Phenotypical modulation of liver fat-storing cells by retinoids

Phenotypical modulation of liver fat-storing cells by retinoids

Journalof~epa?ofogy, 1992; 14: 211-220 @ 992 Elsevier Science Publishers B.V. Ail rights reser.ed. ir168-&!78/92/$05.00 211 HEPAT 00987 Massimo Pin...

2MB Sizes 0 Downloads 42 Views

Journalof~epa?ofogy, 1992; 14: 211-220 @ 992 Elsevier Science Publishers B.V. Ail rights reser.ed. ir168-&!78/92/$05.00

211

HEPAT 00987

Massimo Pinzad, Paolo Gentilini’ and anna E. Abbou& ‘Istiiu~o di ClinIca Medica It - Centro Knteruniversitariodi FisiopatologiaEpatica, Universila’di Firenze, Firenze,

Italyand ‘Deparment Medicine, Universityof Texas Health Science Center at San Anronio, San Antonio, TX, United Sratesof America

of

(Received 25 October 1990)

liver i~~amma~ion, liver fat-storing cells (FSC) differentiate into ‘myofibroblast-like cells’. This transition is characterized by a gradual loss of vitamin A stores, and previous studies suggest a possible relationship between tile intracellular retinoid content and the proliferative potential of this cell type. In the present study, we further characterized this aspect of FSC biology by monitoring ultrastrucPura1 changes and growth characteristics during several serial passages in culture. Our observations suggest that the complete transition to the ‘myofibroblast-like phenotype’ is paralleled by a sudden and remarkable increase in the growth rate. At this stage, cell growth appears rather independent from the presence of mitogens in the culture medium, suggesting cell transformation. Accordingly, the mitogenic effects of platelet-derived growth factor and epidermal growth factor appears reduced when compared to those observed in FSC retaining the original ‘storing’ phenotype. Incubation of vitamin A-depleted FSC with retinol and retinoic acid resulted in the partial recovery of intracellular retinoid stores and in a significant reduction of basal growth rate and basal and growth factor-induced DNA synthesis. In summary, these in vitro observations suggest that intracellular retinoids play a central role in the control of unstimulated and growth factor-induced FSC proliferation and may help understand in vivo mechanisms leading to liver fibrosis. In conditions

of chronic

The naturally occurring retinoids, retinol (vitamin A) and retinoic acid, exert important effects on the maintenance of the differentiated state of both epithelial aad nonepithelial cells (1). In particular, retinoic acid has been shown to act at multiple cellular sites, including the nucleus (2) XG ;>lasma membrane (3). Additional observations suggest a role of retinoids in the regulation of spe. cific gene expression (4). The liver play> a central role ir; retinoids metabolism, holding more than 90% of the vitamin A of the body (3). In adult rats, about 80% of the hepatic retinoids are stored, mostly as retinyl palmitate, in liver-specific perivascular cells known as fat-storing cells (FSC; Ito cells, stellate cells). In addition, FSC have much

higher levels of retinyl palmitate idrolase and acylCoA:retinol acyltransferase than parenchymal cells, indicating a central role of this cell type in hepatic retinoid metabolism (5). Recent studies have suggested that FSC proliferation and differentiation into ‘myofibroblast-like cells’ is strictly related to the development of liver fibrosis (6-10). Interestingly, the progression to the myofibroblast or ‘activated’ phenotype is characterized by a gradual loss of intracellular vitamin A stores. Furthermore, incubation of cultured rat liver FSC with retinol has been shown to affect their collagen producing phenotype and proliferation rate (11-13). Therefore, the intracellular content of retinoids may be an important factor in conditioning FSC phenotype, and particularly in conditions of chronic liver

Correspondence: Massimo Pinzani, M.D., Ph.D., Clinica Medica II, Universita’ di Firenze, Viale Morgagni, 85, I-50134 Firenze, Italy.

M. PINZANI

inflammation, when FSC are likely to be exposed to several proliferative stimuli. Along these lines, we have recently reported that FSC ate potential targets sf several polypeptide growth factors released by activatated macrophages (14-16). In the present study, we examined the relationship betweeu the intracellular retinoid content and the phenotypica1 and functional changes of rat FSC in culture. In addition, we evaluated the effects of several retinoids on the DNA synthesis and cell proliferation induced by plateletderived growth factor (PDGF) and epidermal growth factor (EGF) in this cell type. Our results are consistent with an influence of intracellular retinoid stores on unstimulated and growth factor-induced FSC proliferation.

Materials and Methods Growth factors

Recombinant PDGF (PDGF,,,) was purchased from Amgem Inc. (Thousand Oaks, CA), and purified receptor grade EGF from Collaborative Research Inc. (Lexington, MA). Retinoids Retinol (R), retinoic acid (RA) and retinyl palmitate (RP) were obtained from Sigma Chemical Co. (St. Louis, MO). All retinoids used were initially diluted in 100% ethanol to obtain a 5 mM solution. One mM stock solutions were then immediately obtained by further dilution in PBS with the addition of Tween-80 and nitrogen. These procedures were performed the night preceding the addition of retinoids to the cell cultures, and stock solutions were stored overnight in brown containers in the dark. Fresh aliquots of 5 PM retinoid-containing media were prepared immediately prior to use in serum-free culture medium, and the final concentration of ethanol was always
et al.

free access to water and a standard chow diet. Rats were not injected with retinyl acetate before being used for cell isolatiaz. FEZ were isolated according to the method of Friedman and Roll (17) with minor modifications and extensively characterized as described elsewhere (14). Freshly ise:c?icd FSC were resuspended in Waymouth’s MB 752/l medium (Gibco Laboratories, Grand Island, NY) supplemented with 15 mM Hepes, 0.6 U/ml insulin, 2 mM glutamine, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, antibiotic-antifungal solution (Sigma Chemical Co.), 10% horse serum (y-globulin free; Gibco Laboratories) and 10% calf serum (HyClone Laboratories, Inc., Logan, UT). The cells were plated in plastic tissue culture dishes or flasks (Costar, Cambridge, MA) at a density of 0.5-1.0.106 cells/ml and incubated at 37 “C in 5% CO*. The medium was replaced 24 h after plating and every 72 h thereafter. Primary cultures were allowed to grow to 90% confluence for 7-10 days. For further passage, cell monolayers were washed with Hanks’ balanced salt solution and removed with 0.025% trypsiu/O.5 mM EDNA in calcium- and magnesium-free I9ulbecco’s PBS (Gibco Laboratories) and plated in comp!ete culture medium. C,zll monolayers were consistently passaged using a 1:3 split ratio maintaining the same culture conditions. The observations reported in this paper are relative to two independent cell lines cultured up to the seventh serial passage. Ceil morphology Cell morphology was studied using an inverted microscope IMR2 (Olympus Corporation of America, New Hyde Park, NY) with Hoffman optics, a light microscope BH-2 (Olympus Corporation of America), and an immunofluorescence microscope (E. Leitz, Inc., Rockleigh, NJ). For electron microscopy, cells, grown on plastic culture dishes, were fixed overnight with a 1:l dilution of Kamovsky’s fixative and normal saline, postfixed with 2% osmium tetroxide, and stained with 1% uranyl acetate. After a graded series of ethanol dehydratations, the cell layer was diced and removed from the plastic with propylene oxide. The cells were then embedded in Spurr embedding medium and examined with JEOL 1OOCX transmission electron microscope (JEOL USA, Peabody, MA)* DNA synthesis DNA synthesis was measured as the amount of [methyl3H1thymidine ([3H]TdR) incorporated into TCA-precipitable material. Unless otherwise specified, cells were plated in 24-well dishes at constant densities and incubated in Waymouth’s medium containing 10% horse serum and 10% calf serum until ihey became 70-90% con-

FAT-STORING

CELLS PHENOTYTE

AND RETINOIDS

fluent. Cells were then incubated for 20 h under t ous conditions to be tested and subsequently pulsed for 4 h with 1.0 &i/ml of [3H]TdR (6.7 Ci/mmol; New England Nuclear, Boston, MA). At the end of the pulsing period, medium was carefully aspirated, ice-cold 5% trichloacetic acid was add&, and dishes were kept on ice for 15 min. After two additional washes with 5% TCA, cells were solubilized by the addition of 750~1 of 0.25 M NaOH, 0.1% SDS. The solubilized cell solution (0.5 ml) was then neutralized with 50 ~1 of 6 M HCl and counted in a scintillation counter. Cell number was determined in ;hree separate wells from each dish after trypsinization and counting in a Coulter counter (Coulter Electronics Inc., Hialeah, FL)* Autoradiography Autoradiography, used in some experi?lents to validate the DNA synthesis data, was performed using a technique described elsewhere (14). Two hundred to three hundred cells per each incubation were counted and the percentage of labeled nuclei (labeling index) was determined. Cellproliferation assay Rat FSC were plated in 12-well dishes at a density of l-lo4 cells/well in Waymouth’s medium with 20% serum (10% horse serum plus 10% calf serum, e.g., experiments described in Fig. 4) or with 1% Zeta serum plus retinoids (e.g., experimeats described in Fig. 6). Cell counts were performed on triplicate wells 24 h after plating (day 0) and after an additional 3 (day 3) and 6 days (day 6) by trypsinizing the cells and using a Coulter counter. Fresh medium and test conditions were added to the remaining wells at each time point. Statisticalanalysis Data, expressed as mean + SD., were analyzed by a two-w?y analysis of variance.

ResuIts

Morphologicai and functional changes of FSC in culture As already reported, freshly isolated FSC appeared as round cells with prominent cytoplasmic lipid-like droplets (14-117). Within 2-3 days in culture, the cells rapidly spread on the plastic suriace assuming a flattened and stellate-like appearance (Fig. 1A). The characteristic vitamin A fading green-blue fluorescence was easily de&table by exposing the cells to ultraviolet light at 330 nm. Trausmission electron microscopy studies revealed the presence of numerous large lipid droplets associated with microfilaments and dense bodies (Fig. 2A). These fea-

213

tures remained qualitatively present after three serial passages, although the size and the number of the fat droplets progressively decreased. These changes were associated with an increased number of microfil.ments and dense bodies, and a markedly enlarged rough endoplasmic reticulum containing homogenous material (Fig. 2B). A remarkable phenotypic modulation, however, was observed in cells cultured beyond the third passage. At this stage the cells started to assume a spindle-shape, rather than a stellate appearance, with a general reduction of cell dimensions (Fig. 1B). However, giant cells with multiple nuclei and elongated cytoplasmic processes were occasionally observed. Electron-microscopic observation revealed a complete transition to the ‘myofibroblast-like phenotype’ with complete disappearance of vitamin A droplets and a dramatic increase in the number of microfilaments and dense bodies (Fig. 3A). In addition, the chromatin pattern showed a complete transition from a clumped heterochromatin to a homogeneous euchromatin pattern (Fig. 3B). Interestingly, staining for the intermediate filament desmin, detected by the immunoperoxidase method (14), was still positive in FSC cultured up to the seventh passage. From a functional point of view, FSC cultured up to the third passage did not show appreciable differences. Quiescence, as reflected by basal counts of [3H]TdR and percent of labeled nuclei by autoradiography was easy to achieve by incubating the cells for 48 h in Waymouth medium supplemented with 1% Zeta serum (AMF, Meriden, CT), a very low mitogenicity serum (Table 1). FSC cultured beyond the third passage became highly proliferative, with a marked decrease in their doubling time, and failed to demonstrate density-arrested growth.. This resulted in the development of cell multilayers with consequent early detachment from the plastic surface. Quiescence could not be achieved even after 4 days in 1% Zeta serum (Table 1) or in serum-free/insulin-fee medium (data not shown). Fig. 4 shows the growth curves of FSC up to and beyond the third passage. As expected, a dramatically higher growth rate was observed in FSC cultured beyond the third passage. Effects of retinaidson vitaminA-depleted FSC (myofibroblast-likephenotype) In this set of experiments we evaluated the effects of different retinoids on the phenotypical and growth characteristics of FSC cultured beyond the third passage. In particular we used cells between rhe fourth aad the seventh passage, characterized by the complete disappearance of vitamin A droplets. Given the very high proliferation rate, the cells were plated at low density (0.5-l-lo4 cells/ml) in Waymouth’s medium containing 1% Zeta se-

214

M. PINZANI et al.

Fig. 1. (A) Light microphotograph of rat FSC in primary culture. Hoffman contrast. Fat-storing celis 48 h after plating. Cells are spread 06 plastic witha flattened and stellate-!ike appearance (300x). (B) Light microphotograph of rat FSC after four serial passages (passage 4). Hoti an contrast. At this stage, the ceils start to assume a spindle-shaped rather than stellate appearance and are characterized by a remarkably higher ’ proiiferation rate (300X).

FAT-STORING

CELLS PHENOTYPE AND RETINOfDS 215

Fig. 2. Transmission electron microphotographs of rat FSC in culture. (A) Fat-storing cells in primary culture, 3 days after plating. The ce!!s conrain numerous large hpid droplets (9600~). (B) Fat-storing cells after three subcultures (passage 3). The number and the size of the lipid droplets are considerably reduced. Aggregates of microfilaments (h4F) and dense bodies (DE) are evident in the cytoplasm associated with an enlauged rough endoplasmic reticulum (RER) containing homogeneous material (0,600~ ).

216

M. PINZANI

et al.

FAT-STORING

CELLS PHENOTYPE

AND RETINQIDS

217

Fig. 3. Transmission electron microphotographs of FSC after seven serial passages (passage 7). (A) In this picture is evident the complete transition to the ‘myofibroblast-like phenotype’ with complete disappearance ofvitamin A droplets and a dramatic increase in the number of microfiiaments (MF) and dense bodies (DB) (12 660X). (B) An additional feature is tbe presence of an homogeneous eucbromatin pattern (15 2O@x). (C) Passage 7 fat-storing cells incubated for 24 h with 5 .&I retinoic acid. The reappearance of fat-droplets, particularly at the cell periphery, is associated with a significant reduction of the rough endoplasmic reticulum (RER) (12 660x). e

rum with or without the three different retinoids at concentrations of 0.5, 2.5 and 5.0 PM. Cell morphology was studied until the sixth day of incubation. Addition of ?X and RA tti the culture medium induced the reappearance of vitam:.!i A droplets in the cytoplasm as early as 24 h after the 51,,!rt of incubation (Fig. 3C). The number and the size of tk droplets progressively increased reaching a plateau after 3-4 days. The presence of vitamin A in the droplets ltias confirmed by detecting the characteristic

6.0.

TABLE

I

Quiescence test in rat fat-storing cells cultured up to and beyond the third serial passage Passage 1-3

Passage 4-7-

[3HjTdR incorporation into DNA (cpm/well) 24h 48h 72 h 96h

382 f 186 + 157 + 161 2

2747 3890 6791 7043

Autoradiography c/c of labeled nuclei (labeling index) 24h 48h 72 h 96h

3.0 1.0 1.0 1.0

57 21 64 43

* 1.7 -c 0.7 + 0.4 + 0.9

f 721 + 410 -c 804 + 141

21+4 27 f 6 34 f 5 33 f 5

Ninety percent confluent FSC were incubated in Waymouth’s medium containing 1% Zeta serum for the indicated time periods. Data represent mean f SD. of two experiments done in triplicate

Fig. 4. Growth curves of rat FSC cultured up to (Pl-P3) and beyond the third passage (P4-P7). FSC were plated in 12-well dishes at a density of l-104 ceils/weli in Waymouth’s medium with serum. Cell counts were performed in triplicate wells 24 h after plating (day 0) and after an additional 3 and 6 days (day 3 and day 6, respectively) by rrypsinizing the cells and using a Coulter counter. Fresh medium was added to the remaining wells at each time point. Data are mean + SD. for four experiments performed in triplicate. Changes were statistically significant @ < 0.05 or higher degree of significance) starting at day 0.

Fig. 5. Effect of retinal, retinoic acid and retinyl palmitate on 13H]TdR incorporation into DNA of rat FSC cultured beyond the third passage. Ninety percent confluent ceb in 24-well dishes were incubated for 20 h in 1% Zeta serum with or without different retinoids. Cells were &en harvested after a 4-h pulsing period with (‘H]TdR (1.0 @/ml). Data (mean + SD.), expressed as the percent change relative to control wells incubated with 14 Zeta se~tim alone (CPM = 2317 :? 456), are from three different experiments (passages 5,6 and 7) done in tripiicate.

M. PINZANI et al.

218 fading green-blue autofluorescence under fluorescence microscopy by exposing the cells to ultraviolet light at 330 nm. However, a remarkable individual variance, together with a general inability to recover a full content of vitamin A, were observed. In this respect, fat droplets showed a tendency to be confined at the cell periphery. Additional electron microscopy findings included an overall reduction of the RER (Fig. 3C) and dense bodies, and a shifting to the clumped heterochromatin pattern observed in primary and early passaged FSC. No apparent differences were found using different concentrations of R or RA. Cell cultures incubated with RP did not show appreciable changes. As shown in Fig. 5, a 24-h incubation with R or with RA, but not with RP, significantly reduced DNA synthesis. These findings were confirmed by cell proliferation experiments performed over a period of 7 days (Fig,

6). Cell viability, assessed by Trypan blue exclusion, was consistently 297%. Of interest, growth curves of cells incubated with 1% Zeta serum (Fig. 6, control wells) were not significantly different from those obtained by incubating the cells with complete medium (Fig. 4, P4-P7), confirming that the high growth rate of cells cultured beyond the third passage is independent from the presence of serum in the culture medium. Effects of rehwids on PEW’- and EGF-induced DNA synthesis

For these studies, FSC, between the fourth and the seventh passage, were plated at very low density in Waymouth’s medium supplemented with 1% Zeta serum with or without the three different retinoids at concentrations of 0.5,2.5 and 5.O@I. After 72 h, when the cells were 7090% confluent, the medium was replaced with fresh medium with or without the addition of PDGF or EGF (both 10 @ml). After a 20-h incubation period with the grow!h factors, cells were pulsed for 4 h with [3HjTdR (I.0 &i/ml). Because of the reduction of basal cell proliferation induced by R and RA, the cell number was deter-

B

Ratiw *

l-vi RetinOic Acid

= 8 $-

3.0

X $

1.0

f v 0.5 I

0

1 PDGFlO

Ip/ml

1

I

EGF 10 w/ml

I

0

I

I

3

6

Days in Culture

Fig. 6. Effects of retinol (0), retinoic acid (m) and retinyl palmitate (O), compared to control ( ), 011the growth of rat FSC cultured beyond the third passage. Cells were plated in 12-well dishes at a density of l.lO’?well in Waymouth’s medium containing 1% Zeta serumwith or without different retinoids (all 5pM). Cell counts were performed in triplicate wells 24 h after plating (day 0) and after an additional 3 and 6 days (day 3 and day 6, respectively) by trypsinizing the cells and using a Coulter counter. Fresh medium containirigthe same conditions was added to the remaining wells at each time point. Data are mean f S.D. for three experiments (passages 4, 6 and 7) performed in triplicate. Changes were statistically significant (p < 0.05 or higher degree of significance) for retinol and retinoic acid starting at day 3.

Fig. 7. Effect of retinal, retinoic acid and retinyl paimitate on [3H]TdR incorporation into DNA of rat FSC induced by PDGF (panel A) and EGF (panel 9). FSC (passages $6 and 7) were plated in 26well dishes at a density of 0.51@ cells/well in Waymouth’s medium supplemented with 1% Zeta serum with or without the three different retinoids at concentrations of 0.5,2.5 and .5.O@I. After 72 h, whenthe cells were 70-90% confluent, the medium was replaced with fresh medium with or without the addition of PDGF or EGF (both 10 n&l). After a 20-h incubation period with the growth factors, cells were pulsed for 4 h with i3H]TdR (1 .O@X/ml). C = Control wells, incubated without retinoids or growth factors. The ‘O.O,UM retinoids’ refers to cell cultures incubated with PDGF or EGF without retinoids. Data (mean + SD.), expressed as cpm/l@ cells, are relaiise to three experiments in quadruplicate for each retiuoid. *, significant difference @ e 0.05 or higher degree of significance) from cells incubated with PDGF or EGF without retinoids.

FAT-STORING CELLS PHENOTYPE AND RETINOIDS

mined in three separate wells for each condition studied in each experiment and counts/well were corrected per cell number. As shown in tiig. 7, R and EPAsignificantly reduced PDCF-induced DNA synthesis with a maximal effect between 2.5-5 0~ , whereas RP had no effect. The effects of retinoids on EGF-induced DNA synthesis were less marked and significant only for 2.5 @vl R and 5.0 PM RA. In a similar set of experiments performed on early passaged FSC (passage 1, with strll abundant intracellular retinoid content) we were unable to observe significant effects of retinoids on PDGF- or EGF-induced mitogenicity. In particular, a slight trend to inhibition of PDGF-induced mitogenicity was observed only with 5 ,&I RA (data not shown).

iscussion The infiuence of retinoids in the maintenance of cell differentiation assumes a peculiar importance in liver FSC because of their role in hepatic retinoic metabolism and development of liver fibrosis. Previous studies have snggested a possible relationship between FSC intraceliularetinoids, monitored by serial transmission electron-mrcroscopic observations, is an important factor in condievaluating the correlation between ultrastructurai changes and the proliferative potential of this cell type are still lacking. The results of the present study provide circumstantial evidence that the presence of intracellular retinoids, monitored by serial transmission electron microscopic observations, is an important factor in conditioning the phenotypical modulation of FSC and their proliferation rate. The complete transition to the ‘myofibroblast-like phenotype’, characterized by a loss of intracellular vitamin A droplets, is paralleled by a sudden change in the growth characteristics, Interestingly, we have observed that by maintaining a rigourous culture schedule this transition usually occurs after three or four serial passages. In addition, at this stage cell growth seems rather independent from the presence of serum in the culture medium, indicating a transition to a ‘transformed’ phenotype. In this respect, most of the activities ascribed to retinoids are in the direction of suppressing cell transformation. Thest activities are directed to the control of cssential cell functions such as oncogene expression, RNA processing, protein synthesis and CAMP- and calcium-dependent protein kinases (19). Recent studies by Matsuura et al. (20) using a laser ceil sorting system, have clearly shown that physiological concentrations of R and RA added to the culture medium were transferred to FSC in significant amounts by passive transport and without the participation of retinoi-binding

219 protein (RBP). Remarkably, addition of RP did not result in any significant uptake by FSC (20). Our EM obser-vatiOnS are in full agreement with these results and suggest, along with DNA synthesis and cell proliferation analysis, that rk antiproiiferative effect of a single retinoid is related to the effective re-uptake by the cell. Accordingly, both K and RA induced a signicant inhibitory effect on cell growth, whereas RP had no effect. Two principal mechanisms have been proposed for the transport of retinoids to FSC: binding to specific RBP receptors on the ceil membrane (5) and transport through a concentration gradient involving cellular RBP (20). Since the experiments described in this paper were carried out in culture medium which was virtually serum deprived, it is possible that the exposure to the different retinoids was somehow ‘nonphysiologic’. Indeed, the absence of RBP and/or other factors in the culture medium may explain the inability to recover a full content of vitamin A following prolonged exposure to R and RA. In addition, reduced levels of intracellular RBP, described in pluripassaged FSC (llj, together with the poor solubility of retinoids, and particularly RP, in aqueous medium may contribute to this phenomenon. The potential role of polypeptide growth factors, PDGF in particular, in inducing FSC proliferation has been recently reported (14,16,21). In the present study we have also evaluated the possibility that the intracellular content rof rctinoids may influence the mitogenic response of FSC to polypeptide growth factors such as PDGF and EGF. The mitogenic effects of these growth factors on pluripassaged FSC (myofibroblast-like phenotype) were markedly *educed when compared to those observed in early pass -.ged FSC (14). This observation is not surprising knowing that pluripassaged FSC show a remarkable intrinsic proliferative potential. Alternatively, it is possible that receptor number and/or affinity for the growth factors tested are reduced at this stage of culture. Several in vitro observations have suggested that modulation of ceil proliferation by retinoids might result from an alteration of the response pattern to hormones and growth factors present in the serum. In addition, it has been shown, at least for EGF and insulin, that this effect is not secondary to modulation of membrane receptors but probably due to the control of gene expression exerted by retinoids (19). A:though the biologic relevance of these studies remains specuhtive, our in vitro findings may help in understanding the important aspects of FSC biology in vivo. In particular OUT results suggest a close relationship between intracellular retinoids and the proliferative potential of this cell type. It is possible that in the early stages of liver inflammation, when FSC retain their original ‘storing’

M. PINZANI et al.

220

phenotype, cell proliferation is mainly due to interactions with polypeptide growth factors and other mitogens released locally during the tissue repair and inflammatory processes. In later stages, when the development of liver fibrosis seems rather independent from the presence of an inflammatory infiltrate, it is possible that FSC exert their fibrogenic role autonomously as a consequence of their ‘transformed’ phenotype.

References 1 Lotan R. Effects of vitamin A and its analogs (retinoids) on normal and neoplastic eels. Biochim Biophys Acta 1980; 605: 33-91. 2 Crow >A, Ong DE, Chytil F. Specificity of celhdar retinol-binding protein in the transfer of retinol to nuclei and chromatin. Arch Biochem Biophys 1987; 254: 372-5. 3 Goodman DS, Blaner WS. Biosynthesis, absorption, and hepatic metabolism of retinal. In: Sporn MB, Roberts AB, Goodman DS, eds. The Retinoids, Vol. 2. New York: Academic Press, 1984; l-39. 4 GiIfix BM, Eckert RL. Coordinate control by vitamin A of keratin gene expression in human keratinocytes. J Biol Chem 1985; 260: 14026-g. 5 Blomhoff R, Rasmussen M, N&son A, et al. Hepatic retinol wetabolism, distribution of retincids, enzymes and binding proteins in isolated rat liver cells. J Biol
Acknowledgements The authors thank Dr. Medhat 0. Hassan and the electron microscopy unit at the Veterans Administration Medical Center, Cleveland, OH, U.S.A., for perfcrcning the electron microscopy studies. These studies were supported by Veterans Administration Research funds, National Institutes of Health grants DK-33665, and Consiglio Nazionale delle Ricerche of Italy. H.E. Abboud is an Established Investigator of Lhe American Heart Association.

cellular retinol binding proteins levels. J Biol Chem 1987; 262: 10280-6. 12 Davis BH, Vucic A. The effect of retinol on Ito ceU proliferation. Hepatology 1988; 8: 788-93. 13 Davis BH, Kramer RT, Davidson NO. Retinoic acid modulates rat Ito ceU proliferation, collagen, and plansforming growth factor-/l production. J Clin Invest 1990; 86: 2062-70. 14 Pinzani M, Gesualdo L, Sabbah GM, Abboud HE. Effects of platelet-derived growth factor and other polypeptide mitogens on DNA synthesis and growth of cultured rat liver fat-storing cells. J Clin Invest 1989; 84: 1786-93. 15 Pinzani M, Abboud HE, Aron DC. Secretion of insulin-I&e growth factor-1 and binding proteins bv rat liver fat-storine cells: regulatory role of platelet-&ived growth factor. Endocrinology 1990: 127: 2343-9. 16 Pinz&i M, Knauss TC, Pierce GF, et al. Mitogenic signals for platelet-derived growth factor isoforrrs in liver fat-storing cells. Am J Physiol-Cell Physioll991; 29: C485-91. 17 Friedman SL, RoU FJ. Isolation and culture of hepatic lipocytes, Kupffer cells, and sinusoidal endothelial cells by density gradient centrifugation with stractan. Anal Biochem 1987; 161: 207-18. 18 De Leeuw AM, McCarthy SF, Geerts A, Knook DL. Purified rat liver fat-storing cells in culture divide and contain collagen. Hepatology 1984; 4: 392-403. 19 Roberts AB, Spom MB. Cellular biology and biochemistry of the retinoids. In: Spom MB, Roberts AB, Goodman DS, eds. The Retinoids, Vol. 2. New York: Academic Press, 1984; 209-85. 20 Matsuura T, Nagamcri S, Fujise K, et al. Retinol transport io cuitured fat-storing cells of rat liver. Quantitative analysis by anchored cell analysis and sorting system. Lab Invest 1989; 61: 10715. 21 Friedman SL, Arthur MJP. Activation of cultured rat hepatic Iipocytes by Kupffer ceU conditioned medium. Direct enhancement of matrix synthesis and stirnlrlarion of cell proliferation via induction of platelet-derived growth factor receptors. J Ciin Invest 1989; 84: 1780-5.