Second, anti-TGF-ß1neutralizing antibody was applied to SPC-stimulated collagen gelcontraction and �-SMA expression systems. As expected, neitherSPC-augmented collagen gel contraction (Fig. 2A) nor SPC-induced�-SMA expression (Fig. 2C and D) was affected by anti-TGF-ß1neutralizing antibody while this antibody could neutralize TGF-ß1effect on collagen gel contraction and �-SMA expression (Fig. 2).Third, effect of SPC on Smad3 phosphorylation and transloca-tion was examined. While TGF-ß1 (5 ng/ml = 200 pM) significantlyinduced Smad3 phosphorylation in the cell nuclei, SPC at highestconcentration (10−5 M) for up to one-hour stimulation had no effecton Smad3 phosphorylation and translocation (Fig. 2B).

3.3. S1P2 receptor mediates SPC-augmented collagen gelcontraction and ˛-SMA synthesis

While SPC-specific receptor has not been clearly defined, studieshave indicated that sphingosine-1-phosphate receptors 1–3 (S1P1,S1P2, S1P3) might be involved in mediating SPC effect in certain celltypes [15]. Thus, the effects of a specific antagonist of sphingosine-1-phoshpate receptor W-146 (S1P1), JTE-013 (S1P2), and Suramin

(S1P3) on SPC stimulatory effect were studied in the current study.

Neither S1P1 antagonist W-146 nor S1P3 antagonist Suraminblocks HFL-mediated gel contraction induced by SPC (Fig. 3A). Inter-estingly, JTE-013 not only significantly blocked SPC-augmented

X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30 27

Fig. 4. Effect of SPC on RhoA activity. Panel A and B: concentration-dependent effect. HFL-1 cells were serum-deprived for 24 h, and then treated with various concentrations ofSPC for 5 min. Activated Rho was isolated using the pull-down assay kit as described in Section 2. Activated Rho and total Rho proteins were separated by SDS–polyacrylamidegel electrophoresis and immunoblotted with anti-Rho antibodies. Panel A: one representative immunoblotting. Panel B: densitometric quantification of 3 separate experi-ments. * p < 0.05 compared to control. Panel C and D: effect of JTE013 on SPC-induced RhoA activation. HFL-1 cells were serum-deprived for 24 h, pre-treated with or withoutS was it immuP

cST(

3

src

di15IcbD

astrCac

PC (10−6 M) in presence or absence of JTE-013 (10−6 M) for 5 min. Activated RhoAotal Rho proteins were separated by SDS–polyacrylamide gel electrophoresis andanel D: densitometric qauntification of 3 separate experiments. *p < 0.05.

ollagen gel contraction (Fig. 3A), but also inhibited SPC-induced �-MA expression (Fig. 3B and C). In contrast, JTE-013 had no effect onGF-ß1-augmented collagen gel contraction or �-SMA expressionFig. 3).

.4. RhoA mediates SPC effect on HFL-1 cells

RhoA is known to increase stress fiber formation and has beenhown to increase �-SMA expression in muscle cells. Therefore, theole of Rho signaling pathway in SPC stimulatory effect on HFL-1ells was investigated with the following strategies.

First, activation of RhoA by SPC was examined using the pull-own assay. RhoA activity was very rapidly and significantly

ncreased in a SPC-concentration dependent manner, that is,0−6 M or higher had a dramatic effect on RhoA activation with

min exposure while 10−7 M SPC had little effect (Fig. 4A and B).n addition, consistent with the above observation on collagen gelontraction and �-SMA expression (Fig. 3A and B), RhoA activationy SPC (10−6 M) was completely blocked by JTE-013 (Fig. 4C and).

Second, having demonstrated that SPC could stimulate RhoActivation, we studied the possibility of RhoA mediating SPCtimulatory effect on �-SMA protein expression. To accomplishhis, C3 exoenzyme, which specifically inactivates RhoA by ADP-

ibosylation, was used to block RhoA activation by SPC. As expected,3 exoenzyme completely blocked not only RhoA activation (Fig. 5And B), but also �-SMA synthesis induced by SPC (Fig. 5C and D). Inontrast, pertussis toxin (PTX, 10 ng/mL), a Gi/o specific inhibitor,

solated using the pull-down assay kit as described in Section 2. Activated Rho andnoblotted with anti-Rho antibodies. Panel C: one representative immunoblotting.

did not inhibit SPC-induced RhoA activation (Fig. 5A and B) and�-SMA synthesis (Fig. 5C and D).

Finally, the Rho kinase, which is down stream of RhoA and playsa major role in cytoskeleton reorganization and stress fiber forma-tion, was studied using a Rho kinase inhibitor, Y27632. As expected,Y27632 significantly blocked SPC-stimulated collagen gel contrac-tion (Fig. 6A) and �-SMA expression (Fig. 6B and C).

4. Discussion

Chronic inflammatory airway diseases that cause airway remod-eling [1] such as COPD and asthma are associated with fibroblasts tomyofibroblasts transition and increased fibroblast contractile activ-ity. This transition is marked by increased expression of smoothmuscle actin (�-SMA) in fibroblasts. Furthermore, �-SMA proteinhas been demonstrated to be involved in fibroblast contractileactivity in other labs and our lab [3,4,6]. TGF-�1 is well known forits ability to stimulate transition of fibroblast to myofibroblast [5].A number of other cytokines such as IL-4, IL-13 [9], and throm-bin [8] are also involved and can induce fibroblast-mediated gelcontraction.

SPC is an important sphingomyelin metabolite that plays a rolein many cellular processes, and it has been implicated in a number

of biological processes including smooth muscle contraction andwound healing [10,11]. We have demonstrated that SPC can stim-ulate HFL-1 cells mediated gel contraction, and pretreatment withdifferent concentration of SPC also augments fibroblast-mediated

28 X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30

Fig. 5. Effect of C3 exoenzyme on SPC-induced Rho activation and �-SMA synthesis. Panel A and B: effect on Rho activiation. HFL-1 cells were treated with or without C3exoenzyme (10 �g) as described in the methods. Cells were then serum-deprived followed by treatment with SPC (10−5 M). For comparison, cells were also treated with orwithout pertussis toxin (PTX, 10 ng/ml). Activated Rho was isolated using the pull-down assay kit as described in Section 2. Activated Rho and total protein were separated byS dies.

a ll lysaa ne reps

gwestit

eatentscihlapfitaSstsfg

DS–polyacrylamide gel electrophoresis and immunoblotted with anti-RhoA antibos described above followed by treatment with or without SPC (10−6 M) for 48 h. Cend immunoblotted with anti-�-SMA and anti-�-actin antibodies. Panel A and C: oeparate experiments. *p < 0.05.

el contraction in a dose- and time-dependent manner. Therefore,e speculate that SPC may also stimulate �-SMA expression. As we

xpected, the data in this study confirmed that SPC, like TGF-�1, cantimulate �-SMA protein expression in HFL-1 cells. We also foundhat SPC stimulated �-SMA expression by immunochemistry stain-ng (data not shown). These results proved that SPC can stimulatehe transition of fibroblast to myofibroblast.

In order to test the idea that SPC stimulates �-SMA proteinxpression by a TGF-�1 pathway, we used TGF-�1 neutralizingntibodies to treat the cells. The result showed that TGF-�1 neu-ralizing antibodies did not block SPC-induced �-SMA proteinxpression. We also found that TGF-�1 neutralizing antibodies doot block SPC-induced gel contraction in cells pretreated with neu-ralizing TGF-�1 antibodies and SPC. These results suggest that SPCtimulated �-SMA protein expression and fibroblast-mediated gelontraction were independent of the TGF-�1 release. Smad3 is anmportant signal protein for TGF-�1, and our lab as well as othersave already demonstrated that smad3 mediates TGF-�1 stimu-

ation of �-SMA expression [4,6], and that �-SMA is required tougment fibroblast-mediated collagen gel contraction in that sup-ression of �-SMA by siRNA resulted in significant inhibition ofbroblast-mediated collagen gel contraction [6]. Recent publica-ions suggest that sphingosine 1-phosphate (S1P), which shares

similar structure with SPC, is capable of a cross-activation ofmad3 [16,17]. Thus we considered the possibility that SPC maytimulate Smad3 phosphorylation by affecting fibroblast pheno-

ype. Using TGF-�1 as a positive control, we found that SPC neithertimulates Smad3 phosphorylation nor stimulates translocationrom cytoplasm to nucleus compared to TGF-�1. This result sug-ests that SPC stimulation of �-SMA protein expression may not be

Panel C and D: effect on �-SMA expression. Cells were treated with C3 exoenzymete proteins (5 �g/lane) were separated by SDS–polyacrylamide gel electrophoresisresentative immunoblotting data. Panel B and D: densitometric quantification of 3

mediated by TGF-�1 and its signal protein Smad3. Although it hasbeen reported that cytokines such as IL-4, IL-13 [9], and thrombin[8] are also able to induce �-SMA, it is unlikely that IL-4 or IL-13mediates SPC-induced �-SMA expression in HFL-1 cells in that HFL-1 cells do not synthesize and release IL-4 or IL-13 (unpublisheddata).

Compared to LPA and S1P, SPC was not studied extensively.Although studies have attempted to identify SPC-specific mem-brane receptors, no receptors have been conclusively identified[15]. Previous studies, however, suggest that SPC may share S1Preceptors like S1P1, S1P2, S1P3, S1P4, and S1P5 [11,12]. We havepreviously reported that HFL-1 cells express S1P1, S1P2, S1P3, andS1P5 receptors [14]. Currently, although there are not specific S1Preceptor agonists, there are some S1P receptor antagonists avail-able including S1P1 antagonist (W-146), S1P2 antagonist (JTE-013),and S1P3 antagonist (Suramin). We found that only S1P2 receptorantagonist (JTE-013) blocks �-SMA expression and fibroblast-mediated gel contraction while W-146 and Suramin have no effect.These results suggest that SPC stimulated �-SMA expression andfibroblast mediated gel contraction might be mediated via the S1P2receptor. While HFL-1 cells express the S1P5 receptor, the role ofthe S1P5 receptor in mediating the effect of SPC was not evalu-ated in the current study with pharmacologic inhibitors as thesewere not available to us. However, in order to confirm the role ofthe S1P2 receptor and to evaluate the role of the S1P5 receptor inmediating the SPC effect on HFL-1 cells, RNA interference has been

used. Suppression of S1P2R or S1P5R by siRNA, however, resultedin apoptosis of the cells when they were cast into the collagen gels(data not shown), indicating S1P receptors may be crucial in cellsurvival.

X.Q. Wang et al. / Prostaglandins & other Lipid Mediators 108 (2014) 23–30 29

Fig. 6. Effect of Rho kinase inhibitor (Y27632) on collagen gel contraction and -SMA protein expression in response to SPC. Panel A: Effect of Y27632 on gel contraction.HFL-1 cells were pre-treated with or without Y27632 (10−5 M) and SPC (10−6 M). Cells were then cast into the type I collagen gel. After polymeriztion, gels were releasedand transferred to 60 mm tissue culture dishes containing 5 ml of serum free DMEM. Gel size was measure on day 1 with an image analyzer. Vertical axis: gel size expressedas % of initial size. Horizontal axis: pre-treatment conditions. Data are shown as meansSEM. *p < 0.01. Panel B: Effect of Y27632 on �-SMA expression. HFL-1 cells weres (10−6

g a pres

i[btSeaamc

ncimaiR[smctwRsc

erum-deprived for 24 h, and then treated with or without Y27632 (10−5 M) or SPCel electrophoresis and immunoblotted with anti-SMA or anti-actin antibodies. Dat

It is plausible that exogenous SPC could be further metabolizednto S1P, which is known to augment fibroblast repair functions14,18]. However, this conversion of SPC into S1P is mediatedy autotaxin [15,19]. In this regard, a study has demonstratedhat effects of S1P on motility in NIH 3T3 cells are identical toPC + autotaxin. If autotaxin is heat inactivated, however, theseffects are not observed, suggesting fibroblasts may not expressutotaxin [19], and thus, it is unlikely that the effect of SPC is medi-ted via being metabolized into S1P in our collagen gel contractionodel. Indeed, autotaxin is expressed and secreted by endothelial

ells [20].Rho GTPase is a crucial signal protein for cytoskeleton reorga-

ization in many cell types including fibroblasts, smooth muscleells, and neuronal cells [21,22]. It is now suggested RhoA GTPases also involved in muscle-specific gene transcription in smooth

uscle cells [17,21] and smooth muscle cell phenotypic alter-tion, which was mediated by S1P2R [23]. Microinjection of Rhon fibroblasts stimulates stress fiber formation, and inhibition ofho prevents these effects in response to lysophosphatidic acid24]. In order to demonstrate SPC stimulation of �-SMA expres-ion in HFL-1 cells and fibroblast-mediated gel contraction isediated by the RhoA GTPase, the following experiments were

onducted. We first demonstrated SPC stimulates RhoA activa-ion in a SPC concentration-dependent manner in HFL1 cells. Then,e also found C3 exoenzyme, a RhoA specific inhibitor, can block

hoA activation and �-SMA expression induced by SPC. Using theame method, we found Y27632, a specific Rho kinase inhibitor,an also block �-SMA expression and fibroblast-mediated gel

M) for 48 h. Cell lysate proteins (5 �g/lane) were separated by SDS polyacrylamideented is one representative of 3 separate experiments.

contraction induced by SPC, pertussis toxin (PTX), a Gi/o inhibitor,had no effect on SPC-induced collagen gel contraction or �-SMAexpression despite it has been reported that Gi/o is also a crucialsignal protein for lipid LPA, S1P [25,26]. These results demonstratethat SPC induced �-SMA expression in HFL-1 cells and fibroblastgel contraction is through the Rho/Rho kinase pathway.

While we have demonstrated that SPC augmentation of fibro-blast �-SMA expression and collagen gel contraction is mediatedthrough S1P2 receptor and its downstream mediator RhoA, whichclass of G-proteins that specifically connect S1P2R and Rho wasnot investigated in the current study. In this regard, there are fourclasses of heterotrimeric G proteins, Gs, Gi, Gq, and G12/13 [27]. Ithas been reported that S1P2R is not a Gs-coupled type of receptor[28,29], and that S1P2R induced Rho activation via both the Gq andG12/13 classes [30]. While it is likely that Gq or G12/13 may medi-ate signal transduction of SPC-induced Rho activation, this remainsto be determined. In addition, recently, it has also been reportedthat MAPKs, ERK1/2, p38 and JUK may be involved in mediating SPCstimulation in endothelial cells [31,32]. However, the role of MAPKsin mediating SPC-induced �-SMA and collagen gel contraction bylung fibroblasts remains to be determined.

In conclusion, this study demonstrated that SPC can increase�-SMA protein expression in HFL-1 cells, and pretreatment of HFL-1 with SPC augments fibroblast-mediated gel contraction. Theseeffects were not mediated by TGF-�1 and its signal protein smad3

or the Gi/o pathway. SPC-induced �-SMA protein expression andfibroblast gel contraction were mediated by S1P2 receptor andthe Rho/Rho kinase pathway. Our findings suggest that SPC may

3 other

bmc

A

lUHr

bABCDlFPmRatAaNiM

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

0 X.Q. Wang et al. / Prostaglandins &

e involved in airway remodeling through inducing fibroblasts toyofibroblasts transition, and thus, it may provide new targets for

hronic inflammatory disease intervention.

cknowledgements and disclosure

The authors are grateful for the secretarial support of Ms. Lil-ian Richards. This work was sponsored by the Larson Endowment,niversity of Nebraska Medical Center. X.Q.W., L.J.M., Q.H.F., T.K.,.J.K., H.S., S.K., S.T., and K.K. have nothing to disclosure. X.L. has

eceived research grants from GlaxoSmithKline.In the last three years, SIR has consulted or been a mem-

er of an advisory board for Able Associates, Adelphi Research,lmirall/Prescott, APT Pharma/Britnall, Aradigm, AstraZeneca,oehringer Ingelheim, Chiesi, CommonHealth, Consult Complete,OPDForum, DataMonitor, Decision Resources, Defined Health,ey, Dunn Group, Eaton Associates, Equinox, Gerson, GlaxoSmithK-

ine, Infomed, KOL Connection, M. Pankove, MedaCorp, MDRxinancial, Mpex, Novartis, Nycomed, Oriel Therapeutics, Otsuka,ennside Partners, Pfizer (Varenicline), Pharma Ventures, Phar-axis, Price Waterhouse, Propagate, Pulmatrix, Reckner Associates,

ecruiting Resources, Roche, Schlesinger Medical, Scimed, Sudlernd Hennessey, TargeGen, Theravance, UBC, Uptake Medical, Van-agePoint Mgmt. He has lectured for American Thoracic Society,straZeneca, Boehringer Ingelheim, California Allergy Society, Cre-tive Educational Concept, France Foundation, Information TV,etwork for Continuing Ed, Novartis, Pfizer, SOMA. He has received

ndustry-sponsored grants from AstraZeneca, Biomarck, Centocor,pex, Nabi, Novartis, Otsuka.

eferences

[1] Spurzem JR, Rennard SI. Pathogenesis of COPD. Semin Respir Crit Care Med2005;26(2):142–53.

[2] Kuwano K, Bosken CH, Pare PD, Bai TR, Wiggs BR, Hogg JC. Small airways dimen-sions in asthma and in chronic obstructive pulmonary disease. Am Rev RespirDis 1993;148(5):1220–5.

[3] Gabbiani G. The myofibroblast in wound healing and fibrocontractive diseases.J Pathol 2003;200(4):500–3.

[4] Hinz B, Celetta G, Tomasek JJ, Gabbiani G, Chaponnier C. Alpha-smooth mus-cle actin expression upregulates fibroblast contractile activity. Mol Biol Cell2001;12(9):2730–41.

[5] Grinnell F. Fibroblast biology in three-dimensional collagen matrices. TrendsCell Biol 2003;13(5):264–9.

[6] Kobayashi T, Liu X, Wen FQ, et al. Smad3 mediates TGF-beta1-induced colla-gen gel contraction by human lung fibroblasts. Biochem Biophys Res Commun2006;339(1):290–5.

[7] Bogatkevich GS, Tourkina E, Silver RM, Ludwicka-Bradley A. Thrombin dif-ferentiates normal lung fibroblasts to a myofibroblast phenotype via theproteolytically activated receptor-1 and a protein kinase C-dependent path-

way. J Biol Chem 2001;276(48):45184–92.

[8] Bogatkevich GS, Tourkina E, Abrams CS, Harley RA, Silver RM, Ludwicka-Bradley A. Contractile activity and smooth muscle alpha-actin organizationin thrombin-induced human lung myofibroblasts. Am J Physiol Lung Cell MolPhysiol 2003;285(2):L334–43.

[

Lipid Mediators 108 (2014) 23–30

[9] Saito A, Okazaki H, Sugawara I, Yamamoto K, Takizawa H. Potential action ofIL-4 and IL-13 as fibrogenic factors on lung fibroblasts in vitro. Int Arch AllergyImmunol 2003;132(2):168–76.

10] Meyer zu Heringdorf D, Himmel HM, Jakobs KH. Sphingosylphosphorylcholine-biological functions and mechanisms of action. Biochim Biophys Acta2002;1582(1–3):178–89.

11] Xu Y. Sphingosylphosphorylcholine and lysophosphatidylcholine. G protein-coupled receptors and receptor-mediated signal transduction. Biochim BiophysActa 2002;1582(1–3):81–8.

12] Sun L, Xu L, Henry FA, Spiegel S, Nielsen TB. A new wound healing agent –sphingosylphosphorylcholine. J Invest Dermatol 1996;106(2):232–7.

13] Suhr KB, Tsuboi R, Ogawa H. Sphingosylphosphorylcholine stimulates contrac-tion of fibroblast-embedded collagen gel. Br J Dermatol 2000;143(1):66–71.

14] Hashimoto M, Wang X, Mao L, et al. Sphingosine 1-phosphate potentiateshuman lung fibroblast chemotaxis through the S1P2 receptor. Am J Respir CellMol Biol 2008;39(3):356–63.

15] Nixon GF, Mathieson FA, Hunter I. The multi-functional role of sphingosylphos-phorylcholine. Prog Lipid Res 2008;47(1):62–75.

16] Xin C, Ren S, Kleuser B, et al. Sphingosine 1-phosphate cross-activates the Smadsignaling cascade and mimics transforming growth factor-beta-induced cellresponses. J Biol Chem 2004;279(34):35255–62.

17] Bishop AL, Hall A. Rho GTPases and their effector proteins. Biochem J2000;348(2):241–55.

18] Cooke ME, Sakai T, Mosher DF. Contraction of collagen matrices mediatedby alpha2beta1A and alpha(v)beta3 integrins. J Cell Sci 2000;113(Pt. 13):2375–83.

19] Clair T, Aoki J, Koh E, et al. Autotaxin hydrolyzes sphingosylphosphorylcholineto produce the regulator of migration, sphingosine-1-phosphate. Cancer Res2003;63(17):5446–53.

20] Nam SW, Clair T, Kim YS, et al. Autotaxin (NPP-2), a metastasis-enhancingmotogen, is an angiogenic factor. Cancer Res 2001;61(18):6938–44.

21] Hall A. Rho GTPases and the actin cytoskeleton. Science 1998;279(5350):509–14.

22 Hotchin NA, Hall A. Regulation of the actin cytoskeleton, integrins and cellgrowth by the Rho family of small GTPases. Cancer Surv 1996;27:311–22.

23] Medlin MD, Staus DP, Dubash AD, Taylor JM, Mack CP. Sphingosine 1-phosphate receptor 2 signals through leukemia-associated RhoGEF (LARG),to promote smooth muscle cell differentiation. Arterioscler Thromb Vasc Biol2010;30(9):1779–86.

24] Amano M, Chihara K, Kimura K, et al. Formation of actin stress fibers and focaladhesions enhanced by Rho-kinase. Science 1997;275(5304):1308–11.

25] Grinnell F, Ho CH, Lin YC, Skuta G. Differences in the regulation of fibro-blast contraction of floating versus stressed collagen matrices. J Biol Chem1999;274(2):918–23.

26] Hla T. Sphingosine 1-phosphate receptors. Prostaglandins 2001;64(1–4):135–42.

27] Siehler S. Regulation of RhoGEF proteins by G12/13-coupled receptors. Br JPharmacol 2009;158(1):41–9.

28] Okamoto H, Takuwa N, Yatomi Y, Gonda K, Shigematsu H, Takuwa Y. EDG3 is afunctional receptor specific for sphingosine 1-phosphate and sphingosylphos-phorylcholine with signaling characteristics distinct from EDG1 and AGR16.Biochem Biophys Res Commun 1999;260(1):203–8.

29] Sugimoto N, Takuwa N, Okamoto H, Sakurada S, Takuwa Y. Inhibitory andstimulatory regulation of Rac and cell motility by the G12/13-Rho and Gipathways integrated downstream of a single G protein-coupled sphingosine-1-phosphate receptor isoform. Mol Cell Biol 2003;23(5):1534–45.

30] Takashima S, Sugimoto N, Takuwa N, et al. G12/13 and Gq mediate S1P2-induced inhibition of Rac and migration in vascular smooth muscle in a mannerdependent on Rho but not Rho kinase. Cardiovasc Res 2008;79(4):689–97.

31] Lee HY, Lee SY, Kim SD, et al. Sphingosylphosphorylcholine stimulates

CCL2 production from human umbilical vein endothelial cells. J Immunol2011;186(7):4347–53.

32] Wirrig C, Hunter I, Mathieson FA, Nixon GF. Sphingosylphosphorylcholine isa proinflammatory mediator in cerebral arteries. J Cereb Blood Flow Metab2011;31(1):212–21.