L

Pp

SOFa

b

a

ARAA

KILNOP

1

c

i

R

G

h1

Digestive and Liver Disease 46 (2014) 632–638

Contents lists available at ScienceDirect

Digestive and Liver Disease

journa l h om epage: www.elsev ier .com/ locate /d ld

iver, Pancreas and Biliary Tract

olyunsaturated fatty acids balance affects platelet NOX2 activity inatients with liver cirrhosis

tefania Basili a, Valeria Raparelli a, Laura Napoleonea, Maria Del Bena, Manuela Merlib,liviero Riggiob, Cristina Nocellaa, Roberto Carnevalea, Pasquale Pignatelli a,rancesco Violi a,∗, on behalf of the CALC Group1

First Medical Clinic (Department of Internal Medicine and Medical Specialties), Sapienza-University of Rome, ItalyGastroenterology Unit (Department of Clinical Medicine), Sapienza-University of Rome, Italy

r t i c l e i n f o

rticle history:eceived 22 October 2013ccepted 23 February 2014vailable online 2 April 2014

eywords:soprostanesiver cirrhosisADPH oxidasexidative stressolyunsaturated fatty acids

a b s t r a c t

Background: NADPH-oxidase-2 up-regulation has been suggested in liver damage perpetuation via anoxidative stress-mediated mechanism. n-6/n-3 polyunsaturated fatty acids ratio derangement has beenreported in liver disease.Aim: To explore polyunsaturated fatty acids balance and its interplay with platelet oxidative stress inliver cirrhosis.Methods: A cross-sectional study in 51 cirrhotic patients and sex- and age-matched controls was per-formed. Serum polyunsaturated fatty acids and oxidative stress markers (urinary isoprostanes and serumsoluble NADPH-oxidase-2-derived peptide) were measured. The effect on platelet oxidative stress ofn-6/n-3 polyunsaturated fatty acids ratio in vitro and in vivo (1-week supplementation with 3 g/dailyn-3-polyunsaturated fatty acids) was tested.Results: Compared to controls, cirrhotic patients had significantly higher n-6/n-3 polyunsaturated fattyacids ratio. n-6/n-3 polyunsaturated fatty acids ratio correlated significantly with disease severity andoxidative stress markers. In vitro experiments showed that in Child–Pugh C patients’ platelets incubationwith low n-6/n-3 polyunsaturated fatty acids ratio resulted in dose-dependent decrease of radical oxigen

species (−39%), isoprostanes (−25%) and NADPH-oxidase-2 regulation (−51%). n-3 polyunsaturated fattyacids supplemented patients showed significant oxidative stress indexes reduction.Conclusions: In cirrhosis, n-6/n-3 polyunsaturated fatty acids imbalance up-regulates platelet NADPH-oxidase-2 with ensuing oxidative stress. Further study to evaluate if n-3 supplementation may reducedisease progression is warranted.

© 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

. Introduction

Progressive liver fibrosis is the main cause of organ failure in

hronic liver diseases of any aetiology [1].

Oxidative stress has been suggested to play a major rolen the progression of liver disease by contributing to fibrotic

∗ Corresponding author at: Divisione I Clinica Medica, Viale del Policlinico 155,oma 00161, Italy. Tel.: +39 06 4461933; fax: +39 06 49383333.

E-mail address: francesco.violi@uniroma1.it (F. Violi).1 See Appendix A for the CALC (Coagulation Abnormalities in Liver Cirrhosis)roup members.

ttp://dx.doi.org/10.1016/j.dld.2014.02.021590-8658/© 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All

process [2]. Studies in vitro documented a role for reactive oxidantspecies (ROS). In particular, isoprostanes stimulated a prolifera-tion of hepatic stellate cells responsible for the fibrotic process[3–5]. Isoprostanes are chemically stable eicosanoids derived fromarachidonic acid (AA) interaction with ROS [6]. NADPH oxidase 2(NOX2), which is the most important cellular producer of ROS, isa major determinant for isoprostanes production [7–9]. Upon acti-vation platelets produce isoprostanes that contribute to propagateplatelet aggregation via activation of the glycoprotein IIb/IIIa [7]. A

recent study demonstrated that NOX2 and platelet isoprostanesare more up-regulated in patients with severe cirrhosis com-pared with mild suggesting a potential interplay between plateletoxidative stress and liver disease progression [10]. However, the

rights reserved.

Liver Disease 46 (2014) 632–638 633

mo

i(m3aaao

[tdltn

moTtopctttep

2

otlD

anbb(

clcwepac

twa

eIstt

Table 1Clinical and laboratory characteristics of cirrhotic patients and controls.

Variables Cirrhoticpatients(N = 51)

Controls(N = 51)

P

Age, years (mean ± SD) 65 ± 13 66 ± 10 0.375*

Male sex, n (%) 25 (49) 26 (50) 0.843**

BMI, kg/m2 (mean ± SD) 26 ± 5 24 ± 7 0.949**

Diabetes, n (%) 11 (22) 10 (20) 0.807**

Arterial hypertension, n (%) 11 (22) 13 (25) 0.815**

Smoking habit, n (%) 3 (6) 4 (8) 0.695**

Hypercholesterolemia, n(%)

0 0 –

Renal insufficiency, n (%) 0 0 –Arachidonic acid, ng/ml

Median (IQR)10.0 (6.0–14.7) 16.5

(12.5–20.5)<0.001***

n-6 PUFA, ng/mLMedian (IQR)

8.9 (5.6–13.4) 15.0(12.0–20.0)

<0.001***

n-3 PUFA, ng/mLMedian (IQR)

0.8 (0.5–1.3) 2.9 (2.0–3.8) <0.001***

Abbreviations: BMI, body mass index; n-3 PUFA, �-linolenic acid; n-6 PUFA, linoleicacid; PUFA, polyunsaturated fatty acids; Mean ± SD, Mean ± standard deviation;Median (IQR), Median (interquartile range).

* T-test.** Chi-square statistic.

*** Mann–Whitney U-test.

S. Basili et al. / Digestive and

echanism accounting for NOX2 up-regulation and isoprostanesver-production was not clarified.

NOX2 up-regulation may be a consequence of several pro-nflammatory patterns including tumour necrosis factor-alphaTNF-�) activation [11]. Polyunsaturated fatty acids (PUFA), which

ay up- or down-regulate NOX2 by n-6 PUFA (e.g. AA) or by n- PUFA such as eicosapentaenoic acid (EPA) and docosahexaenoiccid (DHA), respectively, could also be implicated. Thus, an imbal-nce between n-6 and n-3 PUFA is potentially detrimental sincen absolute or relative increase of AA could be a potent trigger forxidative stress and cell damage [12].

Abnormal PUFA pattern has been reported in liver disease13–15]. Several experimental and human studies demonstratedhat non-alcoholic steato-hepatitis and non-alcoholic fatty liverisease were associated with the reduction of hepatic n-3

ong-chain PUFA, liver injury and disease progression, but the rela-ionship between n-3 PUFA depletion and oxidative stress wasever investigated [16–27].

On this basis, the study hypothesis was that liver damageight be associated with PUFA imbalance and eventually with

xidative stress and that platelets contribute to this phenomenon.o explore this issue, PUFA balance, urinary isoprostanes excre-ion and soluble NOX2-derived peptide (sNOX2-dp), a markerf NADPH oxidase activation [28], were measured in a cirrhoticopulation affected by different degrees of liver failure and in aontrol group. Moreover, an in vitro study was performed to testhe hypothesis that PUFA imbalance may affect platelet oxida-ive stress via regulation of NOX2 activity in cirrhosis. Finally,he in vivo effect of a short-term n-3 supplementation on mark-rs of oxidative stress was tested in a subgroup of cirrhoticatients.

. Methods and materials

Consecutive patients with diagnosis of cirrhosis based on biopsyr on clinical, laboratory and ultrasound evidence, aged from 18o 85 years, attending Internal Medicine service or Gastroentero-ogic Unit at Sapienza, University of Rome, from December 2009 toecember 2011, were enrolled in the study.

Patients were excluded if they had: (i) less than 18 years ofge; (ii) hepatocarcinoma or extra-hepatic tumours; (iii) sponta-eous bacterial peritonitis or other infectious diseases diagnosedy clinical (fever and/or abdominal pain) and laboratory (ascitic andlood culture, polymorphonuclear count in ascitic fluid) indexes;iv) cholestatic liver diseases.

According to eligibility criteria, a total of 51 out of 75 consecutiveirrhotic screened patients were enrolled: 6 patients had histo-ogical diagnosis of cirrhosis while 45 patients were diagnosed onlinical, laboratory and ultrasound evidence. Twenty-four patientsere excluded: 11 patients had hepatocarcinoma, 5 patients had

xtra-hepatic tumours (n = 3 colon adenocarcinoma; n = 2 lung neo-lasms), 4 patients had spontaneous bacterial peritonitis, 2 wereffected by primary biliary cirrhosis and 2 did not sign the informedonsent.

All patients recruited underwent a complete physical examina-ion with the purpose of scoring liver failure. Degree of liver failureas defined as low (class A), moderate (class B), or severe (class C)

ccording to Child–Pugh’s classification [29].Pharmacological data were recorded for each subject enrolled to

xclude the use of drugs known to interfere with blood coagulation.

nformed consent was obtained from each patient included in thetudy and the study protocol conforms to the ethical guidelines ofhe 1975 Declaration of Helsinki as reflected in a priori approval byhe institution’s human research committee.

2.1. Subjects

Fifty-one cirrhotic patients (25 men and 26 women; aged68 ± 13 years) entered the study. Thirty-one cirrhotic patients wereclassified as A, 17 as B, and 3 as C class according to Child–Pugh’sclassification. Fifteen patients had ascites.

Alcohol abuse (n = 18), viral hepatitis (n = 31) and cryptogeniccirrhosis (n = 2) were the causes of cirrhosis. Eleven patients haddiabetes mellitus and none had hepato-renal syndrome.

Fifteen patients were hospitalized as part of a liver transplant-screening programme in a single institution (n = 15, Gastroen-terology Unit) at the time of inclusion and they were enrolledafter 7 days from admission. All hospitalized cirrhotic patientswere assuming a normo-caloric and normo-proteic Mediterraneandiet.

Outpatients were also advised to maintain a regular foodintake with a minimum of 30 kcal/kg and 1 g/kg proteins perday.

All patients were asked to complete a medical questionnaire (toexclude if they had consumed fish oil supplements or consumedmore than two seafood servings/weekly) and 24-h diet recall. A dietlow in seafood prior to the study day was recommended. A diet con-taining lipid 45 g/d, protein 60 g/d, and energy 1600–1800 kcal/dwas administrated.

None of the patients reported an active alcohol intake in thelast 6 months. Routine biochemical tests were carried out in allsubjects.

Pharmacological data were recorded for all patients. Twopatients in the Child A group, 3 in the Child B group and 2 in theChild C group were on non-absorbable antibiotics at the time of theblood draw. None of the patients received transfusion of plateletsor plasma or used aspirin or other non-steroidal anti-inflammatorydrugs or antibiotics in the 2 weeks before blood and urine collec-tions. No patients were on antioxidant treatment in the previous2 months.

Fifty-one subjects matched for sex, age and metabolic profilewere selected as control group. The complete clinical and lab-oratory characteristics of all recruited subjects are reported in

Table 1.

6 Liver

2

sdM

stm

2

Emt

2

u[c

2

u(8ai

2

CPtta

2

ata

it(n(tottat

2

[

2

um

34 S. Basili et al. / Digestive and

.2. Sample collection and laboratory analysis

After an overnight fast and a rest period of at least 20 min, bloodamples mixed or not with Na citrate 3.8% (1:9, v:v) were with-rawn, without stasis, from the antecubital vein using a 20 G needle.oreover, a morning urine sample was collected.Measurements were performed blinded to the sample origin. All

amples were assayed in duplicate and those showing values abovehe standard curve were re-tested with appropriate dilutions. No

issing data were detected.

.2.1. Analysis of serum fatty acidsSerum n-3 PUFA (�-linolenic acid), n-6 PUFA (linoleic acid), AA,

PA and DHA were determined as methyl esters after vigorousethanolysis with 0.5 N methoxy-HCL at 100 ◦C for 18 h according

o Alessandri et al. [30].

.2.2. Measurement of serum and platelet sNOX2-dpsNOX2-dp, a marker of NADPH oxidase activation, was detected

sing ELISA method as previously described by Pignatelli et al.28]. Values were expressed, as pg/mL; intra-assay and inter-assayoefficients of variation were 5.2 and 6%, respectively.

.2.3. Urinary and platelet isoprostanesUrinary 8-iso-prostaglandin F2� (8-iso-PGF2�) was measured

sing a previously described and validated enzyme immunoassayCayman Chemical Company, Ann Arbor, MI, USA) [31]. Urinary-iso-PGF2� concentrations were corrected for recovery and cre-tinine excretion and expressed as pg/mg of creatinine. Intra- andnter-assay coefficients of variation were 2.1 and 4.5%, respectively.

.3. In vitro study

In vitro study was carried out in platelets obtained fromhild–Pugh class C patients to evaluate if different ratios of n-6/n-3UFA resulted in changes of ROS production, isoprostanes forma-ion and NOX2 activation. Blood mixed with 3.8% Na citrate wasaken from cirrhotic patients and healthy subjects matched for sexge, who had fasted for 12 h.

.3.1. Preparations of liposomesLiposomes were prepared according to Alessandri et al. [30]. The

mount of liposomes that was added to each sample was adjustedo approximately 50 nM of phosholipids, 20 and 100 ng/mL linoleiccid and 5 ng/mL �-linolenic acid.

Platelets (2 × 108/mL: final concentration in tyrode buffer) werencubated at 37 ◦C for 30 min with liposomes in the form of dimyris-oylphosphatidylcholine (LPC)/18:2 n-6 (20 ng/mL)/LPC/18:3 n-35 ng/mL) (PUFA ratio = 4) or LPC/18:2 n-6 (100 ng/mL)/LPC/18:3-3 (5 ng/mL) (PUFA ratio = 20) in presence or not of apocynin50 �M). Platelet suspension, incubated with liposomes containingwo different n-6/n-3 PUFA ratio values that mimicked the ratiobserved in study population: (i) n-6/n-3 PUFA ratio = 4, similar tohat detected in controls and (ii) n-6/n-3 PUFA ratio = 20 similar tohat found in Child–Pugh class C cirrhotic patients, was used fornalysis of platelet ROS production, platelet 8-iso-PGF2� produc-ion and sNOX2-dp activation.

.3.2. Platelet ROS productionPlatelet ROS production was evaluated according to Basili et al.

32]. Intra-assay coefficient of variation was 5%.

.3.3. Platelet 8-iso-PGF2 ̨ assaysUpon platelet activation, the supernatant was stored at −80 ◦C

ntil measurement. Quantification of isoprostanes was performedeasuring 8-iso-PGF2� by a previously described and validated EIA

Disease 46 (2014) 632–638

assay method [31] and was evaluated in vivo and in vitro study.Intra-assay and inter-assay coefficients of variation were 5.8% and5.0%, respectively.

2.3.4. ELISA detection of platelet sNOX2-dpNOX2-derived peptide was measured in the supernatant by

ELISA method as previously described by Pignatelli et al. [28].

2.4. n-3 PUFA short-term supplementation

In a sub-group of consecutively recruited Child–Pugh class Cpatients (N = 5; 4 M and 1 F, 51 ± 9 years), an interventional studywith n-3 PUFA supplementation was performed.

Patients taking supplements containing fish oil or with highcontent of n-3 PUFA were excluded. All patients had to followdietary rules that included a Mediterranean diet during the preced-ing week. According to previous studies [33,34], a dosage of 3 g (onepill of 1 g t.i.d) of n-3 PUFA (ratio EPA/DHA = 0.9–1.5) was adminis-trated. Blood samples were collected at baseline, after 2 h, 24 h and7 days from PUFA assumption, to measure markers of oxidativestress and n-3 and n-6 PUFA serum levels.

2.5. Statistical analysis

The sample size was computed with respect to a two-tailed Stu-dent’s t-test for independent groups, considering as (i) clinicallyrelevant difference in n-6/n-3 PUFA ratio to be detected betweenpatients and controls |ı| ≥ 2 (ii) standard deviations homogeneousbetween the groups, SDs = 3, (iii) type-I error probability ̨ 0.05 andpower 1 − ̌ 0.90; this resulted in n = 48 per group.

Statistical analyses were performed using chi-square statistics,unpaired t-test and Pearson’s correlation coefficient analysis orwith appropriate nonparametric tests [one-sample Wilcoxon testand Mann–Whitney U-test].

To further analyse the relationship between PUFA balance andoxidative stress, a multiple regression analysis, considering age,sex, ascites, serum albumin, total bilirubin, encephalopathy, PT-INR, cholesterol and n-6/n-3 PUFA ratio as possible determinantsof urinary 8-iso-PGF2� or sNOX2-dp was performed.

In vitro experiments were analysed by ANOVA. Data are pre-sented as mean ± SD or as median and interquartile range (IQR)because they showed skewed distribution. Only two-tailed prob-abilities were used for testing statistical significance. Probabilityvalues <0.05 were regarded as statistically significant. All cal-culations were made with the computer program STATISTICA 7(StatSoft, Tulsa, OK, USA).

3. Results

As shown in Table 1, cirrhotic patients had lower serum levelsof n-6 PUFA, n-3 PUFA and AA compared to controls [P < 0.001].

The n-6/n-3 PUFA ratio was significantly higher in cirrhoticpatients [median (IQR): 11.2 (5.4–20.0)] compared with controls[median (IQR): 4.6 (3.7–6.0)] (Mann–Whitney U-test; P < 0.001)(Fig. 1, Panel A) and increased from A (n = 31) [median (IQR): 8.5(5.0–14.6)] to B + C (n = 20) Child–Pugh classes [median (IQR): 16.7(10.9–24.1)] (Mann–Whitney U-test; P = 0.0038). In addition, n-6/n-3 PUFA ratio significantly correlated with Child–Pugh score(r = 0.30, P = 0.35), PT-INR (r = 0.32, P = 0.022) and serum albumin(r = −0.41, P = 0.003). These data were not influenced by the pres-ence of diabetes or by the liver disease aetiology; in fact, levelsof n-6 and n-3 PUFA, as well as n-6/n-3 PUFA ratio, were similar

in post-alcoholic (n = 18) and post-viral (n = 31) cirrhosis [n-6/n-3PUFA ratio, median (IQR): 10.7 (6.5–17.8) vs. 11.3 (5.3–18.1)].

Urinary excretion of isoprostanes and sNOX2-dp were sig-nificantly higher in cirrhotic patients [median (IQR) 300.1

S. Basili et al. / Digestive and Liver

Fig. 1. n-6/n-3 polyunsaturated fatty acids (PUFA) ratio in cirrhotic patients andcontrols. (Panel A) Scattered point graph of n-6/n-3 PUFA ratio in cirrhotic patientsand in control subjects [data were expressed also as median, interquartile range(box) and range]. *P < 0.001. (Panel B) Correlation between urinary isoprostanesaN

(r(tocpc((4rruB

nd n-6/n-3 PUFA ratio in cirrhotic patients. (Panel C) Correlation between solubleOX2-derived peptide (s-NOX2dp) and n-6/n-3 PUFA ratio in cirrhotic patients.

168.7–492.4) pg/mg creatinine and 30.0 (15.0–40.0) pg/mL,espectively] compared with controls [median (IQR) 82.48.9–282.5) pg/mg creatinine and 8.0 (5.0–30.0) pg/mL, respec-ively] (Mann–Whitney U-test; P < 0.0001). Urinary excretionf isoprostanes (r = 0.43, P < 0.002) and sNOX2-dp significantlyorrelated with Child–Pugh score (r = 0.55, P < 0.0001). Thus,atients of Child–Pugh B + C classes showed significantly higheroncentration of urinary isoprostanes [median (IQR): 250.084.7–427.1) pg/mg creatinine] and sNOX2-dp [median (IQR) 4030.5–58.0) pg/mL] than those of Child–Pugh class A [median (IQR):47.5 (276.5–595.6) pg/mg creatinine and 20.0 (8.0–34.0) pg/mL,

espectively] (Mann–Whitney U-test; P = 0.0078 and P = 0.00012,espectively).n-6/n-3 PUFA ratio significantly correlated withrinary excretion of isoprostanes (r = 0.38, P = 0.006; Fig. 1, Panel) and sNOX2-dp (r = 0.30, P = 0.036; Fig. 1, Panel C).

Disease 46 (2014) 632–638 635

Multivariate analysis showed that higher n-6/n-3 PUFA ratio(beta coefficient = 0.26, P = 0.045) and the presence of ascites (betacoefficient = 0.27, P = 0.035) independently predicted urinary excre-tion of isoprostanes. Similar data were obtained using sNOX2 asdependent variable. In fact, both n-6/n-3 PUFA ratio (beta coeffi-cient = 0.27, P = 0.033) and ascites (beta coefficient = 0.42, P = 0.001)independently predicted sNOX2-dp serum values.

3.1. In vitro study

In vitro experiments showed that n-6/n-3 PUFA ratio signif-icantly affected platelet response to the agonists. Thus, n-6/n-3PUFA ratio = 4 [corresponding to the n-6/n-3 PUFA ratio observedin controls] decreased ROS (P < 0.001) and isoprostanes formation(P < 0.001) and NOX2 activation (P < 0.001) in both cirrhotic patientsand controls (Fig. 2). Conversely, platelet incubation with the n-6/n-3 ratio = 20 (corresponding to that found in the Child–Pugh class Cpatients) increased ROS and isoprostanes formation (P < 0.05) andNOX2 activation (P < 0.05) in both cirrhotic patients and controls(Fig. 2). Also, platelet incubated with n-6/n-3 PUFA ratio = 20 andapocynin, a specific inhibitor of NADPH oxidase, decreased ROSproduction, NOX2 activation and isoprostanes formation (Fig. 2).

3.2. Interventional study with n-3 PUFA

Interventional study with n-3 PUFA supplementation showeda decreased of platelet NOX2 activation, platelet isoprostanes for-mation and platelet ROS production after 2 h from supplementation(Fig. 3, Panels A–C).

At the same time an increase of EPA, DHA and LPC 18:3 n-3 at2 h with a consequent decrease of LPC 18:2 n-6/LPC 18:3 n-3 wereobserved. This determined a ratio [n-6/n-3 PUFA ratio = 4] whichwas similar to that detected in control subjects (Fig. 3, Panels D–H).

4. Discussion

This study shows that n-6/n-3 PUFA imbalance may be impli-cated in NOX2-mediated oxidative stress in cirrhotic patients.

It was previously shown that urinary excretion of 8-epiProstaglandin F2 alpha (8-epi-PGF2�), a marker of oxidative stress,is increased in cirrhotic patients particularly in advanced cirrhosissuggesting that oxidant stress might contribute to the deteriorationof liver disease [35]. Consistently with the pivotal role played byNOX2 in isoprostanes over-expression, cirrhotic patients disclosedan up-regulation of NOX2 corresponding with urinary isoprostanesexcretion elevation. Such changes were associated with deteriora-tion of liver failure, as they were more evident in cirrhotic patientsof Child–Pugh’s classes B and C [10]. In this context platelets mayplay an important role as they express NOX2, produce 8-epi-PGF2�on agonist-induced platelet aggregation and contribute to urinary8-epi-PGF2� excretion [10,36]. Accordingly, platelet 8-epi-PGF2�is over-expressed in cirrhosis and correlated with urinary 8-epi-PGF2� suggesting that systemic oxidative stress, as detected byurinary 8-epi-PGF2�, may in part reflect platelet 8-epi-PGF2� over-expression [36]. However, the mechanism potentially accountingfor NOX2-generated platelet isoprostanes is still undefined.

In the present study these findings were supported andextended because it was demonstrated that the n-6/n-3 PUFAratio is potentially implicated in the NOX2 up-regulation, whichultimately leads to enhancing isoprostanes formation. Hence,the increase of n-6/n-3 PUFA ratio was associated with NOX2-generated oxidative stress; of note, such changes progressively

increased from A to B + C Child–Pugh classes suggesting a rela-tionship between n-6/n-3 PUFA imbalance and oxidative stresstogether with liver failure deterioration. An important implica-tion of this finding is that the relative reduction of n-3 PUFA could

636 S. Basili et al. / Digestive and Liver Disease 46 (2014) 632–638

Fig. 2. Relationship between n-6/n-3 polyunsaturated fatty acids balance and platelet oxidative stress. Reactive oxygen species (ROS), isoprostanes formation and NOX2activation in platelets from controls (n = 3; grey columns; left side of figure) and Child–Pugh C class cirrhotic patients (n = 3; black columns; right side of figure) incubatedw (PUFAr breviai

cwpdiPiCp

smmrdt

ith liposomes containing two different ratio of n6/n3 polyunsaturated fatty acids

atio = 20, similar to that found in Child–Pugh class C cirrhotic patients). *P < 0.05. Abndex; sNOX2-dp, soluble NOX2-derived peptide.

ontribute in favouring oxidative stress in cirrhosis. This agreedith a previous report showing that n-3 PUFA down-regulatedlatelet oxidative stress by reducing the activation of NADPH oxi-ase [30]. Further support to this hypothesis was provided by

n vitro experiments in which the relationship between n-6/n-3UFA balance and platelet oxidative stress was analysed. Thus,ncreasing n-6/n-3 PUFA ratio to the same levels observed inhild–Pugh C class resulted in up-regulation of NOX2-generatedlatelet 8-epi prostaglandin F2 alpha.

Together these data indicate that n-6/n-3 PUFA ratio progres-ively increases jointly with liver failure deterioration and that itight be responsible for NOX2-generated oxidative stress. Various

echanisms, potentially involved in such derangement, have been

eported: (i) decreased uptake of essential fatty acids by the liverue to portacaval shunt; (ii) low absorption from the small intes-ine due to lack of bile; (iii) impaired synthesis from essential fatty

) (n6/n3 PUFA ratio = 4, similar to that detected in control subjects and n6/n3 PUFAtions: AA, arachidonic acid; n-3, �-linolenic acid; n-6, linoleic acid; S.I., stimulation

acid (�5- and �6-desaturases and elongase are in the microsomalfraction of the liver) [37–40].

This finding has potential implication as it suggests thatreducing n-6/n-3 PUFA ratio may lower oxidative stress bydown-regulating NOX2 activation. Interestingly, previous studyin experimental model of hepatitis showed that n-3 supplemen-tation alleviated hepatic injury and inflammation by reduction ofhepatic gene expression of TNF-�, interleukin-1�, interferon-� andinterleukin-6 [41]. Moreover, case reports suggest that n-3 PUFAmight be beneficial in the treatment of neonatal liver injury withparental formulation [42]. PUFA administration to Pten KO mice hasbeen showed to decrease inflammatory activity in chronic hepatitis

by inhibiting ROS production [43].

To corroborate the hypothesis of an n-6/n-3 PUFA ratio effect onNOX2 activity, an interventional study in a subgroup of Child–PughC patients was performed. Such supplementation restored the

S. Basili et al. / Digestive and Liver Disease 46 (2014) 632–638 637

Fig. 3. Platelet soluble NOX2-derived peptide (sNOX2-dp) (Panel A), platelet isoprostanes formation (Panel B), platelet reactive oxygen species (ROS) production (Panel C),serum dimyristoylphosphatidylcholine (LPC) 18:2 n-6 (Panel D), serum LPC 18:3 n-3 (Panel E), ratio LPC 18:2 n-6/LPC 18:3 n-3 (Panel F), eicosapentaenoic acid (EPA) (Panel G)a d 7 da(

ncpwaeo

nd docosahexaenoic acid (DHA) (Panel H) before (T0) and after 2 h (T1), 24 h (T2) anN = 5). Data have been expressed as mean ± standard error. *P < 0.05.

-6/n-3 PUFA ratio to the same ratio found in controls and,oincidentally, lowered oxidative stress; in particular, reducedlatelet ROS and isoprostanes along with NOX2 down-regulation

ere found. Of note, such effect was detected as early as 2 h

fter PUFA supplementation, suggesting an immediate antioxidantffect. These results, however, should be wisely considered becausef small sample size and open design of the study. Moreover, it was

ys (T3) of treatment with n-3 PUFA supplementation in Child–Pugh class C patients

not verified if the n-3 short supplementation could modify the fattyacid profile in liver tissue.

In conclusion, the present study shows that liver cirrhosis is

associated with an increase of n-6/n-3 PUFA ratio that may beresponsible for up-regulation of NOX2-generated oxidative stress.Such association is particularly evident with deterioration of liverfailure, suggesting a possible relationship between n-6/n-3 PUFA

6 Liver

dpta

C

F

o

A

FBSS

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

38 S. Basili et al. / Digestive and

erangement and liver damage progression. Therefore, these datarovide a scientific background to evaluate in a randomized con-rol study if n-3 PUFA supplementation may reduce oxidative stressnd eventually liver disease progression in cirrhosis.

onflict of interest

None declared.

unding

This study was supported by a grant from Sapienza-Universityf Rome (Project no. C26F0934XR, Year 2009, to Stefania Basili).

ppendix A.

CALC (Coagulation Abnormalities in Liver Cirrhosis) Group:rancesco Angelico, Daria Amoroso, Simona Bartimoccia, Vito Leliourgio, Cinzia Calabrese, Alessia Carboni, Olivia De Felice, Serena Dianto, Valerio Giannelli, Cristina Lucidi, Carmen Nigro, Francescaerena Pignataro, Marco Proietti, Roberta Russo, Giovanni Talerico.

eferences

[1] Pinzani M, Rombouts K. Liver fibrosis: from the bench to clinical targets. Diges-tive and Liver Disease 2004;36:231–42.

[2] Gressner OA, Weiskirchen R, Gressner AM. Biomarkers of hepatic fibrosis,fibrogenesis and genetic pre-disposition pending between fiction and reality.Journal of Cellular and Molecular Medicine 2007;11:1031–51.

[3] Gardi C, Arezzini B, Monaco B, et al. F2-isoprostane receptors on hepatic stellatecells. Laboratory Investigation 2008;88:124–31.

[4] Comporti M, Arezzini B, Signorini C, et al. Oxidative stress, isoprostanes andhepatic fibrosis. Histology and Histopathology 2009;24:893–900.

[5] De Minicis S, Seki E, Paik YH, et al. Role and cellular source of nicotin-amide adenine dinucleotide phosphate oxidase in hepatic fibrosis. Hepatology2010;52:1420–30.

[6] Morrow JD, Hill KE, Burk RF, et al. A series of prostaglandin F2-like compoundsare produced in vivo in humans by a non-cyclooxygenase, free radical-catalyzed mechanism. Proceedings of the National Academy of Sciences of theUnited States of America 1990;87:9383–7.

[7] Pignatelli P, Carnevale R, Di Santo S, et al. Inherited human gp91phox deficiencyis associated with impaired isoprostane formation and platelet dysfunction.Arteriosclerosis, Thrombosis, and Vascular Biology 2010;31:423–34.

[8] Violi F, Basili S, Nigro C, et al. Role of NADPH oxidase in atherosclerosis. FutureCardiology 2009;5:83–92.

[9] Seno T, Inoue N, Gao D, et al. Involvement of NADH/NADPH oxidase in humanplatelet ROS production. Thrombosis Research 2001;103:399–409.

10] Basili S, Raparelli V, Riggio O, et al. NADPH oxidase-mediated plateletisoprostane over-production in cirrhotic patients: implication for platelet acti-vation. Liver International 2011;31:1533–40.

11] Morgan MJ, Liu ZG. Reactive oxygen species in TNF alpha-induced signaling andcell death. Molecules and Cells 2010;30:1–12.

12] Calder PC. n-3 Polyunsaturated fatty acids, inflammation, and inflam-matory diseases. American Journal of Clinical Nutrition 2006;83(Suppl.):1505S–19S.

13] El-Badry AM, Graf R, Clavien PA. Omega 3–Omega 6: What is right for the liver?Journal of Hepatology 2007;47:718–25.

14] Lee S, Gura KM, Puder M. Omega-3 fatty acids and liver disease. Hepatology2007;45:841–5.

15] Araya J, Rodrigo R, Videla LA, et al. Increase in long-chain polyunsaturated fattyacid n-6/n-3 ratio in relation to hepatic steatosis in patients with non-alcoholicfatty liver disease. Clinical Science (London) 2004;106:635–43.

16] Sekiya M, Yahagi N, Matsuzaka T, et al. Polyunsaturated fatty acids ame-liorate hepatic steatosis in obese mice by SREBP-1 suppression. Hepatology2003;38:1529–39.

17] Videla LA, Rodrigo R, Araya J, et al. Oxidative stress and depletion ofhepatic long-chain polyunsaturated fatty acids may contribute to nonal-coholic fatty liver disease. Free Radical Biology and Medicine 2004;37:1499–507.

18] Alwayn IP, Andersson C, Zauscher B, et al. Omega-3 fatty acids improve hepatic

steatosis in a murine model: potential implications for the marginal steatoticliver donor. Transplantation 2005;79:606–8.

19] Kurihara T, Adachi Y, Yamagata M, et al. Role of eicosapentaenoic acid in lipidmetabolism in the liver, with special reference to experimental fatty liver.Clinical Therapeutics 1994;16:830–7.

[

Disease 46 (2014) 632–638

20] Capanni M, Calella F, Biagini MR, et al. Prolonged n-3 polyunsaturated fatty acidsupplementation ameliorates hepatic steatosis in patients with non-alcoholicfatty liver disease: a pilot study. Alimentary Pharmacology and Therapeutics2006;23:1143–51.

21] Shapiro H, Tehilla M, Attal-Singer J, et al. The therapeutic potential of long-chain omega-3 fatty acids in nonalcoholic fatty liver disease. Clinical Nutrition2011;30:6–19.

22] Nobili V, Sanyal AJ. Treatment of nonalcoholic fatty liver disease in adultsand children: a closer look at the arsenal. Journal of Gastroenterology2012;47:29–36.

23] Cussons AJ, Watts GF, Mori TA, et al. Omega-3 fatty acid supplementationdecreases liver fat content in polycystic ovary syndrome: a randomized con-trolled trial employing proton magnetic resonance spectroscopy. Journal ofClinical Endocrinology and Metabolism 2009;94:3842–8.

24] Oya J, Nakagami T, Sasaki S, et al. Intake of n-3 polyunsaturated fattyacids and non-alcoholic fatty liver disease: a cross-sectional study inJapanese men and women. European Journal of Clinical Nutrition 2010;64:1179–85.

25] Tanaka N, Sano K, Horiuchi A, et al. Highly purified eicosapentaenoic acidtreatment improves nonalcoholic steatohepatitis. Journal of Clinical Gastro-enterology 2008;42:413–8.

26] Vega GL, Chandalia M, Szczepaniak LS, et al. Effects of N-3 fatty acids onhepatic triglyceride content in humans. Journal of Investigative Medicine2008;56:780–5.

27] Zhu FS, Liu S, Chen XM, et al. Effects of n-3 polyunsaturated fatty acids from sealoils on nonalcoholic fatty liver disease associated with hyperlipidemia. WorldJournal of Gastroenterology 2008;14:6395–400.

28] Pignatelli P, Carnevale R, Cangemi R, et al. Atorvastatin inhibits gp91phoxcirculating levels in patients with hypercholesterolemia. Arteriosclerosis,Thrombosis, and Vascular Biology 2010;30:360–7.

29] Pugh RN, Murray-Lyon IM, Dawson JL, et al. Transection of the oesopha-gus for bleeding oesophageal varices. British Journal of Surgery 1973;60:646–9.

30] Alessandri C, Pignatelli P, Loffredo L, et al. Alpha-linolenic acid-rich wheatgerm oil decreases oxidative stress and CD40 ligand in patients with mildhypercholesterolemia. Arteriosclerosis, Thrombosis, and Vascular Biology2006;26:2577–8.

31] Hoffman SW, Roof RL, Stein DG. A reliable and sensitive enzyme immunoassaymethod for measuring 8-isoprostaglandin F2 alpha: a marker for lipid per-oxidation after experimental brain injury. Journal of Neuroscience Methods1996;68:133–6.

32] Basili S, Pignatelli P, Tanzilli G, et al. Anoxia-reoxygenation enhances plateletthromboxane A2 production via reactive oxygen species-generated NOX2:effect in patients undergoing elective percutaneous coronary intervention.Arteriosclerosis, Thrombosis, and Vascular Biology 2011;31:1766–71.

33] Phang M, Lincz L, Seldon M, et al. Acute supplementation with eicosapen-taenoic acid reduces platelet microparticle activity in healthy subjects. Journalof Nutritional Biochemistry 2012;23:1128–33.

34] Phang M, Sinclair AJ, Lincz LF, et al. Gender-specific inhibition of platelet aggre-gation following omega-3 fatty acid supplementation. Nutrition, Metabolismand Cardiovascular Diseases 2012;22:109–14.

35] Pratico D, Iuliano L, Basili S, et al. Enhanced lipid peroxidation in hepatic cir-rhosis. Journal of Investigative Medicine 1998;46:51–7.

36] Carnevale R, Iuliano L, Nocella C, et al. Relationship between platelet andurinary 8-Iso-PGF2� levels in subjects with different degrees of NOX2regulation. Journal of the American Heart Association 2013;2:e000198,http://dx.doi.org/10.1161/JAHA.113.000198.

37] Watanabe A, Saito S, Tsuchida T, et al. Low plasma levels of docosahexaenoicacid in patients with liver cirrhosis and its correction with a polyunsaturatedfatty acid-enriched soft oil capsule. Nutrition 1999;15:284–8.

38] Cabré E, Periago JL, Abad-Lacruz A, et al. Plasma fatty acid profile in advancedcirrhosis: unsaturation deficit of lipid fractions. American Journal of Gastroen-terology 1990;85:1597–604.

39] González J, Periago JL, Gil A, et al. Malnutrition-related polyunsaturated fattyacid changes in plasma lipid fractions of cirrhotic patients. Metabolism: Clinicaland Experimental 1992;41:954–60.

40] Johnson SB, Gordon E, McClain C, et al. Abnormal polyunsaturated fatty acidpatterns of serum lipids in alcoholism and cirrhosis: arachidonic acid deficiencyin cirrhosis. Proceedings of the National Academy of Sciences of the UnitedStates of America 1985;82:1815–8.

41] Schmöcker C, Weylandt KH, Kahlke L, et al. Omega-3 fatty acids alleviatechemically induced acute hepatitis by suppression of cytokines. Hepatology2007;45:864–9.

42] Gura KM, Duggan CP, Collier SB, et al. Reversal of parenteral nutrition-associated liver disease in two infants with short bowel syndrome

using parenteral fish oil: implications for future management. Pediatrics2006;118:e197–201.

43] Ishii H, Horie Y, Ohshima S, et al. Eicosapentaenoic acid ameliorates steatohep-atitis and hepatocellular carcinoma in hepatocyte-specific Pten-deficient mice.Journal of Hepatology 2009;50:562–71.