International Journal of Molecular Sciences

Article Stimulates Proliferation and Migration of Myofibroblasts Isolated from Myocardial Infarction Model Rats

Akira Sugiyama, Yuka Hirano, Muneyoshi Okada * and Hideyuki Yamawaki

Laboratory of Veterinary Pharmacology, School of Veterinary Medicine, Kitasato University, Higashi 23 bancho 35-1, Towada City, Aomori 034-8628, Japan; [email protected] (A.S.); [email protected] (Y.H.); [email protected] (H.Y.) * Correspondence: [email protected]; Tel.: +81-176-23-4371; Fax: +81-176-24-9456

Received: 27 December 2017; Accepted: 1 March 2018; Published: 6 March 2018

Abstract: Myofibroblasts contribute to the healing of infarcted areas after myocardial infarction through proliferation, migration, and production of (ECM). Expression of endostatin, a cleaved fragment of type XVIII , increases in the heart tissue of an experimental myocardial infarction model. In the present study, we examined the effect of endostatin on the function of myofibroblasts derived from an infarcted area. The myocardial infarction model was created by ligating the left anterior descending artery in rats. Two weeks after the operation, α-smooth muscle actin (α-SMA)-positive myofibroblasts were isolated from the infarcted area. Endostatin significantly increased the proliferation and migration of myofibroblasts in vitro. On the other hand, endostatin had no effect on the production of , a major ECM produced by myofibroblasts. Endostatin activated Akt and extracellular signal-regulated kinase (ERK), and the pharmacological inhibition of these signaling pathways suppressed the endostatin-induced proliferation and migration. A knockdown of the COL18A1 gene in the myocardial infarction model rats using small interference RNA (siRNA) worsened the cardiac function concomitant with wall thinning and decreased the α-SMA-positive myofibroblasts and scar formation compared with that of control siRNA-injected rats. In summary, we demonstrated for the first time that endostatin might be an important factor in the healing process after myocardial infarction through the activation of myofibroblasts.

Keywords: endostatin; myofibroblast; myocardial infarction

1. Introduction Functional disorder of organs induces tissue fibrosis which is characterized by an accumulation of extracellular matrix (ECM), including collagen, , and fibronectin [1–3]. Moderate fibrosis is required for the healing process to maintain the structure and function of the damaged organs [1,4]. During the development of myocardial fibrosis, cardiac fibroblasts differentiate into their contractile phenotype, myofibroblasts, which express α-smooth muscle actin (α-SMA). Myofibroblasts are involved in the fibrosis through the production of ECM (e.g., type I and type III collagen), ECM degrading enzymes (e.g., matrix (MMPs)), pro-inflammatory cytokines (e.g., interleukin (IL)-1,6) and growth factors (e.g., epidermal growth factor and hepatocyte growth factor) [5–10]. Excessive proliferation and ECM production of myofibroblasts by the pathological activation of fibroblasts leads to the deterioration in myocardial compliance, dystelectasis, and cardioplegia [5]. After myocardial infarction, the moderate activation of myofibroblasts is necessary in the wound healing process through proliferation, migration, and ECM production. Therefore, the inadequate activation of myofibroblasts causes a poor scar formation and leads to cardiac

Int. J. Mol. Sci. 2018, 19, 741; doi:10.3390/ijms19030741 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2018, 19, 741 2 of 15 dilatation, contractile dysfunction, and rupture [5]. These observations suggest that the regulation of myofibroblasts function might be an attractive therapeutic target for myocardial infarction. Type XVIII collagen, a member of the family, is mainly localized in the vascular , which is cleaved by various proteases including MMPs, cathepsins, and elastase [11–15]. Endostatin, a 20 kDa cleaved C-terminal fragment of type XVIII collagen, was first separated from the culture medium of a murine hemangioendothelioma cell line [14]. Endostatin exerts an anti-angiogenic effect through the inhibition of migration and proliferation and the induction of apoptosis in vascular endothelial cells [14,16,17]. Subsequently, it inhibits the proliferation of tumor cells by suppressing the supply of nutrition and oxygen from blood vessels [18]. From the anti-tumor effect, zinc binding recombinant endostatin in the N-terminus (ZBP-endostatin; its trade name is Endostar) is approved by the State Food and Drug Administration of China for the treatment of non-small-cell lung cancer [19]. Endostatin also has pleiotropic effects including anti-hypertensive, anti-fibrotic, and anti-inflammatory effects [20–22]. The serum level of endostatin is increased in the patients with cardiovascular diseases [23,24]. Expression level of endostatin in cardiac tissue has been reported to be increased in the experimental cardiac disease models, such as pressure overload-induced cardiac hypertrophy [25,26] and myocardial infarction [27,28]. An anti-endostatin antibody-treatment has been reported to exacerbate the cardiac remodeling of non-infarcted area after myocardial infarction in rats [27]. We have recently reported that endostatin promotes the proliferation, migration, and wound-healing functions of cardiac fibroblasts derived from normal rats [29]. However, the pathological role of endostatin in the healing process of the infarcted area has not been clarified. In the present study, we investigated whether endostatin is responsible for the remodeling of the infarcted area after myocardial infarction through the modulation of biological functions of myofibroblasts.

2. Results

2.1. Effect of Endostatin on Proliferation of Myofibroblasts We first investigated the effect of endostatin (100–3000 ng/mL, 48 h) on cell morphology. Endostatin had no cytotoxicity on myofibroblasts (n = 5). Next, we examined the effect of endostatin on proliferation of myofibroblasts. Endostatin (48 h stimulation) significantly increased the number of cells, and the maximal effect was observed at 300 ng/mL (endostatin 300 ng/mL: p < 0.01 vs. control (Cont); endostatin 3000 ng/mL: p < 0.05 vs. Cont, n = 5) (Figure1A). Additionally, we examined the effect of endostatin (100–3000 ng/mL, 24 h) on bromodeoxyuridine (BrdU) incorporation of myofibroblasts. Endostatin significantly increased BrdU incorporation (DNA synthesis activity) at 3000 ng/mL (p < 0.01 vs. Cont, n = 8) (Figure1B).

2.2. Effect of Endostatin on Migration of Myofibroblasts We next examined the effect of endostatin on migration of myofibroblasts by a Boyden chamber assay. Endostatin (300, 3000 ng/mL, 24 h stimulation) significantly increased cell migration (p < 0.01 vs. Cont, n = 7) (Figure1C,D).

2.3. Effect of Endostatin on ECM Production and MMPs Secretion of Myofibroblasts We next examined secretion of type I collagen, a major ECM component produced by myofibroblasts, and MMP-2, an ECM-degrading enzyme, by Western blotting. Endostatin (100–3000 ng/mL, 48 h stimulation) had no effect on secretion of type I collagen (n = 6) (Figure2A,B). Endostatin (100–3000 ng/mL, 48 h stimulation) also had no effect on MMP-2 secretion (n = 5) (Figure2C,D). Int. J. Mol. Sci. 2018, 19, 741 3 of 15 Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 3 of 14

Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 3 of 14

Figure 1.FigureEndostatin 1. Endostatin induced induced proliferation proliferationand and migration of ofmyofibroblasts. myofibroblasts. (A) After (A) myofibroblasts After myofibroblasts were grown to 30–50% confluence, they were stimulated with endostatin (100–3000 ng/mL) for 48 h. were grown to 30–50% confluence, they were stimulated with endostatin (100–3000 ng/mL) for 48 h. Control (Cont) was treated with a vehicle of endostatin (citrate-phosphate buffer). The number of Controlviable (Cont) cells was was treated counted with by a a colorimetric vehicle of endostatinmethod using (citrate-phosphate a cell counting kit-8,buffer). and the normalized The number cell of viable cells wasnumberFigure counted 1. relativeEndostatin by a colorimetricto inducedCont are proliferation shown method as using andmean migration a± cellstandard counting of myofibroblasts. error kit-8, of the and (meanA the) After normalized(S.E.M.) myofibroblasts (n = cell 5). number relative*,were to p Cont< 0.05,grown are**, top shown <30–50% 0.01 vs. as confluence, Cont; mean (B±) After standardthey thewere cells stimulat error were of seededed the with mean at endostatin a density (S.E.M.) of(100–3000 1 ( n× =10 5).4 cells/well ng/mL) *, p < 0.05,for in a48 96- **,h. p < 0.01 vs. Cont;wellControl (B culture) After (Cont) plate the was and cells treated starved were with (in seeded a0.5% vehicle fetal at aof bovine density endo statinserum of 1(citrate-phosphate (FBS)),× 104 theycells/well were buffer).treated in a withThe 96-well numberendostatin culture of plate viable cells was counted by a colorimetric method using a cell counting kit-8, and the normalized cell and starved(100–3000 (in 0.5%ng/mL) fetal or vehicle bovine for serum 24 h. (FBS)),The cells they were wereincubated treated with with a bromodeoxyuridine endostatin (100–3000 (BrdU) ng/mL) reagentnumber (20relative µL/well) to forCont 24 areh, and shown BrdU as incorporat mean ± ionstandard was detected error byof athe colorimetric mean (S.E.M.) method (n using= 5). or vehicle for 24 h. The cells were incubated with a bromodeoxyuridine (BrdU) reagent (20 µL/well) an*, p anti-BrdU< 0.05, **, p antibody< 0.01 vs. Cont;and tetramethylbenzidine (B) After the cells were solution. seeded atThe a density normalized of 1 × 10BrdU4 cells/well incorporation in a 96- for 24 h,relativewell and culture BrdU to Cont plate incorporation are and shown starved as was(in mean 0.5% detected ±fetal S.E.M. bovine by (n a = colorimetricserum 8). **, (FBS)), p < 0.01 they method vs. were Cont; usingtreated (C, anD with) anti-BrdUMigration endostatin of antibody and tetramethylbenzidinemyofibroblasts(100–3000 ng/mL) was or determined vehicle solution. for by 24 The a h.Boyden The normalized cells chamber were BrdUassay.incubated The incorporation withcells ata bromodeoxyuridinea density relative of 1 × to105 Cont cells/100 (BrdU) are shown as meanµLreagent± serumS.E.M. (20 free µL/well) (n medium= 8). for **, were24p h,< seededand 0.01 BrdU vs. in theincorporat Cont; upper-chamber (Cion,D) was Migration detected coated with ofby myofibroblastsa 2%colorimetric , and method endostatin was using determined by a Boyden(300,an anti-BrdU 3000 chamber ng/mL) antibody assay.or vehicle and The wastetramethylbenzidine cells trea atted a in density the bottom ofsolution. 1 well× 10 for 5The cells/10024 normalizedh; (C) RepresentativeµL serumBrdU incorporation free pictures medium of were relative to Cont are shown as mean ± S.E.M. (n = 8). **, p < 0.01 vs. Cont; (C,D) Migration of seededGiemsa-stained in the upper-chamber migrated coatedcells ar withe shown. 2% gelatin,Scale bar: and 100 endostatin µm. (D) The (300, normalized 3000 ng/mL) migrated or vehiclecell was numbermyofibroblasts relative was to Cont determined are shown by aas Boyden mean ± ch S.E.M.amber (n assay. = 7). **, The p < cells 0.01 atvs. a Cont.density of 1 × 105 cells/100 treated in the bottom well for 24 h; (C) Representative pictures of Giemsa-stained migrated cells are µL serum free medium were seeded in the upper-chamber coated with 2% gelatin, and endostatin shown.(300, Scale 3000 bar: ng/mL) 100 µ orm. vehicle (D) The was normalizedtreated in the migratedbottom well cell for number24 h; (C) Representative relative to Cont pictures are of shown as mean ±Giemsa-stainedS.E.M. (n = 7). migrated **, p < 0.01cells vs.are Cont.shown. Scale bar: 100 µm. (D) The normalized migrated cell number relative to Cont are shown as mean ± S.E.M. (n = 7). **, p < 0.01 vs. Cont.

Figure 2. Endostatin had no effect on secretion of type I collagen and matrix (MMP)-2 in myofibroblasts. Myofibroblasts were stimulated with endostatin (100–3000 ng/mL) or vehicle for 48 h, and culture medium was collected. Secretion of type I collagen and MMP-2 was determined by Western blotting. (A) Representative blot of type I collagen is shown. (B) Levels of Figure 2.secretedFigureEndostatin 2. type Endostatin I hadcollagen no had effect were no effecton corrected secretion on secretion by oftota type l ofcell type I collagenprotein I collagen concentration, and and matrix matrix metalloproteinase and metalloproteinase the normalized (MMP)-2 in myofibroblasts.expression(MMP)-2 in relative myofibroblasts. Myofibroblasts to Cont are Myofibroblasts shown were stimulatedas mean were ± S.E.M. st withimulated (n endostatin = 6); with (C) Representativeendostatin (100–3000 (100–3000 blot ng/mL) of ng/mL)MMP-2 or vehicle oris for 48 h, andshown.vehicle culture for(D) 48Levels medium h, and of culturesecreted was collected.medium MMP-2 waswere Secretion collecte correctedd. of Secretion by type total I collagenofcell type protein I collagen and concentration, MMP-2 and MMP-2 was and determined wasthe determined by Western blotting. (A) Representative blot of type I collagen is shown. (B) Levels of by Westernnormalized blotting. expression (A) Representative relative to Cont are blot shown of type as mean I collagen ± S.E.M. is (n shown. = 5). (B) Levels of secreted type secreted type I collagen were corrected by total cell protein concentration, and the normalized I collagen were corrected by total cell protein concentration, and the normalized expression relative expression relative to Cont are shown as mean ± S.E.M. (n = 6); (C) Representative blot of MMP-2 is to Contshown. are shown (D) Levels as mean of secreted± S.E.M. MMP-2 (n = were 6); ( Ccorrected) Representative by total cell blot protein of MMP-2 concentration, is shown. and the (D ) Levels of secretednormalized MMP-2 expression were corrected relative to by Cont total are cell shown protein as mean concentration, ± S.E.M. (n = 5). and the normalized expression relative to Cont are shown as mean ± S.E.M. (n = 5). Int. J. Mol. Sci. 2018, 19, 741 4 of 15

Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 4 of 14 2.4. Effect of Endostatin on Activation of Akt and Extracellular Signal-Regulated Kinase (ERK) in Myofibroblasts2.4. Effect of Endostatin on Activation of Akt and Extracellular Signal-Regulated Kinase (ERK) in Myofibroblasts We examined the effect of endostatin on phosphorylation of Akt and ERK by Western blotting. We examined the effect of endostatin on phosphorylation of Akt and ERK by Western blotting. From the results of the cell counting assay and Boyden chamber assay (Figure1A,C), 300 ng/mL From the results of the cell counting assay and Boyden chamber assay (Figure 1A,C), 300 ng/mL endostatin seems to be a maximally effective dose for proliferation and migration of myofibroblasts. endostatin seems to be a maximally effective dose for proliferation and migration of myofibroblasts. Thus, we examined the effect of 300 ng/mL endostatin. In a preliminary experiment, we observed that Thus, we examined the effect of 300 ng/mL endostatin. In a preliminary experiment, we observed 300that ng/mL 300 ng/mL endostatin endostatin (10–360 (10–360 min) min) increased increased phosphorylation phosphorylation of Akt of Akt and and ERK ERK maximally maximally at at 60 60 min (Akt:minn (Akt:= 7; ERK:n = 7; nERK:= 9). n = Then, 9). Then, we we examined examined the the dose-response dose-responseof of endostatinendostatin at at 60 60 min. min. Endostatin Endostatin (10–300(10–300 ng/mL, ng/mL, 60 60 min) min) increased increased phosphorylationphosphorylation of of Akt Akt (endostatin (endostatin 300 300 ng/mL: ng/mL: p < 0.01p < 0.01vs. Cont, vs. Cont, n n == 12 12)) andand ERK (endostatin (endostatin 300 300 ng/mL: ng/mL: p < 0.05p < vs. 0.05 Cont, vs. n Cont, = 11) nin =a concentration-dependent 11) in a concentration-dependent manner manner(Figure (Figure 3). 3).

FigureFigure 3. Endostatin3. Endostatin induced induced phosphorylation phosphorylation ofof Akt and extracellular extracellular signal-regulated signal-regulated kinase kinase (ERK) (ERK) in myofibroblasts.in myofibroblasts. Myofibroblasts Myofibroblasts were were stimulated stimulated with with endostatinendostatin (10–300 (10–300 ng/mL, ng/mL, 60 60 min) min) or orvehicle, vehicle, andand cell cell lysate lysate was was collected. collected. Phosphorylation Phosphorylation ofof AktAkt (at Ser473: AA,B,B) )and and ERK ERK (C (,CD,D) was) was detected detected by by WesternWestern blotting. blotting. (A (,CA),C Representative) Representative blots blots of of phosphorylatedphosphorylated proteins proteins (pAkt: (pAkt: AA pERK:pERK: C)C and) and total total proteinsproteins (tAkt: (tAkt:A; tERK1:A; tERK1:C) forC) for Akt Akt and and ERK ERK are are shown; shown. (B, D(B), LevelsD) Levels of phosphorylatedof phosphorylated proteins proteins were correctedwere corrected by total proteins,by total proteins, and the normalizedand the normalized expression expression relative relative to Cont to are Cont shown are shown as mean as ±meanS.E.M . (Akt:± S.E.M.n = 12, (Akt: ERK: n =n 12,= 11). ERK: *, pn <= 11). 0.05, *, **,p

2.5.2.5. Endostatin Endostatin Induced Induced Proliferation Proliferation and and Migration Migration ofof MyofibroblastsMyofibroblasts through Activation Activation of of Phosphoinositide 3-KinasePhosphoinositide (PI3K)/Akt 3-Kinase and Mitogen-Activated (PI3K)/Akt and Mito Proteingen-Activated Kinase (MAPK)/ERK Protein Kinase Kinase (MAPK)/ERK (MEK)/ERK Kinase Pathway (MEK)/ERK Pathway We next investigated whether endostatin induced proliferation and migration of myofibroblasts We next investigated whether endostatin induced proliferation and migration of myofibroblasts through the activation of Akt and ERK pathways by using wortmannin, a PI3K/Akt inhibitor, and through the activation of Akt and ERK pathways by using wortmannin, a PI3K/Akt inhibitor, and PD98059, a MEK/ERK inhibitor [29–31]. Wortmannin (1 µM) or PD98059 (50 µM) significantly PD98059, a MEK/ERK inhibitor [29–31]. Wortmannin (1 µM) or PD98059 (50 µM) significantly inhibited the endostatin (3000 ng/mL, 24 h)-induced BrdU incorporation (endostatin: p < 0.05 vs. inhibited the endostatin (3000 ng/mL, 24 h)-induced BrdU incorporation (endostatin: p < 0.05 vs. Cont; Cont; +Wortmannin: p < 0.01 vs. endostatin; +PD98059: p < 0.05 vs. endostatin, n = 9) (Figure4A). +Wortmannin: p < 0.01 vs. endostatin; +PD98059: p < 0.05 vs. endostatin, n = 9) (Figure 4A). µ WortmanninWortmannin (100 (100 nM) nM) oror PD98059 (50 (50 µM)M) significantly significantly inhibited inhibited the endostatin the endostatin (300 ng/mL, (300 24 ng/mL, h)- 24 h)-inducedinduced migration migration (endostatin: (endostatin: p < 0.01p vs.< 0.01Cont; vs. +Wortmannin: Cont; +Wortmannin: p < 0.01 vs. endostatin;p < 0.01 vs. +PD98059: endostatin; p +PD98059:< 0.01 vs.p endostatin,< 0.01 vs. endostatin,n = 6) (Figuren =4B,C). 6) (Figure 4B,C).

Int. J. Mol. Sci. 2018, 19, 741 5 of 15 Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 5 of 14

FigureFigure 4. 4.Wortmannin Wortmannin(a (aphosphoinositide phosphoinositide 3-kinase inhibitor) inhibitor) or or PD98059 PD98059 (a(a mitogen-activated mitogen-activated protein kinase/ERK kinase inhibitor) inhibited endostatin-induced proliferation and migration of protein kinase/ERK kinase inhibitor) inhibited endostatin-induced proliferation and migration of myofibroblasts. (A) After the cells were seeded at a density of 1 × 104 cells/well in a 96-well culture myofibroblasts. (A) After the cells were seeded at a density of 1 × 104 cells/well in a 96-well culture plate and starved (in 0.5% FBS), they were treated with endostatin (3000 ng/mL) or vehicle for 24 h in plate and starved (in 0.5% FBS), they were treated with endostatin (3000 ng/mL) or vehicle for 24 h the presence or absence of wortmannin (Wort: 1 µM, 30 min pre-treatment) or PD98059 (PD: 50 µM, in the presence or absence of wortmannin (Wort: 1 µM, 30 min pre-treatment) or PD98059 (PD: 30 min pre-treatment). The cells were incubated with a BrdU reagent (20 µL/well) for 24 h, and BrdU 50 µM, 30 min pre-treatment). The cells were incubated with a BrdU reagent (20 µL/well) for 24 h, incorporation was detected by a colorimetric method using an anti-BrdU antibody and andtetramethylbenzidine BrdU incorporation solution. was detected The normalized by a colorimetric BrdU incorporation method using relative an anti-BrdUto Cont are antibody shown as and tetramethylbenzidinemean ± S.E.M. (n = 9). solution. *, p < 0.05 Thevs. Cont, normalized #, p < 0.05, BrdU ##, p incorporation < 0.01 vs. endostatin-alone relative to Cont treatment; are shown (B,C) as meanMigration± S.E.M of.( myofibroblastsn = 9). *, p < 0.05 was vs. determined Cont, #, p by< 0.05,a Boyd ##,enp chamber< 0.01 vs. assay. endostatin-alone The cells at a treatment;density of 1 ( B× ,C) Migration105 cells/100 of myofibroblasts µL serum free medium was determined were seeded by ain Boyden the upper-chamber chamber assay. coated The with cells 2% atgelatin, a density and of 5 1 ×endostatin10 cells/100 (300 µng/mL)L serum or freevehicle medium was treated were seededin the bottom in the upper-chamberwell for 24 h. Wortmannin coated with (Wort: 2% gelatin, 100 andnM) endostatin or PD98059 (300 (PD: ng/mL) 50 µM) or vehiclewas pre-treated was treated in both in the upper bottom and welllower for chambers 24 h. Wortmannin 30 min before (Wort: 100endostatin-stimulation. nM) or PD98059 (PD: (B 50) RepresentativeµM) was pre-treated pictures inof bothGiemsa-stained upper and migrated lower chambers cells are shown. 30 min Scale before endostatin-stimulation.bar: 100 µm. (C) The normalized (B) Representative migrated pictures cell number of Giemsa-stained relative to Cont migrated are shown cells as mean are shown. ± S.E.M. Scale bar:(n 100 = 6).µ **,m. p ( C< )0.01 The vs. normalized Cont, ##, p < migrated 0.01 vs. endostatin-alone cell number relative treatment. to Cont are shown as mean ± S.E.M. (n = 6). **, p < 0.01 vs. Cont, ##, p < 0.01 vs. endostatin-alone treatment. 2.6. A Knockdown of the COL18A1 Gene in the Infarcted Area by COL18A1 Small Interference (si) RNA 2.6.Worsened A Knockdown Cardiac of Function the COL18A1 and Decreased Gene in α the-SMA-Positive Infarcted Area Myofibroblasts by COL18A1 Small Interference (si) RNA WorsenedWe Cardiac used siRNA Function against and Decreasedthe COL18A1α-SMA-Positive gene, the effectiveness Myofibroblasts of which on endostatin expression wasWe previously used siRNA confirmed against in the heartsCOL18A1 of rats [32].gene, COL18A1 the effectiveness siRNA had of no which significant on endostatin effect on the expression body wasweight previously of rats confirmed (Table 1). in hearts of rats [32]. COL18A1 siRNA had no significant effect on the body weight of rats (Table1). Table 1. Body weight (BW) in small interference (si) RNA-injected myocardial infarction model rats.

Table 1. BodyParameters weight (BW) in small interference Cont siRNA (si) RNA-injected (n = 6) myocardialCOL18A1 infarction siRNA model (n = 6) rats. BW (before operation) 265.8 ± 16.4 g 270.9 ± 17.2 g BW (2 weeksParameters after operation) Cont 318.6 siRNA ± 10.0 (n g= 6) COL18A1 310.7siRNA ± 13.2 (ng = 6) ΔBW 52.8 ± 6.9 g 39.8 ± 5.9 g BW (before operation) 265.8 ± 16.4 g 270.9 ± 17.2 g BWBW of (2 Cont weeks siRNA- after or operation) COL18A1 siRNA-injected 318.6 myocardial± 10.0 g infarction model rats. 310.7 The data± 13.2 areg shown as mean ± standard∆BW error of the mean (S.E.M.). ΔBW: 52.8 BW± 6.9 (2 weeks g after operation)—BW 39.8 (before± 5.9 operation). g BW of Cont siRNA- or COL18A1 siRNA-injected myocardial infarction model rats. The data are shown as meanIn ±thestandard echocardiographic error of the mean analysis, (S.E.M.). ∆leftBW: ventricular BW (2 weeks (LV) after operation)—BWmass was not (beforechanged operation). by a COL18A1 siRNA injection. On the other hand, LV internal dimension at end-systole (LVIDs) (n = 6, p < 0.01), interventricular septum (IVS)/LV posterior wall (LVPW) (n = 6, p < 0.05) and end-systolic volume In the echocardiographic analysis, left ventricular (LV) mass was not changed by a COL18A1 (ESV) (n = 6, p < 0.05) were increased by a COL18A1 siRNA injection. In addition, LVPW at end- siRNA injection. On the other hand, LV internal dimension at end-systole (LVIDs) (n = 6, p < 0.01), diastole (LVPWd) (n = 6, p < 0.01), end-systole (LVPWs) (n = 6, p < 0.01) and percentage LVPW interventricular septum (IVS)/LV posterior wall (LVPW) (n = 6, p < 0.05) and end-systolic volume (ESV) (n = 6, p < 0.05) were increased by a COL18A1 siRNA injection. In addition, LVPW at end-diastole Int. J. Mol. Sci. 2018, 19, 741 6 of 15

(LVPWd)Int. J. Mol. ( nSci.= 2018 6, ,p 19<, x0.01), FOR PEER end-systole REVIEW (LVPWs) (n = 6, p < 0.01) and percentage LVPW thickness 6 of 14 n p COL18A1 (LVPW%)thickness ( (LVPW%)= 6, < 0.05)(n = were6, p < decreased 0.05) were by decreased a by siRNAa COL18A1 injection. siRNA These injection. observations These demonstratedobservations that demonstrated the endostatin that knockdownthe endostatin induced knockdown thinning induced and dilatationthinning and of LV.dilatation The decrease of LV. of fractionalThe decrease shortening of fractional (FS) (n shortening= 6, p < 0.01) (FS) and (n = ejection 6, p < 0.01) fraction and (EF)ejection (n = fraction 6, p < 0.01) (EF) indicated(n = 6, p < that0.01) the endostatinindicated knockdown that the endostatin worsened knockdown the cardiac worsened function the (Figure cardiac5 functionand Table (Figure2). 5 and Table 2).

FigureFigure 5. The5. The M-mode M-mode echocardiographic echocardiographic imageimage from parast parasternalernal short short axis axis view view of ofthe the left left ventricle ventricle at theat the level level of of papillary papillary muscles muscles in in siRNA-injected siRNA-injected myocardial infarction infarction model model rats. rats. After After the the ligation ligation of theof the left left anterior anterior descending descending artery (LAD) (LAD) in in rats, rats, sixteen sixteen µg ofµg COL18A1 of COL18A1 siRNAsiRNA or Cont or siRNA Cont siRNAwas wasinjected injected into into the the infarcted infarcted myocardium. myocardium. Two Two we weekseks after after the the operation, operation, echocardiography echocardiography was was performed. Representative images of M-mode on the parasternal short axis view of the left ventricle performed. Representative images of M-mode on the parasternal short axis view of the left ventricle are are shown (Cont siRNA: n = 6, COL18A1 siRNA: n = 6). Scale bar: 100 msec (horizontal) and 5 mm shown (Cont siRNA: n = 6, COL18A1 siRNA: n = 6). Scale bar: 100 msec (horizontal) and 5 mm (vertical). (vertical). LVIDd: left ventricular internal dimension at end-diastole, LVIDs: Left ventricular internal LVIDd: left ventricular internal dimension at end-diastole, LVIDs: Left ventricular internal dimension at dimension at end-systole, LVPWd: Left ventricular posterior wall at end-diastole, LVPWs: Left end-systole, LVPWd: Left ventricular posterior wall at end-diastole, LVPWs: Left ventricular posterior ventricular posterior wall at end-systole. wall at end-systole. Table 2. Echocardiographic parameters in siRNA-injected myocardial infarction model rats. Table 2. Echocardiographic parameters in siRNA-injected myocardial infarction model rats. Cont siRNA COL18A1 siRNA Parameters Unit (n = 6) (n = 6) ParametersIVSd Unit cm Cont 0.18 siRNA( ± 0.01 n = 6) 0.17 ± 0.01COL18A1 siRNA (n = 6) IVSdLVIDd cm cm 0.72 0.18 ± 0.04± 0.01 0.85 ± 0.04 0.17 ± 0.01 LVIDdLVPWd cm cm 0.20 0.72 ± 0.01± 0.04 0.15 ± 0.01 ** 0.85 ± 0.04 LVPWdIVSs cm cm 0.30 0.20 ± 0.01± 0.01 0.29 ± 0.02 0.15 ± 0.01 ** LVIDs cm 0.40 ± 0.02 0.57 ± 0.04 ** IVSs cm 0.30 ± 0.01 0.29 ± 0.02 LVPWs cm 0.29 ± 0.02 0.19 ± 0.02 ** ± ± LVIDsEDV (MM-Teich) cm mL 0.89 0.40 ± 0.120.02 1.38 ± 0.21 0.57 0.04 ** LVPWsIVS/LVPW cm (MM) 0.94 0.29 ± 0.07± 0.02 1.17 ± 0.06 * 0.19 ± 0.02 ** EDV (MM-Teich)LV Mass mL(Cubed) g 1.4 0.89 ± 0.1± 0.12 1.4 ± 0.1 1.38 ± 0.21 IVS/LVPW (MM)IVS% (MM) % 69.0 0.94 ± 9.3± 0.07 69.7 ± 7.6 1.17 ± 0.06 * LV Mass (Cubed)ESV (MM-Teich) g mL 0.16 1.4 ± 0.02± 0.1 0.49 ± 0.11 * 1.4 ± 0.1 IVS% (MM)FS (MM-Teich) % % 44.5 69.0 ± 2.7± 9.3 32.9 ± 2.0 ** 69.7 ± 7.6 ESV (MM-Teich)EF (MM-Teich) mL % 80.0 0.16 ± 2.9± 0.02 66.5 ± 2.7 ** 0.49 ± 0.11 * FS (MM-Teich)LVPW% %(MM) % 51.7 44.5 ± 5.8± 2.7 26.3 ± 8.3 * 32.9 ± 2.0 ** EF (MM-Teich)Heart %rate bpm 405.9 80.0 ± 9.7± 2.9 376.2 ± 7.4 * 66.5 ± 2.7 ** TwoLVPW% weeks after (MM) the operation %of myocardial infa 51.7rction,± 5.8echocardiography was26.3 performed.± 8.3 * IVSd: InterventricularHeart rate septum at end-diastole, bpm LVIDd: left 405.9 ventricular± 9.7 internal dimension376.2 at± end-diastole,7.4 * TwoLVPWd: weeks after Left the ventricular operation posterior of myocardial wall infarction, dimension echocardiography at end-diastole, wasIVSs: performed. Interventricular IVSd: Interventricular septum at septumend-systole, at end-diastole, LVIDs: LVIDd:Left ventricular left ventricular internal internal dimension dimension at end-systole, at end-diastole, LVPWs: LVPWd: Left Leftventricular ventricular posteriorposterior wall wall dimension at end-systole, at end-diastole, EDV: End-diastolic IVSs: Interventricular volume, IVS%: septum percentage at end-systole, IVS thickness, LVIDs: Left ESV: ventricular End- internal dimension at end-systole, LVPWs: Left ventricular posterior wall at end-systole, EDV: End-diastolic volume, IVS%:systolic percentage volume, IVS thickness,FS: Fractional ESV: End-systolic shortening, volume, EF: Ejection FS: Fractional fraction, shortening, LVPW%: EF: Ejection percentage fraction, LVPW LVPW%: percentagethickness. LVPW The thickness.data are shown The data as aremean shown ± S.E.M. as mean *, p ±< 0.05,S.E.M. **, *, pp << 0.01 0.05, vs. **, Contp < 0.01 siRNA. vs. Cont siRNA.

We next investigated the distribution of α-SMA-positive myofibroblasts [33] in the infarcted area We next investigated the distribution of α-SMA-positive myofibroblasts [33] in the infarcted after myocardial infarction. The α-SMA-positive myofibroblasts were observed throughout the area after myocardial infarction. The α-SMA-positive myofibroblasts were observed throughout the infarcted area in Cont siRNA-injected rats (Figure 6A). On the other hand, the α-SMA-positive infarcted area in Cont siRNA-injected rats (Figure6A). On the other hand, the α-SMA-positive myofibroblasts were sparsely expressed in the infarcted area in COL18A1 siRNA-injected rats (Figure myofibroblasts were sparsely expressed in the infarcted area in COL18A1 siRNA-injected rats

Int. J. Mol. Sci. 2018, 19, 741 7 of 15 Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 7 of 14

(Figure6B). 6TheB). Theα-SMA-positiveα-SMA-positive area was area significantly was significantly decreased decreased by a COL18A1 by a COL18A1 siRNA injectionsiRNA injection(Cont (ContsiRNA: siRNA: n = 4;n =COL18A1 4; COL18A1 siRNA:siRNA: n = 6, np =< 0.05) 6, p < (Figure 0.05) (Figure 6C). 6C).

Int. J. Mol. Sci. 2018, 19, x FOR PEER REVIEW 7 of 14

6B). The α-SMA-positive area was significantly decreased by a COL18A1 siRNA injection (Cont siRNA: n = 4; COL18A1 siRNA: n = 6, p < 0.05) (Figure 6C).

FigureFigure 6. Distribution6. Distribution of ofα α-smooth-smooth muscle actin actin (α (α-SMA)-positive-SMA)-positive myofibroblasts myofibroblasts in infarcted in infarcted area of area siRNA-injected myocardial infarction model rats. After ligation of the left anterior descending artery of siRNA-injected myocardial infarction model rats. After ligation of the left anterior descending artery(LAD) (LAD) in rats, in rats, sixteen sixteen µg µofg ofCOL18A1COL18A1 siRNAsiRNA or orCont Cont siRNA siRNA was was injected injected into into the the infarcted infarcted myocardium.Figure 6. Distribution Two weeks of after α-smooth the operation, muscle actin the ( αheart-SMA)-positive was isolated myofibroblasts and then thin in infarcted paraffin area sections of (4 myocardium.siRNA-injected Two weeks myocardial after infarction the operation, model rats. the Afte heartr ligation was isolated of the left and anterior then descending thin paraffin artery sections µm) were made. (A,B) Representative pictures of the myocardium stained with a specific antibody (4 µm) were(LAD) made. in rats, (A sixteen,B) Representative µg of COL18A1 pictures siRNA of or the Cont myocardium siRNA was stainedinjected withinto the a specific infarcted antibody against α-SMA are shown (A: Cont siRNA; B: COL18A1 siRNA). Nuclei were counterstained with against myocardium.α-SMA are Two shown weeks (A after: Cont the siRNA;operation,B the: COL18A1 heart was siRNA).isolated and Nuclei then thin were paraffin counterstained sections (4 with hematoxylin (Cont siRNA: n = 4; COL18A1 siRNA: n = 6). Scale bar: 200 µm. (C) The α-SMA positive hematoxylinµm) were (Cont made. siRNA: (A,B)n Representative= 4; COL18A1 picturessiRNA: of then = myocardium 6). Scale bar: stained 200 µwithm; (aC specific) The α antibody-SMA positive area was calculated, and the normalized area relative to Cont siRNA are shown as mean ± S.E.M. area wasagainst calculated, α-SMA andare shown the normalized (A: Cont siRNA; area B relative: COL18A1 to siRNA). Cont siRNA Nuclei were are showncounterstained as mean with± S.E.M. (Conthematoxylin siRNA: n = (Cont 4, COL18A1 siRNA: nsiRNA: = 4; COL18A1 n = 6). siRNA:*, p < 0.05 n = 6).vs. Scale Cont bar: siRNA. 200 µm. (C) The α-SMA positive (Cont siRNA: n = 4, COL18A1 siRNA: n = 6). *, p < 0.05 vs. Cont siRNA. area was calculated, and the normalized area relative to Cont siRNA are shown as mean ± S.E.M. We finally(Cont siRNA: examined n = 4, COL18A1 the effect siRNA: of endostatin n = 6). *, p < 0.05on scarvs. Cont formation siRNA. by Azan staining. The thickness ofWe the finallyscar tissue examined composed the effectof collagen of endostatin fibers (blu one) scarin the formation infarcted by area Azan was staining.significantly The decreased thickness of We finally examined the effect of endostatin on scar formation by Azan staining. The thickness theby scar a COL18A1 tissue composed siRNA injection of collagen compared fibers (blue) with that in the of infarctedCont siRNA area siRNA-injection was significantly (Cont decreased siRNA: byn a of the scar tissue composed of collagen fibers (blue) in the infarcted area was significantly decreased = 4; COL18A1 siRNA: n = 5, p < 0.05) (Figure 7). COL18A1by siRNAa COL18A1 injection siRNA injection compared compared with thatwith ofthat Cont of Cont siRNA siRNA siRNA-injection siRNA-injection (Cont (Cont siRNA: siRNA: n n = 4; COL18A1= 4;siRNA: COL18A1n =siRNA: 5, p < n 0.05)= 5, p < (Figure 0.05) (Figure7). 7).

Figure 7. Azan staining for scar formation in infarcted area of siRNA-injected myocardial infarction FigureFigure 7. Azan7. Azan staining staining for for scar scar formation formationin in infarctedinfarcted area of siRNA-injected siRNA-injected myocardial myocardial infarction infarction model rats. After ligation of the LAD in rats, sixteen µg of COL18A1 siRNA or Cont siRNA was model rats. After ligation of the LAD in rats, sixteen µgµ of COL18A1COL18A1 siRNA or Cont siRNA was model rats.injected After into ligationthe infarcted of themyocardium. LAD in Two rats, weeks sixteen after gthe of operation, thesiRNA heart was or isolated Cont siRNA and was injected into the infarcted myocardium. Two weeks after the operation, the heart was isolated and injectedthen into thin the infarctedparaffin sections myocardium. (4 µm) were Two made. weeks (A after,B) Representative the operation, Azan the stained heart was pictures isolated of the and then thinthen paraffin infarctedthin paraffin sections area are sections (4 shownµm) were (4(A :µm) Cont made. were siRNA; ( Amade.,B B): RepresentativeCOL18A1 (A,B) Representative siRNA). Azan Blue: stainedcollagen Azan stainedfibers; pictures Red: pictures of muscle the infarctedof the areainfarcted arefibers. shown area Scale ( Aare :bar: Cont shown 100 siRNA; µm. (A (:C Cont)B Scar: COL18A1 siRNA; thickness BsiRNA). :was COL18A1 measured Blue: siRNA). and collagen are Blue: shown fibers; collagen as mean Red: fibers; ± muscle S.E.M. Red: (Cont fibers. muscle Scale bar:fibers. 100siRNA:µ Scalem; (C nbar:) =Scar 4, COL18A1100 thickness µm. (siRNA:C) Scar was n =thickness measured 5). * p < 0.05 was and vs. measured Cont are shownsiRNA. and as are mean shown± S.E.M. as mean (Cont ± S.E.M. siRNA: (Contn = 4,

COL18A1siRNA:siRNA: n = 4, COL18A1n = 5). * siRNA:p < 0.05 n vs.= 5). Cont * p < siRNA.0.05 vs. Cont siRNA.

Int. J. Mol. Sci. 2018, 19, 741 8 of 15

3. Discussion In this study, we revealed for the first time that endostatin induced proliferation and migration of myofibroblasts derived from the infarcted areas of myocardial infarction model rats via the activation of Akt and ERK. A knockdown of endostatin worsened the cardiac function concomitant with wall thinning and decreased α-SMA positive myofibroblasts in the infarcted area. It is thus suggested that endostatin might have an important role in the scar formation and contractile stability after myocardial infarction through the activation of myofibroblasts. In this study, we used 100–3000 ng/mL recombinant mouse endostatin. The blood concentration of endostatin in healthy individuals is ~88.1 ng/mL [34]. In contrast, the serum concentration of endostatin in patients with chronic heart failure is ~229 ng/mL [24]. The concentration of endostatin in coronary sinus serum is significantly elevated in patients with coronary heart disease compared with normal subjects (median 79.7 ng/mL vs. median 49.6 ng/mL) [23]. The serum endostatin level has been reported to be significantly increased at 24 h after myocardial infarction (>100 ng/mL) compared with sham operated rats (<50 ng/mL). In addition, the intensity of immunofluorescent staining for endostatin/collagen XVIII was markedly upregulated in cardiomyocytes in the infarcted area [27]. The myofibroblasts used in this study were isolated from the infarcted area. Since the endostatin level in the infarcted area might be higher than the serum level, the endostatin concentrations (100–3000 ng/mL) used in the present study would be within the pathological ranges. In this study, endostatin induced proliferation and migration of myofibroblasts (Figure3). Endostatin has anti-angiogenic effect through the inhibition of proliferation and migration of vascular endothelial cells via inhibiting phosphorylation of focal adhesion kinase through binding integrin α5β1 or the antagonism of vascular endothelial growth factor receptor [35,36] Alahuhta et al. reported that endostatin induced proliferation of HSC-3 carcinoma cells, a human tongue squamous cell carcinoma cell line [37]. Huang et al. reported that 50 µg/mL recombinant human endostatin inhibited proliferation of synovial fibroblasts in rats with adjuvant arthritis but not normal rats [38]. On the other hand, we previously reported that endostatin induced proliferation and migration of cardiac fibroblasts [29]. Therefore, endostatin might exert different effects on proliferation and migration depending on the cell types. In the present study, endostatin promoted phosphorylation of Akt (Ser473) and ERK (Figure3), and inhibition of PI3K/Akt and MEK/ERK pathways by using wortmannin and PD98059 inhibited the endostatin-induced proliferation and migration, respectively (Figure4). It is known that activation of Akt and ERK is involved in proliferation and migration of myofibroblasts [39–42]. Thus, it is suggested that endostatin induced proliferation and migration via activation of Akt and ERK in cardiac myofibroblasts. Wenzel et al. reported that endostatin accelerated relaxation response via production of nitric monoxide and stimulation of the PI3K/Akt pathway in isolated aortic tissue [43]. Endostatin stimulates reactive oxygen species (ROS) through the increase of intracellular ceramide concentration in bovine coronary artery-derived endothelial cells [44]. Yang et al. reported that plumbagin activated ERK1/2 and Akt via the production of ROS in 3T3-L1 cells, a mouse embryonic fibroblast cell line [45] and that N-acetylcysteine (NAC), an antioxidant, inhibited the plumbagin-induced Akt phosphorylation [45]. Moreover, we previously revealed that NAC inhibited endostatin-induced Akt activation in rat cardiac fibroblasts [29]. Thus, it is supposed that endostatin activates PI3K/Akt and MEK/ERK pathways through the ROS production in myofibroblasts, similar to the case in cardiac fibroblasts. It has been reported that endostatin expression increased in the infarcted area of rat myocardial infarction model [27,28]. Therefore, we examined the effect of endostatin knockdown by COL18A1 siRNA on the remodeling of infarcted area. Interestingly, a knockdown of COL18A1 gene caused a remarkable thinning of the infarcted area, which was characterized by the decrease of LVPWd, LVPWs, and LVPW% and the increase of LVIDs, IVS/LVPW, and ESV (Figure5 and Table2). The endostatin knockdown also worsened cardiac function, which was characterized by the decrease of FS and EF (Figure5 and Table2). In addition, the α-SMA-positive myofibroblasts in the infarcted area of the COL18A1 siRNA-injected rats appeared sparse compared with control siRNA-injected rats (Figure6). Int. J. Mol. Sci. 2018, 19, 741 9 of 15

The insufficient proliferation and migration of myofibroblasts after myocardial infarction is thought to induce thinning of the wall, cardiac dilatation, cardiac dysfunction, and ventricular wall rupture [46]. Furthermore, scar thickness in the infarcted area of COL18A1 siRNA-injected rats were decreased compared with control siRNA-injected rats (Figure7). Therefore, it is suggested that endostatin is necessary for the sufficient proliferation and migration of myofibroblasts in the infarcted area, which contributes to the scar formation. The limitation of this study is that we used a COL18A1 siRNA to knockdown endostatin. Further study is needed to investigate the influence of COL18A1-specific gene knockdown on the healing process after myocardial infarction. In conclusion, we revealed for the first time that endostatin might regulate the healing process in an infarcted area after myocardial infarction through the activation of proliferation and migration of myofibroblasts. These findings indicate endostatin as a new therapeutic target for myocardial infarction.

4. Materials and Methods

4.1. Reagents and Antibodies Reagent sources were as follows: Recombinant mouse endostatin (Sigma-Aldrich, St. Louis, MO, USA), PD98059 (Cayman, Ann Arbor, MI, USA), wortmannin (Wako, Osaka, Japan). Antibodies sources were as follows: anti-phospho-Akt (Ser473), anti-total Akt, anti-phospho-ERK (Cell Signaling Technology, Beverly, MA, USA), anti-total-ERK1, anti-endostatin (Santa Cruz Biotech, Santa Cruz, CA, USA), anti-type I collagen (Rockland immunochemicals, Gilbertsville, PA, USA), anti-MMP-2 (Kyowa pharma chemical, Toyama, Japan), anti-total-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Wako), α-SMA (Dako, Glostrup, Denmark), anti-rabbit immunoglobulin G (IgG) horseradish peroxidase linked whole antibody, anti-mouse IgG horseradish peroxidase linked whole antibody (Amersham Bioscences, Buckinghamshire, UK).

4.2. Myocardial Infarction Model All animal studies were approved by the President of Kitasato University through the judgment by the Institutional Animal Care and Use Committee of Kitasato University (Approval No. 14-126 (13 March 2015), 15-020 (4 March 2016), 16-031 (5 December 2016) and 17-081 (21 April 2017)). Adult male Wistar rats (7–10-week old; CLEA Japan, Tokyo, Japan) were cared for in accordance with the National Institutes of Health Guide for the Care of Laboratory Animals and the Guideline for Animal Care and Treatment of the Kitasato University. Coronary artery ligation of the rats was performed to create a myocardial infarction model as described previously [42]. After the rats were anesthetized with isoflurane by using a vaporizer, endotracheal intubation was connected to a ventilator with a respiratory rate of 100 times per min and tidal volume of 5 mL. Buprenorphine (0.005 mg/100 g) was subcutaneously administered for a preoperative analgesia. Then, a left thoracotomy between third or fourth intercostal space was performed to expose the heart, and the proximal left anterior descending artery was permanently ligated using a 6-0 nylon suture. After the chest was closed, rats recovered. We also performed a knockdown of endostatin in the infarcted area by using an siRNA as described previously [32,47]. It was previously confirmed that the COL18A1 siRNA used in this study decreased endostatin expression in hearts of rats [32]. After the coronary ligation, sixteen µg of COL18A1 siRNA or control siRNA was injected into the infarcted myocardium. After the chest was closed, rats recovered and were cared for 14 days. The sequence of COL18A1 siRNA was as follows: sense 50-UCGUCAACCUGAAGGAUGAdTdT-30 and antisense 50-UCAUCCUUCAGGUUGACG AdTdT-30.

4.3. Isolation of Myofibroblasts from the Areas of Myocardial Infarction Myofibroblasts were isolated from the areas of myocardial infarction as described previously [42]. The myocardial infarction model rats were cared for 2 weeks and the hearts were isolated under Int. J. Mol. Sci. 2018, 19, 741 10 of 15 pentobarbital (80 mg/kg, intraperitoneally injection) anesthesia. The isolated heart was washed with oxygenated Krebs-Henseleit solution (119 mM NaCl, 4.8 mM KCl, 2.5 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4, 24.9 mM NaHCO3, 10.0 mM Glucose) for an exsanguination, and the infarcted area was isolated. The infarcted tissue was cut into 4–5 mm pieces for isolation of myofibroblasts and washed with Tris buffered saline (TBS, pH 7.4) and Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma Aldrich). Then, the piece of tissue was put on a 100 mm-culture dish and incubated in DMEM containing 10% fetal bovine serum (FBS: HyClone/GE Healthcare, Little Chalfont, UK) at 37 ◦C in 5% CO2. The medium was changed every 2–3 days. The piece of tissue was removed 10 days after culture, and the migrated and proliferated cells around the tissue were trypsinized and passaged. By using an immunofluorescence staining, we confirmed that the isolated cells were stained positively with antibody against α-SMA, vimentin, and collagen type I but not CD31, which was consistent with the myofibroblasts phenotype as described previously [42]. Passage 2–9 cells were used and a 24 h starvation in DMEM containing 0.5% FBS was performed before the stimulation.

4.4. Cell Proliferation Assay Cell proliferation was measured by using Cell counting kit-8 (CC8, Dojindo, Kumamoto, Japan) as described previously [42]. After the cells were grown to 30–50% confluence in a 6-well culture plate, they were stimulated with endostatin (100–3000 ng/mL) for 48 h. After washing with TBS, the cells were treated with the CC8 solution (25 µL/0.5 mL medium) for 1 h at 37 ◦C. The absorbance of media at 485 nm was measured using a standard microplate reader (Tristar, Berthold Technologies, Bad Wildbad, Germany). BrdU incorporation assay was also performed to detect the cell proliferation by using a BrdU cell proliferation assay kit (Exalpha biologicals Inc., Shirley, MA, USA) as described previously [48]. After the cells were seeded at a density of 1 × 104 cells/well in a 96-well culture plate and incubated ◦ for 24 h at 37 C and 5% CO2, they were starved with DMEM containing 0.5% FBS for 24 h. After the cells were stimulated with endostatin (100–3000 ng/mL) for 24 h in the presence or absence of wortmannin (1 µM) or PD98059 (50 µM), they were incubated with a BrdU reagent (20 µL/well) for 24 h. After the cells were fixed with a fixing solution (50 µL/well) for 30 min, they were incubated with an anti-BrdU antibody for 1 h and subsequently incubated with a horseradish peroxidase conjugated anti-mouse IgG (1:2000 dilution) for 30 min. After the cells were treated with tetramethylbenzidine solution (50 µL/well) for 30 min, the reaction was stopped by adding a stop solution. The absorbance at 450 nm was measured by using a microplate reader (Berthold Technologies).

4.5. Boyden Chamber Assay Cell migration was detected by a Boyden chamber assay with Transwell chambers (pore size: 8.0 µm, Corning Incorporated, Arizona, AZ, USA) coated with 2% gelatin as described previously [42]. The cells at density of 1 × 105 cells/100 µL were plated in the upper-chamber, and endostatin (300, 3000 ng/mL) was treated in the bottom well. Then they were incubated for 24 h in the presence or absence of wortmannin (100 nM) or PD98059 (50 µM) in both the upper and bottom wells. The migrated cell number in three randomly selected areas was counted with a phase-contrast microscope (CKX-41, OLYMPUS, Tokyo, Japan) equipped with a universal serial bus (USB) camera (MIR-MDCE-5C, Bio Medical Science, Tokyo, Japan), and the normalized cell number relative to vehicle-treated control was evaluated.

4.6. Western Blotting Western blotting was performed to detect the expression and phosphorylation of proteins as described previously [42]. After the stimulation with endostatin (100–3000 ng/mL, 10 min–48 h), the culture media was collected and the total cell lysate was extracted by a cell lysis buffer (Cell Signaling Technology) with protease inhibitors (Nakalai Tesque, Kyoto, Japan). Ten µg of total cell lysate or 20 µL culture media was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis Int. J. Mol. Sci. 2018, 19, 741 11 of 15

(7.5–14%) (80–120 V, 1.5–2 h) and transferred to a nitrocellulose membrane (Pall Corporation, Ann Arbor, MI, USA) (400 mA, 1.5 h). After the membranes were blocked with 0.5% skim milk (for total proteins) or 3% bovine serum albumin (for phosphorylated proteins), the membranes were incubated with a primary antibody at 4 ◦C overnight and then reacted with horseradish peroxidase-conjugated secondary antibody (1:10,000 dilution in TBS, 1 h). The chemiluminescence signal was detected by the EZ-ECL system (Biological Industries, Kibbutz Beit Haemek, Israel). Equal loading of protein was confirmed by measuring total-protein or total-actin. The detected bands were analyzed using CS analyzer 3.0 software (AE-6972; ATTO, Tokyo, Japan).

4.7. Echocardiography Echocardiography was performed under isoflurane anesthesia with iE33 (Philips, WA, USA) as described previously [27,32]. The heart rate of rats was constantly regulated by anesthetic depth, and the M-mode image was obtained from the parasternal short axis view. IVS, LVID, LVPW, EDV, ESV, LV mass, FS, EF, and heart rate were measured in both diastolic and systolic phases from the M-mode images.

4.8. Immunohistochemical Staining After the isolated heart was fixed in 10% neutral buffered formalin, thin paraffin sections (4 µm) were immunohistochemically stained as described previously [42]. First, the deparaffinized section was heated by using a microwave for antigen activation and incubated in methanol with 0.3% H2O2 for 20 min to block endogenous peroxidase activity. Next, the section was blocked with 5% normal goat serum and incubated with primary antibody against α-SMA (1:100 dilution) at 4 ◦C overnight. After washing with TBS, the section was incubated in biotinylated link (Dako) for 60 min and next in streptavidin-horseradish peroxidase (Dako) for 30 min at room temperature. Then, expression of α-SMA was visualized by a liquid DAB+substrate chromogen system (Dako). The images were obtained using a light microscope (BX-51, OLYMPUS) equipped with a microscope digital camera (DP74, OLYMPUS). The α-SMA-positive area to infarcted area was calculated by ImageJ software.

4.9. Azan Staining After the isolated heart was fixed in 10% neutral buffered formalin, thin paraffin sections (4 µm) were applied with Azan staining as described previously [49]. The deparaffinized section was soaked in 5% potassium dichromate solution for 60 min and stained with Azocarmine G (Waldeck, Division Chroma, Münster, Germany) over night. Next, the section was soaked in 3% 12-tungsto-(IV)-phosphoric acid n-hydrate solution for 15 min and stained with aniline blue and orange G solution (Waldeck, Division Chroma) for 60 min. The images were obtained using a light microscope (BX-51) equipped with a microscope digital camera (DP74). The scar thickness was measured in the point of the thickest scar tissue by cellSens Imaging Software (OLYMPUS).

4.10. Statistical Analysis Data are presented as means ± the standard error of the mean. In two-group comparison, statistical analyses were performed by using Student’s t test. In multi-group comparison, statistical analyses were performed by one-way ANOVA followed by Dunnet’s post hoc test (single treatment of endostatin) and Bonferroni’s post hoc test (co-treatment of endostatin and inhibitors). A value of p < 0.05 was considered statistically significant.

Acknowledgments: This research was supported by Kitasato University Research Grant for Young Researchers and JSPS KAKENHI Grant Number 26850186 (Grant-in-Aid for Young Scientists B). Author Contributions: Akira Sugiyama, Yuka Hirano, Muneyoshi Okada, and Hideyuki Yamawaki conceived and designed the experiments; Akira Sugiyama and Yuka Hirano performed the experiments; Akira Sugiyama, Yuka Hirano, and Muneyoshi Okada analyzed the data; Muneyoshi Okada and Hideyuki Yamawaki contributed Int. J. Mol. Sci. 2018, 19, 741 12 of 15 reagents/materials/analysis tools; Muneyoshi Okada acquired the funding; Akira Sugiyama and Yuka Hirano wrote the original draft; Muneyoshi Okada and Hideyuki Yamawaki reviewed and edited the manuscript. Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations

ECM extracellular matrix α-SMA α-smooth muscle actin MMPs matrix metalloproteinases IL interleukin S.E.M. standard error of the mean ERK extracellular signal-regulated kinase PI3K phosphoinositide 3-kinase MAPK mitogen-activated protein kinase MEK MAPK/ERK kinase siRNA small interference RNA LAD left anterior descending artery GAPDH glyceraldehyde 3-phosphate dehydrogenase BW body weight LV left ventricular LVIDd LV internal dimension at end-diastole LVIDs LV internal dimension at end-systole IVSd interventricular septum at end-diastole IVSs interventricular septum at end-systole IVS% percentage IVS thickness LVPWd LV posterior wall at end-diastole LVPWs LV posterior wall at end-systole LVPW% percentage LVPW thickness EDV end-diastolic volume ESV end-systolic volume FS fractional shortening EF ejection fraction ROS reactive oxygen species NAC N-acetylcysteine TBS tris buffered saline DMEM Dulbecco’s Modified Eagle’s Medium FBS fetal bovine serum CC8 cell counting kit-8

References

1. Goldsmith, E.C.; Bradshaw, A.D.; Spinale, F.G. Cellular Mechanisms of Tissue Fibrosis. 2. Contributory pathways leading to myocardial fibrosis: Moving beyond collagen expression. Am. J. Physiol. Cell Physiol. 2013, 304, 393–402. [CrossRef][PubMed] 2. Sproul, E.P.; Argraves, W.S. A cytokine axis regulates elastin formation and degradation. Matrix Biol. 2013, 32, 86–94. [CrossRef][PubMed] 3. Aziz-Seible, R.S.; Casey, C.A. Fibronectin: Functional character and role in alcoholic liver disease. World J. Gastroenterol. 2011, 17, 2482–2499. [CrossRef][PubMed] 4. Wynn, T.A. Cellular and molecular mechanisms of fibrosis. J. Pathol. 2008, 214, 199–210. [CrossRef][PubMed] 5. Ma, Y.; Brás, L.E.D.C.; Toba, H.; Iyer, R.P.; Lindsey, M.L. Myofibroblasts and the extracellular matrix network in post-myocardial infarction cardiac remodeling. Pflügers Arch. Eur. J. Physiol. 2014, 446, 1113–1127. [CrossRef][PubMed] 6. Creemers, E.E.; Pinto, Y.M. Molecular mechanisms that control interstitial fibrosis in the pressure-overloaded heart. Cardiovasc. Res. 2011, 89, 265–272. [CrossRef][PubMed] Int. J. Mol. Sci. 2018, 19, 741 13 of 15

7. Porter, K.E.; Turner, N.A. Cardiac fibroblasts: At the heart of myocardial remodeling. Pharmacol. Ther. 2009, 123, 255–278. [CrossRef][PubMed] 8. Miró, L.; Pérez-bosque, A.; Maijó, M.; Amat, C.; Naftalin, R.J.; Moretó, M. Aldosterone induces myofibroblast EGF secretion to regulate epithelial colonic permeability. Am. J. Physiol. Cell Physiol. 2017, 304, 918–926. [CrossRef][PubMed] 9. Miyagi, H.; Thomasy, S.M.; Russell, P.; Murphy, C.J. The role of hepatocyte growth factor in corneal wound healing. Exp. Eye Res. 2018, 166, 49–55. [CrossRef][PubMed] 10. Neaud, V.; Faouzi, S.; Guirouilh, J.; Le Bail, B.; Balabaud, C.; Bioulac-Sage, P.; Rosenbaum, J. Human hepatic myofibroblasts increase invasiveness of hepatocellular carcinoma cells: Evidence for a role of hepatocyte growth factor. Hepatology 1997, 26, 1458–1466. [CrossRef][PubMed] 11. Felbor, U.; Dreier, L.; Bryant, R.A.; Ploegh, H.L.; Olsen, B.R.; Mothes, W. Secreted cathepsin L generates endostatin from collagen XVIII. Eur. Mol. Biol. Organ. J. 2000, 19, 1187–1194. [CrossRef][PubMed] 12. Heljasvaara, R.; Nyberg, P.; Luostarinen, J.; Parikka, M.; Heikkilä, P.; Rehn, M.; Sorsa, T.; Salo, T.; Pihlajaniemi, T. Generation of biologically active endostatin fragments from human collagen XVIII by distinct matrix metalloproteases. Exp. Cell Res. 2005, 307, 292–304. [CrossRef][PubMed] 13. Ma, D.H.K.; Yao, J.Y.; Kuo, M.T.; See, L.C.; Lin, K.Y.; Chen, S.C.; Chen, J.K.; Chao, A.S.; Wang, S.F.; Lin, K.K. Generation of endostatin by matrix metalloproteinase and cathepsin from human limbocorneal epithelial cells cultivated on amniotic membrane. Investig. Ophthalmol. Vis. Sci. 2007, 48, 644–651. [CrossRef][PubMed] 14. O’Reilly, M.S.; Boehm, T.; Shing, Y.; Fukai, N.; Vasios, G.; Lane, W.S.; Flynn, E.; Birkhead, J.R.; Olsen, B.R.; Folkman, J. Endostatin: An endogenous inhibitor of and tumor growth. Cell 1997, 88, 277–285. [CrossRef] 15. Wen, W.; Moses, M.A.; Wiederschain, D.; Arbiser, J.L.; Folkman, J. The generation of endostatin is mediated by elastase. Cancer Res. 1999, 59, 6052–6056. [PubMed] 16. Dhanabal, M.; Ramchandran, R.; Volk, R.; Simons, M.; Sukhatme, V.P. Endostatin: Yeast Production, Mutants, and Antitumor Effect in Renal Cell Carcinoma. Cancer Res. 1999, 59, 189–197. [PubMed] 17. Dhanabal, M.; Ramchandran, R.; Waterman, M.J.F.; Lu, H.; Knebelmann, B.; Segal, M.; Sukhatme, V.P. Endostatin Induces Endothelial Cell Apoptosis. J. Biol. Chem. 1999, 274, 11721–11726. [CrossRef][PubMed] 18. Lim, J.; Duong, T.; Lee, G.; Seong, B.L.; El-Rifai, W.; Ruley, H.E.; Jo, D. The effect of intracellular protein delivery on the anti-tumor activity of recombinant human endostatin. Biomaterials 2013, 34, 6261–6271. [CrossRef][PubMed] 19. Fu, Y.; Tang, H.; Huang, Y.; Song, N.; Luo, Y. Unraveling the mysteries of endostatin. IUBMB Life 2009, 61, 613–626. [CrossRef][PubMed] 20. Sunshine, S.B.; Dallabrida, S.M.; Durand, E.; Ismail, N.S.; Bazinet, L.; Birsner, A.E.; Sohn, R.; Ikeda, S.; Pu, W.T.; Kulke, M.H.; et al. Endostatin lowers blood pressure via nitric oxide and prevents hypertension associated with VEGF inhibition. Proc. Natl. Acad. Sci. USA 2012, 109, 11306–11311. [CrossRef][PubMed] 21. Yamaguchi, Y.; Takihara, T.; Chambers, R.A.; Veraldi, K.L.; Larregina, A.T.; Feghali-Bostwick, C.A. A Peptide Derived from Endostatin Ameliorates Organ Fibrosis. Sci. Transl. Med. 2012, 4, 136ra7. [CrossRef][PubMed] 22. Kurosaka, D.; Yoshida, K.; Yasuda, J.; Yokoyama, T.; Kingetsu, I.; Yamaguchi, N.; Joh, K.; Matsushima, M.; Saito, S.; Yamada, A. Inhibition of arthritis by systemic administration of endostatin in passive murine collagen induced arthritis. Ann. Rheum. Dis. 2003, 62, 677–679. [CrossRef][PubMed] 23. Mitsuma, W.; Kodama, M.; Hanawa, H.; Ito, M.; Ramadan, M.M.; Hirono, S.; Obata, H.; Okada, S.; Sanada, F.; Yanagawa, T.; et al. Serum Endostatin in the Coronary Circulation of Patients With Coronary Heart Disease and Its Relation to Coronary Collateral Formation. Am. J. Cardiol. 2007, 99, 494–498. [CrossRef][PubMed] 24. Gouya, G.; Siller-Matula, J.M.; Fritzer-Szekeres, M.; Neuhold, S.; Storka, A.; Neuhofer, L.M.; Clodi, M.; Hülsmann, M.; Pacher, R.; Wolzt, M. Association of endostatin with mortality in patients with chronic heart failure. Eur. J. Clin. Investig. 2014, 44, 125–135. [CrossRef][PubMed] 25. Givvimani, S.; Munjal, C.; Gargoum, R.; Sen, U.; Tyagi, N.; Vacek, J.C.; Tyagi, S.C. Hydrogen sulfide mitigates transition from compensatory hypertrophy to heart failure. J. Appl. Physiol. 2011, 110, 1093–1100. [CrossRef] [PubMed] 26. Nikolova, A.; Ablasser, K.; Ballmoos, M.C.W.; Von Poutias, D.; Kaza, E.; Mcgowan, F.X.; Moses, M.A.; Nido, P.J.; Friehs, I. Endogenous Angiogenesis Inhibitors Prevent Adaptive Capillary Growth in Left Ventricular Pressure Overload Hypertrophy. Ann. Thorac. Surg. 2012, 94, 1509–1517. [CrossRef][PubMed] Int. J. Mol. Sci. 2018, 19, 741 14 of 15

27. Isobe, K.; Kuba, K.; Maejima, Y.; Szuki, J.; Kubota, S.; Isobe, M. Inhibition of Endostatin/Collagen XVIII Deteriorates Left Ventricular Remodeling and Heart Failure in Rat Myocardial Infarction Model. Circ. J. 2010, 74, 109–119. [CrossRef][PubMed] 28. Qipshidze, N.; Metreveli, N.; Mishra, P.K.; Lominadze, D.; Tyagi, S.C. Hydrogen sulfide mitigates cardiac remodeling during myocardial infarction via improvement of angiogenesis. Int. J. Biol. Sci. 2012, 8, 430–441. [CrossRef][PubMed] 29. Okada, M.; Oba, Y.; Yamawaki, H. Endostatin stimulates proliferation and migration of adult rat cardiac fibroblasts through PI3K/Akt pathway. Eur. J. Pharmacol. 2015, 750, 20–26. [CrossRef][PubMed] 30. Chen, J.; Liu, W.; Che, H.; Liu, J.; Sun, H. Adenosine-5J-triphosphate up-regulates proliferation of human cardiac fibroblasts. Br. J. Pharmacol. 2012, 16, 1140–1150. [CrossRef][PubMed] 31. Yasuda, J.; Fukui, K.; Okada, M. T3 peptide, a fragment of , stimulates proliferation and migration of cardiac fibroblasts through activation of Akt signaling pathway. N.-S. Arch. Pharmacol. 2017, 390, 1135–1144. [CrossRef][PubMed] 32. Imoto, K.; Kumatani, S.; Okada, M.; Yamawaki, H. Endostatin is protective against monocrotaline-induced right heart disease through the inhibition of T-type Ca2+channel. Pflügers Arch. Eur. J. Physiol. 2016, 468, 1259–1270. [CrossRef][PubMed] 33. Aljinovi´c,J.; Vukojevi´c,K.; Košta, V.; Gui´c,M.M.; Saraga-Babi´c,M.; Grkovi´c,I. Histological differences in healing following experimental transmural infarction in rats. Histol. Histopathol. 2010, 25, 1507–1517. [PubMed] 34. Kłubo-Gwie´zdzinska,J.; Junik, R.; Kopczy´nska,E. Serum endostatin levels in patients with metastatic and non-metastatic well-differentiated thyroid cancer. Endokrynol. Pol. 2010, 61, 7–12. [PubMed] 35. Sudhakar, A.; Sugimoto, H.; Yang, C.; Lively, J.; Zeisberg, M.; Kalluri, R. Human tumstatin and human endostatin exhibit distinct antiangiogenic activities mediated by alpha v beta 3 and alpha 5 beta 1 integrins. Proc. Natl. Acad. Sci. USA 2003, 100, 4766–4771. [CrossRef][PubMed] 36. Kim, Y.M.; Hwang, S.; Kim, Y.M.; Pyun, B.J.; Kim, T.Y.; Lee, S.T.; Gho, Y.S.; Kwon, Y.G. Endostatin blocks vascular endothelial growth factor-mediated signaling via direct interaction with KDR/Flk-1. J. Biol. Chem. 2002, 277, 27872–27879. [CrossRef][PubMed] 37. Alahuhta, I.; Aikio, M.; Väyrynen, O.; Nurmenniemi, S.; Suojanen, J.; Teppo, S.; Pihlajaniemi, T.; Heljasvaara, R.; Salo, T.; Nyberg, P. Endostatin induces proliferation of oral carcinoma cells but its effect on invasion is modified by the . Exp. Cell Res. 2015, 336, 130–140. [CrossRef][PubMed] 38. Huang, X.-Y.; Zhang, X.-M.; Chen, F.-H.; Zhou, L.-L.; Deng, X.-F.; Liu, Y.-J.; Li, X.-J. Anti-proliferative effect of recombinant human endostatin on synovial fibroblasts in rats with adjuvant arthritis. Eur. J. Pharmacol. 2014, 723, 7–14. [CrossRef][PubMed] 39. Benelli, R.; Venè, R.; Minghelli, S.; Carlone, S.; Gatteschi, B.; Ferrari, N. Celecoxib induces proliferation and Amphiregulin production in colon subepithelial myofibroblasts, activating erk1-2 signaling in synergy with EGFR. Cancer Lett. 2013, 328, 73–82. [CrossRef][PubMed] 40. Lee, C.C.; Wang, C.N.; Lee, Y.L.; Tsai, Y.R.; Liu, J.J.; Yang, C. High mobility group box 1 induced human lung myofibroblasts differentiation and enhanced migration by activation of MMP-9. PLoS ONE 2015, 10, e0116393. [CrossRef][PubMed] 41. Uyama, N.; Iimuro, Y.; Kawada, N.; Reynaert, H.; Suzumura, K.; Hirano, T.; Kuroda, N.; Fujimoto, J. Fascin, a novel marker of human hepatic stellate cells, may regulate their proliferation, migration, and collagen gene expression through the FAK-PI3K-Akt pathway. Lab. Investig. 2012, 92, 57–71. [CrossRef][PubMed] 42. Sugiyama, A.; Okada, M.; Yamawaki, H. Pathophysiological roles of canstatin on myofibroblasts after myocardial infarction in rats. Eur. J. Pharmacol. 2017, 807, 32–43. [CrossRef][PubMed] 43. Wenzel, D.; Schmidt, A.; Reimann, K.; Hescheler, J.; Pfitzer, G.; Bloch, W.; Fleischmann, B.K. Endostatin, the proteolytic fragment of collagen XVIII, induces vasorelaxation. Circ. Res. 2006, 98, 1203–1211. [CrossRef] [PubMed] 44. Zhang, A.Y.; Teggatz, E.G.; Zou, A.; Campbell, W.B.; Li, P.-L. Endostatin uncouples NO and Ca2+ response to − bradykinin through enhanced O2 production in the intact coronary endothelium. Am. J. Physiol. Heart Circ. Physiol. 2005, 288, 686–694. [CrossRef][PubMed] 45. Yang, S.J.; Chang, S.C.; Wen, H.C.; Chen, C.Y.; Liao, J.F.; Chang, C.H. Plumbagin activates ERK1/2 and Akt via superoxide, Src and PI3-kinase in 3T3-L1 Cells. Eur. J. Pharmacol. 2010, 638, 21–28. [CrossRef][PubMed] Int. J. Mol. Sci. 2018, 19, 741 15 of 15

46. Turner, N.A.; Porter, K.E. Function and fate of myofibroblasts after myocardial infarction. Fibrogenesis Tissue Repair 2013, 6, 5. [CrossRef][PubMed] 47. Huang, K.; Liu, J.; Zhang, H.; Wang, J.; Li, H. Intramyocardial Injection of siRNAs Can Efficiently Establish Myocardial Tissue-Specific Renalase Knockdown Mouse Model. BioMed Res. Int. 2016, 2016.[CrossRef] [PubMed] 48. Usui, T.; Sakatsume, T.; Nijima, R.; Otani, K.; Kazama, K.; Morita, T.; Kameshima, S.; Okada, M.; Yamawaki, H. Death-associated protein kinase 3 mediates vascular structural remodelling via stimulating smooth muscle cell proliferation and migration. Clin. Sci. (Lond.) 2014, 127, 539–548. [CrossRef][PubMed] 49. Sakamoto, Y.; Kameshima, S.; Kakuda, C.; Okamura, Y.; Kodama, T.; Okada, M.; Yamawaki, H. Visceral adipose tissue-derived serine protease inhibitor prevents the development of monocrotaline-induced pulmonary arterial hypertension in rats. Pflügers Arch. Eur. J. Physiol. 2017, 1425–1432. [CrossRef][PubMed]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).