microscopy (Lux Permanox no. 5213) from Nunc, Inc., Naperville, IL; Sepharose CL-6B from Pharmacia, Uppsala, Sweden; all other reagents were of the highest grade commercially available.

Human urinary epidermal growth factor (hEGF) was kindly provided by Hitachi Chemical Co., Ltd, Hitachi, Ibaraki, Japan.

Cultures of pericardial cells from rabbit pericardial cavity

Japanese normal white rabbits (3.0–3.5 kg) were anesthetized with sodium pentobarbital (39 mg kg-1, intravenously). After exposing the heart by a transverse thoractomy, pericardial fluid containing pericardial cells was aspirated from the pericardial cavity using a syringe. Then, sterilized phosphate-buffered saline without calcium and magnesium (PBS(–)) (about 1ml) was injected into the pericardial space. After 10 min incubation at room temperature, during which time the heart was massaged, the solution in the pencardial space was recovered using a syringe. The cell suspensions were combined and the cells were collected by centrifugation at 800g for 5 min at 4°C. The cells were resuspended in 2 ml of 20 % FBS-MEM containing penicillin (100 i.u. ml-1) and streptomycin (100 μg ml-1) and seeded onto 35mm diameter plastic dishes (Falcon Primaria). Cells were cultured at 37 °C in a humidified atmosphere containing 5 % CO2. Once the cells had become confluent (about 2 weeks after seeding), they were subcultured by dispersing them with a 5-min incubation in 0.025% trypsin and 0.025% K2EDTA in PBS(–), pH 7.3. The cells were split at a ratio of 1:3 and cultured in 10% FBS-MEM containing penicillin (100i.u. ml-1) and streptomycin (100μgml-1). Culture medium was changed every three days.

Experiments for morphological studies were done using cells from first and second passages, and those on hyaluronic acid synthesis were performed on cells of less than eight passages.

Morphological methods

Morphological observations of cultured pericardial cells were made using an Olympus CK2 inverted microscope (Olympus Optical Co., Ltd, Tokyo).

Electron microscopy

Pieces of pericardial tissue and, also cells grown on glass coverslips, were fixed in 2 % paraformaldehyde–2.5 % glutaraldehyde in 0.1 M cacodylate buffer solution at room temperature. After washing in the buffer solution at 4 °C, they were postfixed in osmium tetroxide in the same buffer solution at 4°C. They were dehydrated in a graded ethanol series. For transmission electron microscopy, pericardial tissues were passed through n-butyl glycidyl ether and then embedded in epoxy resin. Ultrathin sections were stained with uranyl acetate and lead citrate and observed with a JEM-100B electron microscope (JEOL, Tokyo). For scanning electron microscopy, dehydrated cells on coverslips were passed through isoamyl acetate and subjected to critical point drying. After coating with evaporated gold—platinum in a high vacuum, they were examined with a JSM-35 scanning electron microscope (JEOL, Tokyo).

Indirect immunofluorescence microscopy

Fragments of pericardial tissue were fixed in 95% cold ethanol overnight at 4°C. They were dehydrated and embedded in paraffin wax. Sections were cut at 4 /an, and after wax removal, they were air-dried. Cultured cells on glass coverslips were also fixed in 95 % cold ethanol at 4°C for 20 min. They were then frozen at —20°C and thawed at room temperature, three times. After washing in phosphate-buffered saline (PBS) three times, they were air-dried. Both tissue sections and cells were immersed in PBS and were then treated with guinea pig antibody against keratin (Bio-Yeda, Israel) diluted 1:16 with PBS in a moist chamber for 2 h at room temperature. After washing three times in PBS, they were inoculated in fluorescein isothiocyanate-conjugated rabbit anti-guinea pig IgG (MBL, Tokyo) diluted 1:30 with PBS in a moist chamber for 40 min at room temperature. They were rinsed in PBS, mounted in glycerol-PBS (9:1, v/v) and examined with a Zeiss microscope. Control staining was done using non-immune guinea pig serum as the first antibody.

Assay for glycosaminoglycan synthesis

Pericardial cells were inoculated at about 2×106cells ml-1 in 35mm diameter plastic culture dishes or in plastic microwells and maintained in MEM supplemented with 10 % FBS. After growth had ceased at confluency, the medium was replaced with fresh MEM. After preincubation for 16–18h at 37°C, mEGF and/or IGF-I at various concentrations (1–1000 ng ml-1) or saline as a vehicle was added to the dishes or the microwells. After preincubation for 2h, 10μCi of [6-3H]glucosamine and/or 10μCi of H235SO4 was added to the dishes or the microwells. After a 6-h or 24-h incubation at 37°C either with or without mEGF and/or IGF-I, the medium (medium fraction) was removed from each dish or microwell and the cells (cell fraction) were rinsed with cold PBS(–). The cells, to which 0.5 M NaOH (1ml) was added, were harvested with a rubber policeman.

In case of treatment with a tyrosine-specific protein kinase inhibitor, genistein, the confluent pericardial cells pretreated with serum-free MEM for 16 h at 37 °C were treated for 30 min at 37 °C with various concentrations of genistein (5–50 μg ml-1) before 24h incubation with [6-3H]glucosamine (10μCiml-1) at 37 °C with or without IGF-I and/or EGF.

Preparation of glycosaminoglycans and enzymic digestion of glycosaminoglycans

Labeled glycosaminoglycans from medium and cells were prepared after Pronase digestion and cetylpyridinium chloride (CPC) treatment (Iwama et al. 1985). Digestion with Streptomyces hyaluronidase has been described (Ohya and Kaneko, 1970).

Gel chromatography

Gel chromatography was done on a Sepharose CL-6B column (1.4cm×60cm) equilibrated and eluted with 50mM Tris-HCl buffer (pH 7.4) containing 0.15 M NaCl at a flow rate of 15 ml h-1 at 4°C. Fractions (1ml) were collected and assayed for radioactivity.

Assay for hyaluronic acid synthase activity

Pericardial cells were incubated at 2×105 cells ml-1 in 60 mm diameter plastic culture dishes and cultured for 5 days in 10% FBS-MEM until the stationary phase of growth. Medium was replaced with serum-free MEM. After preincubation for 16 h at 37°C, medium was replaced with MEM containing IGF-I and/or EGF, or saline only. After cells were incubated for 6h at 37 °C, they were harvested from the dishes with a rubber policeman. For treatment with genistein, confluent pericardial cells pretreated with serum-free MEM for 16 h at 37 °C were treated for 30 min at 37 °C with genistein before a 6-h incubation with or without IGF-I and/or EGF. Hyaluronic acid synthase activity was measured as described in our previous paper (Honda et al. 1989). Briefly, the harvested cells were sonicated twice for 5 s each at 50 W with an ultrasonic cell disruptor (Heat systems-Ultrasonics, Inc., Farmingdale, NY). The cell sonicate was centrifuged at 800 g for 5 min at 4°C, and the supernatant separated and recentrifuged at 20 000 g for 10 min at 4°C. More than 80% of the hyaluronic acid synthase activity was recovered in the 20 000 g cell sonicate pellet. To this pellet was added 100 μl of reaction buffer (pH 7.0) containing 40mM NaH2PO4, 2MM MgCl2, 0.4mM uridine 5′-diphospho (UDP)-N-acetylglucosamine (Sigma Chemical Co., St Louis, MO), and 0.1 μCi of UDP-D-[14C]glucuronic acid (328.2 mCi mmol-1, New England Nuclear Research Products, Boston, MA). After incubation for 1 h at 37°C, 100 μl of 2 % SDS was added to the reaction mixture, which was then boiled for 2 min to stop the reaction. The sample was subjected to descending paper chromatography on Whatman no. 3MMChr filter paper in isobutyric acid/1M NH4OH (5:3, v/v). After development for 24 h, the spotted origins of the chromatogram were cut out, and radioactivity was measured in a liquid scintillation counter. Specific activity of this enzyme was expressed as disintsmin 1 of UDP-D-[14C]glucuronic acid incorporated into hyaluronic acid h-1 μg-1 of DNA.

Assay for protein synthesis

After the preincubation of confluent pericardial cells in microwells with MEM for 16–18h at 37°C, medium was replaced with 1 ml MEM containing 3H-labeled amino acid mixture (1 μCiml-1) with or without IGF-I and/or EGF. After a 2-h incubation at 37°C, the medium was removed from the microwells. Cells were washed with PBS(–) solution and harvested with 0.025% trypsin solution. To the cell suspension, 60% trichloroacetic acid (TCA) was added to a final concentration of 10 %. Pellets were washed with 10 % TCA then dissolved in 1 ml of 0.5 M NaOH. The alkali solution was neutralized in 1ml of 0.5 M HC1 and radioactivity was then measured,

Assay for radioactivity

An 8 ml portion of ACS-II (Amersham) and 0.5–1 ml of the sample solution to be assayed were mixed in a vial bottle, then radioactivity was measured with a liquid scintillation counter (Aloka LSC-900).

Statistics

A significant level of P<0.05 was calculated by Student’s t-test.

Morphological and immunological characterization of rabbit pericardial cells

Transmission electron micrographs (Fig. 1A and B) show that rabbit pericardium consists of thin connective tissue lined on both surfaces by mesothelial cells that have numerous microvilli on their free surface. Cilial structure was noted in one mesothelial cell (Fig. 1A). Many microvilli projected from the cytoplasm of the mesothelial cell into the cavity and the basement membrane was demonstrated between the cell and the connective tissue (Fig. 1B).

Observation of pericardial cells in monolayer culture by phase-contrast light microscopy shows that the cells exhibit a polygonal shape at confluence and have a pavement-like appearance (Fig. 1D).

Localization of keratin between rabbit pericardial tissue and pericardial cells in culture was compared by indirect immunofluorescent staining with anti-keratin antibody (Fig. 1C and E). Mesothelium of parietal pericardium was specifically stained with anti-keratin antibody, but submesothelium including fibroblasts and connective tissue was not, suggesting that a positive reaction with antibody against keratin occurs only in mesothelial cells (Fig. 1C). Cytoplasm of pericardial cells isolated from the pericardial cavity was stained with anti-keratin antibody (Fig. 1E).

A scanning electron micrograph of pericardial cells in culture shows numerous microvilli on the cell surface (Fig. 1F).

These morphological and immunological results strongly suggest that pericardial cells isolated from the rabbit pericardial cavity by our culture methods are mesothelial cells.

Stimulation of hyaluronic acid synthesis by EGF in pericardial cells

We previously examined incorporation of [6-3H]glucosa-mine into hyaluronic acid and other sulfated glycosaminoglycans in pericardial mesothelial cells. More than 70 % of the total glycosaminoglycans biosynthesized in stationary growth-pbase cultures was hyaluronic acid, and the remaining minor components were chondroitin sulfate and heparan sulfate (Ohashi et al. 1988). Fig. 2 shows the dose dependence of the effect of mEGF on glycosaminoglycan synthesis in pericardial cells after an 8-h exposure to mEGF. Stimulation of 3H-labeled glycosaminoglycan synthesis by mEGF was observed at concentrations above 10 ng ml-1. Maximum synthesis occurred at 30–100 ngml-1 concentration where the level was about twice that of the control. However, mEGF did not stimulate 35S-labeled sulfated glycosaminoglycan synthesis in pericardial cells. These results suggest that mEGF stimulates 3H-labeled glycosaminoglycan synthesis, mainly as a result of hyaluronic acid synthesis in cells.

To confirm this, [3H]glycosaminoglycan stimulated by mEGF was subjected to Sepharose CL-6B chromatography and Streptomyces hyaluronidase digestion. Fig. 3 shows Sepharose CL-6B column chromatographs of [3H]glycosa-minoglycan synthesized by pericardial cells either with or without mEGF (100 ng ml-1). The material that eluted in the void volume was identified as hyaluronic acid, because it was completely digested by Streptomyces hyaluronidase. This result suggests that synthesis of hyaluronic acid in pericardial cells is enhanced by mEGF. Increased hyaluronic acid synthesis induced by mEGF (100 ng ml-1) was reduced to the control level by concomitant addition of anti-EGF antibody (50μgml-1, anti-epidermal growth factor IgG, Collaborative Research Inc., Bedford, MA). Also, anti-EGF antibody itself lacked the capacity to stimulate hyaluronic acid synthesis (data not shown). Further, we obtained a similar stimulatory effect of human urinary EGF (100 ng ml-1) on hyaluronic acid synthesis in pericardial cells (data not shown). These results suggest that EGF induces specific stimulation of hyaluronic acid synthesis in pericardial cells.

Cooperative stimulation of hyaluronic acid synthesis by the combined effect of IGF-1 and EGF in pericardial cells

Because we found that IGF-I (Honda et al. 1989) or EGF (see Fig. 3) stimulates hyaluronic acid synthesis in pericardial cells, we examined the combined effect of IGF-I and EGF on hyaluronic acid synthesis.

Fig. 4 shows the time course of glycosaminoglycan-synthesizing activity from [3H]glucosamine in pericardial cells after the addition of EGF (100ngml-1), IGF-I (100ngml-1), and EGF (100ngml-1) plus IGF-I (100ngml-1), or vehicle to cultured pericardial cells at confluence. Significant stimulatory effects of EGF, IGF-I, and the combination of EGF and IGF-I, on glycosaminoglycan synthesis were observed after a 6-h exposure to growth factors, and the total activity induced by each of these growth factors was about 1.8 times, 1.5 times and 3.1 times the control level after 48 h culture periods.

Fig. 5 shows the dose dependence of the combined effect of EGF plus IGF-I on hyaluronic acid synthesis in pericardial cells after a 24-h exposure to the growth factors. Maximal increase in hyaluronic acid synthesis occurred at each concentration (100ngml-1) of EGF and IGF-I, and the synthesis was about six to seven times higher compared with that of untreated cells. However, treatment with EGF, IGF-I, and EGF plus IGF-I, did not cause significant stimulation of sulfated glycosaminoglycan synthesis in the pericardial cells.

To clarify the stimulatory mechanisms of these growth factors that act on hyaluronic acid synthesis in pericardial cells, we measured hyaluronic acid synthase activity in pericardial cells treated with or without EGF and/or IGF-I, Fig. 6 shows that hyaluronic acid synthase activity in confluent cells after a 6-h exposure to EGF (100ngml-1), IGF-I (100ngml-1), and the combined exposure to EGF (100ngml-1) plus IGF-I (100ngml-1), increased to a level about 4.5 times, 2.3 times, and 8 times, respectively, that of controls. These results suggest that both EGF and IGF-I may cause induction of hyaluronic acid synthase, because the activity in pericardial cells after pretreatment with actinomycin D (0.1μgml-1) or cycloheximide (lμgml-1) for 30 min was inhibited, to the control level of the activity (data not shown).

Inhibitory effects of genistein on hyaluronic acid synthesis and hyaluronic acid synthase activity induced by IGF-I and EGF

We examined whether or not a specific tyrosine protein kinase inhibitor, genistein, affects hyaluronic acid synthesis (including hyaluronic acid synthase) in rabbit pericardial cells, because both EGF- and IGF-I receptors on the cell membrane have tyrosine kinase activity in their cytoplasmic domains. Fig. 7 shows that genistein caused significant partial inhibition of glycosaminoglycan synthesis, i.e. hyaluronic acid synthesis induced by EGF and IGF-I (100ngml-1) after pretreatment with genistein for 30 min at concentrations of 5–10 μgml-1, and complete inhibition to the control level at a concentration of 30μgml-1.

Fig. 8 shows that the activity of hyaluronic acid synthase induced by EGF (100ngml-1) plus IGF-I (100ngml-1) in pericardial cells after pretreatment with genistein (10μgml-1) for 30min was inhibited, to the control level of activity, though cells treated with genistein (10μgml-1) remained at about 70% of the general protein synthesis levels of controls (100%) as determined by 3H-labeled amino acid incorporation into cells. However, genistein itself added to a cell-free assay system did not cause a direct inhibitory effect on the activity of hyaluronic acid synthase of pericardial cell membrane fraction at an even higher concentration of genistein (100–150μgml-1).

These results suggest that the specific tyrosine kinase inhibitor, genistein, causes inhibition of hyaluronic acid synthesis induced by EGF and IGF-I, probably due to inhibition of tyrosine kinase activity in the cytoplasmic domain of the cell-membrane receptors of EGF and IGF-I, rather than to direct inhibition of hyaluronic acid synthase activity in plasma membrane.

This study demonstrates that IGF-I and EGF cooperatively stimulate hyaluronic acid synthesis and hyaluronic acid synthase activity in rabbit pericardial mesothelial cells (Figs 4, 5 and 6) and that the stimulatory effects induced by these growth factors can be blocked after pretreatment with the tyrosine-specific protein kinase inhibitor, genistein (Figs 7 and 8).

There have been few studies on hyaluronic acid synthesis induced by a combination of growth factors, but we found that combined treatment of pericardial cells with IGF-I and EGF under serum-free culture conditions resulted in a marked increase in hyaluronic acid synthesis and the level of hyaluronic acid synthase activity, compared to that achieved by IGF-I or EGF alone (Figs 4,5 and 6). The effect induced by the combination of the two growth factors seems synergistic rather than additive. Also, increased hyaluronic acid synthesis induced in pericardial cells by these growth factors occurs before DNA synthesis begins (Honda et al. 1989). This process could be one of many prerequisite steps in cells progressing to mitosis.

Recent studies on peptide growth factors have clearly shown that the initial step of interaction of the growth factor’s signals with target cells is the binding to receptors present on plasma membrane. An EGF receptor molecule is a single polypeptide chain containing both a ligand binding domain and a growth factor-sensitive tyrosine kinase activity domain (Carpenter, 1987). An IGF-I receptor molecule has a heterotetrameric structure consisting of two α- and two β-subunits and a tyrosine kinase domain in the cytoplasmic portion of the β-subunits (Ullrich et al. 1986). Growth factor-dependent tyrosine kinase activity of the cytoplasmic domain is regarded as the primary, though not necessarily exclusive, mechanism for generation of intracellular second messages (Morgan and Roth, 1987). An isoflavone compound, genistein, is a highly specific inhibitor of tyrosine-specific protein kinases such as the EGF receptor tyrosine kinase, but scarcely inhibits the activity of serine- and threoninespecific kinases such as cyclic AMP-dependent protein kinases, protein kinase C and phosphorylase kinase (Akiyama et al. 1987). Therefore, we used genistein to probe the involvement of tyrosine kinase in the process of cooperatively increased hyaluronic acid synthesis in pericardial cells induced by IGF-I and EGF. Genistein pretreatment inhibited IGF-I and EGF-induced hyaluronic acid synthesis and hyaluronic acid synthase activity (Figs 7 and 8). This, therefore, raised an intriguing possibility of the involvement of tyrosine kinase activation of signal transduction events in the stimulation of hyaluronic acid synthesis by IGF-I and EGF. This notion is supported by our findings that IGF-II, whose receptor lacks tyrosine-specific protein kinase in the cytoplasmic domain (Morgan et al. 1987), did not stimulate hyaluronic acid synthesis in this pericardial cell system (Honda et al. 1989).

We established a method to isolate and culture the mesothelial cells that line the surface of the rabbit pericardial cavity, using morphological (Kluge and Hovig, 1967), immunological (Sun et al. 1979; Schlegel et al. 1980) and biochemical criteria (Sato and Prescott, 1987; Ohashi et al. 1988). This culture system will be useful for the study of pericardial physiology at the biochemical or cell biological level.

This paper presents the first evidence that the combined use of IGF-I and EGF causes cooperative stimulation of hyaluronic acid synthesis in rabbit pericardial cells and that the stimulation mechanism may be involved in the tyrosine kinase-mediated transmembrane signaling process.

We thank Mr T. Okada of Tokyo College of Pharmacy for technical assistance in part of this work. This study was supported in part by a Grant-m-Aid from the Ministry of Education, Science and Culture of Japan (no. 02671014).

Akiyama
,
T.
,
Ishida
,
J.
,
Nakagawa
,
S.
,
Ogawara
,
H
,
Watanabe
,
S.
,
Itoh
,
N.
,
Shibuya
,
M.
and
Fukami
,
Y.
(
1987
).
Genistein, a specific inhibitor of tyrosine-specific protein kinases
.
J. biol. Chem
.
262
,
5592
5595
.
Carpenter
,
G.
(
1987
).
Receptors for epidermal growth factor and other polypeptide mitogens
.
A. Rev. Biochem
.
56
,
881
914
.
Holt
,
J. P.
(
1970
).
The normal pericardium
.
Am. J. Cardiol
.
26
,
455
465
.
Honda
,
A.
,
Iwai
,
T.
and
Mori
,
Y.
(
1989
).
Insulin-like growth factor I (IGF-I) enhances hyaluronic acid synthesis in rabbit pericardium
.
BiocmM. biophys. Acta
1014
,
305
312
.
Honda
,
A.
,
Ohashi
,
Y.
and
Mori
,
Y.
(
1986a
).
Hyaluronic acid in rabbit pericardial fluid and its production by pericardium
.
FEBS Lett
.
203
,
273
278
.
Honda
,
A.
,
Ohashi
,
Y.
and
Mori
,
Y.
(
1986b
).
Effect of high-molecular-weight hyaluronic acid on the viscosity of rabbit pericardial fluid measured with a cone and plate viscometer
.
Chem. Pharm. Bull
.
34
,
4844
4847
.
Iwama
,
M.
,
Honda
,
A.
,
Ohashi
,
Y.
,
Sakai
,
T.
and
Mori
,
Y
(
1985
).
Alterations in glycosaminoglycans of the aorta of vitamin E-deficient rats
.
Atherosclerosis
55
,
115
123
.
Kluge
,
T.
and
Hovig
,
T.
(
1967
).
The ultrastructure of human and rat pericardium. 1. Parietal and visceral mesothelium. 2. Intercellular spaces and junctions
.
Acta path, microbiol. scand
.
71
,
529
546
, 547-563.
Morgan
,
D. O.
,
Edman
,
J. C.
,
Standring
,
D. N.
,
Fried
,
V. A.
,
Smith
,
M. C.
,
Roth
,
R. A
and
Rutter
,
W. J.
(
1987
).
Insulin-like growth factor II receptor as a multi-functional binding protein
.
Nature
329
,
301
307
.
Morgan
,
D. O.
and
Roth
,
R. A.
(
1987
).
Acute insulin action requires insulin receptor kinase activity: introduction of an inhibitory monoclonal antibody into mammalian cells blocks the rapid effects of insulin
.
Proc. natn. Acad. Sci. U.S.A
.
84
,
41
45
.
Ohashi
,
Y.
,
Honda
,
A.
,
Iwai
,
T.
and
Mori
,
Y.
(
1988
).
Stimulatory effect of vanadate on hyaluronic acid synthesis in mesothelial cells from rabbit pericardium
.
Biochem. Int
.
16
,
293
302
.
Ohya
,
T.
and
Kaneko
,
Y.
(
1970
).
Novel hyaluronidase from Streptomyces
.
BiocmM. biophys. Acta
198
,
607
609
Sato
,
K.
and
Prescott
,
S. M.
(
1987
).
Culture of mesothelial cells from bovine pericardium and characterization of their arachidonate metabolism
.
BiocmM. biophys. Acta
930
,
283
296
.
Schlegel
,
R.
,
Banks-Schlegel
,
S
and
Pinkus
,
G. S.
(
1980
).
Immunohistochemical localization of keratin in normal human tissues
.
Lab. Invest
.
42
,
91
96
.
Sun
,
T.-T.
,
Shih
,
C.
and
Green
,
H.
(
1979
).
Keratin cytoskeletons in epithelial cells of internal organs
Proc. natn. Acad. Sci. U.S.A
.
76
,
2813
2817
.
Ullrich
,
A.
,
Gray
,
A
,
Tam
,
A W
,
Yang-Feng
,
T.
,
Tsubokawa
,
M
,
Collins
,
C.
,
Henzel
,
W.
,
Lebon
,
T.
,
Kathuria
,
S.
,
Chen
,
E.
,
Jacobs
,
S.
,
Francke
,
U.
,
Ramachandran
,
J.
and
Fugita-Yamaguchi
,
Y.
(
1986
).
Insuhn-hke growth factor I receptor primary structure: comparison with insulin receptor suggests structural determinants that define functional specifics
.
EMBO J
.
5
,
2503
2512
.