The induction of glycogenolysis in the perfused liver by platelet activating factor is mediated by prostaglandin D2 from Kupffer cells

Vol. 157, No. 3,1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS Pages ]288-]295

December 30,1988

THE INDUCTION VATING

OF GLYCOGENOLYSIS

FACTOR IS MEDIATED

IN THE PERFUSED

BY PROSTAGLANDIN

LIVER BY PLATELET ACTID2 FROM K U P F F E R

CELLS

JOHAN KUIPER, YOLANDA B. DE RIJKE, FREEK J. ZIJLSTRA, ~ MARIEKE P. VAN WAAS AND THEO J.C. VAN BERKEL

Division

of Biopharmaceutics,

Laboratories,

Center for Bio-Pharmaceutical

Sciences,

Sylvius

University of Leiden, P.O. Box 9503, 2300 RA Leiden The Netherlands

Dept.

of Pharmacology,

Erasmus University Rotterdam,

3000 DR Rotterdam,

P.O.

Box 1738,

The Netherlands

Received October 20, 1988

Induction of glycogenolysis in the perfused liver by platelet activating factor (PAF) was blocked by the cylcooxygenase inhibitor indomethacin. 3Hlabeled PAF was shown to interact in the perfused liver primarily with Kupffer cells. The addition of PAF to Kupffer cells resulted in a dose-dependent stimulation of prostaglandin D2 (PGD=) production, which was identified as the main eicosanoid formed after PAF stimulation of the Kupffer cells. PGD2 was able to induce a dose-dependent stimulation of glycogenolysis both in the perfused liver and in isolated parenchymal cells. The time-dependency of the PGD= production and the glucose output by the perfused liver is consistent with a primary interaction of PAF with the Kupffer cells, followed by PGD2 formation, which subsequently stimulates glucose production in parenchymal cells. © 1988 Academic Press, Inc.

Platelet choline) (1,2),

activating

factor

(PAF,

l-0-alkyl-2-acetyl-sn-glycero-3-phospho-

is not only a potent mediator in inflammatory and allergic reactions

but PAF also exerts major effects on liver metabolism:

to a perfused

liver

Ca2+-flux

vasoconstriction

and

induced

a change

in oxygen

(3-6).

consumption,

Interestingly

PAF

addition of PAF glycogenolysis,

does

not

elicit

a

change in oxygen consumption and glycogenolysis

in isolated parenchymal cells,

although

still

the

isolated

increased

breakdown

genolytic

effect

increased

level liver

indicating

can

effects that

suggested

leads

to

of glycogenolysis

bromide,

reported

PAF was

which

perfused

metabolic

cells

could

respond

of phosphatidylinositol-4,5-bisphosphate

of

vasoconstriction,

parenchymal

however

of

that PAF

prostanoids

be

inhibited

the

are

by

from

inducers

0006-291X/88 $1.50 1288

effects

liver

in

effects

subsequently

indomethacin

of PAF

and

in

in the

of

(6). Recently

of glycogenolysis

of an

bromophenacyl

transduction

cells

an

(7). The glyco-

secondary

results

involved

parenchymal

are potent

Copyright © 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.

and

(5). The metabolic

prostanoids to

to result

hypoxia

to PAF with

the we

in isolated

Vol. 157, No. 3, 1988

parenchymal the liver,

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

cells,

due

their

Kupffer and endothelial

and prostaglandin

D2

(PGDz)

representing

of

the

(i0). We mediate

isolated se. We

50Z

therefore

PGD2

cells.

The

produced

suggest

that

the transduction parenchymal

cells

to activate

amount

of PAF.

in the PGD2

phosphorylase

cells are the main producers be identified of

on

via

its

present a primary

turn

cells and in the perfused the glycogenolytic

response

of a signal

from Kupffer

(metabolic

site) by PGD2.

cells

is

a

(8,9).

produced

eicosanoid

in

the

show

interaction to

induce

liver a glycogenolytic of the liver

liver

that PGDz might

work we

capable

In

of eicosanoids

liver

on the possibility

In the

liver

as the main

eicosanoids

our attention

effects

production

parenchymal

could

total

focussed

the metabolic

stimulates Kupffer

capacity

that

PAF

with

the

both

in

respon-

to PAF involves

(primary interaction

site)

to

MATERIALS AND METHODS

PGD2, collagenase type I and IV, bovine serum albumin (BSA, fraction V) were from Sigma. Indomethacin was from Merck, Sharp and Dome. PAF (l-0-alkyl2-acetyl-sn-glycero-3-phosphocholine) and ABTS were from Boehringer Mannheim. Nycodenz was from Nycomed. l-0-[3H]alkyl-2-acetyl-sn-glycero-3-phosphocholine (81 Ci/mmol), l-[m4C]-arachidonic acid and a radioimmunoassay specific for PGD2 was from Amersham. Male Wistar rats fed ad libitum, weighing 250-300 g were used. For liver perfusion experiments the portal vein of anaesthetized rats was cannulated and the liver was perfused with non-circulating Krebs-Ringer bicarbonate buffer, containing 1.3 mM CaCI2, at 35 ml/min. The perfusion buffer was saturated with 02/C0z (95Z/5Z), pH 7.4. Effluent was collected at 1 min time intervals and glucose was determined in the effluent by the glucoseoxidase ABTS method (ii). PGDz was determined with a specific radioimmunoassay. ~H-labeled PAF (1.10 -9 M) was perfused recirculating through the liver and ~Ster 5 passages through the liver a cell isolation procedure was started at ~°C using pronase and collagenase as described before (12). To determine the ~otal liver uptake a lobule was tied off before the start of the cell isolation procedure. After the isolation of parenchymal, endothelial and Kupffer c ~ i s the contribution of the various liver cell types to total liver uptake ~f PAJ could be calculated using the determined PAF-uptake (% of the added dose.104/per mg cell protein) and the established contribution of the various liver cell types to total liver protein (13). This method has proven to be accurate for the determination of the uptake of various ligands. Parenchymal cells were isolated by perfusion of the rat liver with collagenase (type IV, 0.05%) by the method of Seglen (14). Parenchymal cells were kept at 37°C under constant shaking (5 mg cell protein/ml) in Krebs-Ringer bicarbonate buffer saturated with 0=/C0= (95%/5%), pH 7.4. i0 min after the addition of PGD2 or PAF (2.10 -B M) cells suspensions were rapidly cooled and centrifuged (500 x g, 5 min) and glucose was determined in the supernatant by the glucose-oxidase ABTS method (ii). Kupffer cells were isolated by collagenase (0.05Z, type I) perfusion of the liver and subsequent counterflow centrifugation, exactly as described before (14). Kupffer cells were cultured on 24-wells culture dishes in RPMI 1640 containing i0% foetal calf serum, 2 mM L-glutamine, i00 iU penicillin/ml and i00 ~g streptomycin/ml. After 20 hrs culture Kupffer cells were challenged for one hour with various doses PAF in the presence of 0.25Z Bovine serum albumin (BSA) and subsequently the amount of PGD= was measured in the medium with a specific radioimmnunoassay. For identification of the various arachidonate metabolites produced by Kupffer cells after stimulation with PAF, Kupffer

1289

Vol. 157, No. 3, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

cells were loaded w i t h 1.25 ~Ci x4C-arachidonic acid and challenged for 10 min with 2.10 -7 M PAF. Arachidonic metabolites were separated by reverse-phase HPLC. Recovery and identification of the formed eicosanoids was performed by coelution with a tritiated standard eicosanoid. The formed eicosanoids were also identified by immunoassays as described in detail earlier (i0). PAF was dissolved in Krebs-Ringer bicarbonate buffer containing 0.25Z BSA in the appropriate concentration. Before any addition livers were preperfused for 45 min with 0.25Z BSA.

RESULTS

Fig

1

shows

induced

a

2.2

stimulation was (i0 ~M) was

that fold

the

addition

increase

observed

present

in

of

the

PAF

(2.10 -s M)

hepatic

to

glucose

the

output

perfused and

liver

a maximal

3 to 4 min after addition of PAF. W h e n indomethacin

in the perfusion medium,

the glycogenolytic

response

of

determined

by

the perfused liver to PAF (2.10 -8 M) was inhibited for 80%. The

primary

perfusion various cells this and

cell

analyze

of

the

site of PAF

liver with

liver cell types

and

14%

interaction

of

77~

type. to

the

to what

the

3H-labeled

Parenchymal liver

amount cells

and

of liver

subsequent

associated

and endothelial

association

extent PAF was able

the effect on eicosanoid production 2.10 -7 M PAF,

PAF

liver was

isolation

of

the

(table i). PAF associated primarily with the Kupffer

total

total

in the perfused

of

PAF,

cells

PAF was

contributed

respectively.

to activate Kupffer

recovered

In

in

only

9%

order

to

cells, we determined

(fig. 2) After challenge of the cells with

several arachidonate metabolites were produced and PGDz appeared

220 <¢ u) <

~

180

n I0

u~ 140 0 0

3 100

~'~

PAF ~ " ~

TIME (MIN) FiKure i. Influence of PAF on the glucose output by the perfused liver in the presence and absence of indomethacin. The liver was perfused for 45 min in the presence or absence of indomethacin (i0 ~M), after which PAF (2.10-SM) was added for 8 min. Glucose output was monitored in the effluent in 1 min time intervals. Data represent a typical experiment out of 3.

1290

Vol. 157, No. 3, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABLE 1 Uptake of 3H-labeled PAF by the various liver cell types Z of added dose /mg cell protein

Liver Parenchymal cells Endothelial cells Kupffer cells

contribution of the various cell types to total liver uptake

.10 -~

13.3 1.4 56.0 421.0

i00 Z 9.2% 13.6Z 77.3~

SH-labeled PAF was perfused through the liver and after 5 passages a liver cell isolation procedure at 8°C was started exactly as described before (12).

tO

be

the

arachidonic production output

eicosanoid

acid to

PAF

lower

in vit!Q.

genolysis

formed,

metabolites

by K u p f f e r

centrations cells

major

(fig

3)

cells was than

Addition

in a dose

~"

elicit

The

indicates found

i0 -a M

of PGD~

dependent

that

of

not

elicit

with

total

curve

stimulation

amount the

PGD~

of

the

PGDz

production

parenchymal

cells

effect

in i s o l a t e d

PAF con-

by

induced

at a PGDz

Kupffer glyco-

concentra-

(fig 4).

PAF

parenchymal

(2.10 -s

cells

15

z

O.J

z

Qw

OO

O 10

8s mg

--

g .

~

~J

-~-~

o

~'

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

,

.

.

.

.

.

.

.

r P A F (M)

Figure 2. Eicosanoid production by PAF-stimulated Kupffer cells. Kupffer cells (loaded with 1,25 ~Ci x4C-arachidonic acid) were challenged for i0 min with PAF (2.10-7M). Eicosanoids were separated and identified as previously described (i0). Data are mean ± S.D. from 3 experiments. Figure 3. Effect of PAP" on the PGD2 production by primary cultured Kupffer cells. Kupffer cells were incubated for 1 hr with increasing amounts of PAF. PGD~ was determined using a radioimmunoassay. Broken line represents PGDz production in unstimulated Kupffer cells.

1291

of

of

of 2.10 -7 M.

PGD=

at 2.10 -6 M PGD2

response

16

any

a maximal

effect

the

response

maximal

to isolated

way,

a glycogenolytic

66Z

dose

at a PAF c o n c e n t r a t i o n

did

tion of 5.10 -7 M and h a l f - m a x i m a l M) did not

representing

detected.

(not

Vol. 157, No. 3, 1988

BIOCHEMICAL A N D BIOPHYSICAL RESEARCH COMMUNICATIONS

220

75

,.J <~ (/) In = o

== o

180

I

50,

a. I2~ O ~ O O

"6

._g -~

2s

140

3 100 i

Q

0.5

i

i

i

1.0

1.5

2,0

;o

75

;o T I M E (MIN)

PGD 2 concentration (,~M)

Figure 4. Influence of PGD= on the glucose production by isolated parenchymal cells. PGD= was added to isolated parenchymal cells and after i0 min the glucose output was determined. Data express the stimulation over control and are mean ± S.D. of 3 experiments. Figure 5. Influence of PGD= on the glucose output by the perfused liver in the absence and presence of indomethacin. PGD= (10-~M) was added to the perfused liver in the presence or absence of indomethacin (i0 ~M). Glucose output was determined in the effluent (I min time interval). Data are from a typical experiment out of 3.

shown). In the perfused liver PGD= also induced a glucose output and a maximal effect is already reached shortly

(within 2 min) after addition

effect is not influenced by the presence of indomethacin. as

a mediator

induction

of

the

PAF

effect

of PGDz precedes

on glycogenolysis

the glucose

(fig 5). This

In order to function

it is necessary

production by the liver.

As

that

the

indicated

in fig 6 a sharp increase in PGD2 production by the perfused liver was observed directly after the addition of PAF and the increase of PGDz production in response to PAF preceded the increase in hepatic glucose output.

DISCUSSION

The

finding

was originally

that

stimulates

glucose

formation

reported by Buxton and coworkers

vasoconstriction increased

PAF

of the liver,

glycogenolysis

(5).

accompanied However

the

in

we

analyzed

that

the

Kupffer

cells

form

perfused

(3-5). It was

cell

the

major

type

liver

speculated that

by local hypoxia might primary

the interaction of the liver with PAF was not identified.

which

the

lead to an

responsible

for

In the present study

intrahepatic

cell

type,

interacts with PAF (more than 75% of the total liver-associated PAF was

recovered isolated

in liver

this

cell

cells

in

type). order

In

addition

to analyze

1292

we

performed

on a cellular

experiments

level

with

the metabolic

Vol.

1 5 7 , N o . 3, 1 9 8 8

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

250 90

ir

A _1

m

m

200

It

0

60

g I-el I.-

LU

(I) -% W_ I.Ll

0 W

co

D.

150

O O _1 (.-3

30

Z

c4 a (:3

a.

I I I 100 ~'~,.'X~.X'~.~'~ PA F ~ . ~

,

~o

~

~'o

&

TIME (MIN)

Figure 6. Influence of PAF on the PGD= and glucose production in the perfused liver. PAF (2.10-8M) was added to the perfused liver and PGD2 and glucose output was monitored in 1 min time intervals.

events during PAF interaction with the total liver. With isolated parenchymal cells

addition

therefore cells

we

of

PAF

did

determined

leading

not

lead

to what

to release

to

glucose

extent

of factors,

PAF

production

could

which might

(see

activate influence

also

isolated

7)

and

Kupffer

glycogenolysis

in

parenchymal cells. Our data indicate that PAF increases the production of PGD2 in Kupffer cells in a dose dependent way. PGDz, which was identified before as the main eicosanoid ionophore Kupffer

produced

by the Kupffer

cells after

(i0), is now also identified as the main eicosanoid

cells

in response

to PAF stimulation.

The capacity

duce the glycogenolytic effect of PAF to the parenchymal fact

that PGD=

induced

perfused liver. by PGD2,

stimulation with Ca-

an immediate

stimulation

produced by the

of PGD= to trans-

cells is shown by the

of the glycogenolysis

in the

Indomethacin did not affect the stimulation of glycogenolysis

indicating that indomethacin does not affect the ability of the liver

to perform glycogenolysis.

The ability of PGDz to stimulate glycogenolysis

in

the perfused liver could be ascribed to a direct dose-dependent effect of PGDz on glycogenolysis

in isolated

parenchymal

cells.

This may

indicate

that PGDz

can regulate the glucose output by the liver parenchymal cells as a consequence of production PGDz production

in the Kupffer

cells in response to PAF.

and the induction of glycogenolysis

1293

The time course of

in response

to PAF is in

Vol. 157, No, 3, 1988

agreement with

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

such an

intermediary

role

of PGDm.

Addition

of PAF

to

the

perfused liver led to a direct stimulation of the PGDz production (peak after 1-2 min), while the glucose output by the perfused liver showed a peak at 3-4 min after PAF addition. It should be noted that the amount of PGD~ measured in the perfusate may be

underestimated because PGD2 is rapidly converted in the

liver by parenchymal cells (16). Our results

suggest that the induction of hepatic glycogenolysis by PAF

may be explained by a direct action of PGD= on parenchymal cells rather than from haemodynamic effects, as suggested before (5). The present data indicate that

PGDe,

produced

by

PAF-stimulated Kupffer

cells

is

able

to

stimulate

directly glycogenolysis in isolated parenchymal cells. Such a direct effect is also consistent with the recent finding (6) that PAF mobilizes the same Ca z÷ pool

in the

liver as phenylephrine, which is known to influence Ca =÷ flux

inside parenchymal

cells

(17). Furthermore vasoconstriction was

reported to

play a minor role in the stimulation of glycogenolysis in the perfused liver by various types of eicosanoids (18). In conclusion our data are consistent with the following mechanism for the

induction of glycogenolysis

in the liver by PAF:

PAF interacts in the

liver primarily with the Kupffer cells and induces a release of eicosanoids, of which PGD= is the main eicosanoid. After release from Kupffer cells PGD2 stimulates subsequently glycogenolysis in parenchymal cells. A major role for PGDz is in agreement with our recent finding that PGDe among several eicosanoids tested was the most potent inducer of glycogenolysis in isolated parenchymal liver cells

(8). Our present data with PAF are therefore consistent

with a function of PGDz as metabolic messenger between the various cell types inside the liver.

REFERENCES

i. Sirganian, R.P. and Osier, A.G. (1971) J. Immunol. 106, 1244-1251. 2. Beneviste, J., Henson, P.M. and Cockvane, C.G. (1972) J. Expo Med. 136, 1356-1377 3. Shukla, S.D., Buxton, D.B., Olson, M.S. and Hanahan, D.J. (1983) J. Biol. Chem. 258, 10212-10214. 4. Buxton, D.B. Shukla, 8.D., Hanahan, D.J. and Olson, M.S. (1986) J. Biol. Chem. 259, 1468-1471. 5. Buxton, D.B., Fisher, R.A., Hanahan, D.J. and Olson, M.S. (1986) J. Biol. Chem. 261, 644-649. 6. Altin, J.G., Dieter, P. and Bygrave, F.L. (1987) Biochem. J. 245, 145-150. 7. Fisher, R.A., Shukla, S.D., Debuysere, M.S., Hanahan, D.S. and 01son, M.S. (1984) J. Biol. Chem. 259, 8685-8688 8. Casteleijn, E., Kuiper, J., Van Rooij, H.C.J., Kamps, J.A.A.M., Koster, J.F. and Van Berkel, Th.J.C. (1988) J. Biol. Chem. 263, 2699-2703. 9. Casteleijn, E., Kuiper, J., Van Rooij, H.C.J., Koster, J.F. and Van Berkel, Th.J.C. (1988) Biochem. J. 252, 601-605. i0. Kuiper, J., Zijlstra, F.J., Kamps, J.A.A.M. and Van Berkel, Th.J.C. (1988) Biochim. Biophys. Acta 959, 143-152. ii. Werner, W., Rey, H.G. and Wielingen, H. (1970) Zeitschr. Anal. Chem. 252, 224-228.

1294

Vol. 157, No. 3, 1988

BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

12. Nagelkerke, J.F., Barto, K.P. and Van Berkel, Th.J.C. (1983) J. Biol. Chem. 258, 12221-12227. 13. Blouin, A., Bolender, R.P. and Weibel, E.R. (1977) J. Cell. Biol. 72, 441455. 14. Seglen, P.O. (1976) Methods Cell. Biol. 13, 29-83 15. Fisher, R.A., Kumar, R., Hanahan, D.J. and Olson, M.S. (1986) J. Biol. Chem. 261, 8817-8823. 16. Pugliere, G., Spokas, E.G., Marcinkiewicz, E. and Wong, P.Y.-K. (1985) J. Biol. Chem. 260, 14621-14625. 17. Reinhart, P.H., Taylor, W.M. and Bygrave, F.L. (1982) Biochem. J. 108, 619-630. 18. Iwai, M. and Jungermann, K. (1988) Biochem. Biophys. Res. Commun. 151, 283-290.

1295