Fat storing cells as liver specific pericytes Spatial dynamics of agonist stimulated intracellular calcium transients

Fat-storing cells as liver-specific pericytes.

Spatial dynamics of agonist-stimulated

intracellular calcium transients.

M Pinzani, … , A Giotti, P Gentilini

J Clin Invest.

1992;

90(2)

:642-646.

https://doi.org/10.1172/JCI115905

.

Liver perisinusoidal fat-storing cells (FSC) show morphological and ultrastructural

characteristics similar to pericytes regulating local blood flow in other organs. In the present

study we have analyzed whether FSC respond to local vasoconstrictors such as thrombin,

angiotensin-II, and endothelin-1 with an increase in intracellular free calcium concentration

([Ca2+]i) coupled with effective cell contraction. All agonists tested induced a rapid and

dose-dependent increase in [Ca2+]i followed by a sustained phase lasting several minutes

in confluent monolayers of Fura-2-loaded human FSC. Pharmacological studies performed

using different Ca2+ channel blockers indicated that, at least for thrombin and

angiotensin-II, the sustained phase is due to the opening of voltage-sensitive membrane Ca2+ channels.

To analyze the temporal and spatial dynamics of Ca2+ release in response to these

agonists, we performed experiments on individual Fura-2-loaded human FSC using a dual

wavelength, radiometric video imaging system. The rise in [Ca2+]i was exclusively

localized to the cytoplasm, particularly in the branching processes. Increases in [Ca2+]i

more than four-fold were associated with a simultaneous and transient reduction of cell area

indicating reversible cell contraction. Our results indicate that the Ca(2+)-dependent

contraction of human FSC in vitro may reflect a potential role in regulating sinusoidal blood

flow in vivo.

Research Article

Find the latest version:

Rapid

Publication

Fat-storing

Cells

as

Liver-specific Pericytes

Spatial Dynamics

of

Agonist-stimulated

Intracellular Calcium Transients

Massimo Pinzani, Paola Failli,*Carlo Ruocco,* Alessandro Casini,t Stefano

Milani,*

Elisabetta

Baldi,O

Alberto Giotti,* and Paolo Gentilini

Istitutodi ClinicaMedicaII;*Centro InteruniversitarioIpossie;tUnitddi Gastroenterologia; and Unitd diEndocrinologia-Dipartimento

diFisiopatologia Clinica, Universita di Firenze, I-50134 Firenze, Italy

Abstract

Liver perisinusoidalfat-storingcells(FSC)show morphologi-cal andultrastructural characteristics similartopericytes regu-lating local blood flowinotherorgans.Inthepresentstudywe

haveanalyzedwhether FSCrespondtolocal vasoconstrictors suchasthrombin,angiotensin-II,and endothelin-1 withan in-creaseinintracellular free calcium concentration

(ICa2I1I)

cou-pled with effective cell contraction. Allagoniststested induced

arapidand dose-dependent increase in

ICa2`i1

followed bya

sustained phase lasting several minutes in confluent

mono-layers ofFura-2-loaded human FSC. Pharmacologicalstudies performed using differentCa2+channel blockers indicated that,

atleast for thrombin andangiotensin-II, the sustained phase is duetotheopeningofvoltage-sensitive membrane

Ca2"

chan-nels. To analyze the temporal and spatial dynamicsof

Ca2"

release in response to these agonists, we performed

experi-ments onindividual Fura-2-loaded human FSC using a dual wavelength, ratiometric video imaging system. The rise in

ICa2"],

was

exclusively

localizedtothe

_{cytoplasm, particularly}

inthe branchingprocesses.Increases in

ICa2"1,

morethan four-foldwereassociated withasimultaneous andtransient

reduc-tion of cellareaindicating reversible cell contraction. Our

re-sults indicate that the

Ca2"-dependent

contraction of human FSC in vitromayreflectapotentialrole inregulatingsinusoidal

blood flow in vivo. (J. Clin. Invest. 1992. 90:642-646.) Key words:fat-storingcells*intracellular free calcium*cell

contrac-tion*thrombin*angiotensin-II *endothelin-1

Introduction

Liverfat-storingcells(FSC1;alsoknownasperisinusoidal

stel-latecells, lipocytes,Itocells) have recently beenshowntoplay importantroles in retinolmetabolism and hepatic fibrogenesis

Portions of this workwerepresented in abstract formattheannual meetingof theAmerican Association for the Study of Liver Diseases, Chicago, IL,2-5November 1991.

AddresscorrespondencetoMassimoPinzani,M.D., Ph.D.,Istituto

di Clinica Medica II, Universita di Firenze, Viale Morgagni, 85,

1-50134Firenze, Italy.

Receivedfor publication24March 1992 andinrevisedform4 May

1992.

(forreview see 1, 2). A possiblerole of FSC as liver-specific pericyteshas also beenhypothesized (1-3). Indeed,FSC show morphological and ultrastructural characteristics similar to

pericytes regulatingblood flow in other organs. These include: perisinusoidal and interhepatocellular branching processes containing massive 5-nm actin-like filaments and encircling neighboringsinusoids (4), a contact surface between stellate cells and nerveendings (5),and theexpressionofthe a-smooth muscle actin gene(6).

The presentstudywasundertaken toverifywhetherFSC, like otherperivascularcontractilecells, respondtolocal

vaso-constrictors such asthrombin,endothelin-1, and angiotensin-II.For this purpose, we have analyzed the variations of intra-cellular free calcium concentration

_([Ca2]ji)

inmonolayersand inindividualagonist-stimulatedFSC isolated from normal hu-man liver. Inaddition,wehave studied the reversiblechanges ofcell areacoupled withagonist-stimulated intracellular cal-cium transientsby usingadynamicvideoimaging techniqueof cellular fluorescence.

Methods

Isolation, culture, and characterization of human FSC. HumanFSC wereisolated from wedge sections of normal humanliver unsuitable for transplantation.After extensivewashingsinsaline,liver tissuewas

finelymincedusinga razorblade, placed in asterile flaskcontaining

0.5% pronase(103 proteolyticU/mg; Calbiochem Corp., SanDiego, CA),0.05% typeIVcollagenase(SigmaChemical Co., St.Louis, MO),

and 10

_Ag/ml

of DNAse(bovinepancreas;Calbiochem)in 100 mlof

HBSSwithout calcium and magnesium, and agitated at370Cfor 30 min. Theresultingcellsuspensionwasfilteredthrough a105-Mmnylon gauze. Theundigestedtissue retained into the gauze was further di-gestedusing0.05% pronase,filtered,and pooledwith theremainder of the cellsuspension.Thecombineddigestwaswashedfourtimesat450 g for 10 min in HBSScontaining 10Mg/ml ofDNAse andthe final

pelletwasfinally resuspendedin 25 mlofthe samesolution.FSC were separated fromotherliver nonparenchymalcellsbyultracentrifugation over gradients of stractan (Larex-LO; Larex International Co., Ta-coma,WA)asdescribed elsewhere (7, 8). FSC were recovered from the interface between the 1.053 stractan gradient and the medium. Cells recoveredatthis level (1.4X 106cells/g of tissue) werehighlyviable and

- 90% pure.

Cells were cultured in Iscove's modified Dulbecco's medium

(Gibco Laboratories,Grand Island, NY) supplemented with 0.6 U/ml

insulin,2mM glutamine,0.1 mMnonessential amino acids, 1 mM

1.Abbreviations used in thispaper: AII, angiotensin-II;

_[Ca2"],,

extra-cellularcalcium;

[Ca2+]i,

intracellular free calciumconcentration; ET-1,endothelin-l;FSC,fat-storingcells; KHH,Krebs-Henseleit-Hepes; SFIF,serum-freeinsulin-free;THR, thrombin.

642 M.Pinzani,P.Failli, C.Ruocco,A.Casini, S. Milani, E.Baldi,A.Giotti, and P. Gentilini

J.Clin. Invest.

© The AmericanSocietyfor Clinical Investigation, Inc.

0021-9738/92/08/0642/05 $2.00

sodium pyruvate,antibioticantifungal solution,and 20%fetalbovine serum. Experimentsdescribed in thisstudywereperformedoncells betweenfirstand second passageusingtwoindividual celllines. Hu-manFSC inprimary culture wereidentifiedbyimmunostaining for intermediatefilaments, surface antigens, and by transmission electron microscopy. Cells platedontosterile tissue culture chambers (Lab-tek Div., MilesLaboratories Inc.,Naperville,IL) were washedtwice with PBS, driedovernight at room temperature, and fixed in acetone at 4VC for 5 min. Monoclonalantibodies specific forvimentin (V9; Dako-patts, Glostrup,Denmark),a-smooth muscleactin (1A4; BioGenex,

San Ramon,CA)humandesmin(D-33;Dakopatts), pan-cytokeratin

(Lu5; Boehringer MannheimGmbH,Mannheim, Germany), Ki-M IP,

directedagainst themonocyte/macrophageantigenCD68 (9), and poly-clonalantibodiesagainst factorVIII-relatedantigen(Dakopatts), and

porcinedesmin(SigmaChemicalCo.)wereappliedontocells and de-tected with the alkaline anti-phosphatase anti-alkaline phosphatase (APAAP)method(10). Cells demonstrated intense staining for

vimen-tin,andslightly fora-smooth muscleactinandporcinedesmin. The negativestainingfor CD-68, factorVIII-relatedantigens,and

cytokera-tin,demonstratedtheabsence ofcontaminatingmono/macrophagic,

endothelial,andepithelial cells, respectively. Transmissionelectron mi-croscopystudies, performedasdescribedelsewhere (1 1), revealed the presenceofnumerouslargelipiddroplets in the cytoplasmassociated

with abundant bundlesofmyofilaments.

Fluorimetric analysisof intracellular freecalcium inhumanFSC

monolayers. HumanFSCweregrowntoconfluenceon 14X 14mm

plastic (Aclar,Pottsville, PA) cover slips in complete culture medium, and thenincubated inserum-free insulin-free (SFIF)mediumfor24h. Loading of the cells with the fluorescentCa2" indicator Fura-2-AM

(Calbiochem)wasachievedbyincubatingthecellsfor 45 min at370C

with 1MMFura-2-AM. The loading medium was then replaced with 1 mloffreshSFIFmedium and the cells were incubated for additional 20 minat370C,followedbyrinsing with ice-cold Krebs-Henseleit-Hepes

(KHH)containing(inmM):Na+140.7, K+5.3,Cl- 132.4,Ca2+ 1.0,

Mg2'0.81,glucose 5.5, Hepes20.3, with0.1%fatty acid-free bovine

serumalbumin, pH7.4. Thecoverslipswereplaceddiagonallyina

square quartzcuvette sothattheexcitationandemission pathswere at a450 angletothecoverslip.The cuvette, containing2 mlofKHH

buffer, was maintained at 37°C. Thrombin (THR; from human plasma, Boehringer Mannheim) endothelin-l

(ET-l;

Novabiochem AG,Llufelfingen,Switzerland) orangiotensin-II (All;Novabiochem) weredirectly added to the cuvette, andfluorescencewascontinuously recorded, under constantstirring,byaJohnsonFoundation Biomedi-calInstrumentationGroup fluorometer(Philadelphia,PA) using a sin-gle-wavelengthexcitation(340nm)/emission(500 nm). In some

exper-iments,cell monolayers were pretreated with 3mMEGTA(Sigma)or with twoCa2+channelblockers,namelybepridil andnifedipine(both

purchased from Calbiochem). Calibration was performed for each

coverslipasindicatedelsewhere(12).Cellautofluorescencewas evalu-atedinparallelcoverslipsbothbymeasuring fluorescenceof unloaded cells andbyquenchingFura-2fluorescencewith 1 mMMnCI2 afterthe addition ofionomycin.In bothcasescellautofluorescencewas

negligi-ble and, inaddition,itwas notmodified by addition oftheagonists,as

measured inmonolayersnotloadedwith Fura-2.

Digital video imagingof intracellularfreecalcium in individual

FSC. For theseexperimentshuman FSCwereseededatlowdensity

and grown incompleteculture mediumonroundglasscoverslips (25

mmdiameter, 0.2mmthick) for72h,and thenincubated for 24hin SFIFmedium. Cellswereloadedwith 10,MFura-2-AM in

Hepes-NaHCO3 buffer containing 140 mM NaCl, 3 mM KCI, 0.5 mM NaH2PO4, 12mMNaHCO3, 1.2mM MgCI2, 10mMHepes, 10 mM glucose, with 0.1%fattyacid-free bovineserumalbumine, pH 7.4,at

37°Cfor 45 min(13).Afterloading,thecoverslipswerewashed and storedat roomtemperature untilused(always<45min afterloading).

Thecoverslipswerethenplaced inaperfusionchamber mountedon

the stage ofaDiaphot-TMDepifluorescenceinvertedmicroscope

(Ni-konCo., Tokyo, Japan) equippedwitha xenon lamp. Fura-2-loaded cellswere visualized with a Nikon CF x100 oil immersion orx40

objectives (Nikon Co.).Afilter cassette with adichroicmirror(DM 400)and abarrier filter(BA 510)wereused. Excitationwavelengths

werealternated between 340 and 380 nm with an automated filter

changer.Neutral density filters were used to diminish Fura-2

photob-leaching.Videoimageswereobtained with an extended ISIS-M camera

(Photonic Science,RobertsBridge,EastSussex, UK)andthe resultant

analogicvideo signalfrom the camera was digitalized with an 8-bit analogue-to-digital converter. Images were collected every 5 s for a standard time of 5-6 min.Agonistswereaddeddirectlyto the

perfu-sion chamber immediately after recording the [Ca2]ij basal value.

[Ca2+]i

wascalculatedusing "Tardis" software(Joyce Loebl,

Newcas-tle, UK)after the creation ofratioimages (340:380)by dividingoriginal images from whichthebackground (cell removed) had been subtracted onapixel-to-pixel basis. The calibration was performed according to Cheung et al. (14), calculating A,

Rn,,,,,

andR,.,foreach preparation andusingaKdfor Fura-2 of 224accordingtoGrynkiewicset al.(15).

Tomeasure cell area, spatial calibration was performed by measuring division on a graticule under the same optical conditions as the rest of theexperiments.

Results and Discussion

Fig. 1 illustrates the changes in

[Ca2+]i

induced by exposure of human FSC monolayers to THR, All, and ET- I in a series of representative experiments. For all agonists tested, the onset of

responses was virtually immediate and peak increments of

[Ca2+]i,

over a resting level of 90-100 nM, were transiently reached within 15-20 s,decliningrapidly to a sustained phase which was maintained for several minutes. When otherwise identical experiments were performed in virtual absence of

ex-tracellular calcium

([Ca2-`)

the peak height of the calcium transient induced by all agonistswas reduced onaverageby 5-10% and the sustained phase was almost completely abro-gated. These observations indicate that the peak effect induced by these vasoconstrictors is mainly due to intracellular release ofCa2+ from cytosolic stores, whereas the sustained phase de-pendsonstimulated influx. In the presenceof1.0mM

_[Ca2k]e

the sustained phase induced by exposuretoTHR and Allwas

abolished by pretreating cell monolayers with two different membrane Ca2+ channel blockers, namely bepridil (a pheneth-ylamine type of

Ca2'

channelblocker)and nifedipine (a 1 ,4-di-hydropyridine type of Ca2+ channel blocker), suggesting that

thestimulated entryofextracellular calciumresponsiblefor the sustained phase is likely duetotheopeningof

voltage-depen-dent transmembrane Ca2+ channels. Indeed, the two Ca2+ channel blockers used,belongingtoseries with different

molec-ularstructure,have been shown tospecificallyblock thesame

putativevoltage-sensitive

Ca21

channels (16).Conversely,

pre-treatment with nifedipine or bepridil did not affect the sus-tainedphaseinducedby exposuretoET-l in thepresence of 1.0mM

_[Ca2J]i

failingtosupportarole for voltage-dependent Ca2+channelsor,atleast,forthosephenethylamine-and 1.4-dihydropyridine-sensitive.

InFura-2-loaded human FSCmonolayersthe

[Ca2+]i

in-creaseinducedby THR,All,andET-I wasdose-dependentas

showninFig.2.

The results of this first set ofexperiments indicated that

human FSC respond tolocal vasoconstrictoragonists with a

rapidincrease in

[Ca2+]i

asotherperivascularcontractilecells.

Inaddition,themorphologyand thedynamicsof

[Ca2+]i,

and

1 min

j893

~~~~~771

701_-674A

I1

a

0L 1011w. 97L~ 100o o"m

t

Thrombin

5.0 U/ml

Lca2+].

1.0mM

t

Thrombin

5.0U/ml

LCa2+].

0.0mM

1 min 800

+ff 400

0c

200-OL

_t

Al

0.1 pM

t

t _I

EGTA Thrombin

3 mM 5.0 Ut/mA

[Ca2+].

0.0_mM

Bepridil Thrombin 1dM 5.0 U/ml

Eca

].

1.0

mM

tt ,t

EGTA All Nifedipine Aff

3 mM 0.1 _JpM 10 pM 0.1 PM

[Ca2+]e

1.0_mM [Co2

+]j

00_mM

[Eco"+]

_1.0mM

1min

Lj

800 600

400 200

ET-1 0.1 pM

LCU2+]e 1.0_mM

t

,t

t

EGTA ET-1 Nifedipine ET-1 3 mM 0.1 jpM 10pM 0.1 pM

ca2+]e 0.0mM Lca ]. 1.0mM

tt

Bepridil ET-1

10i6M

0.1 JjM

LCa2+e 1.0mM

Figure1.Changesincytosolicfree calcium concentration([Ca2+]i) in-ducedbyexposureofFura-2-loaded human FSCmonolayerstodifferent vasoconstrictoragonists.Human FSCmonolayerswere loadedwith Fura-2-AM as described inMethods.

(A)Addition of thrombin at a final concentration of 5.0 NIHunits/ml

induced arapid[Ca(j] peak incre-mentfollowedbyasustainedphase lastingseveral minutes. Thisplateau

wasvirtually abolished by repeating

theexperimentsinthe absence of extracellularcalcium

([Ca2"],)

with orwithoutpretreatmentwith 3.0 mMEGTA.Similarly,the sustained phase was abolishedbypretreating

the cells with 10-6Mbepridil,a

phenethylamine typeof calcium channelblocker, in thepresenceof 1.0 nM

_[Ca2'J%.

Similar results were obtainedpretreatingthe cells with

10MM nifedipine,a

dihydropyri-dine-sensitive

Ca2"

channel blocker

(not shown).(B)Similar[Ca2+]i in-creases wereinducedbythe addition ofangiotensinII at a final concen-tration of 0.1MM. Analogously, the

sustainedphase wasabolished either

by performingtheexperimentin thevirtualabsenceof

_[Ca2"],

orby

pretreatingthe cells with 10 M

ni-fedipine. (C) Additionof0.1 M endothelin-I inducedanalogous [Ca2]iincreases withabrogationof thesustained phasein the absence of

_[Ca2`],.

However,bothnifedipine

andbepridilwereineffectivein

abolishingthesustainedphase.

Sinceanelevation of

[Ca2J]i

insmooth musclecells results

in activation of contractile proteins (22) and FSC have been hypothesizedtoplayarole in the local regulation of sinusoidal

hemodynamics,wehave also analyzed cell contractility in

re-sponse to thesameagonists able toincrease

[Ca2+]i.

Forthis

purpose we employed a dual wavelength, ratiometric video

imagingsystemthat allowstostudy variations of

[Ca2+]i

along with changesofcellareain individual Fura-2-loadedcells.

Sim-ilarlytowhatwasobservedin the experiments performed using

cellmonolayers,exposureof individual FSCtoTHR,AII,and

ET-1resulted inarapid and transient increase of

[Ca2J]i

over a

resting level of 140-150nM.Whenmorethanonecellperfield wasobserved, itwasevidentthat the number of cells

respond-ing and the extentofresponsewerevariable,ranging from 0

(nonresponders)to 10-fold increase in

[Ca2J]i.

Inaddition,

ago-nist-inducedCa2+ transients did notbegin simultaneouslyin

differentcells butvariedupto60s,indicating different

activa-tiontimes.Video-imaging analysis revealed thattheincrease of

[Ca2+]i

begins in discrete areas located atthe cell periphery, particularly in the branching processes, with subsequent

spreadingtothe cellcytoplasm, often withatypical"calcium wave"pattern.Amongthe three agonists tested, THRwasthe

mosteffective intermsof the number of responding cellsand

extent of

[Ca2"Ji

increase. Indeed, a clear response was

ob-served in 20outofthe 24 cellcells analyzed, withanaverage

delta increase of- 600 nM.Fig. 3 shows four selected

time-se-quence frames from a representative experiment performed

using 0.3 NIH units of THR. By analyzing the frame-to-frame variations of cell area versusthe changesof

[Ca2+]i

(Fig. 4)it

wasevident that the calcium increase induced by THRwas

coupled witha simultaneous and transient reduction of cell

area,indicating reversible cell contraction. Similarly, reversible

changes of cellareacoupled with

[Ca2+]i

increases (more than fourfold)wereobserved when FSCwerestimulated with both

ET-1andAII, although the number of responding cells and the

extentofresponse werelower than thoseinduced by THR.

Insummary,the results of thisstudy demonstratethatin

humanFSCvasoconstrictoragonistsinduce intracellular

cal-cium transientscoupled with cell contraction analogouslyto

whatwasobserved in other bettercharacterized pericytes.

Al-644 M. Pinzani,P.Failli, C.Ruocco,A. Casini, S. Milani,E.Baldi,A.Giotti,andP. Gentilini

L~1

c

.)

x

c

._

0o

0.1 1.0 5.0 10.0

IThrombinl U/mi

700

600

500 400

300F

200 100

-9 -8 -7 -6 tog10

[Angiotenin-ull

M

700

X 600

c

500

v.-

400-a 300

200

t9oo o3~

0

-12 -11 -10 -9 -8 -7 -6

log10

Endothelin-11

M

Figure2.Dose-responsecurvesfor the effect ofdifferent vasoconstrictor agonistson[Ca,]ipeakincrease inmonolayersof Fura-2-loadedhuman FSC. Human FSC monolayerswereloaded with Fura-2asdescribedinMethods. (A)Thrombin.(B) AngiotensinII.(C)Endothelin-1. Dataare

means±SD for three individual determinations.

though the in vivo biologic relevance of these observations

re-mains speculative, the results of this in vitro study strongly

supportthehypothesis that FSC, beyond their morphological

appearance, mayfunction as perisinusoidal contractile cells.

Thepresenceofafunctionally active contractileapparatus

en-circling the sinusoids suggests that these vascular structures

might constituteamajor regulatory site of intrahepatic- blood

flow. Theresponsesinduced inFSC by endothelium-derived

Figure3.Time-sequence changesin intracellularcalcium concentrationina_singleFura-2-loaded human FSCrespondingto 0.3 NIHunits/ml of

thrombin(representative experiment).Fura-2loadingandexperimental proceduresaredescribedinMethods. Frame numberandtime(s) after

the additionof theagonistareshown in the left lowercornerof eachframe. Frame 3(0.0 s)showsrestinglevels of[Ca+2]i.Frame 25(61.4 s)shows

that increase in

[Ca+2]i

(indicated bythe shift fromgreen-bluetoorange-red)starts at the cellperipheryand_{rapidly progresses}throughthe

cyto-plasm (frame 33,79.4s)as a"calcium wave." Frame 43(101.9 s)showstheprogressivereturnto basalvalues. Note the reductionof cellarea

associated with the increasein[Ca2]i.

a

L2J

Iq

800 700

600

500 400 300 200

i 1400 ---- C

C

1000 20008

8600 1500.

200---0

---'---

1000

I Time

s.L=5s

Figure4.Increaseinintracellular calcium concentration

[Ca2"i

is

coupledwithreversiblecellcontraction inanindividual

Fura-2-loaded human FSC. Data from the same representative experiment

shown inFig.3. 0.3 NIHunits/ml thrombinwereaddedatthetime

pointindicatedby the black arrow.Ratioframeswerecollectedevery 5 s.Closed circles,

[Ca2i]

nM; opencircles, cell area RM2.

mediators, suchasET-1 and AII, raisethe possibility of alocal regulation of sinusoidal resistanceoperated byFSC-sinusoidal endothelium interactions.

It should be noted, however, that early passaged human FSC, as those employed in the present study,arecharacterized by an "activated" phenotype resembling "transitional" or "myofibroblast-like" cells rather than quiescentFSCretaining the original "storing" phenotype.Thisphenotypical transition, normally observed in cultures onplastic or glass (23) and in vivo during active fibrogenesis(24), is characterized by a pro-gressively more intense staining for a-smooth muscle actin (25). Myofibroblast-like cells with prominenta-smoothmuscle actin filaments have been described in fibroussepta, around sinusoids, and terminal hepatic venules of cirrhoticlivers(26). Inthis clinical condition they are believedto be responsible for the contraction of maturing scar tissueand to contribute, by maintaining a contractile state, to the increased resistance to portal flow (27). Although the relevance of our findingstothe situation in normal liver remains to be established, they are likely to be more representative of FSC contractile statusin

fibrotic

liver. The potent effect of THR describedin thisstudy provides an example of the effect of vasoconstricting agents

possibly

involved in this process. In conclusion,our observa-tions open new perspectives in the interpretation of mecha-nisms regulating intrahepatic blood flow andmaycontributeto thedevelopment of

pharmacological

strategies able to affect intrasinusoidal blood pressure.

Acknowledgments

Theauthors wish to thank Chiara Sali and Renata Salzano for excellent technicalassistance, and Fabio Marra, M.D., for his expert advice.

This work was supported by grants from ConsiglioNazionaledelle Ricerche (Rome, Italy), and Ministero Italiano

dell'Universiti

edelta

Ricerca Scientifica eTecnologica-ProgettoNazionale Cirrosi Epatica (Rome, Italy). Financial support was also provided by Fondazione Ita-liana per lo Studio del Fegato (Italian Liver Foundation, Florence, Italy).

References

1.Wake, K. 1980. Perisinusoidal stellate cells (fat-storing cells, interstitial cells, lipocytes), their related structure in and around the liver sinusoids, and vitamin A-storing cells in extrahepatic organs.Int.Rev. Cytol. 66:303-353.

2.Blomhoff,R., and K. Wake. 1991.Perisinusoidalstellate cells of theliver importantroles in retinol metabolism and fibrosis. FASEB (Fed. Am.Soc.Exp.

Biol.)J.5:271-277.

3. Ramadori,G. 1991. The stellate cell(Ito-cell, fat-storingcell,lipocyte,

perisinusoidalcell) ofthe liver.Virchows Archiv.BCell. Pathol. 61:147-158.

4. Wake, K. 1988. Liver perivascular cells revealed by gold and silver impreg-nation methods and electron microscopy. InBiopathology oftheLiver,an

Un-trastructural Approach. P. Motta,editor. KluwerAcademicPublishers, Dor-drecht,Netherlands.23-26.

5.Lafon,M. E., P. Bioulac-Sage, and N.LeBail.1989. Nerves and

perisinusoi-dal cells in human liver. In Cells of theHepatic Sinusoid.E.Wisse,D. L.Knook,

and K. Decker,editors.KuppferCellFoundation,Riswijk,Netherlands. 20:230-234.

6.Ramadori,G., T. H.Veit,S.Shwogler, H. P.Dienes,T.Knittel, H. Rieder, and K. H. Meyer zumBushenfelde. 1990.Expression ofthe geneof a-smooth

muscle actinisoform in ratliver and in ratfat-storing (ITO)cells. Virchows Archiv.BCell.Pathol. 59:349-357.

7. Friedman, S. L., and F. J. Roll. 1987. Isolationand culture ofhepatic

lipocytes,Kuppfercells, andsinusoidal endothelialcells bydensity gradient

cen-trifugationwith stractan.Anal. Biochem. 161:207-218.

8. Pinzani,M., H. E.Abboud,L.Gesualdo,and S. L. Abboud. 1992. Regula-tionof macrophage-colony stimulatingfactor in liverfat-storingcells by peptide

growth factors.Am.J.Physiol. (Cell. Physiol.). 262:C876-C881.

9. Wacker, H. H., M. L. Hansmann, H.Lumbeck,H. J.Radzun, and M. R. Parwaresch. 1990. A newpan-macrophages antibodyKi-M lPstains plasmacy-toidcells inparaffin sectionsonlymphnodes.Verh. Dtsch.Ges.Pathol.

74:159-164.

10.Cordell, J. L., B. Falini, W. N. Erber, A. K. Ghosh, Z.Abdulaziz,S.

MacDonald,K.A.F.Pulford,H.Stein,and D. Y. Mason. 1984. Immunoenzy-maticlabelingofmonoclonalantibodiesusingimmunecomplexes ofalkaline

phosphataseand monoclonalanti-alkaline phosphatase(APAAPcomplexes).J.

Histochem. Cytochem. 32:219-229.

11.Pinzani,M., L. Gesualdo, G.M.Sabbah, and H. E. Abboud. 1989. Effects

ofplatelet-derivedgrowth factorandotherpolypeptidemitogensonDNA synthe-sisandgrowth ofculturedratliverfat-storingcells. J.Clin.Invest.84:1786-1793.

12.Pinzani,M., T. C. Knauss, G. F.Pierce,P.Hsieh,W.Kenney, G. R.

Dubiak,and H. E. Abboud. 1991.Mitogenic signalsforplatelet-derivedgrowth

factor isoforms in liver fat-storing cells. Am. J. Physiol. (Cell. Physiol.).

260:C485-C491.

13.Poenie,M., J. Alderton, R.Steinhardt,and R. Y.Tsien. 1986.Calcium

rises abruptly andbrieflythroughout the cell at the onsetofanaphase. Science

(Wash.DC). 233:886-889.

14.Cheung, J. Y., D. L.Tillotson,R. V.Yelamary, and R. C. Scaduto. 1989.

Cytosolic freecalciumconcentrationinindividualcardiacmyocytes inprimary

culture.Am.J. Physiol. (Cell. Physiol.).256:C1120-1130.

15.Grynkiewicz,G., M.Poenie,and R. Y. Tsien. 1985.Anewgenerationof

Ca2"indicators withgreatly improved fluorescenceproperties.J. Biol. Chem. 260:3440-3450.

16.Galizzi,J.-P., M. Borsotto, J.Barhanin,M.Fosset, and M.Lazdunski.

1986.Characterizationandphotoaffinitylabeling ofreceptorsites fortheCa2"

channelinhibitorsd-cis-diltiazem,(±)-bepridil, desmethoxyverapamil,and

(+)-PN200-110 inskeletalmuscle transverse tubule membranes. J. Biol. Chem. 261:1393-1397.

17.Neylon, C. B., J. Hoyland, W. T. Mason, and R. F. Irvine. 1990.Spatial

dynamicsofintracellularcalciuminagonist-stimulatedvascular smooth muscle

cells.Am.J.Physiol. (Cell. Physiol.). 28:C675-C686.

18.Kreisberg,J. I., M. Venkatachalam, and D. Troyer. 1985.Contractile properties ofglomerularmesangialcells. Am.J.Physiol. (Renal Fluid Electrolyte Physiol.). 18:F457-F463.

19.Simonson,M.S., and M. J. Dunn. 1990. Endothelin- 1 stimulates contrac-tionofratglomerular mesangialcellsandpotentiatesfl-adrenergic-mediated cy-clicadenosinemonophosphateaccumulation.J.Clin.Invest. 85:790-797.

20. Shultz, P., T. C. Knauss, P. Mene, and H. E. Abboud. 1989.Mitogenic signalsfor thrombin inmesangialcells:regulation ofphospholipaseC and PDGF genes.Am.J.Physiol. (Renal Fluid Electrolyte PhysiolJ.26:F366-F374.

21. Mene, P., M. S.Simonson,and M. J. Dunn. Physiologyofmesangialcells. 1989.Physiol.Rev. 69:1347-1424.

22.Somlyo, A. V., M. Bond, A. P. Somlyo, and A. Scarpa. 1985. Inositol

trisphosphate-induced calciumreleaseandcontractioninvascular smooth mus-cle. Proc. Natl.Acad.Sci. USA. 82:5231-5235.

23.Geerts, A., R.Vrijsen,J.Rauteberg, P.Schellinck,and E. Wisse. 1989. In vitro differentiation of fat-storing cells parallels marked increase of collagen syn-thesis and secretion. J.Hepatol.(Amst.).9:59-68.

24.Mak,K.M., and C. S. Lieber. 1988. Lipocytes and transitional cells in alcoholic liver disease:amorphometricstudy.Hepatology.8sl027-1033.

25. Schmitt-Graf, A., S. Kruger, F. Borchard, G. Gabbiani, and H. Denk. 1991.Modulation ofa-smooth muscle actin and desminexpressionin

perisinu-soidalcellsofnormal and diseased human livers. Am.J.Pathol. 138:1233-1242.

26.Nakano, M., T. M. Worner, and C. S. Lieber. 1982. Perivenular fibrosis in alcoholic liverinjury: ultrastructure and histological progression. Gastroenterol-ogy.83:777-785.

27.Bhathal, P.5.,and H. J. Grossman. 1985. Reductionof theincreased