Genomic and functional characterization of stellate

cells isolated from human cirrhotic livers

Pau Sancho-Bru1, Ramon Bataller1,*, Xavier Gasull2, Jordi Colmenero1, Valeriya Khurdayan1,

Arcadi Gual2, Josep M. Nicolas3, Vicente Arroyo1, Pere Gines1

1Liver Unit, Hospital Clınic, Institut d’Investigacions Biomediques August Pi i Sunyer (IDIBAPS), Barcelona, Catalonia, Spain2Laboratory of Neurophysiology, University of Barcelona School of Medicine, IDIBAPS, Barcelona, Catalonia, Spain

3Department of Internal Medicine, Hospital Clınic, IDIBAPS Barcelona, Catalonia Spain

Background/Aims: Hepatic stellate cells (HSCs) are believed to participate in liver fibrogenesis and portal

hypertension. Knowledge on human HSCs is based on studies using HSCs isolated from normal livers. We investigated

the phenotypic, genomic and functional characteristics of HSCs from human cirrhotic livers.

Methods: HSC were obtained from normal and cirrhotic human livers. Cells were characterized by

immunocytochemistry and gene microarray analysis. Cell proliferation, Ca2+ changes and cell contraction were

assessed by 3H-thymidine incorporation and by using an epifluorescence microscope.

Results: HSCs freshly isolated from human cirrhotic livers showed phenotypical features of myofibroblasts. These

features were absent in HSCs freshly isolated from normal human livers and become prominent after prolonged

culture. HSCs from cirrhotic human livers markedly express genes involved in fibrogensis, inflammation and apoptosis.

HSCs from normal livers after prolonged culture preferntially expressed genes related to fibrogenesis and contractility.

Agonists induced proliferation, Ca2+ increase and cell contraction in HSCs isolated from human cirrhotic livers.

Response to agonists was more marked in culture-activated HSCs and was not observed in HSCs freshly isolated from

normal livers.

Conclusions: HSCs from human cirrhotic livers show fibrogenic and contractile features. However, the current

model of HSCs activated in culture does not exactly reproduce the activated phenotype found in cirrhotic human

livers.

q 2005 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved.

Keywords: Liver fibrosis; Cell activation; Collagen; Microarray; Portal hypertension

0168-8278/$30.00 q 2005 European Association for the Study of the Liver. Pub

doi:10.1016/j.jhep.2005.02.035

Received 28 September 2004; received in revised form 17 January 2005;

accepted 23 February 2005; available online 11 May 2005* Corresponding author.

E-mail address: bataller@clinic.ub.es (R. Bataller).

Abbreviations: Ang II, angiotensin II; AT1, angiotensin receptor type 1;

BKCa, big potassium-dependent calcium channel; ECM, extracellular

matrix; FBS, fetal bovine serum; HSCs, hepatic stellate cells; PDGF,

platelet-derived growth factor; aSMA, smooth muscle alpha actin; TGF-b1,

transforming growth factor b1; TIMP-1, tissue inhibitor of metalloprotei-

nase type 1; PAI, plasminogen activator inhibitor type 1; MCAM,

melanoma cell adhesion molecule.

1. Introduction

In chronic liver diseases, hepatic stellate cells (HSCs)

acquire myofibroblastic properties including collagen syn-

thesis and contractility [1], playing a role in the pathogen-

esis of liver fibrosis and portal hypertension [2]. This

assumption is based on experimental studies, while

evidence from human studies is scarce. Early studies

showed that collagen-expressing activated HSCs accumu-

late in patients with diseased livers [3–5]. However, the

pathogenic role of HSCs in chronic liver diseases is unclear.

Functional characteristics of human HSC have been

extensively studied using the model of activation in culture

Journal of Hepatology 43 (2005) 272–282

www.elsevier.com/locate/jhep

lished by Elsevier B.V. All rights reserved.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282 273

(i.e. HSCs isolated from normal human livers in prolonged

culture on uncoated plastic dishes) [6,7]. The functional and

genomic characteristics of myofibroblastic HSCs from

damaged human livers have not been assessed.

Because of their high content in vitamin A, HSCs can be

isolated from normal human livers using a discontinuous

gradient [7,8]. In early culture, human HSCs show a

quiescent phenotype that consists of a round cell shape and

numerous vitamin A-rich fat droplets [9]. After prolonged

culture on plastic, human HSCs acquire characteristics of

myofibroblast-like cells [7,10]. Cell activation in culture

involves: (1) increased cellular proliferation in response to

agonists such as platelet-derived growth factor (PDGF) and

angiotensin II (Ang II) [11–14]; (2) a dramatic increase in

extracellular matrix (ECM) protein expression [3]; (3)

expression of metalloproteinases [15]; (4) expression of cell

adhesion molecules and secretion of pro-inflammatory

cytokines [16–18] as well as features of antigen-presenting

cells [19]; and (5) de novo expression of cytoskeletal

proteins, receptors for vasoactive substances and membrane

channels [7,20].

It has been proposed that activation of human HSCs in

culture reproduces the phenotypical changes that occur in

chronic liver diseases [21]. However, no studies have

tested this hypothesis. It is conceivable that the degree of

phenotypic activation of culture-activated human HSCs

exceeds the actual cell activation that occurs in patients

with chronic liver diseases. Here, we characterize the

phenotypic, genomic and functional profile of HSCs

freshly isolated from cirrhotic human livers. Moreover,

we compare the features of HSCs freshly isolated from

cirrhotic human livers, with those of HSCs isolated from

normal livers in early culture and after prolonged culture

on plastic.

2. Materials and methods

2.1. Isolation and culture of human HSCs

HSCs were isolated from human cirrhotic livers obtained from patientssubmitted to liver transplantation (nZ10). Eight patients had hepatitis Cvirus-induced liver cirrhosis and two had alcohol-induced liver cirrhosis.Twenty-five grams of liver tissue were digested by two enzymaticsolutions. First digestion was performed in Gey’s Balanced Salt Solution(GBSS, Sigma Chemical Co., St Louis, MI) containing 0.33% pronase,0.053% collagenase, and 0.003% DNase (Roche Diagnostics, Mannheim,Germany) for 45 min at 37 8C. Second digestion was performed in GBSScontaining 0.08% pronase, 0.04% collagenase, and 0.003% DNase for30 min at 37 8C. The resulting cell pellet was centrifuged over a gradient ofNycodenz 13% (Sigma). Cells obtained from the upper layer were seededfor 15 min in serum free medium to allow Kupffer cell attachment. Toremove contaminating leucocytes, non-attached cells were recovered andpurified using magnetic anti-CD45 beads (Dynal Biotech, Compiegne,France). Average yield per isolation was 2.5!105 cells/g liver. HSCs werealso isolated from fragments of normal human livers obtained fromresections of liver metastasis (nZ5). Briefly, liver tissue was digested bytwo enzymatic solutions. First digestion was performed in GBSS containing0.33% pronase, 0.035% collagenase, and 0.001% DNase for 30 min at37 8C. Second digestion was performed in GBSS containing 0.06%pronase, 0.035% collagenase and 0.001% DNase for 30 min at 37 8C.

The resulting cell pellet was centrifugated over a gradient of 10%Nycodenz. Average yield per isolation was 5!105 cells/g liver. HSCsisolated from both cirrhotic livers were studied at 24 h after isolation. HSCsfrom normal livers were studied at 24 h after isolation (quiescentphenotype) and after the second serial passage (culture-activatedphenotype). In all cultures, no staining was found for CD45, factor VIIIrelated-antigens, and Cam 5.2 (Dako, Glostrup, Denmark), indicating theabsence of mono/macrophagic, endothelial, and epithelial cells. Cells werecultured in Iscove’s Modified Dulbecco’s Medium (IMDM, BioWhittaker,Verviers, Belgium) containing 15% fetal bovine serum. The protocol wasapproved by the Investigational Review Board of the Hospital Clinic ofBarcelona.

2.2. Cell proliferation assay

DNA synthesis was estimated by methyl-3H-thymidine (AmershamBiosciences, Buckinghamshire, UK) incorporation, as described in detailpreviously [22].

2.3. Immunocytochemistry studies

Cells were fixed in methanol at K20 8C for 10 min, blocked in PBScontaining goad 2% FCS for 30 min, and incubated with primary antibodiesanti-smooth muscle a-actin (aSMA) (1:100) (Dako), vimentin (1:200),fibronectin (1:100) (Sigma), nerve growth factor (1:100) (Santacruz),Synaptophysin (1:50) (Dako), neural cell adhesion molecule (1:100)(Sigma) for 1 h. Cells were incubated with fluorescent secondary antibodyfor 1 h.

2.4. Gene expression analysis

Total RNA was isolated from HSCs freshly isolated from cirrhoticlivers, HSCs isolated from normal livers after prolonged culture and totalnormal human livers with Trizol (Life Technologies, Inc., Rockville, MD).RNA integrity and concentration was assessed with a microfluidic glasschip platform (Bioanalizer 2100, Agilent, Palo Alto, CA). Six microgramsof total RNA were used for microchip hybridizations. Preparation of cRNAprobes, hybridization, and scanning of arrays were performed according tomanufacturer’s protocol (Affimetrix, Santa Clara, CA). CELL filesgenerated by Affymetrix Microarray Suit Software were imported intothe dChip Software to calculate the perfect match/mismatched differencemodel expression value [23]. Arrays were normalized against the medianintensity array. Approximately 8400 genes included in HG-Focus arraywere analyzed. Genes with a coefficient of variation of less than 0.5 acrossall samples, and a P call of at least 30% were filtered, yielding 2142 genesthat were considered for further analysis. To identify differentiallyexpressed genes, a combined comparison of the different group of sampleswas performed. A fold change exceeding 2 and a t-test P-value less than0.01 were considered significant. Unsupervised hierarchical clustering ofthe 740 resulting genes was performed. Gene Ontology annotations wereused to assess enriched functional clusters (P!0.001 was consideredsignificant) [24].

2.5. Quantitative polymerase chain reaction (PCR)

Pre-designed Assays-on-Demand TaqMan probes and primer pairs forcollagen a1(I), transforming growth factor type 1 (TGF-b1), tissue inhibitorof metalloproteinase type 1 (TIMP-1), intercellular adhesion molecule type1 (ICAM-1), lysyl oxidase, thrombospondin type 1, plasminogen activatorinhibitor type 1 (PAI), and melanoma cell adhesion molecule (MCAM)were obtained from Applied Biosystems (Foster City, CA) (Table 1).Information on these Assay-on-Demand is available in: http://myscience.appliedbiosystems.com/cdsEntry/Form/gene_expression_keyword.jsp.TaqMan reactions were carried out in duplicate on an ABI PRISM 7900machine (Applied Biosystems).

Table 1

Information on selected genes studied by quantitative PCR

Gene Cat. number Ref. sequence Exon boundary Assay location

collagen, type I, alpha 1 (COL1A1) Hs00164004_m1 NM_000088 Exon 1jExon 2 225

tissue inhibitor of metalloproteinase 1 (TIMP-1) Hs00171558_m1 NM_003254 Exon 5jExon 5 517

transforming growth factor, beta 1 (TGF-B1) Hs00171257_m1 NM_000660 Exon 1jExon 2 1198

intercellular adhesion molecule 1 (ICAM-1) Hs00164932_m1 NM_000201 Exon 2jExon 3 389

lysyl oxidase (LOX) Hs00184700_m1 NM_002317 Exon 2jExon 3 1017

melanoma cell adhesion molecule (MCAM) Hs00174838_m1 NM_006500 Exon 5jExon 6 584

thrombospondin 1 (THBS1) Hs00170236_m1 NM_003246 Exon 7jExon 8 1301

plasminogen activator inhibitor type 1 Hs00167155_m1 NM_000602 Exon 2jExon 3 346

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282274

2.6. Measurement of changes in cytosolic free Ca2C

([Ca2C]i) and cell area

Changes in [Ca2C]i and cell area were measured in Fura-2(Calbiochem, San Diego, CA) loaded cells using an inverted epifluores-cence microscope as described in detail previously [22]. Cells wereconsidered as responders when [Ca2C]i increased more than 50% abovethe resting value. Cell contraction was defined as reduction in cell area ofO8%, a figure greater than meanGSD of spontaneous variations of cellarea in control cells.

2.7. Patch-clamp procedures

Solutions and general procedures to characterize high-conductancecalcium-activated potassium channels (BKCa) were previously described[25]. Single-channel experiments in the cell-attached mode were performedin physiological solution. Cells were clamped at K60 mV and depolarizingpulses were applied in 10 mV steps to record outward KC currents. Whole-cell recordings were performed in physiological solution in the bath and‘high KC’ solution in the pipette. Iberiotoxin 10 nM (Alomone Labs,Israel), a specific BKCa channel blocker, was applied by bath perfusion toblock KC currents.

2.8. Data analysis

Results are expressed as meanGSD, unless otherwise indicated.Statistical analysis was performed by Mann–Whitney test, Student’s t-testand analysis of variance.

3. Results

3.1. Characterization of HSCs isolated from normal

and human cirrhotic livers

Cells isolated from both normal and cirrhotic livers

contained vitamin A-rich fat droplets and expressed neural

crest markers such as synaptophysin, nerve growth factor

and neural cell adhesion molecule, which are indicative of

HSCs (Fig. 1) [4]. The expression of markers of mesench-

ymal origin (vimentin) and myofibroblastic transition

(aSMA and fibronectin) was also assessed. HSCs isolated

from human cirrhotic livers in early culture highly

expressed vimentin, aSMA and fibronectin. The presence

of stress fibers of aSMA and fibronectin was more

prominent in human HSCs from normal livers after

prolonged culture and were barely expressed in cells freshly

isolated from normal livers. HSCs isolated from cirrhotic

livers in early culture showed morphological characteristics

of myofibroblasts with a spindle-like shape and low vitamin

A droplets content (Fig. 2A). When cultured for 7 days,

these cells showed marked proliferation and loss of vitamin

A droplets (Fig. 2B). In contrast, HSCs freshly isolated from

normal human livers showed a quiescent phenotype

consisting of a round shape and numerous vitamin A-rich

droplets (Fig. 2C). Following 7 days in culture, HSCs from

normal human livers spread out and showed morphological

transition to myofibroblast-like cells (Fig. 2D). Overall,

these findings indicate that HSCs freshly isolated from

cirrhotic livers show an activated phenotype, although it is

less marked than that observed in culture-activated HSCs.

3.2. Gene expression studies

We performed an extensive microarray analysis of HSCs

freshly isolated from cirrhotic human livers and HSCs

isolated from normal livers activated after prolonged

culture. Whole normal human liver was used as baseline

and genes only expressed in the liver tissue were discarded.

cDNA was obtained from three independent samples for

each group. We identified 740 genes differentially expressed

in at least one of the study groups. Unsupervised

hierarchical analysis clustered the cell samples into two

different groups. Four groups of interest were identified

(Fig. 3). Group 1 comprises genes with low expression

levels in HSCs and similar expression among cell

phenotypes. Enriched gene ontology clusters related to

oxidoreductases (e.g. catalase, superoxide dismutase) and

blood coagulation (e.g. fibrinogen, coagulation factors)

were identified. Group 2 consists of genes up-regulated in

culture-activated HSCs compared to HSCs freshly isolated

from cirrhotic livers. This group includes enriched func-

tional clusters related to skeletal development, ECM (e.g.

thrombospondin-2, procollagen I), and tumor suppressors

(e.g. caveolin, vinculin), among others. Group 3 comprises

genes similarly expressed in both cell phenotypes and up-

regulated compared to normal human liver. These genes

include enriched clusters such as cytoskeleton (e.g.

vimentin, tubulin) and enzyme regulators (e.g. annexin,

TIMP-1). Finally, group 4 consists of genes up-regulated in

HSCs freshly isolated from cirrhotic livers compared to

cultured-activated HSCs. This group includes enriched

Fig. 1. Immunocytochemistry study of hepatic stellate cells (HSCs) freshly isolated from normal and cirrhotic livers and HSCs from normal human

livers after prolonged culture. HSCs freshly isolated from cirrhotic and normal human livers show vitamin A-rich droplets fluorescence under UV

light, which are not present in culture-activated HSCs. Moreover, HSCs isolated from cirrhotic and normal livers show positive immunostaining for

neural cell adhesion molecule (N-CAM), nerve growth factor (NGF) and synaptophysin (SYN). HSCs from cirrhotic livers show immunostaining for

aSMA and fibronectin, which is increased after 1 week in culture. Well-developed vimentin filaments are expressed in all HSC phenotypes.

Magnifications 200! and 400!.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282 275

functional clusters related to apoptosis regulation (e.g. TNF-

induced protein, apoptosis antagonizing transcription fac-

tor), cell adhesion molecules (e.g. melanoma cell adhesion

molecule, CD58) and other inflammatory genes. Table 2

describes genes similarly and differentially expressed by

different cell phenotypes, respectively. To confirm these

results, expression of well-known genes related to liver

fibrogenesis was assessed by TaqMan PCR. Expression of

procollagen a1(I), TGF-b1, TIMP-1, ICAM-1, lysil oxi-

dase, thrombospondin-1, PAI-1 and MCAM confirmed the

results obtained in the microarray analysis (data not shown).

These results suggest that in chronic liver diseases HSCs

express genes related to ECM, inflammation and apoptosis.

However, the gene expression profile in these cells differs

from that of culture-activated HSCs, which predominantly

express genes involved in fibrogenesis and contractility.

Fig. 2. Phase contrast microscope examination of primary cultures of

human hepatic stellate cells (HSCs). HSCs obtained from cirrhotic

livers at 24 h after cell isolation (A) and 7 days after culture on plastic

(B). HSCs obtained from normal livers at 24 h after cell isolation (C)

and 7 days after culture on plastic (D). Note that HSCs freshly isolated

from cirrhotic livers already show long cell processes similar to those

shown in (B) and (D).

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282276

3.3. Proliferation studies

We next investigated the mitogenic response to PDGF-

BB, Ang II and fetal bovine serum (FBS). These mitogens

induced proliferation of HSCs freshly isolated from

cirrhotic livers, indicating that these cells have proliferative

properties in damaged livers. However, no proliferative

response was found in HSCs freshly isolated from normal

human livers. The mitogenic response to PDGF-BB and

FBS, but not Ang II, was less intense in HSCs isolated from

cirrhotic livers than in culture-activated HSCs (Fig. 4). The

intensity of the mitogenic effect of Ang II in HSCs from

cirrhotic livers was similar to that of PDGF and FBS,

suggesting that Ang II may be a powerful mitogen in

chronic liver diseases. These results indicate that phenotypic

transformation of HSCs in vivo results in the acquisition of

mitogenic properties.

3.4. Electrophysiology, calcium mobilization and cell

contraction studies

To assess whether HSCs isolated from cirrhotic human

livers display contractile properties, we first investigated the

expression of BKCa channels, which regulate intracellular

calcium levels and cell contraction in culture-activated

human HSCs [25]. Membrane channels with conductance of

101G4 pS in cell-attached and 208G4 pS in inside-out

configurations were recorded (Fig. 5). The number of

channel openings increased with bath [Ca2C] (1–20 mM;

data not shown). Moreover, depolarization of HSCs elicited

outward KC currents were blocked by iberotoxin, demon-

strating that HSCs from cirrhotic livers contain functional

BKCa channels. However, the number of channels found in

HSCs from cirrhotic livers was lower than that found in

culture-activated HSCs (54 vs 82%, respectively; nZ50,

P!0.006).

We next studied the calcium increase and cell contraction

in response to vasoactive substances. Ang II induced a

concentration-dependent calcium mobilization in HSCs

isolated from human cirrhotic livers. This response was

characterized by a quick peak of [Ca2C]i followed by a

rapid recovering to resting values. The percentage of

responding cells and the mean [Ca2C]i peak were lower

compared to those of HSCs activated in culture (Table 3 and

Fig. 6). Ang II did not induce calcium increase in HSCs

freshly isolated from normal liver. We next studied whether

[Ca2C]i increase was associated with a reduction in cell

area. Ang II decreased cell area in 33% of HSCs from

cirrhotic livers compared to 52% of culture-activated HSCs

(P!0.05). Similar results were obtained when different

HSC phenotypes were incubated with endothelin-1 and

thrombin (data not shown). These data indicate that HSCs

acquire contractile properties in chronic liver diseases.

4. Discussion

Studies investigating human HSCs have been performed

using passaged culture-activated cells obtained from normal

livers. Few studies in liver specimens from patients with

chronic liver diseases suggest that HSCs express genes/

proteins involved in fibrogenesis [3,4]. Although HSCs

isolated from rodents with experimentally induced liver

cirrhosis show myofibroblastic properties [26,27], no

studies have directly assessed the features of HSCs from

human cirrhotic livers. We provide extensive evidence that

HSCs freshly isolated from cirrhotic human livers display

fibrogenic, mitogenic and contractile properties. These data

confirm the current assumption that HSCs participate in the

pathogenesis of liver fibrosis and portal hypertension.

Moreover, we compare the phenotypic, genomic and

functional features of these cells with those from HSCs

freshly isolated from normal livers and cells activated after

prolonged culture.

The main finding of the current study is that HSCs

isolated from cirrhotic livers display morphological and

immunophenotypical features of myofibroblast-like cells.

Cells isolated from normal livers showed a typical vitamin

A-rich quiescent phenotype. Cells from cirrhotic livers

contained less amount of vitamin A droplets, yet they

showed a spindle-like shape and express aSMA and

fibronectin. The HSC origin of these latter cells isolated is

supported by their vitamin A content and the gene

expression of neural crest markers. These data indicates

that HSCs undergo a phenotypical transition to myofibro-

blasts in chronic liver diseases. The expression of

myofibroblast markers, however, was more patent in

culture-activated HSCs, suggesting that prolonged culture

on plastic results in a non-physiological cell phenotype.

Several reasons may account for this latter finding,

Fig. 3. Pattern of gene expression, as assessed by microarrays analysis. Hierarchical clustering of 740 genes differentially expressed in samples from

total normal human livers, HSCs isolated from cirrhotic livers and HSCs isolated from normal liver activated in culture. Each column represents an

independent sample. Red bars denote overexpression, black bars denote equally expression and green bars denote underexpression. Significant gene

ontology clusters are shown on the right column (P!0.001).

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282 277

including the lack of cell–cell contact, the need for ECM

interaction, and the progressive loss of vitamin A [28,29].

Moreover, it is conceivable that HSCs with more activated

phenotype, which probably contain less amounts of vitamin

A fat droplets, were underestimated in the procedure used to

obtain cells from cirrhotic livers. The gradient density was

not further augmented, since it would have facilitated the

presence of contaminating cells.

The activated phenotype of HSCs from cirrhotic human

livers is supported by gene expression studies. HSCs freshly

isolated from human cirrhotic livers express genes encoding

key molecules implicated in the wound healing process.

Interestingly, gene expression profile of HSCs from

cirrhotic livers and culture-activated HSCs was not

identical. An extensive group of genes implicated in

inflammation and apoptosis regulation were preferentially

expressed in HSCs from cirrhotic livers. This finding

suggests that HSCs may play a more active pro-inflamma-

tory role in vivo than previously suggested by studies using

culture-activated HSCs. Secretion of inflammatory

mediators and expression of cell adhesion molecules is

low in culture-activated HSCs, and is strongly stimulated by

Table 2

Gene expression in hepatic stellate cells (HSCs) freshly isolated from cirrhotic human livers and HSCs activated after prolonged culture, as assessed

by microarray analysis

Accession

number

Gene HSCs from

cirrhotic livers

Culture-acti-

vated HSCs

Accession

number

Gene HSCs from

cirrhotic livers

Culture-acti-

vated HSCs

Extracellular matrix Vasoactive substances

NM_000088 collagen I, a1 245.9 985.3* NM_000961 prostaglandin I2

synthase

1.2 9.7*

AA788711 collagen I, a2 5.4 22.1* NM_001124 adrenomedullin 2.2 4.6*

NM_000138 fibrillin 1 5.1 16* NM_000722 VOCC 1.7 7.3*

NM_001856 collagen XVI, a1 2.4 8.2* NM_002560 purinergic recep-

tor P2X4

4.4 2.4

M33653 collagen XIII, a1 98.1 522.82 NM_000029 angiotensinogen K13.7 K29.6

NM_001845 collagen, IV, a1 9.5 9.7 D32201 adrenergic a1A

receptor

K2.6 K2.7

AF018081 collagen XVIII, a1 K2.4 K4.2 NM_001992 thrombin recep-

tor

4.1 7

NM_002291 laminin, b1 5.5 6.8 Integrins

NM_002293 laminin, g1 7 10.7 NM_007036 ESM1 53.6 1.7*

NM_003254 TIMP-1 7.2 6.9 BG532690 integrin, a4

CD49D

5 7.5

NM_003255 TIMP-2 4 3.4 AI093579 vitronectin

receptor, CD51

5 5.3

M85289 heparan sulfate

proteoglycan 2

14.8 4.1 NM_001795 cadherin 5, type 2 7 K64.2*

NM_002421 matrix metalloproteinase 1 4710 719* Cell cycle/apoptosis

NM_004530 matrix metalloproteinase 2 3 11.3 M73554 cyclin D1 1.2 5.2*

NM_000917 P4HA K1.5 2.4* AW134535 cyclin G2 3 3.8

L16895 lysyl oxidase 7 70.9* AF083208 AATF 4 2.9*

M92934 CTGF 12.1 13.9 BC000324 granulin 3.7 2.5

NM_006182 DDR2 3.6 12.1* NM_000321 retinoblastoma 1 4 2.8

Z48199 syndecan 1 K6.5 K23.4 AI721219 TNF receptor,

factor 3

8 5.4

NM_000638 vitronectin K21.6 K56.5 NM_005256 growth arrest-

specific 2

K29.2 K93.2*

Growth factors/fibrogenic cytokines Cytoskeleton

M60485 FGF receptor 1 4.2 7.51 NM_001613 a2SMA 6.9 16.8

NM_002006 FGF 2 3.7 17.5* Z24727 tropomyosin 1a 7.2 12.7*

NM_003246 thrombospondin 1 4.9 4.5 NM_003289 tropomyosin 2b 589.6 4080.3*

NM_000875 insulin-like GF 1 receptor 4.8 8.3* NM_014000 vinculin 3.7 7.7

X59065 FGF 1 105.8 611.5 AF141347 tubulin, alpha 3 11.2 12

U58111 VEGF C 3.2 7.8* AI922599 vimentin 5.7 6.4

NM_002607 PDGF-alpha 4.6 8.7 Others

NM_000118 endoglin 16.1 3.9* NM_006870 destrin 1.8 3.5

BC000125 TGF, b1 8.3 3.6 BF197655 caveolin 2 1.6 3.6*

NM_000930 plasminogen activator, tis-

sue

4.5 41.1 NM_001753 caveolin 1 7.9 21.3*

NM_002658 plasminogen activator,

urokinase

27.6 3.4 NM_000801 FK506 binding

protein 1A

4.6 2.9

NM_000602 plasminogen activator

inhibitor 1

3.2 4.2 NM_000714 benzodiazapine

receptor

4 6.7

AA485908 insulin receptor K2.4 K3.8 NM_001769 CD9 antigen

(p24)

4.8 6.4

Inflammation NM_005328 hyaluronan

synthase 2

4 4

M24915 CD44 antigen 11.4 10.4 NM_002444 moesin 5 4.3

NM_000611 CD59 antigen 3.9 4.6 NM_004390 cathepsin H 2 K5.6

NM_002184 interleukin 6 signal trans-

ducer

2.8 3.3 NM_004877 glia maturation

factor, g

5.5 K2.04*

D28586 CD58 antigen 9 4* NM_002309 leukemia inhibi-

tory factor

33.2 11.5*

NM_000609 chemokine (C-X-C motif)

12

1.3 K3.5 BE620457 neuropilin 1 5.7 2.1

NM_000873 intercellular adhesion

molecule 1

4.5 1* AI700518 nuclear factor I/B 1.2 K2.6

(continued on next page)

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282278

Table 2 (continued)

Accession

number

Gene HSCs from

cirrhotic livers

Culture-acti-

vated HSCs

Accession

number

Gene HSCs from

cirrhotic livers

Culture-acti-

vated HSCs

NM_002189 interleukin 15 receptor,

alpha

2.4 K4.6* NM_005045 reelin K1.2 K6.1

NM_000418 interleukin 4 receptor 1.2 K2.6* NM_000376 vitamin D

receptor

282 190.3

NM_000600 interleukin 6 (interferon,

beta 2)

44.9 27.26 L03203 peripheral myelin

protein 22

25.1 13.5

M28882 melanoma cell adhesion

molecule

23.1 3.8* NM_000454 superoxide dis-

mutase 1

K2.9 K2.4

AL574096 tissue factor inhibitor 2 25.8 13.6 NM_001752 catalase K5.2 K10

The table shows fold expression of representative genes in hepatic stellate cells (HSCs) vs total normal human livers. Results are the mean of three different

samples. *P!0.01 culture-activated HSCs vs HSCs from cirrhotic livers. DDR2, discoidin domain receptor family, member 2; CTGF, connective tissue

growth factor; TIMP-1, tissue inhibitor of metalloproteinase 1; MMP1, matrix metalloproteinase 1; FGF, fibroblast growth factor; VEGF, vascular endothelial

growth factor C; TGF, transforming growth factor; VOCC, voltage-dependent calcium channel; ESM1, endothelial cell-specific molecule 1; a2SMA, actin,

alpha 2, smooth muscle.

Fig. 4. 3H-thymidine incorporation of hepatic stellate cells (HSCs)

isolated from normal liver (,), HSCs isolated from cirrhotic livers

(G), and HSCs isolated from normal liver activated in prolonged

culture (&). Cells were cultured in 24-well plates and incubated for

24 h with buffer, platelet-derived growth factor (PDGF-BB) (20 ng/ml),

angiotensin II (Ang II) (10K7 M), and fetal bovine serum (FBS) (15%)

for 24 h and 3H-thymidine incorporation was assessed as described in

Section 2. Results are expressed as meanGSD (nZ6). *P!0.05 vs

buffer; #P!0.05 vs HSCs from cirrhotic liver.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282 279

cytokines [17,30]. It is conceivable that cytokines present in

cirrhotic human livers stimulate the expression of inflam-

matory mediators in HSCs. Moreover, it has been shown

that culture-activated HSCs are extremely resistant to

apoptosis [31]. The different expression of genes implicated

in apoptosis regulation among HSCs freshly isolated from

cirrhotic livers and culture-activated HSCs suggest that their

apoptotic properties may differ. Further studies are needed

to test this hypothesis. On the other hand, expression of

genes implicated in fibrogenesis and cell contractility was

higher in culture-activated HSCs than in HSCs from

cirrhotic livers. As demonstrated in functional studies,

culture-activated HSCs display more powerful contractile

properties than HSCs from cirrhotic livers. These data

indicate that the model of HSCs activation in prolonged

culture does not exactly reproduce the activated phenotype

found in chronic liver diseases. Several studies have

revealed that culture on plastic induces up-regulation of

mesenchymal and myofibroblast-related genes and down-

regulation of cell–cell and cell–matrix interacting genes in

many cell types [32,33]. Therefore, the different genomic

profile found in culture-activated HSCs could be considered

as an artifact due to prolonged culture. It would have been

interesting to compare HSCs isolated from cirrhotic livers

with HSCs freshly isolated from normal human livers.

Unfortunately, we could not obtain enough RNA from

quiescent cells to perform the microarray experiments. This

was due to the paucity of normal human livers available for

cell isolations as well as the RNA content in these cells.

A relevant finding of the current study is that HSCs

freshly isolated from cirrhotic human livers, but not from

normal livers, display mitogenic and contractile properties.

Although studies were performed in vitro, all experiments

were carried out within the first 24–48 h after cell isolation.

Therefore, it is likely that the results closely reproduce the

cell behavior in vivo. Our results confirm previous

observations in rats showing that HSCs isolated from

experimentally induced cirrhosis display myofibroblastic

properties [26,34]. Proliferation of HSCs in areas of tissue

repair is a common pathogenic step in liver fibrosis [2].

Following activation in culture, HSCs increase the

expression of cell membrane receptors and develop

intracellular pathways to respond to mitogens such as

PDGF-BB and Ang II [14,35,36]. These factors are locally

secreted at the areas of tissue repair and may induce the

accumulation of fibrogenic HSCs [37]. Here, we demon-

strate that HSCs from cirrhotic, but not normal, livers

respond to mitogens. The mitogenic response was more

accentuated in culture-activated HSCs, indicating that

prolonged culture strongly stimulates the acquisition of

mitogenic properties. The mitogenic response to Ang II in

HSCs from cirrhotic livers was similar to that induced by

PDGF-BB and FBS, suggesting that Ang II is a major

mediator in human liver fibrogenesis. Because human

activated HSCs de novo generate Ang II, an autacrine

mitogenic loop is likely [6]. Our genomic and functional

10 pA50 ms

o

o

oo

o

o

o

o

o

o

o

o

c

c

c

c

c

c

c

c

–80

–70

–60

–50

–40

–30

–20

–10

Vm (mV)

500 pA50 ms –60 mV

+80 mV

–100 –80 –60 –40 –20 0 20 40 60 80 100

–25

–20

–15

–10

–5

0

5

10

15

20

25

Vm (mV)

I (pA)

Inside-out [K+]out = [K+]in

Inside-out [K+]out > [K+]in Cell-attached

A

C

B

Fig. 5. Characterization of Ca2C-dependent KC channel (BKCa) in hepatic stellate cells (HSCs) freshly isolated from cirrhotic livers. (A) Single-

channel KC currents at different voltages in the inside-out configuration. Asymmetric KC solutions ([KC]outO[KC]in) were used. A voltage-

dependent KC currents were detected. Open (o) and closed (c) states of the channel are shown. Patch pipette voltage (Vm) in mV. (B) Representative

experiment showing whole-cell KC currents in an HSC after stimulation with depolarizing voltage pulses (K50 to C80 mV). Holding potential in the

whole-cell configuration was K60 mV (Vm). Experiments were done in physiological solution in the bath and high KC solution in the pipette. (C)

Current/voltage relationships in cell-attached and inside-out configurations. Experiments using the cell-attached configuration were performed with

140 mM [KC]out (pipette) (:, nZ6). Linear fitting from cell-attached experiments is shown as a dotted line. Experiments using the inside-out

configuration were performed with 140 mM [KC]outZ[KC]in (C, nZ7) or 140:4.3 mM [KC]outO[KC]in (B, nZ6). Linear and Goldman–Hodgkin–

Katz fittings from inside-out experiments are shown as continuous lines. Data from both cell-attached and inside-out configurations are consistent

with the presence of a high-conductance BKCa in HSCs from cirrhotic livers.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282280

data suggest that HSCs display contractile properties in

chronic liver diseases, thereby reinforcing their potential

role in the pathogenesis of portal hypertension. However, it

should be pointed out that the contractile properties found in

HSCs from cirrhotic livers were less pronounced than in

culture-activated HSCs. Whether the degree of cell

contraction shown by HSCs from cirrhotic livers is strong

enough to modulate sinusoidal resistance to blood flow is

unknown and deserves further investigation. It should be

noted that cell types other than HSCs regulate intrahepatic

vascular resistance in chronic liver diseases.

In conclusion, we demonstrate that HSCs isolated from

cirrhotic livers show characteristics of fibrogenic and

Table 3

Effect of angiotensin II (10K7 M) on [Ca2C]i and cell contraction in hepatic s

livers and HSCs from normal liver activated in culture

Responder cells (%)a

HSCs from normal human livers (nZ50) 0

HSCs from cirrhotic human livers (nZ102) 73

HSCs activated in culture (nZ89) 87

*P!0.001 vs HSCs freshly isolated from normal liver. **P!0.001 vs HSCs frea Responder cells are those showing at least a 2-fold increase in [Ca2C]i compb Responder cells only.c Contraction is defined as a reduction area of R8% (see Section 2).

contractile myofibroblasts. However, the genomic and

functional characteristics of HSCs from cirrhotic human

livers differ from those observed in culture-activated HSCs.

Therefore, the most-widely used in vitro model to

investigate liver fibrogenesis (i.e. HSCs activation in

culture) does not exactly reproduce the cell behavior in

chronic liver diseases. It is conceivable that some

phenotypic, genomic and functional characteristics

described in culture-activated HSCs are influenced by

prolonged culture on plastic. We propose that future studies

investigating the efficacy of antifibrotic agents in human

HSCs should include a subset of experiments using cells

freshly isolated from cirrhotic livers.

tellate cells (HSCs) freshly isolated from normal and cirrhotic human

Peak [Ca2C]i (nmol/l)b Cells with contraction (%)c

– –

323G67* 33

538G217*/** 52**

shly isolated from cirrhotic liver.

ared to resting values.

Intr

acel

lula

r ca

lciu

m(%

res

pect

to r

estin

g va

lues

)

0

100

200

300

400

500

600

1 90 180

270

360

HSCs freshly isolated from cirrhotic livers.

HSCs freshly isolated from normal livers.

HSCs isolated from normal livers andactivated in culture.

sec

Fig. 6. Intracellular calcium mobilization induced by angiotensin II

(Ang II) in hepatic stellate cells (HSCs) freshly isolated from cirrhotic

and/or normal livers and cultured-activated HSCs. Cells were loaded

with fura-2 and studied in an epifluorescence microscope. Pictures

were taken every 4 s. Figure shows normalized [Ca2C]i of a

representative experiment. Ang II (10K7 M) induced a marked increase

in [Ca2C]i in both HSCs from cirrhotic liver and cultured-activated,

but not HSCs from normal liver.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282 281

Acknowledgements

The authors thank Dr Pedro Jares from the Unitat de

Genomica, IDIBAPS, for microarray technical support, and

Sergio Lario for his help on real time PCR experiments.

Supported by grants from Ministerio de Ciencia y

Tecnologia, Direccion General de Investigacion:

SAF2002-03696, BFI2002-01202, and by a grant from

Instituto de Salud Carlos III (CO3/02). Pau Sancho-Bru has

a grant from Institut Investigacions Biomedicas August Pi i

Sunyer. Valeriya Khurdayan has a grant from the Instituto

Reina Sofia de Investigacion Nefrologica.

References

[1] Reeves HL, Friedman SL. Activation of hepatic stellate cells—a key

issue in liver fibrosis. Front Biosci 2002;7:d808–d826.

[2] Bataller R, Brenner DA. Hepatic stellate cells as a target for the

treatment of liver fibrosis. Semin Liver Dis 2001;21:437–451.

[3] Milani S, Herbst H, Schuppan D, Surrenti C, Riecken EO, Stein H.

Cellular localization of type I III and IV procollagen gene transcripts

in normal and fibrotic human liver. Am J Pathol 1990;137:59–70.

[4] Cassiman D, Libbrecht L, Desmet V, Denef C, Roskams T. Hepatic

stellate cell/myofibroblast subpopulations in fibrotic human and rat

livers. J Hepatol 2002;36:200–209.

[5] Kweon YO, Goodman ZD, Dienstag JL, Schiff ER, Brown NA,

Burchardt E, et al. Decreasing fibrogenesis: an immunohistochemical

study of paired liver biopsies following lamivudine therapy for

chronic hepatitis B. J Hepatol 2001;35:749–755.

[6] Bataller R, Sancho-Bru P, Gines P, Lora JM, Al Garawi A, Sole M,

et al. Activated human hepatic stellate cells express the renin–

angiotensin system and synthesize angiotensin II. Gastroenterology

2003;125:117–125.

[7] Pinzani M, Milani S, De Franco R, Grappone C, Caligiuri A,

Gentilini A, et al. Endothelin 1 is overexpressed in human cirrhotic

liver and exerts multiple effects on activated hepatic stellate cells.

Gastroenterology 1996;110:534–548.

[8] Friedman SL, Rockey DC, McGuire RF, Maher JJ, Boyles JK,

Yamasaki G. Isolated hepatic lipocytes and Kupffer cells from normal

human liver: morphological and functional characteristics in primary

culture. Hepatology 1992;15:234–243.

[9] Sato M, Suzuki S, Senoo H. Hepatic stellate cells: unique

characteristics in cell biology and phenotype. Cell Struct Funct

2003;28:105–112.

[10] Sohara N, Znoyko I, Levy MT, Trojanowska M, Reuben A. Reversal

of activation of human myofibroblast-like cells by culture on a

basement membrane-like substrate. J Hepatol 2002;37:214–221.

[11] Pinzani M. Platelet-derived growth factor receptor expression in

hepatic stellate cells: how too much of a good thing can be bad.

Hepatology 1995;22:997–999.

[12] Marra F, Grandaliano G, Valente AJ, Abboud HE. Thrombin

stimulates proliferation of liver fat-storing cells and expression of

monocyte chemotactic protein-1: potential role in liver injury.

Hepatology 1995;22:780–787.

[13] Schwabe RF, Bataller R, Brenner DA. Human hepatic stellate cells

express CCR5 and RANTES to induce proliferation and migration.

Am J Physiol Gastrointest Liver Physiol 2003;285:G949–G958.

[14] Bataller R, Gines P, Nicolas JM, Gorbig MN, Garcia-Ramallo E,

Gasull X, et al. Angiotensin II induces contraction and proliferation of

human hepatic stellate cells. Gastroenterology 2000;118:1149–1156.

[15] Benyon RC, Arthur MJ. Extracellular matrix degradation and the role

of hepatic stellate cells. Semin Liver Dis 2001;21:373–384.

[16] Hellerbrand C, Wang SC, Tsukamoto H, Brenner DA, Rippe RA.

Expression of intracellular adhesion molecule 1 by activated hepatic

stellate cells. Hepatology 1996;24:670–676.

[17] Schwabe RF, Schnabl B, Kweon YO, Brenner DA. CD40 activates

NF-kappa B and c-Jun N-terminal kinase and enhances chemokine

secretion on activated human hepatic stellate cells. J Immunol 2001;

166:6812–6819.

[18] Marra F. Hepatic stellate cells and the regulation of liver

inflammation. J Hepatol 1999;31:1120–1130.

[19] Vinas O, Bataller R, Sancho-Bru P, Gines P, Berenguer C, Enrich C,

et al. Human hepatic stellate cells show features of antigen-presenting

cells and stimulate lymphocyte proliferation. Hepatology 2003;38:

919–929.

[20] Bataller R, Nicolas JM, Ginees P, Gorbig MN, Garcia-Ramallo E,

Lario S, et al. Contraction of human hepatic stellate cells activated in

culture: a role for voltage-operated calcium channels. J Hepatol 1998;

29:398–408.

[21] Friedman SL. Molecular regulation of hepatic fibrosis, an integrated

cellular response to tissue injury. J Biol Chem 2000;275:2247–2250.

[22] Bataller R, Nicolas JM, Gines P, Esteve A, Nieves GM, Garcia-

Ramallo E, et al. Arginine vasopressin induces contraction and

stimulates growth of cultured human hepatic stellate cells. Gastro-

enterology 1997;113:615–624.

[23] Li C, Wong WH. Model-based analysis of oligonucleotide arrays:

expression index computation and outlier detection. Proc Natl Acad

Sci USA 2001;98:31–36.

[24] Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM,

et al. Gene ontology: tool for the unification of biology. The gene

ontology consortium. Nat Genet 2000;25:25–29.

[25] Gasull X, Bataller R, Gines P, Sancho-Bru P, Nicolas JM, Gorbig MN,

et al. Human myofibroblastic hepatic stellate cells express Ca(2C)-

activated K(C) channels that modulate the effects of endothelin-1 and

nitric oxide. J Hepatol 2001;35:739–748.

[26] Bataller R, Gasull X, Gines P, Hellemans K, Gorbig MN, Nicolas JM,

et al. In vitro and in vivo activation of rat hepatic stellate cells results

in de novo expression of L-type voltage-operated calcium channels.

Hepatology 2001;33:956–962.

[27] Rockey DC, Boyles JK, Gabbiani G, Friedman SL. Rat hepatic

lipocytes express smooth muscle actin upon activation in vivo and in

culture. J Submicrosc Cytol Pathol 1992;24:193–203.

[28] Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a

modulator of hepatic fibrogenesis. Semin Liver Dis 2001;21:

351–372.

[29] Senoo H, Wake K. Suppression of experimental hepatic fibrosis by

administration of vitamin A. Lab Invest 1985;52:182–194.

P. Sancho-Bru et al. / Journal of Hepatology 43 (2005) 272–282282

[30] Bataller R, Schwabe RF, Choi YH, Yang L, Paik YH, Lindquist J, et al.

NADPH oxidase signal transduces angiotensin II in hepatic

stellate cells and is critical in hepatic fibrosis. J Clin Invest 2003;112:

1383–1394.

[31] Zamara E, Bonacchi A, Novo E, Gentilini E, Tamagno E,

Dianzani MU, et al. Activated human hepatic stellate cells

are resistant to classic pro-apoptotic stimuli. J Hepatol 2003;38:21.

[32] Lim YS, Kim KA, Jung JO, Yoon JH, Suh KS, Kim CY, et al.

Modulation of cytokeratin expression during in vitro cultivation of

human hepatic stellate cells: evidence of transdifferentiation from

epithelial to mesenchymal phenotype. Histochem Cell Biol 2002;118:

127–136.

[33] Cassiman D, Roskams T. Beauty is in the eye of the beholder:

emerging concepts and pitfalls in hepatic stellate cell research.

J Hepatol 2002;37:527–535.

[34] Rockey DC, Housset CN, Friedman SL. Activation-dependent

contractility of rat hepatic lipocytes in culture and in vivo. J Clin

Invest 1993;92:1795–1804.

[35] Wong L, Yamasaki G, Johnson RJ, Friedman SL. Induction of beta-

platelet-derived growth factor receptor in rat hepatic lipocytes during

cellular activation in vivo and in culture. J Clin Invest 1994;94:

1563–1569.

[36] Mallat A, Gallois C, Tao J, Habib A, Maclouf J, Mavier P, et al.

Platelet-derived growth factor-BB and thrombin generate positive and

negative signals for human hepatic stellate cell proliferation. Role of a

prostaglandin/cyclic AMP pathway and cross-talk with endothelin

receptors. J Biol Chem 1998;273:27300–27305.

[37] Ikeda K, Wakahara T, Wang YQ, Kadoya H, Kawada N, Kaneda K. In

vitro migratory potential of rat quiescent hepatic stellate cells and its

augmentation by cell activation. Hepatology 1999;29:1760–1767.