Down-Regulation of LPCAT Expression Increases Platelet-Activating Factor Level in Cirrhotic Rat Liver: Potential Antiinﬂammatory Effect of Silybin

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Biochimica et Biophysica Acta 1832 (2013) 2019–2026

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

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

Down-regulation of LPCAT expression increases platelet-activating factor level in cirrhotic rat liver: Potential antiinﬂammatory effect of silybin

Eleonora Stanca a, Gaetano Serviddio b, Francesco Bellanti b, Gianluigi Vendemiale b, Luisa Siculella a, Anna Maria Giudetti a,⁎

a Laboratory of Biochemistry and Molecular Biology, Department of Biological and Environmental Sciences and Technologies, University of Salento, Lecce, Italy b Department of Medical and Surgical Sciences, University of Foggia, Italy

article info abstract

Article history: Cholestasis is one of the major causes of liver diseases. A chronic accumulation of toxic bile acids in the liver, Received 20 February 2013 which occurs in this condition, can induce fibrosis and cirrhosis. Inflammation is a fundamental component of Received in revised form 24 June 2013 acute and chronic cholestatic liver injury. Accepted 3 July 2013 Platelet-activating factor (PAF) is a proinflammatory lipid which may be generated by two independent path- Available online 12 July 2013 ways called the de novo and remodeling pathway being the last responsible for the synthesis of PAF during inflammation. In recent years a key role in PAF remodeling has been attributed to lysophosphatidylcholine Keywords: Biliary cirrhosis acyltransferase (LPCAT) enzymes. Although the knowledge on their characteristic is growing, the exact Lysophosphatidylcholine acyltransferase mechanism of LPCAT in pathological conditions remains still unknown. Platelet-activating factor Here, we reported that the level of lyso-PAF and PAF significantly increased in the liver of cirrhotic vs. control Silybin rats together with a significant decrease in both mRNA abundance and protein level of both LPCAT1 and LPCAT2. Acyltransferase activities of both LPCAT1 and LPCAT2 were parallel decreased in the liver of cirrhotic animals. Interestingly, treatment with silybin strongly decreased the level of both pro-inflammatory lipids and restored the activity and expression of both LPCAT1 and LPCAT2 of cirrhotic liver. Silybin effect was specific for LPCAT1 and LPCAT2 since it did not affect LPCAT3 mRNA abundance of cirrhotic liver. © 2013 Elsevier B.V. All rights reserved.

1. Introduction diabetes [5] and nervous system diseases [6] wheretheyactivateinflam- matory process and oxidative stress. Cholestasis is a pathological condition caused by injury to the bile Recently, several observations have indicated increased plasma [9] ducts induced by immunologic, toxic, or genetic causes [1]. Cholesta- and liver [10] PAF levels in cirrhotic rats suggesting a crucial role of sis results in intrahepatic accumulation of potentially toxic bile acids PAF in liver diseases progression [7–9]. leading to hepatocyte apoptosis and necrosis [1]. Moreover, the com- PAF can be synthesized through two different pathways: the de bined effects of progressive inflammation and toxicity of bile trapped novo and a recovery pathway [4]. The latter involves the hydrolysis within hepatocytes can culminate in cirrhosis [2]. Excessive bile accu- of alkyl-PC at sn-2 position by PLA2 with the production of lyso-PAF mulation into the liver has been reported to markedly alter the ex- which can be acetylated, in presence of acetyl-CoA, in a reaction cat- pression of various genes involved in cholesterol and phospholipid alyzed by the key enzyme lyso-PAF:acetyl-CoA acetyltransferase homeostasis [3]. (PAF-AT, EC 2.3.1.67) [11]. While the de novo synthesis is linked to Platelet-activating factor (1-O-alkyl-2-acetyl-sn-glycero-3- a constitutive production of PAF, the remodeling pathway seems to phosphocholine or PAF) and alkyl-lysophosphatidylcholine (lyso-PAF) be involved in inflammatory responses of PAF [12]. Interestingly, are phosphatidylcholine-derived molecules with important roles in the two biosynthetic pathways are associated to distinct activities inflammatory processes [4]. It has been demonstrated that PAF and thus allowing targeting antiinflammatory intervention to specific en- lyso-PAF are hyper-produced in many inflammatory conditions such as zymes [13]. The catabolism of PAF is normally made by PAF-acetylhydrolases (PAF-AH) which catalyze the degradation of PAF during antiinflammatory response [14]. In fact, treatment of diabetic rats with recombi- ⁎ Corresponding author at: Department of Biological and Environmental Sciences and nant PAF-AH reduces insulitis and the frequency of diabetes [15]. Technologies, Via Monteroni, University of Salento, 73100 Lecce, Italy. Tel.: +39 832298679. In addition to the well characterized PAF-AT and PAF-AH, a grow- E-mail address: [email protected] (A.M. Giudetti). ing number of enzymes have been recognized, in the last few years, in

0925-4439/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbadis.2013.07.005 2020 E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026 the remodeling pathway of PAF. Four different isoforms of with AST, ALT and ALP assay kits, respectively (Catachem, Bridgeport, lysophosphatidylcholine acyltransferases (LPCAT 1–4) are now consid- CT). ered to be involved in PAF remodeling during inflammation [16–20]. Although a significant progress has been made in the study of 2.5. Isolation of microsomal fraction from rat liver LPCAT enzymes, their role in pathological conditions remains still unknown. For microsome isolation liver were homogenized in a cold buffer Aim of the present study was to investigate the role of LPCATs in a containing 100 mM Tris–HCl (pH 7,4), 250 mM sucrose, 5 mM rodent model of secondary biliary cirrhosis. The effect of silybin, a 2-mercaptoethanol, 20 μM 4-APMSF and a cocktail of protease inhibi- well known hepatoprotective compound [21–23] on the activity and tors as reported in [17]. Microsomes were isolated by differential centri- expression of LPCAT was also evaluated. fugation as reported in [24]. The most relevant results were that the level of proinflammatory lyso-PAF and PAF dramatically increased and the mRNA expression of 2.6. Phospholipase A2 activity LPCAT1, LPCAT2 and LPCAT3 significantly reduced in the liver of cirrhotic rats. Silybin was able to restore normal level of inflammatory lipids Phospholipase A2 (PLA2) activity was assayed as reported in [25] in and to normalize gene expression of both LPCAT1 and LPCAT2 but not liver homogenates by using 4N3OBA as substrate. Reaction mixture LPCAT3. A role for LPCAT1 and LPCAT2 in the antiinflammatory effect contained 150 mM KCl, 10 mM CaCl2, 50 mM Tris–HCl, pH 7.5 and of silybin, in the liver, was then discussed. 4N3OBA 2 mM. The reaction was started by addition of samples and after 1 h, at room temperature; the absorbance was read at two differ- 2. Materials and methods ent wavelengths (425 nm and 600 nm to correct for any turbidity in the sample). Specific activity was expressed as μmol/h × mg protein. 2.1. Materials

2.7. Assay of lyso-PAF acyltransferase activity All lipids, acyl-CoAs, 4-amidinophenylmethanesulfonyl fluoride (APMSF), 7-diethylaminocoumarin-3-carbonylazide, 7-methoxycou- The acyltransferase activity of LPCAT1 was assayed essentially as re- marin-4-acetic acid, 1,2-dimethoxyethane, N,N′-dicyclohexyl- ported in [18]. Briefly, 5 μg of microsomal fractions was used to start the carbodiimide, phospholipase C (Bacillus cereus,typeIXV)were reaction in an assay mix containing 100 mM Tris–HCl (pH 7.4), 1 mM purchased from Sigma. [14C]acetyl-CoA, [14C]oleoyl-CoA, [14C]palmitoyl- EDTA, 1 mg/ml PC, 20 μM lyso-PAF (1-O-palmityl-sn-glycero-3- CoA and [14C]arachidonyl-CoA were supplied by PerkinElmer phosphocholine), 25 μM[14C]oleoyl-CoA (2.035 GBq/mmol) or 25 μM (PerkinElmer, Monza, Italy). 4N3OBA was from Vinci-Biochem s.r.l., [14C]palmitoyl-CoA (293 MBq/mmol). After 5 min at 30 °C reaction LPCAT2 and LPCAT1 antibodies were from DBA Italy s.r.l. RNAlater solu- was stopped adding 600 μl of a cold mixture of chloroform/methanol tion and SV Total RNA Isolation System kit from Promega. SuperScriptTM 1:2 (v/v). LPCAT2 acyltransferase activity was measured as in [19] incu- III RNase H-Reverse transcriptase was from Invitrogen. Primers were bating 5 μg of microsomal proteins in a buffer containing 20 mM Tris– purchased from Sigma. All other chemicals and reagents were of analyt- HCl (pH 7.4), 2 mM CaCl , 1 mg/ml PC, 20 μM APMSF, 5 mM ical grade. 2 2-mercaptoethanol, 20 μM lyso-PAF (1-O-palmityl-sn-glycero-3- phosphocholine) and 20 μM[14C]arachidonyl-CoA (1.11–2.035 GBq/ 2.2. Animal model mmol). After 10 min at 37 °C reaction was stopped adding 600 μlofa cold mixture of chloroform/methanol 1:2 (v/v). After extraction, lipids Male Wistar rats, weighing 200–250 g, supplied by Harlan Italy Srl were separated by thin layer chromatography. Bands at positions corre- (S. Pietro al Natisone, UD), housed at 22 ± 1 °C, have been used sponding to expected product were scraped from the plate and radioac- throughout the study. The animals were maintained and sacrificed tivity counted using liquid scintillation. according to the Italian Official Statement N.116/92. The animals were divided into three groups: a group of rats subjected to bile duct ligation (BDL), a group of control rats (SHAM), subjected to the 2.8. Assay of LPCAT1 and LPCAT2 acetyltransferase activities entire surgical procedure except for the ligation, and finally a group of BDL rats treated with silybin (BDL + Sil). The duct ligation was Both LPCAT1 and LPCAT2 have been demonstrated to catalyze performed as previously described by Serviddio et al. [1]. The induc- acetyltransferase reaction [26]. The acetyltransferase activity of tion of chronic cholestasis was demonstrated by increases in the con- LPCAT1 was assayed as reported in [17].Briefly, 5 μg of microsomal pro- centration of bilirubin in the serum of rats BDL which was found tein was incubated in a buffer containing 20 mM Tris–HCl (pH 7.4), 7.3 ± 3.3 mg/dl vs. 0.5 ± 0.3 mg/dl of the controls (Fig. 1). Histolog- 5 mM 2-mercaptoethanol, 20 μM APMSF, 1 mg/ml PC, 1 mM EDTA, 3 ical analysis after 4 weeks of bile duct ligation in rats BDL shows that 20 μM lyso-PAF and 200 μM[H]acetyl-CoA. Acetyltransferase activity the liver developed the characteristic of liver cirrhosis, with an inho- of LPCAT2 was assayed following the procedure described in [17].In mogeneous structure due to the presence of fibrous septa highlighted this case microsomal proteins were incubated in a buffer containing by Masson's trichrome (Fig. 1). 20 mM Tris–HCl (pH 7.4), 5 mM 2-mercaptoethanol, 20 μMAPMSF, 3 1 mg/ml PC, 2 mM CaCl2,20μM lyso-PAF and 100 μM[H]acetyl-CoA. 2.3. Histological analysis After 10 min at 37 °C reactions were stopped by chloroform/methanol 1:2 (v/v). After lipid extraction and separation by TLC radioactivity asso- Small blocks of liver tissue from all mice were immediately fixed ciated to the radiolabeled products was measured by liquid scintillation. in 4% neutral formalin and embedded in paraffin. Sections (4 μm thick) were stained by the hematoxylin and eosin or Masson's 2.9. Assay of liver lyso-PAF acetyltransferase activity trichrome method. At least three discontinuous liver sections were evaluated for each mouse. We also assayed the lyso-PAF acetyltransferase activity as in [27]. In this case, 50 μg of liver homogenate was incubated in a buffer con- 2.4. Biochemical analysis taining 250 mM Tris–HCl, 25 mM BSA at pH 6.9 in presence of 20 μM [14C]acetyl-CoA and 20 μM lyso-PAF. Reaction lasted 20 min at 37 °C Serum activities of aspartate aminotransferase (AST), alanine ami- and was terminated by adding chloroform/methanol 1:2 (v/v). Lipid notransferase (ALT) and alkaline phosphatase (ALP) were measured extracted was loaded on TLC plate and after separation the band E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026 2021

Fig. 1. Panel A: histological analysis of liver specimens stained with haematoxylin and eosin (HE) or Masson trichrome (MT) from the same experimental groups (original magni- fication, 100×). Panel B: serum aspartate and alanine aminotransferases (AST, ALT) and alkaline phosphatase (AP) levels in rats undergoing sham operation (SHAM), bile duct ligation (BDL) and BDL fed a diet containing silybin (BDL + Sil) 21 days after bile duct ligation. Data are expressed as means ± SD of five experiments for each group. corresponding to the expected product, stained with iodine, was 2.11. HPLC separation and quantification of PAF and lyso-PAF scraped and radioactivity was counted by liquid scintillation. Lyso-PAF amount was quantified as reported in [29]. After tissue ho- mogenization, lipids were extracted by chloroform/methanol (1:2, v/v). 2.10. TLC analysis of lipids 2-lysophospholipid species were derivatized with 100 μl of 1 mg/ml solution of 7-diethylaminocoumarin-3-carbonylazide in anhydrous Total lipids were extracted using the Bligh-Dyer procedure [28] toluene. After 3 h at 80 °C the vials were cooled and fluorescent deriv- and separated by TLC. The mobile phase was represented by atives were analyzed by HPLC using a C18 reversed-phase column chloroform/methanol/water (65:25:4, v/v/v). At the end of the (Phenomenex). The elution was performed with a gradient of solvent lipid separation bands corresponding to different lipids were identi- A(methanol–water, 80:20, v/v, containing 250 mg/l choline chloride) fied by comparison with known lipids. and solvent B (chloroform) at a flow-rate of 1.0 ml/min. Detection 2022 E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026 was made fluorescence spectrometer (excitation 400 nm; emission 3.2. PAF and lyso-PAF level increased during cirrhosis 480 nm). PAF was quantified following the method reported in [30].Briefly, Inflammation plays a pivotal role in the pathophysiology of liver extracted lipids were hydrolyzed by phospholipase C. The hydro- cirrhosis [2]. The amount of proinflammatory PAF and its precursor lyzed products were derivatizated overnight at 4 °C in presence lyso-PAF was measured in the liver of sham and BDL animals. The of 7-methoxycoumarin-4-acetic acid, 1,2-dimethoxyethane and level of lyso-PAF (19.8 ± 1.4 pmol/mg protein in cirrhotic vs. 12 ± N,N′-dicyclohexylcarbodiimide. Derivatives were separated by 1.0 pmol/mg protein in normal rats; P b 0.05, Fig. 2A) and PAF HPLC using a C18 reverse-phase column (Phenomenex) by isocratic (10.9 ± 1.1 fmol/mg protein in cirrhotic vs. 3.2 ± 0.8 fmol/mg pro- elution with 90% methanol at 1 ml/min. Fluorescence was monitored tein in normal rats; P b 0.05, Fig. 2B) significantly increased in the at 330 nm excitation and 415 nm emission wavelengths. liver of cirrhotic rats. The treatment with silybin significantly re- Lyso-PAF and PAF amount were evaluated by comparing peak area duced both the content of PAF and particularly of lyso-PAF (Fig. 2). of unknown samples with that of known amount of lyso-PAF or PAF. 3.3. Activity of phospholipase A2 during cirrhosis 2.12. Reverse transcriptase and Real-time PCR It has been demonstrated that PLA2 activity increases after several inflammatory stimuli [31]. This enzyme accounts for the break of the A piece of the liver was dissected out and collected in RNAlater Solu- sn-2 bond in the phospholipids which generates lyso-PAF that, in tion for RNA extraction. Total RNA was extracted from 0.3 g of tissue turn, can be utilized for PAF synthesis. We hypothesized that the in- using the SV Total RNA Isolation System kit following the creased amount of PAF occurring in the liver of BDL animals was de- manufacturer's protocol. Only RNA without DNA contamination was pendent on the activation of PLA2. As expected, PLA2 activity was used for subsequent preparation of cDNA synthesis. The RT (reverse significantly increased (~140%) in cirrhotic liver, as compared to nor- transcriptase) reaction (20 μl) was carried out using 5 μg of total RNA, mal rats (Fig. 3). Once again, the administration of silybin was able to 100 ng of random hexamers and 200 units of SuperScriptTM III RNase reduce the PLA2 activity that, however, did not reach normal values. H-Reverse transcriptase. Quantitative gene expression analysis was performed (SmartCycler System, Cepheid) using SYBR Green technolo- 3.4. Down-regulation of LPCAT gene and protein expression during gy (FluoCycle, Euroclone). The sequences of primers used in Real-time chronic cholestasis PCR were the following: Lpcat1 forward: 5′-gACCCTCgAgTgATgTAA CC-3′ and Lpcat1 reverse: 5′-gATggATgCAgATTTCTAAC-3′ (GenBank™ LPCAT1, LPCAT2 and LPCAT3 mRNA abundance was measured in accession number NM_001100735.1); Lpcat2 forward 5′-ATTATTCTg liver from BDL and sham operated rat livers. As reported in Fig. 4, CTTCCAgTCCg-3′, Lpcat2 reverse 5′-AgAAgAACATggCACgACC-3′ mRNA of all LPCAT isoforms was reduced in the liver of BDL with re- (Gen-Bank™ accession number XM_001064713); and Lpcat3 forward spect to sham rats. This was particularly evident for LPCAT1 and 5′-TTgCTTCCAgATggCCTAC-3′, Lpcat3 reverse 5′-ATgAggggACACCCA LPCAT2 whose expression appeared almost completely suppressed GTATg-3′ (Gen-Bank™ accession number NM_001012189). Rplp0 was used as an internal control for normalization [2].

2.13. Western blot analysis

Microsomal proteins were loaded onto 12% SDS-PAGE gel and transferred to nitrocellulose membrane for immunoblot analysis with anti-LPCAT1 and LPCAT2 antibodies at room temperature for 2 h, then membrane was incubated for 1 h with secondary horserad- ish peroxidase-conjugated IgG (dilution, 1:2000). Signals were detected by enhanced chemiluminescence. For signal normalization, alpha-tubulin was used. The intensity band was evaluated densitom- etry with Molecular Analyst Software.

2.14. Statistical analysis

Data were expressed as means ± standard deviation (SD). The unpaired t test was used to verify the signiﬁcance of differences be- tween means. In all instances P b 0.05 was taken as the lowest level of signiﬁcance. The GraphPad Prism 4 for Windows (GraphPad Soft- ware Inc., San Diego, CA, USA) was used to perform all the analyses.

3. Results

3.1. Liver histology and biochemical analysis

Serum AST, ALT, ALP and bilirubin levels were measured together with liver histology and the results are reported in Fig. 1. Liver from BDL group showed cholestasis and hepatocellular necrosis, as well Fig. 2. Lyso-PAF and PAF liver levels. Total lipids were extracted from liver of control as extensive bile duct proliferation, neutrophils and lymphocytes in- (SHAM), cirrhotic (BDL) and BDL rats treated with silybin (BDL + Sil). After separation by HPLC and fluorimetric detection, lyso-PAF (A) and PAF (B) were quantified compar- filtration, and fibrosis. Silybin significantly reduced liver necrosis, in- ing the area value of identified peaks with that of known amount of standards. Data are fl fi ammation, brosis and bile duct proliferation (Fig. 1A) as well as the means ± SD of three different experiments for each group. Statistical differences serum cholestasis (Fig. 1B) as compared to sham rats. were assessed using Student's t test. E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026 2023

Fig. 3. Measurement of liver phospholipase A2 activity. PLA2 activity was measured in liver homogenates from control (SHAM), cirrhotic (BDL) and BDL rats treated with silybin (BDL + Sil). PLA2 activity is expressed as μmol/h × mg protein. Results reported in the figure are the means ± SD of six independent experiments. Statistical differences were assessed using Student's t test. whereas LPCAT3 mRNA level was significantly reduced by about 30% as compared to sham rats. Very interestingly, in silybin treated BDL rats the expression of LPCAT1 and LPCAT2 was completely restored. Of note, silybin did not impact the expression of LPCAT3. Taken together these results suggested a specific role for LPCAT1 and LPCAT2 in the modulation of liver PAF level during cirrhosis development. Western blotting analysis confirmed that inhibition of LPCAT oc- curred also in terms of protein expression and that silybin was able to restore physiological level of proteins (Fig. 5).

3.5. LPACT1 and LPCAT2 lyso-PAF acyltransferase activities

Since LPCAT1 has antiinflammatory properties by catalyzing the acyl- ation of lyso-PAF to alkyl-PC [5] we verify the hypothesis that the higher level of PAF observed in BDL animals, was due to a low acyltransferase activity which give lyso-PAF more available for PAF synthesis. Both [14C]palmitoyl-CoA and [14C]oleoyl-CoA were used as acyl-CoA donors to assay acyltransferase activity of LPCAT1. We also measured the acyltransferase activity of LPCAT2 using [14C]arachidonyl-CoA as sub- Fig. 4. LPCAT1, LPCAT2 and LPCAT3 mRNA abundance. Quantitative Real time PCR was strate. As reported in Fig. 6A, the acyltransferase activity of LPCAT1 was used to determine LPCAT1, LPCAT2, and LPCAT3 gene expression levels in the liver of control (SHAM), cirrhotic (BDL) and cirrhotic rats treated with silybin (BDL + Sil). higher in the controls when palmitoyl-CoA was used as a substrate as mRNA content was normalized using Rplp0. Data shown are the means ± SD of four compared to oleyl-CoA. When cholestatic livers were compared to different experiments. sham operated animals a significant inhibition of the incorporation of both fatty acids in lyso-PAF was observed (Fig. 6A). The result was similar when acyltransferase activity of LPCAT2 was measured in the presence of 4. Discussion [14C]arachidonyl-CoA (Fig. 6B). Once again, silybin was able to restore normal conditions. Platelet-activating factor (1-O-alkyl-2-acetyl-sn-glycero-3- phosphocholine or PAF) is a potent pro-inflammatory lipid mediator 3.6. LPACT1 and LPCAT2 lyso-PAF acetyltransferase activities which plays an important role in metabolic and pathologic changes observed in cirrhosis [10]. Under inflammatory stimuli LPCAT2 may increase the rate of In the last few years progresses have been realized in the identifica- acetyl-CoA incorporation into lyso-PAF to form PAF [18] whereas tion of different enzymes involved in PAF remodeling during inflamma- LPCAT1 has been demonstrated to be involved in a non-inflammatory tion. Indeed, four different acyltransferase (LPCAT 1–4) isoforms have synthesis of PAF [17]. LPCAT activity involved in PAF synthesis during been recognized to be involved in this pathway [16–20]. Despite the cirrhosis was assayed by using the method described in [17]. large interest on these findings, the role of LPCAT during pathological Acetyltransferase activity of LPCAT2 was hardly detected in all three conditions remains still unknown. groups of rats whereas no activity of LPCAT1 was found (data nor Silybin, the main constituent of milk thistle extract, has been shown shown). to have hepatoprotective effects in several in vitro and animal models As a result we supposed that the increased level of PAF found in [32–34] and to inhibit the hepatic production of pro-inflammatory the liver of BDL animals could be due to liver acetyltransferase activ- and pro-fibrogenic factors [35]. ities not dependent on LPCAT enzymes. The acetyltransferase activity The present study was addressed to dissect the molecular was then measured in the whole liver homogenate using the method mechanisms underlying the increased PAF level in cirrhotic liver reported in [27] and the results are reported in Fig. 7. The activity was pointing out the role of LPCAT pathways. The molecular mechanism of observed to be significantly higher in BDL with respect to control rats the antiinflammatory effect of silybin in cirrhotic liver was also and not prevented by the administration of silybin. investigated. 2024 E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026

Fig. 6. LPCAT1 and LPCAT2 lyso-PAF acyltransferase activities. Lyso-PAF acyltransferase activity was measured in liver microsomal fractions from control (SHAM) cirrhotic (BDL) and cirrhotic rats treated with silybin (BDL + Sil). A) LPCAT1 acyltransferase activ- Fig. 5. LPCAT1 and LPCAT2 protein levels. Proteins from microsomal fractions were ity was measured by following the incorporation of [14C]oleoyl-CoA (18:1) or [14C] separated by gel electrophoresis. After transfer on a nitrocellulose paper, ﬁlters were palmitoyl-CoA (16:0) into lyso-PAF to form alkyl-PC (B) acyltransferase activity of incubated with LPCAT1 and LPCAT2 antibodies. Band intensity corresponding to recog- LPCAT2 was followed using [14C]arachidonyl-CoA as acyl-CoA donors. Results are the nized proteins was quantiﬁed by densitometric analysis. Signals were normalized with means ± SD of four different experiments. alpha-tubulin.

Phosphatidylcholine (PC) is the most representative phospholipid of fibrate administration was observed for LPCAT1 and LPCAT2 as com- cellular membranes and then the main substrate for the production of pared to LPCAT3 and LPCAT4 [40]. Further studies should be encouraged lysophospholipids following breakdown of sn-2 bond by PLA2 activity. to better identify the promoter structures of these enzymes in order to Accordingly to [10,36] an increased level of both lyso-PAF and PAF explain the response to silybin. We should take into account that silybin was found in the liver of BDL with respect to control animals. Moreover, silybin treatment significantly decreased the BDL liver level of both proinflammatory lipids bringing to values even lower than controls. We then supposed a regulation by silybin on PLA2 activity. In our experiments, in agreement with [31], an increase in the PLA2 activity was measured in the livers of BDL compared to control rats (Fig. 3). Al- though silybin had a moderate resetting effect of PLA2 activity (Fig. 3) we considered this not enough to lead lyso-PAF to such low values as we measured. We then hypothesize another mechanisms involved in the increased PAF and lyso-PAF levels in BDL liver. Previous observations reported that LPCAT enzymes are able to re- model membrane phospholipids during inflammation [37]. Among the four LPCAT isoforms so far studied, LPCAT1 and LPCAT2 are the well defined. A major liver LPCAT isoform, named LPCAT3, has been recently characterized [20,38] which is structurally different from other members of the glycerolipid acyltransferase family [39]. In our study the amount of LPCAT1–2 mRNA was drastically reduced in cirrhotic as compared to the control livers, while the inhibitory effect was less evident on LPCAT3. Interestingly, treatment of cirrhotic animals with silybin almost completely restored the mRNA levels of Fig. 7. Liver lyso-PAF acetyltransferase activity. Liver homogenate from control (SHAM), LPCAT1 and LPCAT2 but not of LPCAT3 (Fig. 4). We then speculated cirrhotic (BDL) and cirrhotic rats treated with silybin (BDL + Sil) were used for the that the expression of LPCAT3 is under a different mechanism of regula- assay. The enzymatic activity was measured as [14C]acetyl-CoA incorporation into tion than LPCAT1–2. In fact, in a recent report, a different response to lyso-PAF to form PAF. Data shown are the means ± SD of four different experiments. E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026 2025

Fig. 8. Schematic representation of PAF synthesis and remodeling during cirrhosis. In liver cirrhosis membrane alkyl-PC is hydrolyzed, by activated cytoplasmic PLA2, in lyso-PAF. This latter is converted to PAF by lyso-PAF acetyltransferase (Lyso-PAF-AT) whose activity in increased in cirrhotic liver. Activity and expression of both LPCAT1 and LPCAT2 isoforms were down-regulated under this condition, as highlighted by dashed lines. Silybin treatment, activating both LPCAT1 and LPCAT2 enzymes, subtracted lyso-PAF to acetyltransferase reaction, to form alkyl-PC, thus decreasing liver PAF synthesis.

also inhibits NF-kB activation [23] being NF-kB a transcription factor im- acetyltransferase forms. In any cases silybin was not able to affect plicated in inflammatory pathway PAF-mediated [41]. the acetyltransferase activity in cirrhotic livers (Fig. 7). The analysis of LPCAT protein level confirmed the gene expression results thus strengthening our hypothesis for a specific control, by 5. Conclusions these enzymes, on liver PAF level. To understand the LPCAT contribution in PAF synthesis during cir- To our knowledge, our results demonstrated that both LPCAT1 and rhosis the activity of both LPCAT1 and LPCAT2 was measured. LPCAT2 act as acyltransferase enzymes in cirrhotic livers and a novel LPCAT1 plays an essential role in alveolar type II cells in produc- mechanism for the antiinflammatory effect of silybin during cirrhosis tion of dipalmitoyl-PC, the major phosphatidylcholine in pulmonary development giving its ability to decrease both lyso-PAF and PAF surfactant [18,26]. Furthermore, an antiinflammatory role to LPACT1 levels. Moreover we also demonstrated that, almost in cirrhotic ani- in diabetic rats has been demonstrated [5]. This effect was attributed mals, silybin regulates acyltransferase activity by elevating the ex- to the capacity of LPCAT1 to catalyze acyl-CoA transfer to lyso-PAF pression of genes encoding for LPCAT1 and LPCAT2. thus generating alkyl-PC, an inactive form of PAF. The acyltransferase Fig. 8 depicts, in a schematic way, the role of LPCAT enzyme in activity of LPCAT1 expressed in CHO-K1 cells has been demonstrated liver cirrhosis. Silybin restoring the acyltransferase activity of two to be specific for saturated fatty acids [17] and palmitoyl-CoA was LPCAT enzymes reduces the amount of lyso-PAF available for the syn- used to kinetically characterize the acyltransferase of LPCAT1 [18]. thesis of PAF. In COS-7 cells LPCAT1 was also demonstrated to efficiently catalyzed Although further studies are required to better understand the role oleoyl transfer into lyso-PAF [5]. Moreover, in murine model, LPCAT1 of silybin in the modulation of LPCAT in liver inflammation and regen- is able to catalyze the synthesis of physiological doses of PAF [26]. The eration, these data provide new insights into the antiinflammatory acetyltransferase activity of LPCAT1 is non-inducible, Ca++ indepen- mechanism of silybin in chronic liver diseases. dent and needs high acetyl-CoA concentrations [26]. LPCAT2 may have both acetyltransferase and arachidonyltransferase activities [19] but, contrary to LPCAT1, both acyl- and acetyltransferase References activities of LPCAT2 require Ca++. Moreover, the acetyltransferase activity of LPCAT2 was induced under inflammatory stimulus [19]. [1] G. Serviddio, J. Pereda, F.V. Pallardó, J. Carretero, C. Borras, J. Cutrin, G. Vendemiale, G. Poli, J. Viña, J. Sastre, Ursodeoxycholic acid protects against secondary biliary cirrho- In the liver of cirrhotic rats we found a strong decrease in sis in rats by preventing mitochondrial oxidative stress, Hepatology 39 (2004) acyltransferase activity, associated with both LPCAT1 and LPCAT2 711–720. (Fig. 6). The acyltransferase activities were reported almost to control [2] A. Arduini, G. Serviddio, A.M. Tormos, M. Monsalve, J. Sastre, Mitochondrial dysfunction in cholestatic liver diseases, Front. Biosci. (Elite Ed.) 4 (2012) 2233–2252. values following silybin treatment. The fact that the changes of activity [3] N. Tanaka, T. Matsubara, K.W. Krausz, A.D. Patterson, F.J. Gonzalez, Disruption of were the same as those seen for the mRNA reinforced our idea that the phospholipid and bile acid homeostasis in mice with nonalcoholic steatohepatitis, acyltransferase activities measured were identified in LPCAT1 and Hepatology 56 (2012) 118–129. [4] F. Snyder, V. Fitzgerald, M.L. Blank, Biosynthesis of platelet-activating factor and LPCAT2. enzyme inhibitors, Adv. Exp. Med. Biol. 416 (1996) 5–10. Since values of LPCAT acyltransferase activities were roughly of [5] L. Cheng, X. Han, Y. Shi, A regulatory role of LPCAT1 in the synthesis of inflamma- the same order of magnitude of the amount of inflammatory lipids tory lipids, PAF and LPC, in the retina of diabetic mice, Am. J. Physiol. Endocrinol. Metab. 297 (6) (2009) E1276–E1282. (Fig. 2) we may suggest that the increased amount of lyso-PAF and [6] R.M. Kramer, D.T. Stephenson, E.F. Roberts, J.A. Cytosolic, phospholipase A2 PAF in cirrhotic liver could be almost in part due to a decreased (cPLA2) and lipid mediator release in the brain, J. Lipid Mediat. Cell Signal. 14 LPCAT1 and LPCAT2 acyltransferase activities. (1–3) (1996) 3–7. In good agree with previous data from Harayama et al. [17] LPCAT1 [7] N.P. Karidis, G. Kouraklis, S.E. Theocharis, Platelet-activating factor in liver injury: a relational scope, World J. Gastroenterol. 12 (23) (2006) 3695–3706. acetyltransferase activity resulted practically undetectable in all liver [8] H.C. Karantonis, G. Gribilas, I. Stamoulis, C. Giaginis, C. Spiliopoulou, G. microsomal samples whereas the acetyltransferase activity associated Kouraklis, C. Demopoulos, S.E. Theocharis, Platelet-activating factor involve- fi to LPCAT2 was barely detectable in the three groups of animals. ment in thioacetamide-induced experimental liver brosis and cirrhosis, Dig. Dis. Sci. 55 (2) (2010) 276–284. On the other hand, acetyltransferase activity was found to increase [9] Y.Y. Lu, C.P. Wang, L. Zhou, Y. Chen, S.H. Su, Y.Y. Feng, Y.P. Yang, Synthesis of in the whole homogenate from cirrhotic liver as compared to controls platelet-activating factor and its receptor expression in Kupffer cells in rat carbon by using the protocol reported in [27] avoiding Ca++ and a pH inap- tetrachloride-induced cirrhosis, World J. Gastroenterol. 14 (5) (2008) 764–770. [10] Y. Yang, E.M. Nemoto, S.A.K. Harvey, V.M. Subbotin, C.R. Gandhi, Increased hepatic propriate for LPCAT acetyltransferase activities [17]. Thus we can sup- platelet activating factor (PAF) and PAF receptors in carbon tetrachloride induced port the suggestion that the activity measured became from other liver cirrhosis, Gut 53 (6) (2004) 877–883. 2026 E. Stanca et al. / Biochimica et Biophysica Acta 1832 (2013) 2019–2026

[11] R.L.Wykle,B.Malone,F.Snyder,Enzymaticsynthesis of 1-alkyl-2-acetyl-sn-glycero- [26] H. Shindou, D. Hishikawa, T. Harayama, K. Yuki, T. Shimizu, Recent progress on 3-phosphocholine, a hypotensive and platelet-aggregating lipid, J. Biol. Chem. 255 acyl CoA: lysophospholipid acyltransferase research, J. Lipid Res. 50 (2009) (1980) 10256–10260. S46–S51. [12] S.M. Prescott, G.A. Zimmerman, D.M. Stafforini, T.M. McIntyre, Platelet-activating [27] A.J. Tonks, A. Tonks, R.H.K. Morris, K.P. Jones, S.K. Jackson, Regulation of platelet- factor and related lipid mediators, Annu. Rev. Biochem. 69 (2000) 419–445. activating factor synthesis in human monocytes by dipalmitoyl Phosphatidylcholine, [13] A.B. Tsoupras, E. Fragopoulou, T. Nomikos, C. Iatrou, S. Antonopoulou, C.A. J. Leukocyte Biol. 74 (2003) 95–101. Demopoulos, Characterization of the de novo biosynthetic enzyme of platelet ac- [28] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification, Can. tivating factor, DDT-insensitive cholinephosphotransferase, of human mesangial J. Med. Sci. 37 (1959) 911–917. cells, Mediators Inflamm. 2007 (2007) 27683. [29] L. Servillo, E.L. Iorio, L. Quagliuolo, G. Camussi, C. Balestrieri, A. Giovane, Simulta- [14] D.M. Stafforini, S.M. Prescott, G.A. Zimmerman, T.M. McIntyre, Mammalian platelet- neous determination of lysophospholipids by high-performance liquid chroma- activating factor acetylhydrolases, Biochim. Biophys. Acta 1301 (1996) 161–173. tography with fluorescence detection, J. Chromatogr. B 689 (1997) 281–286. [15] E.S. Lee, J. Jiang, G.C. Sund, W.T. Simonson, J. Graham, G. Dietsch, B. Schimpf, S. [30] H. Mita, H. Yasueda, T. Hayakawa, T. Shida, Quantitation of platelet-activating fac- Bieg, G. Peterman, A. Lernmark, Recombinant human platelet-activating factor tor by high-performance liquid chromatography with fluorescent detection, Anal. acetylhydrolase reduces the frequency of diabetes in the diabetes-prone BB rat, Biochem. 180 (1989) 131–135. Diabetes 48 (1999) 43–49. [31] B.S.Vishwanath,F.J.Frey,G.Escher,J.Reichen, B.M., Liver cirrhosis induces [16] X. Chen, B.A. Hyatt, M.L. Mucenski, R.J. Mason, J.M. Shannon, Identification and renal and liver phospholipase A2 activity in rats, J. Clin. Invest. 98 (2) (1996) characterization of a lysophosphatidylcholine acyltransferase in alveolar type II 365–371. cells, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 11724–11729. [32] P. Ferenci, B. Dragosics, H. Dittrich, H. Frank, L. Benda, H. Lochs, S. Meryn, W. Base, [17] T. Harayama, H. Shindou, R. Ogasawara, A. Suwabe, T. Shimizu, Identification of a B. Schneider, Randomized controlled trial of silymarin treatment in patients with novel noninflammatory biosynthetic pathway of platelet-activating factor, J. Biol. cirrhosis of the liver, J. Hepatol. 9 (1989) 105–113. Chem. 283 (2008) 11097–11106. [33] P. Muriel, T. Garciapiña, V. Perez-Alvarez, M. Mourelle, Silymarin protects against [18] H. Nakanishi, H. Shindou, D. Hishikawa, T. Harayama, R. Ogasawara, A. Suwabe, R. paracetamol-induced lipid peroxidation and liver damage, J. Appl. Toxicol. 12 (1992) Taguchi, T. Shimizu, Cloning and characterization of mouse lung-type acyl-CoA: 439–442. lysophosphatidylcholine acyltransferase 1 (LPCAT1). Expression in alveolar type [34] K. Wellington, B. Jarvis, Silymarin: a review of its clinical properties in the man- II cells and possible involvement in surfactant production, J. Biol. Chem. 281 agement of hepatic disorders, BioDrugs 15 (2001) 465–489. (2006) 20140–20147. [35] G. Serviddio, F. Bellanti, A.M. Giudetti, G.V. Gnoni, A. Petrella, R. Tamborra, A.D. [19] H. Shindou, D. Hishikawa, H. Nakanishi, T. Harayama, S. Ishii, R. Taguchi, T. Romano, T. Rollo, G. Vendemiale, E. Altomare, A silybin-phospholipid complex Shimizu, A single enzyme catalyzes both platelet-activating factor production prevents mitochondrial dysfunction in a rodent model of nonalcoholic and membrane biogenesis of inflammatory cells. Cloning and characterization steatohepatitis, J. Pharmacol. Exp. Ther. 332 (3) (2010) 922–932. of acetyl-CoA:LYSO-PAF acetyltransferase, J. Biol. Chem. 282 (2007) 6532–6539. [36] M. Mathonnet, B. Descottes, D. Valleix, V. Truffinet, F. Labrousse, Y. Denizot, [20] Y. Zhao, Y.Q. Chen, T.M. Bonacci, D.S. Bredt, S. Li, W.R. Bensch, D.E. Moller, M. Platelet-activating factor in cirrhotic liver and hepatocellular carcinoma, World Kowala, R.J. Konrad, G. Cao, Identification and characterization of a major liver J. Gastroenterol. 12 (17) (2006) 2773–2778. lysophosphatidylcholine acyltransferase, J. Biol. Chem. 283 (2008) 8258–8265. [37] J. Chen, L. Yang, J.M. Foulks, A.S. Weyrich, G.K. Marathe, T.M. McIntyre, Intracellu- [21] K. Kropacova, E. Misurova, H. Hacova, Protective and therapeutic effect of silymarin lar PAF catabolism by PAF acetylhydrolase counteracts continual PAF synthesis, J. on the development of latent liver damage, Biol. Radioecol. 38 (1998) 411–415. Lipid Res. 48 (11) (2007) 2365–2376. [22] A.DiSario,E.Bendia,S.Taffetani,A.Omenetti,C.Candelaresi,M.Marzioni,S. [38] D. Hishikawa, H. Shindou, S. Kobayashi, H. Nakanishi, R. Taguchi, T. Shimizu, De Minicis, A. Benedetti, Hepatoprotective and antifibrotic effect of a new Discovery of a lysophospholipid acyltransferase family essential for mem- silybin-phosphatidylcholine-Vitamin E complex in rats, Dig. Liver Dis. 37 braneasymmetryanddiversity,Proc.Natl.Acad.Sci.U.S.A.105(2008) (11) (2005) 869–876. 2830–2835. [23] C. Loguercio, D. Festi, Silybin and the liver: from basic research to clinical practice, [39] R.A. Coleman, D.P. Lee, Enzymes of triacylglycerol synthesis and their regulation, World J. Gastroenterol. 17 (18) (2011) 2288–2301. Prog. Lipid Res. 43 (2004) 134–176. [24] A.M. Giudetti, A.C. Beynen, A.G. Lemmens, G.V. Gnoni, M.J. Geelen, Hepatic lipid [40] T. Yamazaki, M. Wakabayashi, E. Ikeda, S. Tanaka, T. Sakamoto, A. Mitsumoto, N. and carbohydrate metabolism in rats fed a commercial mixture of conjugated Kudo, Y. Kawashima, Induction of 1-acylglycerophosphocholine acyltransferase linoleic acids (Clarinol G-80), Eur. J. Nutr. 44 (1) (2005) 33–39. genes by fibrates in the liver of rats, Biol. Pharm. Bull. 35 (9) (2012) 1509–1515. [25] N. Petrovic, C. Grove, P.E. Langton, N.L. Misso, P.J. Thompson, A simple assay for a [41] B.O. Ibe, L.S. Renteria, Nuclear factor-kappa beta (NF-kB) regulates platelet acti- human serum phospholipase A2 that is associated with high-density lipoproteins, vating factor (PAF) signaling in fetal ovine pulmonary vascular smooth muscle J. Lipid Res. 42 (10) (2001) 1706–1713. cells (PVSMC), FASEB J. 626 (9) (2010).