L

Lam

SIa

b

c

d

a

ARAA

KCFIIKN

1

dds[t(itp

C

1h

Digestive and Liver Disease 46 (2014) 433–439

Contents lists available at ScienceDirect

Digestive and Liver Disease

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

iver, Pancreas and Biliary Tract

ack of interleukin-1� in Kupffer cells attenuates liver inflammationnd expression of inflammatory cytokines in hypercholesterolaemicice

arita Olteanua,1, Michal Kandel-Kfira,1, Aviv Shaisha, Tal Almoga, Shay Shemesha,ris Barshackb,c, Ron N. Apted, Dror Haratsa,c, Yehuda Kamaria,c,∗

The Bert W. Strassburger Lipid Center, Sheba Medical Center, Tel Hashomer, IsraelPathology Department, Sheba Medical Center, Tel Hashomer, IsraelSackler Faculty of Medicine, Tel-Aviv University, IsraelDepartment of Microbiology and Immunology, Ben Gurion University, Beer Sheva, Israel

r t i c l e i n f o

rticle history:eceived 28 November 2013ccepted 20 January 2014vailable online 26 February 2014

eywords:ytokineatty liverL-1nflammationupffer cellsASH

a b s t r a c t

Background: The role of Kupffer cell interleukin (IL)-1 in non-alcoholic steatohepatitis developmentremains unclear.Aims: To evaluate the role of Kupffer cell IL-1�, IL-1� or IL-1 receptor type-1 (IL-1R1) in steatohepatitis.Methods: C57BL/6 mice were irradiated and transplanted with bone marrow-derived cells from WT, IL-1�−/−, IL-1�−/− or IL-1R1−/− mice combined with Kupffer cell ablation with Gadolinium Chloride,and fed atherogenic diet. Plasma and liver triglycerides and cholesterol, serum alanine aminotransferase(ALT), liver histology and expression levels of inflammatory genes were assessed.Results: The ablation and replacement of Kupffer cells with bone marrow-derived cells was confirmed.The atherogenic diet elevated plasma and liver cholesterol, reduced plasma and liver triglycerides andincreased serum ALT levels in all groups. Steatosis and steatohepatitis were induced, but without liverfibrosis. A reduction in the severity of portal inflammation was observed only in mice with Kupffer cell

deficiency of IL-1�. Accordingly, liver mRNA levels of inflammatory genes encoding for IL-1�, IL-1�,TNF�, SAA1 and IL-6 were significantly lower in mice with Kupffer cell deficiency of IL-1� compared toWT mice.Conclusion: Selective deficiency of IL-1� in Kupffer cells reduces liver inflammation and expression ofinflammatory cytokines, which may implicate Kupffer cell-derived IL-1� in steatohepatitis development.

Gast

© 2014 Editrice

. Introduction

Nonalcoholic fatty liver disease (NAFLD) is a major cause of liverisease throughout the world [1]. It reflects a large spectrum of liveriseases ranging from benign fatty liver (steatosis) to non-alcoholicteatohepatitis (NASH), which can lead to cirrhosis and liver cancer2]. NASH is accompanied by histological changes of fat accumula-ion, hepatocellular injury, inflammation (hepatitis) and scarringfibrosis). The pathogenesis of NASH involves various hits, includ-

ng lipotoxicity, gut-derived signals such as endotoxin, signals fromhe innate immune system such as toll like receptors (TLRs) orro-inflammatory cytokines, oxidative and endoplasmic reticulum

∗ Corresponding author at: The Bert Strassburger Lipid Center, Sheba Medicalenter, Tel Hashomer 52621, Israel. Tel.: +972 3 5302940; fax: +972 3 5343521.

E-mail address: yehuda.kamari@sheba.health.gov.il (Y. Kamari).1 Both are first authors.

590-8658/$36.00 © 2014 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevierttp://dx.doi.org/10.1016/j.dld.2014.01.156

roenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.

stress [3]. The lipotoxic molecules implicated in the pathogen-esis of NASH have not been fully clarified, and currently, mostin favour are non-triglyceride lipids such as free fatty acids andfree cholesterol [4,5]. Hepatocyte apoptosis due to lipotoxicity andrelease of inflammatory content of secondary necrotic apoptoticcells initiate sterile inflammation in the liver [6,7]. Kupffer cells,the liver-resident macrophages, release a broad panel of inflam-matory cytokines such as TNF�, IL-6 and IL-1 [3]. IL-1� and IL-1�induce the expression of a variety of pro-inflammatory genes uponbinding to IL-1 receptor type 1 (IL-1R1) [8,9]. Both IL-1� and IL-1�are synthesized as precursors (31 kDa) and are processed to matureforms (17 kDa) via specific cellular proteases [8,10]. IL-1� is gen-erated upon inflammatory signals and is only active as a maturesecreted protein after cleavage by caspase-1. In contrast, IL-1�

exerts its effects both in the mature and the precursor forms whenbinding to IL-1R1 [8,10,11]. IL-1� belongs to a newly recognizedgroup of dual-function cytokines that play a role in the nucleusapart from their extracellular, receptor-mediated effects [15,16].

Ltd. All rights reserved.

4 d Live

Wt[ofcdNIHtralioic1a1l

2

2

dfdcaBmcDIaWI

fWt1

2a

rwBcoohtv4f

2

l

34 S. Olteanu et al. / Digestive an

hen released from dying cells, IL-1� initiates sterile inflamma-ion, acting as a damage-associated molecular pattern molecule8,13]. The link between Kupffer cell-derived IL-1� and regulationf hepatic lipid accumulation has been recently suggested. Kupf-er cells are the main source of IL-1� and depletion of Kupfferells reduced IL-1� expression and liver steatosis under high fatiet [10,12,13]. Activation of Kupffer cells through TLR-4 promotesASH by producing inflammatory cytokines, including TNF� and

L-1� [13] and IL-1R1 deficiency attenuates steatohepatitis [12].owever, the contribution of Kupffer cell-derived IL-1� and IL-1�

o the development of NASH has not been directly addressed. Weecently reported that both IL-1� and IL-1�-deficient mice werelmost entirely protected from atherogenic diet-induced NASH andiver fibrosis, suggesting that these cytokines might be cruciallynvolved in the development of liver inflammation in the contextf fatty liver [14]. In the present study we sought to replace the res-dent Kupffer cells with gene-targeted bone marrow (BM)-derivedells in order to examine whether deficiency of IL-1�, IL-1� or IL-RI in Kupffer cells protects against atherogenic diet-induced NASHnd liver fibrosis. Our results suggest that Kupffer cell-derived IL-� and not IL-1� contributes to the inflammatory milieu of fatty

iver disease in mice.

. Materials and methods

.1. Animals and experimental procedures

The generation of IL-1�, IL-1� and IL-1RI KO mice, has beenescribed previously [15,16]. WT C57BL/6 mice were purchasedrom Harlan, Israel. Mice were maintained on a 12-h light/12-hark cycle and fed a chow diet containing 4.5% fat by weight (0.02%holesterol) (TD19519: Koffolk, Israel). The studies were carried outccording to the protocols approved by the Sheba Medical Centeroard for Studies in Experimental Animals. To induce fatty liver,ice were fed the atherogenic diet containing 17% total fat, 1.25%

holesterol and 0.5% sodium-cholate (Teklad Premier Laboratoryiet No. TD 90221). To study whether deficiency of IL-1�, IL-1� or

L-1RI in Kupffer cells protects against NASH development, we cre-ted the different transplanted groups. We used 10-week-old maleT C57BL6/J mice as recipients and female WT C57BL6/J, IL-1�−/−,

L-1�−/− and IL-1RI−/− mice as donors.The experimental groups: WT cells transplanted to WT mice

ed with chow diet (WT → WT chow) and atherogenic diet-fedT → WT, IL-1�−/− cells to WT (IL-1�−/− → WT), IL-1�−/− cells

o WT (IL-1�−/− → WT) and IL-1RI−/− cells to WT mice (IL-RI−/− → WT).

.2. Bone marrow transplantation (BMT) and Kupffer cellblation

Ten-week-old male WT C57BL/6 mice underwent the bone mar-ow transplantation procedure as previously described [14]. Miceere fed with a standard chow diet for 5 weeks. When using theMT procedure in mice it is estimated that the majority of Kupfferells are replaced after a recovery period of 9 weeks [17]. More-ver, chemical depletion of Kupffer cells speeds up the replacementf host cells by BM derived cells [18]. To ensure replacement ofost cells, we used Gadolinium Chloride (GC) for Kupffer cell abla-ion; Five weeks after the BMT procedure, mice were injected (tailein) with saline containing GC (25 mg/kg in saline, 0.2 mL) (Cat#39770, Lot# 16169EJ, Sigma–Aldrich) and fed standard chow dietor additional 5 weeks.

.3. Chemical analysis of serum and tissue

Animals were sacrificed by inhaling CO2 and blood was col-ected after a 16-h fast, by a puncture of the inferior vena

r Disease 46 (2014) 433–439

cava. In some experiments, blood was obtained from a punc-ture of the retro orbital sinus under mild isoflurane anaesthesia.Livers were immediately collected and weighed, and tissue sam-ples were frozen in liquid nitrogen and stored at −80 ◦C. Serumanalysis of ALT levels was performed using an automated enzy-matic technique (Boehringer Mannheim). Total plasma cholesteroland TG levels were determined using an automated enzymatictechnique (Boehringer Mannheim, Germany). Hepatic lipids wereextracted according to the method of Folch et al. [19]. The extractwas dissolved in 2-propanol and subsequently analyzed for totalcholesterol, and TGs using an automated enzymatic technique(Boehringer Mannheim, Germany).

2.4. Tissue preparation and histology

Livers were harvested, fixed overnight in 4% paraformalde-hyde then embedded in paraffin and sectioned (6-�m thickness).All samples were routinely stained for general morphology withhaematoxylin-eosin and for collagen with Masson’s trichrome. Thehistopathological score was performed in a blinded fashion.

2.5. Analysis of gene expression by real-time PCR

Total RNA was isolated from frozen tissues, treated with DNaseand purified using the Qiagen Rneasy Mini Kit, following the man-ufacturer’s protocol (Qiagen, Hilden, Germany). RNA was reverse-transcribed using the High-Capacity cDNA Reverse TranscriptionKits according to manufacturer’s instructions (Applied Biosystems,Foster City, CA, USA). Real time PCR was carried out using a 7500sequence detection System (Applied Biosystems). Gene expres-sion profiling was achieved using the comparative cycle threshold(ct) method of relative quantification using Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the endogenous gene.Analysis of real-time PCR data was done using the ��Ct method.PCR primers for Taqman/Probe Library assays (Appendix A) weredesigned with the Probe Library Assay Design Center (http://www.roche-applied-science.com/sis/rtpcr/upl/adc.jsp).

2.6. Statistical analyses

The values reported are the mean ± SE. Student’s t-test was used,where applicable, and an ANOVA analysis was applied to assessinteractions between groups and differences between means.P < 0.05 is accepted as statistically significant.

3. Results

We first established the experimental mouse model of replacingKupffer cells with transplanted cells, by using the BMT procedurecombined with Kupffer cell ablation (Figs. S1–S5). We used thismodel to study whether deficiency of IL-1�, IL-1� or IL-1RI inKupffer cells protects from the development of NASH. After irradia-tion, 10-week-old male WT C57BL6/J mice were transplanted withBM-derived cells from WT (WT → WT), IL-1�−/− (IL-1�−/− → WT),IL-1�−/− (IL-1�−/− → WT) and IL-1RI−/− (IL-1RI−/− → WT) mice.Following the BMT and Kupffer cell ablation procedures, the athero-genic diet was given to the mice for 11 weeks. A control group ofWT → WT mice was fed chow diet.

3.1. Serum ALT levels and total plasma cholesterol and TGs inchow-fed mice

We measured serum ALT as a parameter of liver injury and totalplasma cholesterol and TG levels, 10 weeks after the BMT procedureprior to atherogenic diet feeding. Serum ALT and plasma cholesteroland TG levels were similar in all groups (Fig. 1A–C).

S. Olteanu et al. / Digestive and Live

A

B

C

AL

T (

un

its/L

)

WT to

WT c

how

WT to

WT

IL-1

a-/-

to W

T

IL-1

b-/- to

WT

IL-1

RI-/

- to W

T0

20

40

60

80

WT to

WT c

how

WT

toW

T

IL-1

a-/-

to W

T

IL-1

b-/- to

WT

IL-1

RI-/

- to W

T0

20

40

60

80

100

120

140

160

180

Pla

sm

a c

ho

leste

rol

(mg

/dl)

WT to

WT c

how

WT to

WT

IL-1

a-/-

toW

T

IL-1

b-/-to

WT

IL-1

RI-/

- to W

T0

20

40

60

80

100

120

Pla

sm

a T

G (

mg

/dl)

Fig. 1. IL-1�, IL-1� or IL-1RI deficiency in Kupffer cells does not affect serum ALT lev-els and total plasma cholesterol and TGs in mice fed chow diet. Serum ALT (A), plasmacholesterol (B) and TG (C) were measured in mice fed chow diet following the BMTand ablation procedure. WT → WT chow (n = 7), WT → WT (n = 7), IL-1�−/− → WT(f

3

1tiWidTw

compared to WT → WT mice on atherogenic diet. To summarize,

n = 7), IL-1�−/− → WT (n = 7) and IL-1RI−/− → WT (n = 7). Blood was obtained fromasted animals. Data represent mean ± SE of each group.

.2. Liver inflammation

All the experimental groups given the atherogenic diet for1 weeks developed fatty liver, manifested by increased livero body weight ratio, increased liver and plasma cholesterol,ncreased serum ALT, and decreased plasma TG, compared to

T → WT on chow diet (Fig. 2). Liver weight was slightly highern IL-1�−/− → WT compared to IL-1�−/− → WT mice with no

ifference in liver TGs and cholesterol, serum ALT and plasmaGs in all the atherogenic diet-fed groups (Fig. 2). Liver TGsere lower in IL-1�−/− → WT and IL-1RI−/− → WT compared to

r Disease 46 (2014) 433–439 435

WT → WT chow-fed mice and plasma cholesterol was higherin IL-1�−/− → WT compared to WT → WT fed the atherogenicdiet (Fig. 2). Liver cholesterol and TG content was similar inIL-1�−/− → WT compared to WT → WT fed the atherogenic diet(Fig. 2). We assessed the severity of NASH in this experimentalmodel with a histopathology score for steatosis, liver inflammationand fibrosis. The morphology of livers from WT → WT mice follow-ing the atherogenic diet had been dramatically altered comparedto WT → WT on chow diet with extensive parenchymal steatosis(about 90% of the lobule) and moderate to severe lobular and portalinflammation (Fig. 3). These morphological changes are compatiblewith severe steatohepatitis. However, in this experimental model,severe steatosis and inflammation were not accompanied by liverfibrosis. This is in contrast to WT mice given atherogenic dietwithout the BMT and ablation procedures, which develop severeNASH with extensive fibrosis. The degree of steatosis was similarin all the atherogenic diet-fed groups (Fig. 3B and Table 1). Thedegree of lobular and portal inflammation after the atherogenicdiet was similar in IL-1�−/− → WT and IL-1RI−/− → WT comparedto WT → WT, whereas IL-1�−/− → WT mice exhibited a reductionin severity of portal inflammation (Fig. 3B and Table 1).

3.3. mRNA levels of inflammatory cytokines

We assessed the expression of inflammation and liver fibrosis-related genes in livers from WT → WT fed chow diet and WT → WT,IL-1�−/− → WT, IL-1�−/− → WT and IL-1RI−/− → WT mice that werefed atherogenic diet for 11 weeks. First, compared to chow-fed,WT → WT mice fed atherogenic diet showed a 2.7, 4.3, 25.5 and 2.8fold increases in mRNA transcript levels of TNF�, the acute phaseproteins SAA1, IL-1Ra and the macrophage marker CD68, respec-tively (Fig. 4A–D). There was also a 1.7 and 5.4 fold increase in mRNAlevels of TGF� and procollagen type I alpha1 (Col1a1) (Fig. 4E andF). The mRNA levels of IL-1� (Fig. 4G) and IL-6 (Fig. 4I) were sim-ilar in WT → WT mice on atherogenic diet compared to WT → WTfed chow diet, and mRNA levels of IL-1� were about 2 fold lowerin WT → WT on atherogenic diet (Fig. 4H). The IL-1� mRNA levelswere undetectable in IL-1�−/− → WT mice, indicating that Kupffercells are the primary source of IL-1� in this experimental model andthat the BMT and ablation procedure was successful. The expres-sion of IL-1� in livers of IL-1�−/− → WT mice was reduced by 60%compared to WT → WT on atherogenic diet, suggesting that othercell types besides Kupffer cells and bone marrow-derived cells alsoproduce IL-1� in this model (Fig. 4G). In addition, the expressionof IL-1� was unaffected in IL-1�−/− → WT compared to WT → WTmice fed atherogenic diet (Fig. 4G), indicating no regulatory effectof IL-1� on expression of IL-1� in bone marrow derived cells. IL-1�−/− → WT mice did not show reduced expression of TNF�, IL-6,SAA1 or TGF� and in fact IL-1Ra, CD68 and collagen were 2.2, 1.5and 1.9 fold higher compared to WT → WT mice on atherogenicdiet (Fig. 4C, D and F). In IL-1RI−/− → WT mice, the mRNA levelsof SAA1, IL-1Ra, collagen. IL-1�, IL-1� and IL-6 were similar com-pared to WT → WT mice (Fig. 4B, C, F, G, H and I) whereas TNF�and TGF� were 46 and 45% lower (Fig. 4A and E). IL-1� deficiencyin Kupffer cells had the most effect on expression of inflammatorygenes in the liver with 50%, 68% and 41% reduction in the expressionof TNF�, SAA1 andIL-6, respectively compared to WT → WT mice(Fig. 4A, B and I). In addition, IL-1� mRNA levels were 67% lowerin IL-1�−/− → WT mice compared to WT → WT mice (Fig. 4H) sug-gesting the possible regulation of IL-1� expression by IL-1�. ThemRNA levels of TGF� and collagen were similar in IL-1�−/− → WT

IL-1�−/− → WT mice uniquely showed a significant and consistentdown regulation of inflammatory genes following feeding with theatherogenic diet compared to WT → WT mice.

436 S. Olteanu et al. / Digestive and Liver Disease 46 (2014) 433–439

A B

C D

E F

WT to

WT c

how

WT to

WT

IL-1

a-/-

to W

T

IL-1

b-/- to

WT

IL-1

RI-/

- to W

T

% L

iver/

Bo

dy w

eig

ht

0

2

4

6

8

10

*

**

WT

to W

T chow

WT to

WT

IL-1

a-/-

toW

T

IL-1

b-/-to

WT

IL-1

RI-/

- to

WT

Liv

er

ch

ole

ste

rol

(mg

/g)

0.0

0.5

1.0

1.5

2.0

*

WT to

WT c

how

WT to

WT

IL-1

a-/-

toW

T

IL-1

b-/-to

WT

IL-1

RI-/

- to

WT

Liv

er

TG

(m

g/g

)

0

1

2

3

4

5

6

a

b b

ab ab

12 22 19 14 15 6

WT to

WT c

how

WT

toW

T

IL-1

a-/-

to W

T

IL-1

b-/-to

WT

IL-1

RI-/

- toW

T

AL

T (

un

it/L

)

0

100

200

300

400

500

*

WT

toW

Tch

ow

WT to

WT

IL-1

a-/-

toW

T

IL-1

b-/-to

WT

IL-1

RI-/

- to

WT

0

50

100

150

200

250

300

Pla

sm

a C

ho

les

tero

l (m

g/d

l)

*

**

WT to

WT c

how

WT

toW

T

IL-1

a-/-

to W

T

IL-1

b-/-to

WT

IL-1

RI-/

- to W

T

Pla

sm

a T

G (

mg

/dl)

0

20

40

60

80

100 *

Fig. 2. IL-1�, IL-1� or IL-1RI deficiency in Kupffer does not affect the degree of fatty liver formation and serum ALT levels. Liver to body weight ratio (A), liver cholesterol (B),liver TG (C), (WT → WT chow (n = 6), WT → WT (n = 8), IL-1�−/− → WT (n = 10), IL-1�−/− → WT (n = 9) and IL-1RI−/− → WT (n = 8)), serum ALT (D), plasma cholesterol (E) andp −/− -1�−/− −/−

o ± SE og

4

1t

lasma TG (F) (WT → WT chow (n = 6), WT → WT (n = 17), IL-1� → WT (n = 19), ILf atherogenic diet. Blood was obtained from fasted animals. Data represent meanroups).

. Discussion

The aim of the study was to assess the role of Kupffer cell IL- and IL-1R1, in fatty liver disease. We used the bone marrowransplantation procedure combined with Kupffer cell ablation to

→ WT (n = 12) and IL-1RI → WT (n = 15)) levels were measured after 11 weeksf each group (a,bP < 0.05 compared to all other groups, **P < 0.05 between selected

generate mice with selective deficiency of IL-1�, IL-1� or IL-1R1 in

myeloid cells including Kupffer cells. The atherogenic diet inducedsteatosis and steatohepatitis but did not induce liver fibrosis. Ourdata show that Kupffer cell deficiency of IL-1�, IL-1� or IL-1RI didnot affect the extent of hepatic lipid accumulation. Histology score

S. Olteanu et al. / Digestive and Liver Disease 46 (2014) 433–439 437

Table 1Haematoxylin and eosin histopathology score.

WT to WT WT to WT IL-1�−/− to WT IL-1�−/− to WT IL-1R−/− to WT Scale

Chow Atherogenic

Steatosis parenchymal involvement 0.3 ± 0.2 3.0 ± 0.0 3.0 ± 0.0 2.9 ± 0.1 3.0 ± 0.0 0–3Fibrosis 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0–4InflammationLobular 1.0 ± 0.3 2.4 ± 0.2 1.9 ± 0.2 2.3 ± 0.3 2.7 ± 0.2 0–3Portal 0.8 ± 0.3 2.8 ± 0.2 1.8 ± 0.3* 2.0 ± 0.2 2.8 ± 0.2 0–3Liver cell injuryBallooning 0.0 ± 0.0 1.4 ± 0.2 1.7 ± 0.2 1.8 ± 0.2 1.5 ± 0.2 0–2Acidophil bodies 0.0 ± 0.0 0.9 ± 0.0 0.9 ± 0.1 1.0 ± 0.0 0.9 ± 0.1 0–1

Key: WT, wild-type; steatosis parenchymal involvement (<5% 0, 5–33% 1, 33–66% 2, >66% 3); fibrosis (none 0, perisinusoidal/periportal 1, perisinusoidal and periportal 2,bridging fibrosis 3, cirrhosis 4); portal inflammation (none/minimal 0, mild 1, moderate 2, severe 3); lobular inflammation (none 0, mild 1, moderate 2, and severe 3); livercell injury: ballooning (none 0, few 1, many 2); acidophil bodies (none/rare 0, many 1).

* P < 0.05 compared to “WT to WT”.

Fig. 3. Histological score of fat accumulation and inflammation. Representative macroscopic pictures of livers from WT → WT chow, WT → WT, IL-1�−/− → WT, IL-1�−/− → WTand IL-1RI−/− → WT mice (A). Liver sections were prepared for histology and stained with haematoxylin and eosin. Magnification 200× (B). Circle indicating fat accumulationin hepatocytes and arrow indicates inflammatory cells.

438 S. Olteanu et al. / Digestive and Live

Fig. 4. Gene expression analysis of inflammation and fibrosis related genes. Quan-titative real-time PCR analysis of mRNA levels in liver samples of WT → WT onchow diet (n = 6) and WT → WT (n = 9), IL-1�−/− → WT (n = 10), IL-1�−/− → WT (n = 9)and IL-1RI−/− → WT (n = 7) mice after 11 weeks of atherogenic diet. Data representmean ± SE. For statistical analysis, t-test between the specified groups (WT → WTchow compared to WT → WT samples (*P < 0.05) and IL-1�−/− → WT, IL-1�−/− → WTand IL-1RI−/− → WT compared to WT → WT samples (**P < 0.05)) was performed anda P-value was calculated (selected gene/GAPDH).

r Disease 46 (2014) 433–439

analysis revealed a reduction in portal inflammation in mice withKupffer cell deficiency of IL-1�. In line with the histology find-ings, liver mRNA levels of the inflammatory genes encoding forIL-1�, IL-1�, TNF�, SAA1 and IL-6 were significantly lower in micewith selective deficiency of IL-1� in Kupffer cells compared to WTmice.

In addition to activation of inflammatory responses in the liver,cytokine signalling pathways have also been reported to increaseliver lipid accumulation which further amplifies the fat-inducedinflammatory response [20–22]. Yet, very little is known about therole of IL-1� in lipid metabolism and the role of Kupffer cells-derived IL-1� in lipid metabolism and fatty liver has not beenstudied. Previous studies in our laboratory employing IL-1� KOmice showed that deficiency of IL-1� resulted in slightly morecholesterol accumulation during the steatosis phase [14]. In thepresent study, accumulation of liver cholesterol was not affectedby Kupffer cell IL-1� deficiency suggesting that IL-1� derived fromother hepatic cell types (hepatocytes or stellate cells) plays a rolein liver cholesterol accumulation. The role of Kupffer cell IL-1� infatty liver has been investigated in a few studies. Deficiency ofIL-1Ra which was accompanied by increased gene expression ofIL-1� in the liver, deteriorated cholesterol metabolism and fattyliver under atherogenic diet, suggesting that increased IL-1� sig-nalling mediated the deterioration of lipid metabolism and liverinflammation [14]. In line with this, a few more recent reportsshowed that IL-1� signalling aggravates accumulation of choles-terol [22] and TGs [20]. Miura et al. examined IL-1� productionin Kupffer cells in the choline-deficient amino acid (CDAA) dietarymodel of steatohepatitis [12]. Intravenous injection of liposomalclodronate, which exclusively depleted Kupffer cells, reduced IL-1�mRNA levels. In addition, mRNA analysis of isolated hepatocytes,Kupffer cells, HSCs and sinusoidal endothelial cells from WT micefed the CDAA diet, showed high levels of IL-1� only in the Kupffercell fraction indicating that this cell type is the main source of IL-1� in experimental steatohepatitis [12]. Cell culture experimentsshowed that IL-1� that binds to IL-1RI increases lipid accumulationin hepatocytes. In line with the in vitro results, IL-1RI deficient miceexhibited reduced steatosis indicating that IL-1R signalling con-tributes to steatosis development [12]. Kupffer cell depletion wasassociated in another study, with decreased expression of IL-1� andreduced liver steatosis via IL-1�-dependent inhibition of PPAR�expression and activity [20]. We found that IL-1�-deficient miceaccumulate similar liver cholesterol levels as WT mice at the steato-sis stage and are protected from steatohepatitis and liver fibrosis.At the steatohepatitis phase, IL-1�-deficient mice had significantlyreduced inflammation and also less steatosis compared to WT mice[14]. This suggests that inflammation through IL-1� promotes fur-ther accumulation of cholesterol in the liver which contributes tosteatohepatitis and liver fibrosis development. Although these datasupport a possible scenario where Kupffer cell-derived IL-1� bindsto IL-1RI in hepatocytes and promotes hepatic lipid accumulation,the results of the present study show that deficiency of IL-1� orIL-1RI in Kupffer cells does not affect the degree of hepatic lipidaccumulation.

We previously found that deficiency of IL-1� or IL-1� protectsmice from developing atherogenic diet-induced severe NASH [14].Bone marrow transplantation experiments revealed that IL-1� orIL-1� from parenchymal liver cells, rather than BM-derived cells,are critical for development of steatohepatitis and liver fibrosisin hypercholesterolaemic mice [14]. Because the atherogenic dietwas administered only 2 weeks after the bone marrow transplan-tation procedure, our previous experiments could not distinguish

whether the protective effect of IL-1 deficiency was due to lack of itsexpression in Kupffer cells or in hepatocytes. In the current work,to ensure complete repopulation of Kupffer cells with the trans-planted cells, the experimental model included BMT, Kupffer cell

d Live

aoopb

IdaTITsIKmIK1mro

KtwglwotatfcBttawc

ttIm

CN

A

L

A

t

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

S. Olteanu et al. / Digestive an

blation and a recovery period of 10 weeks before administrationf atherogenic diet. Kupffer cell deficiency of IL-1� and not IL-1�r IL-1RI, resulted in a reduction of portal inflammation raising theossibility that Kupffer cell-derived IL-1� promotes steatohepatitisy acting in a paracrine fashion on IL-1RI in other liver cell types.

Kupffer cell IL-1� deficiency also lowered liver mRNA levels ofL-1�, IL-1�, TNF�, IL-6 and SAA1. In contrast, Kupffer cell IL-1�eficiency did not attenuate the expression of inflammatory genes,nd Kupffer cell IL-1RI deficiency only reduced mRNA levels ofNF�. These results suggest that the reduced expression of IL-1�,L-6 and SAA due to Kupffer cell IL-1� deficiency is not mediated byNF�. Previous work has shown that IL-1 upregulates the expres-ion of acute phase proteins and inflammatory cytokines includingL-6 [23,24]. Moreover, these results emphasize the central role ofupffer cell IL-1� in regulation of inflammatory pathways that pro-ote the progression of steatosis to steatohepatitis. Kupffer cell

L-1� deficiency decreased liver mRNA levels of IL-1� whereasupffer cell deficiency of IL-1� did not affect mRNA levels of IL-�. Importantly, Kupffer cell deficiency of IL-1RI did not reduceRNA levels of IL-1�. These in vivo results are in line with previous

eports in cultured macrophages, showing the possible regulationf IL-1� transcription by IL-1� independent of IL-1R1 [25,26].

The complex use of irradiation, bone marrow transplantation,upffer cell ablation and atherogenic diet altered a few features of

he steatohepatitis model. For instance, liver mRNA levels of IL-1�ere not elevated and IL-1� mRNA levels were lower, in athero-

enic versus chow-fed WT mice. Despite the lack of increase iniver IL-1 mRNA levels, liver inflammation was attenuated in mice

ith Kupffer cell IL-1� deficiency suggesting that even basal levelsf IL-1� in Kupffer cells are sufficient to affect liver inflamma-ion. In addition, the development of liver inflammation was notccompanied by liver fibrosis. This can possibly be explained byhe lack of increase in IL-1�. Alternatively, the use of GC for Kupf-er cell ablation possibly reduced the capability of hepatic stellateells to produce fibrotic factors. In addition, although the combinedMT and ablation enabled the repopulation of Kupffer cells withhe transplanted myeloid cells, it is not clear that these cells havehe same characteristics as the replaced Kupffer cells. In order tonswer these questions, we are in the process of generating miceith specific deficiency of IL-1� in myeloid cells including Kupffer

ells, by using the Cre-lox system.In conclusion, by using the BMT combined with GC-Kupffer abla-

ion, we demonstrated that Kupffer cell IL-1� contributes to theransformation of steatosis to steatohepatitis possibly by binding toL-1RI on other liver cell-types and thereby promoting liver inflam-

ation.

onflict of interestone declared.

cknowledgments

This study was supported in part by the Talpiot Sheba Medicaleadership grant (Yehuda Kamari).

ppendix A. Supplementary data

Supplementary material related to this article can be found, inhe online version, at http://dx.doi.org/10.1016/j.dld.2014.01.156.

[

r Disease 46 (2014) 433–439 439

References

[1] Satapathy SK, Sanyal AJ. Novel treatment modalities for nonalcoholic steato-hepatitis. Trends in Endocrinology and Metabolism 2010;21:668–75.

[2] Tilg H, Moschen AR. IL-1 cytokine family members and NAFLD: neglected inmetabolic liver inflammation. Journal of Hepatology 2011;55:960–2.

[3] Tilg H, Moschen AR. Evolution of inflammation in nonalcoholic fatty liver dis-ease: the multiple parallel hits hypothesis. Hepatology 2010;52:1836–46.

[4] Alkhouri N, Dixon LJ, Feldstein AE. Lipotoxicity in nonalcoholic fatty liver dis-ease: not all lipids are created equal. Expert Review of Gastroenterology &Hepatology 2009;3:445–51.

[5] Min HK, Kapoor A, Fuchs M, et al. Increased hepatic synthesis and dysregulationof cholesterol metabolism is associated with the severity of nonalcoholic fattyliver disease. Cell Metabolism 2012;15:665–74.

[6] Alkhouri N, Carter-Kent C, Feldstein AE. Apoptosis in nonalcoholic fatty liverdisease: diagnostic and therapeutic implications. Expert Review of Gastroen-terology & Hepatology 2011;5:201–12.

[7] Malhi H, Kaufman RJ. Endoplasmic reticulum stress in liver disease. Journal ofHepatology 2011;54:795–809.

[8] Dinarello CA. Interleukin-1 in the pathogenesis and treatment of inflammatorydiseases. Blood 2011;117:3720–32.

[9] Kamari Y, Werman-Venkert R, Shaish A, et al. Differential role and tissuespecificity of interleukin-1alpha gene expression in atherogenesis and lipidmetabolism. Atherosclerosis 2007;195:31–8.

10] Dinarello CA. Immunological and inflammatory functions of the interleukin-1family. Annual Review of Immunology 2009;27:519–50.

11] Dinarello CA. Biologic basis for interleukin-1 in disease. Blood 1996;87:2095–147.

12] Miura K, Kodama Y, Inokuchi S, et al. Toll-like receptor 9 promotessteatohepatitis by induction of interleukin-1beta in mice. Gastroenterology2010;139:323e7–34e7.

13] Rivera CA, Adegboyega P, van Rooijen N, et al. Toll-like receptor-4 signaling andKupffer cells play pivotal roles in the pathogenesis of non-alcoholic steatohep-atitis. Journal of Hepatology 2007;47:571–9.

14] Kamari Y, Shaish A, Vax E, et al. Lack of interleukin-1alpha or interleukin-1beta inhibits transformation of steatosis to steatohepatitis and liver fibrosisin hypercholesterolemic mice. Journal of Hepatology 2011;55:1086–94.

15] Horai R, Asano M, Sudo K, et al. Production of mice deficient in genes for inter-leukin (IL)-1alpha, IL-1beta, IL-1alpha/beta, and IL-1 receptor antagonist showsthat IL-1beta is crucial in turpentine-induced fever development and glucocor-ticoid secretion. Journal of Experimental Medicine 1998;187:1463–75.

16] Glaccum MB, Stocking KL, Charrier K, et al. Phenotypic and functional charac-terization of mice that lack the type I receptor for IL-1. Journal of Immunology1997;159:3364–71.

17] Aparicio-Vergara M, Shiri-Sverdlov R, de Haan G, et al. Bone marrow transplan-tation in mice as a tool for studying the role of hematopoietic cells in metabolicand cardiovascular diseases. Atherosclerosis 2010;213(2):335–44.

18] Bradshaw G, Gutierrez A, Miyake JH, et al. Facilitated replacement of Kupf-fer cells expressing a paraoxonase-1 transgene is essential for amelioratingatherosclerosis in mice. Proceedings of the National Academy of Sciences ofthe United States of America 2005;102:11029–34.

19] Folch J, Lees M, Sloane Stanley GH. A simple method for the isolation andpurification of total lipids from animal tissues. Journal of Biological Chemistry1957;226:497–509.

20] Stienstra R, Saudale F, Duval C, et al. Kupffer cells promote hepatic steatosisvia interleukin-1beta-dependent suppression of peroxisome proliferator-activated receptor alpha activity. Hepatology 2010;51:511–22.

21] Chen Y, Zhao L, Mei M, et al. Inflammatory stress exacerbates hepaticcholesterol accumulation via disrupting cellular cholesterol export. Journal ofGastroenterology and Hepatology 2012;27:974–84.

22] Ma KL, Ruan XZ, Powis SH, et al. Inflammatory stress exacerbates lipid accu-mulation in hepatic cells and fatty livers of apolipoprotein E knockout mice.Hepatology 2008;48:770–81.

23] Prowse KR, Baumann H. Interleukin-1 and interleukin-6 stimulate acute-phaseprotein production in primary mouse hepatocytes. Journal of Leukocyte Biology1989;45:55–61.

24] Yap SH, Moshage HJ, Hazenberg BP, et al. Tumor necrosis factor (TNF) inhibitsinterleukin (IL)-1 and/or IL-6 stimulated synthesis of C-reactive protein (CRP)and serum amyloid A (SAA) in primary cultures of human hepatocytes.Biochimica et Biophysica Acta 1991;1091:405–8.

25] Kamari Y, Shaish A, Shemesh S, et al. Reduced atherosclerosis and inflammatorycytokines in apolipoprotein-E-deficient mice lacking bone marrow-derived

interleukin-1alpha. Biochemical and Biophysical Research Communications2011;405:197–203.

26] Shemesh S, Kamari Y, Shaish A, et al. Interleukin-1 receptor type-1 in non-hematopoietic cells is the target for the pro-atherogenic effects of interleukin-1in apoE-deficient mice. Atherosclerosis 2012;222:329–36.