Food and Chemical Toxicology 68 (2014) 23–29

Contents lists available at ScienceDirect

Food and Chemical Toxicology

journal homepage: www.elsevier.com/locate/foodchemtox

Withaferin A is an inhibitor of endothelial protein C receptor shedding in vitro and in vivo ⇑ Sae-Kwang Ku a, Min-Su Han b, Jong-Sup Bae c, a Department of Anatomy and Histology, College of Korean Medicine, Daegu Haany University, Gyeongsan 712-715, Republic of Korea b Laboratory for Arthritis and Bone Biology, Fatima Research Institute, Daegu Fatima Hospital, Daegu 701-724, Republic of Korea c College of Pharmacy, CMRI, Research Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 702-701, Republic of Korea article info abstract

Article history: Withaferin A (WFA), an active compound from , has been widely researched for its Received 28 October 2013 anti-inflammatory and cardioactive properties and effects on the central nervous system. The endothelial Accepted 5 March 2014 cell protein C receptor (EPCR) plays important roles in blood coagulation and inflammation. EPCR activity Available online 13 March 2014 is markedly changed by ectodomain cleavage and release as the soluble EPCR. EPCR is shed from the cell surface, mediated by -a converting enzyme (TACE). In this study, we investigated Keywords: the effects of WFA on the EPCR shedding in human umbilical vein endothelial cells (HUVECs) and in mice Withaferin A and the associated signaling pathways. WFA was found to induce inhibition of phorbol-12-myristate 13- EPCR shedding acetate (PMA), tumor necrosis factor (TNF)- , interleukin (IL)-1b, and on cecal ligation and puncture PMA a CLP (CLP)-induced EPCR shedding and WFA suppressed the expression and activity of TACE. In addition, treat- ment with WFA resulted in reduced PMA-stimulated phosphorylation of p38, extracellular regulated kinases (ERK) 1/2, and c-Jun N-terminal kinase (JNK). These results demonstrate a therapeutic potential- ity of WFA as an anti-sEPCR shedding reagent against PMA and CLP-mediated EPCR shedding. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction sequence (Fukudome and Esmon, 1994; Villoutreix et al., 1999). The gene is located on chromosome 20 (Hayashi et al., 1999), it The protein C (PC) system is an important natural anticoagulant spans 8 kilobase (kb) and comprises 4 exons (Hayashi et al., mechanism, involving proteolytic degradation of the procoagulant 1999; Simmonds and Lane, 1999). cofactors factor Va and VIIIa, through activated protein C (APC) A soluble form of EPCR (sEPCR) is generated from the mem- (Fulcher et al., 1984). PC, a vitamin K-dependent zymogen, is acti- brane bound EPCR through proteolytic cleavage by metallopro- vated at the endothelial surface when thrombin binds to thrombo- tease activity that can be induced by thrombin and other modulin, a protein that transforms the procoagulant enzyme into a inflammatory mediators (Xu et al., 2000), a process that is called potent activator of PC (Mosnier et al., 2007). Another endothelial shedding. sEPCR circulates in plasma and retains its ability to bind factor that contributes to the PC anticoagulant pathway is the both PC and APC but does not enhance protein C activation endothelial cell protein C receptor (EPCR) (Fukudome and Esmon, (Kurosawa et al., 1997). It inhibits APC anticoagulant activity by 1994). This receptor, which can bind to PC or APC with the same forming a complex that involves phospholipid membranes (Liaw affinity (dissociation constant [Kd] = 30 nM), is mainly expressed et al., 2000). sEPCR resulting from shedding of membrane EPCR, on endothelial cells of large vessels (Mosnier et al., 2007). EPCR can be detected in plasma concentration of approximately is a 46-kDa type 1 transmembrane glycoprotein homologous to 100 ng/m; high levels of sEPCR have been reportedly associated major histocompatibility complex class I family proteins with systemic inflammatory diseases (Kurosawa et al., 1998). In vi- (Fukudome and Esmon, 1994; Villoutreix et al., 1999). This tro studies have described a dramatic increase in EPCR shedding 221-amino-acid (aa) protein comprises an extracellular domain, a from the by a wide variety of inflammatory mediators

25-aa transmembrane domain, and a short (3 aa) intracytoplasmic (IL-1b,H2O2, and phorbol myristate acetate) and thrombin, and EPCR shedding is potentiated by the microtubule disrupting agent, nocodazole (Xu et al., 2000). In addition, phosphorylation of mito- ⇑ Corresponding author. Address: College of Pharmacy, CMRI, Research Institute gen-activated protein (MAP) kinases such as p38, ERK1/2, and JNK of Pharmaceutical Sciences, Kyungpook National University, 80 Daehak-ro, Buk-gu, Daegu 702-701, Republic of Korea. Tel.: +82 53 950 8570; fax: +82 53 950 8557. were shown to increased by stimulation with PMA (Han et al., E-mail address: [email protected] (J.-S. Bae). 2010; Leng et al., 2004; Menschikowski et al., 2009) and activation http://dx.doi.org/10.1016/j.fct.2014.03.009 0278-6915/Ó 2014 Elsevier Ltd. All rights reserved. 24 S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29 of TACE was shown to occurs upon activation of ERK or p38 (Huov- by treatment with PMA, tumor necrosis factor (TNF)-a, or interleukin (IL)-1b for ila et al., 2005; Murphy, 2008). Therefore, the shedding of EPCR 1 h. Media were then removed, and cells were washed with PBS and fixed with 50 ll of 1% paraformaldehyde for 15 min at room temperature. After washing, might be connected with the activation of MAPK or TACE directly 100 ll of rabbit polyclonal EPCR antibodies (Abnova, Taipei City, Taiwan) was or indirectly. added, and, 1 h (37 °C, 5% CO2) later, cells were washed 3 times, followed by treat- The search for anticancer drugs and anti-inflammatory agents ment with 100 ll of 1:2000 peroxidase-conjugated anti-rabbit IgG antibodies (Sig- from natural products represents an area of interest worldwide ma, St. Louis, MO, USA) for 1 h. Cells were then washed 3 times and developed using (Aggarwal et al., 2006). Withania somnifera has been used to pre- o-phenylenediamine substrate (Sigma, St. Louis, MO, USA). Colorimetric analysis was performed by measurement of absorbance at 490 nm. All measurements were vent infections of burns, wounds, and dermatological disorders performed in triplicate wells. (Essawi and Srour, 2000). Withaferin A (WFA, Fig. 1) is a steroidal derived from W. somnifera, a plant that has been used for 2.5. Competitive competitive ELISA for sEPCR and TACE centuries to treat several inflammatory disorders (Kaileh et al., 2007). However, the effect of WFA on EPCR shedding and its under- Ninety-six-well flat microtiter plates (Corning, NY) were coated overnight at lying mechanisms in both cellular system and animal model have 4 °C with sEPCR protein for determination of sEPCR or TACE protein for determina- tion of TACE in 20 mM carbonate–bicarbonate buffer (pH 9.6) containing 0.02% so- not yet been elucidated. Anti-inflammatory effects of WFA in HU- dium azide. Lyophilized culture media were prepared for sEPCR, and total cell VECs and mice have recently been reported (Lee et al., 2012). Pre- lysates were prepared for TACE using lysis buffer containing (mM): Tris–HCl (20) vious report showed that PMA-stimulated EPCR shedding is pH 7.5, EGTA (0.5), EDTA (2), dithiothreitol (2), p-methylsulfonyl fluoride (0.5), mediated by tumor necrosis factor-a converting enzyme/ADAM17 and 10 lg/ml leupeptin. Plates were then rinsed 3 times in PBS-T and kept at 4 °C. Prepared samples from cell culture media and mice plasma for sEPCR or from (TACE) (Qu et al., 2007). Noting that sEPCR serves as a marker of cell lysates for TACE were pre-incubated with anti-EPCR antibodies (rabbit poly- vascular barrier integrity in vascular inflammatory disease, and clonal, 1:500, Abnova, Taipei City, Taiwan) or anti-TACE antibodies (goat polyclonal, sEPCR is involved in the pathophysiology of sepsis (Borgel et al., 1:500, Santa Cruz, Dallas, Texas, USA) in 96-well plastic round microtiter plates for 2007; Kurosawa et al., 1998); we hypothesized that WFA may have 90 min at 37 °C, transferred to pre-coated plates, and incubated for 30 min at room anti-sEPCR shedding activity. Therefore, in the present study, we temperature. Plates were then rinsed 3 times with PBS-T, incubated for 90 min at room temperature with peroxidase-conjugated anti-rabbit or anti-goat IgG anti- investigated the effect of WFA on the expression and activity of bodies (1:2000, Amersham Pharmacia Biotech, Uppsala Sweden), rinsed three times TACE and against PMA-induced EPCR shedding in human endothe- in PBS-T, and incubated for 60 min at room temperature in the dark with 200 llof lial cells and in a cecal ligation and puncture (CLP) model of septi- substrate solution (100 lg/ml o-phenylenediamine containing 0.003% H2O2). The cemia in mice. reaction was then stopped by the addition of 50 llof8NH2SO4, and absorbances at 490 nm was measured.

2. Materials and methods 2.6. TACE activity assay

2.1. Reagents For TACE activity assay, commercially available TACE activity kit (Innozyme TACE activity assay kit, EMD Millipore, Billerica, MA, USA) was used as described WFA (Fig. 1) was purchased from Biomol (Plymouth Meeting, PA, USA). sEPCR previously (Miller et al., 2013). Confluent monolayers of HUVECs in 12-well culture and TNF-a were purchased from Abnova (Taipei City, Taiwan). Phorbol-12-myris- plates were treated with or without WFA for 6 h, followed by treatment with PMA tate 13-acetate (PMA) and IL-1b were purchased from Sigma (St. Louis, MO, USA). (1 lM) for 1 h. Media were then removed, and total cell lysates were prepared using the CytoBuster™ Protein Extraction Reagent (EMD Millipore, Darmstadt, Germany). 2.2. Cell culture Total cell lysates were applied to plates pre-coated with a monoclonal antibody specific for human TACE. TACE activity was measured using an internally quenched

Primary human umbilical vein endothelial cells (HUVECs) were obtained from fluorescent substrate, MCA-KPLGL-Dpa-AR-NH2. The cleavage rates were detected Cambrex Bio Science (Charles City, IA) and maintained as previously described within the dynamic range of the assay according to a standard curve. (Bae and Rezaie, 2011). HUVECs of passage numbers 3 or 4 were used in the experiments. 2.7. Cecal ligation and puncture (CLP)

2.3. Animals and husbandry For induction of sepsis, male mice were anesthetized with zoletil 50 and rom- pun. The CLP-induced sepsis model was prepared as previously described (Wang Male C57BL/6 mice (6–7-weeks old, weighting 18–20 g) were purchased from et al., 2004). In brief, a 2 cm midline incision was made in order to allow exposure Orient Bio Co. (Sungnam, KyungKiDo, Korea), and used after a 12-day acclimatiza- of the cecum and adjoining intestine. The cecum was then tightly ligated using a tion period. Five animals per polycarbonate cage were housed under controlled con- 3.0-silk suture (SK34510, Ailee, South Korea) at 5.0 mm from the cecal tip, punc- ditions (20–25 °C/RH 40–45%) under a 12-h light/dark cycle, and supplied a normal tured once with a 22-gauge needle, gently squeezed in order to extrude a small rodent pellet diet and water ad libitum. All animals were treated in accordance with amount of feces, and returned to the peritoneal cavity. The laparotomy site was the Guidelines for the Care and Use of Laboratory Animals issued by Kyungpook Na- then stitched with 4.0-silk suture (SK430, Ailee, South Korea). In sham controls, tional University. the cecum was exposed but not ligated or punctured and then returned to the abdominal cavity. After 12 h or/and 50 h CLP surgery, WFA (1.882 lg per mouse, 2.4. Enzyme-linked immunosorbent assay (ELISA) for cellular EPCR expression n = 5) was administrated intravenously. In each day after CLP, blood was collected via intravenously. Or, after 4 days CLP, mouse was sacrificed by cervical dislocation Modified whole-cell ELISA was performed as previously described to determine for immunohistochemistry. This protocol was approved in advance by the Animal expression levels of EPCR on HUVECs (Kim et al., 2011; Lee et al., 2013). Briefly, con- Care Committee at Kyungpook National University (KNU 2012–13). fluent monolayers of HUVECs were treated with or without WFA for 6 h, followed 2.8. Immunohistochemistry

To analyze the expression pattern of EPCR, aortas from CLP-induced septic- (Day 4) and sham operated mice were removed and fixed in 4% formaldehyde solu- tion (Junsei, Japan) in PBS for 20 h at 4 °C. After fixation, the aortas were dehydrated through an ethanol series, embedded in paraffin, and cut into 3 lm sections. Depa-

raffinized sections were quenched in 3% H2O2 in methanol; washed in PBS; placed in boiled 1 mM Tris solution (pH 9.0), supplemented with 0.5 mM EGTA solution in order to reveal the antigens; and blocked for 1 h at RT in PBS, supplemented with 1% bovine serum albumin, 0.2% gelatin, and 0.05% saponin. Sections were incubated with anti-rabbit EPCR antibody (Abcam, Cambridge, MA, USA) diluted 1:500 in PBS, and supplemented with 0.1% BSA and 0.3% Triton X 100 for 16 h at 4 °Cina humidified chamber. After washing in PBS, supplemented with 0.1% BSA, 0.2% gel- atin, and 0.05% saponin, the sections were incubated with peroxidase-conjugated anti-rabbit IgG antibody (DAKO, Glostrup, Denmark) for 1 h at RT and then devel- Fig. 1. Structure of Withaferin A (WFA). oped using the Liquid DAB + Substrate-Chromogen System (DAKO, Glostrup, Den- S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29 25

Table 1 acid (B). The initial composition was increased to 95% solvent (B) for 3 min. A gra- Pharmacokinetic parameters of WFA after administration of 5 mg/kg (n = 5). dient program was used for HPLC at a flow rate of 230 ml/min. Electrospray ioniza- tion was performed in positive mode at a spray voltage of 3500 V. Nitrogen was Parameters Intravenous injection used as the sheath and auxiliary gas at optimum values of 45 and 20 (arbitrary AUCa 8.4 lg h/ml units), respectively. Vaporizer and capillary temperatures were 150 and 300 °C, respectively. Multiple reaction monitoring (MRM) detection was employed using Half-life (t1/2) 1.9 h nitrogen as the collision as each transition monitored was 479 ? 317 for isorham- Cmax 3.69 lg/ml (7.83 lM) netin 3-O-galactoside, 465 ? 303 for hyperoside, and 609 ? 174 for reserpine (IS). a AUC = area under the curve.

2.11. Statistical analysis mark). Counterstaining was performed with 0.5% methyl green in ddH2O. Non-im- Results are expressed as mean ± standard error of the mean (SEM) of at least mune rabbit IgG (DAKO, at the same concentration instead of EPCR antibody) and three independent experiments with duplicate determination. Statistical signifi- anti-rabbit CD31 antibody (1:200, Abcam, Cambridge, MA, USA) was used as nega- cance was defined to be based on a p value smaller than 0.05 (SPSS, version 14.0, tive and positive control for immunohistochemistry, respectively. SPSS Science, Chicago, Il, USA).

2.9. ELISA for total and phospho p-38MAPK, ERK1/2, and JNK 3. Results and discussion HUVECs were cultured in 96-well microtiter plates for quantitative determina- tion of p38MAPK, ERK1/2, and JNK phosphorylation. On the day of experiments, cul- In this study, the effects of WFA (Fig. 1) on the shedding of EPCR ture medium was replaced by serum-free growth medium. Cells were then treated were determined in vitro and in vivo. The pharmacokinetic values with or without WFA for 6 h, followed by treatment with PMA (1 lM) for 1 h. Acti- in plasma of WFA are shown in Table 1. vation of p38MAPK, ERK 1/2, and JNK was quantified in nuclear lysates using ELISA kits for total/phosphorylated p38MAPK (Life Technologies, Carlsbad, California, for total p38MAPK or Cell Signaling Technology, Danvers, MA, USA, for phosphory- 3.1. Effect of WFA on PMA, TNF-a, or IL-1b-induced EPCR shedding lated-p38MAPK), total/phospho ERK1/2 and JNK (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. Previous studies have reported that PMA stimulates EPCR shed- ding from HUVECs (Qu et al., 2006, 2007). In agreement with the 2.10. Pharmacokinetic parameters previous results, we found that as little as 0.1 lM PMA (Fig. 2A and B) could fully stimulate EPCR shedding from HUVECs After an intravenous penile vein injection of male C57BL/6 mice with WFA at 5 mg/kg, approximately 100 ll of blood was obtained from the tail vein immedi- (Fig. 2A) and that cellular EPCR on HUVECs showed a PMA- ately before and 5, 15, 30, 60, and 180 min after administration of the dose. Follow- mediated dose-dependent decrease (Fig. 2B). EPCR shedding by ing addition of 80 ll acetonitrile and internal standard solution (reserpine), 10 ll TNF-a or interleukin (IL)-1b also showed an increase (Fig. 2C and plasma samples were prepared by centrifugation of the blood at 13,000 rpm for D), in agreement with a previous study (Menschikowski et al., 15 min at 4 °C; 2 ll supernatants were then injected into an HPLC column. The pharmacokinetic parameters were determined using the standard non-compart- 2009). mental method. Serum AUC was calculated using WinNonlin (version 2.0, Scientific To investigate the effect of WFA on PMA-mediated EPCR shed- Consulting, KY, USA) with a log linear trapezoidal method. ding, endothelial cells were pretreated with increasing concentra- The samples were analyzed using an Accela™ LC system coupled to a TSQ Van- tions of WFA for 6 h, followed by stimulation with 1 lM PMA for tage triple quadrupole mass spectrometer (Thermo Fisher Scientific Inc., USA) 1 h. As shown in Fig. 3A and B, treatment with WFA resulted in equipped with a HESI-II Spray source. An ACEÒ 5C18, 3 lm (2.1 Â 50 mm, ACE) col- umn was used for the LC analysis. The mobile phases consisted of LC gradeÒ water inhibition of EPCR shedding induced by PMA in endothelial cells, containing 0.1% formic acid (A) and LC grade acetonitrile containing 0.1% formic with an optimal effect at 1–2 lM. WFA alone (2 lM) did not affect

Fig. 2. Effect of PMA, TNF-a, and IL-1b on EPCR shedding. The effects of various concentrations of PMA (1 h) on EPCR shedding were monitored by measurement of sEPCR (A) or cellular EPCR on HUVECs (B). (C and D) The same as A and B, except that HUVECs were incubated with TNF-a (1 h, white bar) or IL-1b (1 h, black bar). Results indicate the mean ± SEM of 3 separate experiments with duplicate determination. *p < 0.05 or **p < 0.01 compared to ‘‘0’’ control. 26 S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29

Fig. 3. Effect of WFA on PMA, TNF-a, and IL-1b-induced EPCR shedding. The effects of various concentrations of WFA on PMA (1 lM, 1 h)-induced EPCR shedding were monitored by measurement of sEPCR (A) or cellular EPCR on HUVECs (B). (C and D) The same as A and B, except that HUVECs were incubated with TNF-a (25 ng/ml for 1 h, white bar) or IL-1b (25 ng/ml for 1 h, black bar). Results indicate the mean ± SEM of 3 separate experiments with duplicate determination. **p < 0.01 vs. PMA alone (A and B) or TNF-a/IL-1b alone (C and D). the shedding of EPCR. Therefore, WFA alone (2 lM) did not affect twice (1.882 lg per mouse, once 12 h, then 50 h after CLP), result- the total levels of membrane bound EPCR (Fig. 3B). ing in a decrease in EPCR shedding (Fig. 5C). Assuming an average To confirm the inhibitory effects of WFA on EPCR shedding, body weight of 20 g and an average blood volume of 2 ml if no TNF-a or interleukin (IL)-1b was used because previous reports plasma proteins bind to WFA, the amounts of WFA produced a con- showed that EPCR was shed by TNF-a or IL-1b in HUVECs centration of approximately 2 lM in peripheral blood. This marked (Menschikowski et al., 2009). In agreement with previous results, benefit achieved by administration of WFA suggested that inhibi- we found that EPCR shedding by TNF-a or interleukin (IL)-1b in- tion of EPCR shedding provides a therapeutic strategy for manage- creased and WFA also suppressed TNF-a or IL-1b-mediated EPCR ment of severe vascular diseases. shedding in HUVECs (Fig. 3C and D). Because TNF-a and IL-1b have been shown to be important mediators during endotoxemia 3.4. Effects of WFA on PMA-stimulated phosphorylation of p38MAPK, (Kremer et al., 1996; Michie et al., 1988), current finding showing ERK1/2, and JNK that WFA inhibited TNF-a or IL-1b-mediated EPCR shedding could support our recent finding that WFA have anti-inflammatory re- Previous studies have reported the involvement of p38MAPK, sponses in human endothelial cells and in CLP-induced septic mice ERK1/2, and JNK in cytokine-induced EPCR shedding and phos- (Lee et al., 2012). phorylation of p38MAPK, ERK1/2, and JNK was known to be in- creased by stimulation with PMA (Han et al., 2010; Leng et al., 3.2. Effects of WFA on PMA-stimulated expression and activity of TACE 2004; Menschikowski et al., 2009). Therefore, in order to deter- mine the molecular mechanisms of suppression of PMA-induced A previous study reported that PMA-stimulated EPCR shedding EPCR shedding by WFA, the effects of WFA on PMA-stimulated is mediated by tumor necrosis factor-a converting enzyme/ phosphorylation of p38MAPK, ERK1/2, and JNK were tested. Treat- ADAM17 (TACE) (Qu et al., 2007). In order to determine whether ment with WFA (1–2 lM) resulted in reduction of PMA-stimulated WFA could inhibit stimulation of TACE expression and activity, phosphorylation of p38MAPK (Fig. 6A), ERK1/2 (Fig. 6B), and JNK endothelial cells were pretreated with increasing concentrations (Fig. 6C). To confirm the involvement of MAPK on the EPCR shed- of WFA for 6 h, followed by stimulation with 1 lM PMA for 1 h. ding, a panel of known pharmacological inhibitors, SB-203580 Data showed that WFA inhibited TACE expression (Fig. 4A) and (an inhibitor of p38MAPK), PD-98059 (an inhibitor of ERK 1/2), activity (Fig. 4B) induced by PMA in endothelial cells. and SP-600125 (an inhibitor of JNK), respectively, was cross-inves- tigated by using HUVECs. Distinct attenuation of sEPCR release in 3.3. Effect of WFA on CLP-induced EPCR shedding HUVECs was observed after treatment with PD-98059, SB-203580, and SP-600125 (Fig. 6D). These results indicate that To confirm the inhibitory effects of WFA on EPCR shedding WFA inhibited PMA-stimulated phosphorylation of p38, ERK1/2, in vivo, we used a CLP mouse model, because this model more clo- and JNK. Therefore, WFA might inhibit the EPCR shedding by sup- sely resembles human sepsis (Buras et al., 2005; Yang et al., 2004). pressing PMA-stimulated activation of MAP kinases including p38, In CLP-induced septic mice, immunohistochemical analysis ERK and JNK directly or indirectly. showed that lower expression of cellular EPCR than that of the con- Metalloproteinase-mediated ectodomain shedding has been re- trol in aorta of mouse (Fig. 5A). Administration of WFA at a single ported for many cellular receptors (Moss and Lambert, 2002). TACE dose (1.882 lg, 12 h after CLP) did not result in prevention of CLP- (ADAM17) is an important member of the ADAM (a disintegrin and induced EPCR shedding (Fig. 5B); therefore, it was administered metalloproteinase) family (Blobel, 2005). TACE and closely related S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29 27

Fig. 4. Effect of WFA on PMA-stimulated TACE expression and TACE activity. The effects of various concentrations of WFA on PMA (1 lM, 1 h)-induced expression (A) or TACE activity (B) were monitored by TACE ELISA (A) or TACE activity assay kit. All results indicate the mean ± SEM of 3 separate experiments with duplicate determination. *p < 0.05 or **p < 0.01 vs. PMA alone.

Fig. 5. Effect of WFA on CLP-induced EPCR shedding. (A) Immunohistochemical stains of blood vessel for cellular EPCR from CLP-operated mice (four days after CLP), as indicated in the text. Staining results from sham operated mice are compared. Results are representative of 3–6 stainings from 2 independent experiments per condition. (B and C) Serum was obtained from sham-operated (white bar) or CLP-induced septic mice (gray bar) on the indicated day after CLP surgery (n = 5). Alternatively, WFA was administered intravenously once (B, black bar, 1.882 lg per mouse via i.v., once 12 h after CLP) or twice (C, black bar, 1.882 lg per mouse via i.v., once 12 h, then 50 h after CLP). EPCR shedding was then monitored by measurement of sEPCR. CD31 is the endothelial cell (EC) marker. All results indicate the mean ± SEM of five separate experiments. **p < 0.01 vs. CLP alone. matrix metalloproteinases work together as sheddases to cleave There is increasing evidence that diminished EPCR expression hundreds of diverse transmembrane substrates, including TNF-a and function can contribute to chronic inflammation and autoim- (Black et al., 1997), transforming growth factor-a, L-selectin (Lee mune disease. For example, individuals with active inflammatory et al., 2003), b-amyloid precursor protein (Buxbaum et al., 1998), bowel disease exhibit depleted EPCR expression and increased and growth hormone receptor (Zhang et al., 2000). Unfortunately, EPCR shedding on their colonic mucosal microvasculature, caused little is known regarding which inhibitors modulate the activity by local generation of TNF-a and IL-1b in the inflamed local envi- or expression of sheddase and how such a broad palette of proteo- ronment (Scaldaferri et al., 2007). The deleterious effect of vascular lytic activity integrates to modulate behaviors. Furthermore, ther- EPCR depletion in vivo was highlighted by mice with dextran- apeutics have targeted sheddases and their substrates, yet many of sodium sulfate-induced colitis, who had lost the ability to these inhibitors have failed in clinical trials (Fingleton, 2008). effectively generate APC. Replacement of the missing APC by re- Therefore, a need exists for finding new inhibitors of TACE- combinant APC administration reduced disease activity, weight mediated degradation, which integrates multiple layers of signal- loss, and mucosal inflammation by inhibition of chemokine pro- ing networks to coordinately influence cell behavior in various duction and leukocyte adhesion to the colonic microvascular endo- vascular inflammatory diseases. thelium (Scaldaferri et al., 2007). Furthermore, the levels of plasma 28 S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29

Fig. 6. Effect of WFA on PMA-induced phosphorylation of p38MAPK, ERK1/2, and JNK. PMA (1 lM, 1 h)-mediated phosphorylation of phospho-p38MAPK (A, white bar) or total p38MAPK (A, black bar), phospho-ERK1/2 (B, white bar) or total ERK1/2 (B, black bar) and phospho-JNK (C, white bar) or total JNK (C, black bar) were analyzed after treatment of cells with the indicated concentrations of WFA. Results are expressed as fold increase over control values. (D) Cells were pre-incubated with 50 lM PD-98059, 10 lM SB-203580, or 20 lM SP-600125 as indicated for 30 min and thereafter exposed to PMA at a final concentration of 1 lM for an additional hour. EPCR shedding was monitored by sEPCR ELISA. All results indicate the mean ± SEM of 3 separate experiments with duplicate determination. **p < 0.01 vs. PMA alone.

sEPCR were elevated in patients with inflammatory diseases HUVECs and mice (Lee et al., 2012). We reported that WFA (Kurosawa et al., 1998; Sesin et al., 2005; Ware et al., 2006). These inhibited lipopolysaccharide (LPS)-induced HMGB1 release and previous findings may support the fact that increased shedding of HMGB1-mediated barrier disruption, expression of cell adhesion EPCR is associated with the hypercoagulopathy frequently ob- molecules (CAMs) and adhesion/transendothelial migration of leu- served in inflammatory conditions (Esmon, 2005). EPCR can con- kocytes to human endothelial cells. WFA also suppressed acetic tribute to anti-inflammatory mechanisms that involve acid-induced hyperpermeability and carboxymethylcellulose-in- activated receptor-1 on the surface of endothelial cells because duced leukocytes migration in vivo. Further studies revealed that sEPCR effectively inhibits the anti-inflammatory and anti-apopto- WFA suppressed the production of IL-6, TNF-a and activation of tic functions of APC in human endothelial cells (Bae et al., 2007). nuclear factor-jB (NF-jB) by HMGB1 (Lee et al., 2012). In addition, given the established link between chronic inflamma- Collectively, the results of this study show that WFA induced tion and vascular disease, the impairment of the EPCR-dependent potent inhibition of PMA, TNF-a, IL-1b and CLP-induced EPCR APC cytoprotective function may also contribute to the deleterious shedding and it suppressed the expression and activity of TACE. effect of anti-EPCR auto-antibodies in vascular disease. In this Noting that EPCR shedding is an important step involved in the study, we demonstrated the inhibitory effects of WFA on EPCR pathophysiological pathway of vascular inflammatory diseases; shedding by PMA-, TNF-a-, IL-1b-, and CLP-mediated EPCR shed- the hypothesis that WFA could be used as a candidate therapeutic ding. Therefore, WFA could be another potential candidate for for treatment of vascular inflammatory diseases has strengthened treatment of vascular inflammatory diseases. by the finding of our previous study (Lee et al., 2012) and the cur- Phosphorylation levels of p38MAPK, ERK1/2, and JNK are in- rent findings. Even though the use of WFA for therapeutic purposes creased by stimulation with PMA (Han et al., 2010; Leng et al., could have non-specific effects, the results of this study provide 2004; Menschikowski et al., 2009), and activation of TACE occurs novel information on the role of WFA on EPCR shedding. Therefore, upon activation of ERK or p38 MAPK (Huovila et al., 2005; Murphy, our findings suggest that WFA be a potential candidate for treat- 2008). Therefore, PMA-activated TACE is the downstream signaling ment against severe vascular inflammatory diseases, such as sepsis molecules of MAP kinases. In order to define the processes respon- and septic shock. sible for inhibition of PMA-stimulated shedding of EPCR and expression of TACE by WFA, we investigated the involvement of Conflict of Interest MAPK signaling pathways under PMA-stimulated conditions. MAP- Ks comprise a family of highly conserved serine/threonine protein The authors declare that there are no conflicts of interest. kinases implicated as having key regulatory roles in mediating inflammation (Thalhamer et al., 2008). Three major classes of MAP- Transparency document Ks are represented by ERK 1/2 and the 2 stress-activated protein kinase families JNK and p38 MAPK. As shown in Figs. 4 and 6, The Transparency document associated with this article can be WFA remarkably inhibited PMA-stimulated expression of TACE found in the online version. and phosphorylation of p38 MAPK, ERK1/2, and JNK to show WFA as an inhibitor of PMA-stimulated EPCR shedding. Therefore, WFA inhibited PMA-stimulated shedding of EPCR via indirectly Acknowledgement inhibiting the expression and activity of TACE and MAPK. We recently reported an anti-inflammatory property of WFA This study was supported by the National Research Foundation against high mobility group box 1 protein (HMGB1) and the asso- of Korea (NRF) funded by the Korea government [MSIP] (Grant No. ciated signaling pathways-induced inflammatory response using 2013-067053). S.-K. Ku et al. / Food and Chemical Toxicology 68 (2014) 23–29 29

References Lee, W., Kim, T.H., Ku, S.K., Min, K.J., Lee, H.S., Kwon, T.K., Bae, J.S., 2012. Barrier protective effects of Withaferin A in HMGB1-induced inflammatory responses in both cellular and animal models. Toxicol. Appl. Pharmacol. 262, 91–98. Aggarwal, B.B., Ichikawa, H., Garodia, P., Weerasinghe, P., Sethi, G., Bhatt, I.D., Lee, W., Ku, S.K., Bae, J.S., 2013. Emodin-6-O-beta-D-glucoside down-regulates Pandey, M.K., Shishodia, S., Nair, M.G., 2006. From traditional Ayurvedic endothelial protein C receptor shedding. Arch. Pharm. Res. 36, 1160–1165. medicine to modern medicine: identification of therapeutic targets for Leng, Y., Steiler, T.L., Zierath, J.R., 2004. Effects of insulin, contraction, and phorbol suppression of inflammation and . Expert Opin. Ther. Targets 10, 87–118. on mitogen-activated protein kinase signaling in skeletal muscle from Bae, J.S., Rezaie, A.R., 2011. Activated protein C inhibits high mobility group box 1 lean and ob/ob mice. Diabetes 53, 1436–1444. signaling in endothelial cells. Blood 118, 3952–3959. Liaw, P.C., Neuenschwander, P.F., Smirnov, M.D., Esmon, C.T., 2000. Mechanisms by Bae, J.S., Yang, L., Rezaie, A.R., 2007. Receptors of the protein C activation and which soluble endothelial cell protein C receptor modulates protein C and activated protein C signaling pathways are colocalized in lipid rafts of activated protein C function. J. Biol. Chem. 275, 5447–5452. endothelial cells. Proc. Natl. Acad. Sci. USA 104, 2867–2872. Menschikowski, M., Hagelgans, A., Eisenhofer, G., Siegert, G., 2009. Regulation of Black, R.A., Rauch, C.T., Kozlosky, C.J., Peschon, J.J., Slack, J.L., Wolfson, M.F., Castner, endothelial protein C receptor shedding by cytokines is mediated through B.J., Stocking, K.L., Reddy, P., Srinivasan, S., Nelson, N., Boiani, N., Schooley, K.A., differential activation of MAP kinase signaling pathways. Exp. Cell Res. 315, Gerhart, M., Davis, R., Fitzner, J.N., Johnson, R.S., Paxton, R.J., March, C.J., Cerretti, 2673–2682. D.P., 1997. A metalloproteinase disintegrin that releases tumour-necrosis Michie, H.R., Manogue, K.R., Spriggs, D.R., Revhaug, A., O’Dwyer, S., Dinarello, C.A., factor-alpha from cells. Nature 385, 729–733. Cerami, A., Wolff, S.M., Wilmore, D.W., 1988. Detection of circulating tumor Blobel, C.P., 2005. ADAMs: key components in EGFR signalling and development. necrosis factor after endotoxin administration. N. Engl. J. Med. 318, 1481–1486. Nat. Rev. Mol. Cell Biol. 6, 32–43. Miller, M.A., Meyer, A.S., Beste, M.T., Lasisi, Z., Reddy, S., Jeng, K.W., Chen, C.H., Han, Borgel, D., Bornstain, C., Reitsma, P.H., Lerolle, N., Gandrille, S., Dali-Ali, F., Esmon, J., Isaacson, K., Griffith, L.G., Lauffenburger, D.A., 2013. ADAM-10 and -17 C.T., Fagon, J.Y., Aiach, M., Diehl, J.L., 2007. A comparative study of the protein C regulate endometriotic via concerted ligand and receptor pathway in septic and nonseptic patients with organ failure. Am. J. Respir. Crit. shedding feedback on kinase signaling. Proc. Natl. Acad. Sci. USA 110, E2074– Care Med. 176, 878–885. E2083. Buras, J.A., Holzmann, B., Sitkovsky, M., 2005. Animal models of sepsis: setting the Mosnier, L.O., Zlokovic, B.V., Griffin, J.H., 2007. The cytoprotective protein C stage. Nat. Rev. Drug Discov. 4, 854–865. pathway. Blood 109, 3161–3172. Buxbaum, J.D., Liu, K.N., Luo, Y., Slack, J.L., Stocking, K.L., Peschon, J.J., Johnson, R.S., Moss, M.L., Lambert, M.H., 2002. Shedding of membrane proteins by ADAM family Castner, B.J., Cerretti, D.P., Black, R.A., 1998. Evidence that tumor necrosis factor . Essays Biochem. 38, 141–153. alpha converting enzyme is involved in regulated alpha-secretase cleavage of Murphy, G., 2008. The ADAMs: signalling scissors in the tumour microenvironment. the Alzheimer amyloid protein precursor. J. Biol. Chem. 273, 27765–27767. Nat. Rev. Cancer 8, 929–941. Esmon, C.T., 2005. The interactions between inflammation and coagulation. Br. J. Qu, D., Wang, Y., Song, Y., Esmon, N.L., Esmon, C.T., 2006. The Ser219 Gly Haematol. 131, 417–430. dimorphism of the endothelial protein C receptor contributes to the higher Essawi, T., Srour, M., 2000. Screening of some Palestinian medicinal plants for soluble protein levels observed in individuals with the A3 haplotype. J. Thromb. antibacterial activity. J. Ethnopharmacol. 70, 343–349. Haemost. 4, 229–235. Fingleton, B., 2008. MMPs as therapeutic targets-still a viable option? Semin. Cell Qu, D., Wang, Y., Esmon, N.L., Esmon, C.T., 2007. Regulated endothelial protein C Dev. Biol. 19, 61–68. receptor shedding is mediated by tumor necrosis factor-alpha converting Fukudome, K., Esmon, C.T., 1994. Identification, cloning, and regulation of a novel enzyme/ADAM17. J. Thromb. Haemost. 5, 395–402. endothelial cell protein C/activated protein C receptor. J. Biol. Chem. 269, Scaldaferri, F., Sans, M., Vetrano, S., Graziani, C., De Cristofaro, R., Gerlitz, B., Repici, 26486–26491. A., Arena, V., Malesci, A., Panes, J., Grinnell, B.W., Danese, S., 2007. Crucial role of Fulcher, C.A., Gardiner, J.E., Griffin, J.H., Zimmerman, T.S., 1984. Proteolytic the protein C pathway in governing microvascular inflammation in inactivation of human factor VIII procoagulant protein by activated human inflammatory bowel disease. J. Clin. Invest. 117, 1951–1960. protein C and its analogy with factor V. Blood 63, 486–489. Sesin, C.A., Yin, X., Esmon, C.T., Buyon, J.P., Clancy, R.M., 2005. Shedding of Han, H., Du, B., Pan, X., Liu, J., Zhao, Q., Lian, X., Qian, M., Liu, M., 2010. CADPE endothelial protein C receptor contributes to vasculopathy and renal injury in inhibits PMA-stimulated gastric carcinoma cell invasion and matrix lupus: in vivo and in vitro evidence. Kidney Int. 68, 110–120. metalloproteinase-9 expression by FAK/MEK/ERK-mediated AP-1 activation. Simmonds, R.E., Lane, D.A., 1999. Structural and functional implications of the Mol. Cancer Res. 8, 1477–1488. intron/exon organization of the human endothelial cell protein C/activated Hayashi, T., Nakamura, H., Okada, A., Takebayashi, S., Wakita, T., Yuasa, H., protein C receptor (EPCR) gene: comparison with the structure of CD1/major Okumura, K., Suzuki, K., 1999. Organization and chromosomal localization of histocompatibility complex alpha1 and alpha2 domains. Blood 94, 632–641. the human endothelial protein C receptor gene. Gene 238, 367–373. Thalhamer, T., McGrath, M.A., Harnett, M.M., 2008. MAPKs and their relevance to Huovila, A.P., Turner, A.J., Pelto-Huikko, M., Karkkainen, I., Ortiz, R.M., 2005. arthritis and inflammation. Rheumatology (Oxford) 47, 409–414. Shedding light on ADAM metalloproteinases. Trends Biochem. Sci. 30, 413–422. Villoutreix, B.O., Blom, A.M., Dahlback, B., 1999. Structural prediction and analysis Kaileh, M., Vanden Berghe, W., Heyerick, A., Horion, J., Piette, J., Libert, C., De of endothelial cell protein C/activated protein C receptor. Protein Eng. 12, 833– Keukeleire, D., Essawi, T., Haegeman, G., 2007. Withaferin A strongly elicits 840. IkappaB kinase beta hyperphosphorylation concomitant with potent inhibition Wang, H., Liao, H., Ochani, M., Justiniani, M., Lin, X., Yang, L., Al-Abed, Y., Metz, C., of its kinase activity. J. Biol. Chem. 282, 4253–4264. Miller, E.J., Tracey, K.J., Ulloa, L., 2004. Cholinergic agonists inhibit HMGB1 Kim, D.C., Lee, W., Bae, J.S., 2011. Vascular anti-inflammatory effects of curcumin on release and improve survival in experimental sepsis. Nat. Med. 10, 1216–1221. HMGB1-mediated responses in vitro. Inflamm. Res. 60, 1161–1168. Ware, L.B., Camerer, E., Welty-Wolf, K., Schultz, M.J., Matthay, M.A., 2006. Bench to Kremer, J.P., Jarrar, D., Steckholzer, U., Ertel, W., 1996. Interleukin-1, -6 and tumor bedside: targeting coagulation and fibrinolysis in acute lung injury. Am. J. necrosis factor-alpha release is down-regulated in whole blood from septic Physiol. Lung Cell Mol. Physiol. 291, L307–L311. patients. Acta Haematol. 95, 268–273. Xu, J., Qu, D., Esmon, N.L., Esmon, C.T., 2000. Metalloproteolytic release of Kurosawa, S., Stearns-Kurosawa, D.J., Hidari, N., Esmon, C.T., 1997. Identification of endothelial cell protein C receptor. J. Biol. Chem. 275, 6038–6044. functional endothelial protein C receptor in human plasma. J. Clin. Invest. 100, Yang, H., Ochani, M., Li, J., Qiang, X., Tanovic, M., Harris, H.E., Susarla, S.M., Ulloa, L., 411–418. Wang, H., DiRaimo, R., Czura, C.J., Roth, J., Warren, H.S., Fink, M.P., Fenton, M.J., Kurosawa, S., Stearns-Kurosawa, D.J., Carson, C.W., D’Angelo, A., Della Valle, P., Andersson, U., Tracey, K.J., 2004. Reversing established sepsis with antagonists Esmon, C.T., 1998. Plasma levels of endothelial cell protein C receptor are of endogenous high-mobility group box 1. Proc. Natl. Acad. Sci. USA 101, 296– elevated in patients with sepsis and systemic lupus erythematosus: lack of 301. correlation with thrombomodulin suggests involvement of different Zhang, Y., Jiang, J., Black, R.A., Baumann, G., Frank, S.J., 2000. Tumor necrosis factor- pathological processes. Blood 91, 725–727. alpha converting enzyme (TACE) is a growth hormone binding protein (GHBP) Lee, D.C., Sunnarborg, S.W., Hinkle, C.L., Myers, T.J., Stevenson, M.Y., Russell, W.E., sheddase: the metalloprotease TACE/ADAM-17 is critical for (PMA-induced) GH Castner, B.J., Gerhart, M.J., Paxton, R.J., Black, R.A., Chang, A., Jackson, L.F., 2003. receptor proteolysis and GHBP generation. Endocrinology 141, 4342–4348. TACE/ADAM17 processing of EGFR ligands indicates a role as a physiological convertase. Ann. NY. Acad. Sci. 995, 22–38.