Hepatic arachidonic acid metabolism following in vitro perfusion and endotoxin administration

Prostaglandins Leukotrienes and Medicine 9:

363-371, 1982

HEPATIC ARACHIDONIC ACID METABOLISM FOLLOWING IN VITRO PERFUSION AND ENDOTOXIN ADMINISTRATION John T. Flynn Department of Physiology Jefferson Medical College Thomas Jefferson University Philadelphia, Pennsylvania 19107 U.S.A. ABSTRACT Bacterial endotoxins have been shown to stimulate prostanoid production &J vivo. This study ltompares the effect of endotoxin administered in vivo and E C arachidonic acid (AA) metabolism. * on hepatic with either buffer, buffer containing endotoxin, or buffer c plus meclofenamate. Liver slices were then incubated with and chromatographed. In buffer perfused livers, 72 + 8 percent of the post incubation was associated with arachidonic acid. This was in contrast to nonperfused, fresh liver slices where 98 _I 1 percent of the 14C label remained as arachidonic acid. Slices taken from perfused livers which received endotoxin & vitro had 72 f 9% of the label associated with AA. Meclofenamate treatment Eased this percentage to 97 + 1%. Fresh liver slices incubated with endotoxin in vitro for 10 minutes demonstrated no significant increase in AA metabolism -compared to fresh slices. These data demonstrate that fresh liver sIices metabolize very little exogenous AA. In contrast, the perfusion procedure alone is a significant stimulus for enhancing arachidonic acid metabolism. Subjecting liver tissue to endotoxin in vitro or in vivo for a one hour period is not a significant stimulus for enhancemetabolism. INTRODUCTION A number of investigators have documented increased eicosanoid concentrations in blood and lymph of intact animals subjected to circulatory shock and tissue injury. The precise mechanism by which this enhanced arachidonic metabolism occurs is unclear, but appears to be an interaction between several different factors. In our studies of the role of prostanoids in cellular injury, we have used both intact animal models and in vitro models to differentiate those variables which influence prostanoid prod= Endotoxin administration to conscious sheep (1) or anesthetized rabbits (2) resulted in significant increases in plasma and/or lymph concentrations of thromboxane B2, 6-keto pGF1 Q and PCF o which were associated with pulmonary hypertension, systemic hypotension In contrast, the in vitro administration of large quantities of and Teukopenia.

363

gram negative endotoxin to the perfusate of an isolated, perfused rabbit liver preparation had no significant effect on either cellular integrity or hepatic prostanoid production (3). The present study was undertaken to directly compare the effects of endotoxin administration in vivo versus in vitro on hepatic prostaglandin production from exogenouslyxistered arachidonic acid. In addition, this study documents an alteration in exogenous arachidonic acid metabolism which directly results from the in vitro perfusion of the isolated liver. MATERIALS AND METHODS Liver Perfusion. Livers were removed from pentobarbital anesthetized, heparinized rabbits and placed in a non-recirculating perfusion apparatus as previously described (4). The livers were perfused at 39 C with a modified KrebsHenseleit-bicarbonate buffer (pH 7.4) at a flow rate of between 1.5 and 2.0 ml/gram/min. The perfusate was equilibrated with 95% 02 - 5% C02. Following a 30 minute stabilization period, the livers were perfused for an additional 90 minute experimental period consisting of a 30 minute treatment segment and a 60 minute post-treatment segment. At he end of the 90 minute experimental In two period, liver slices were prepared for 14 C arachidonic acid incubation. additional studies, livers were directly removed from anesthetized animals and slices were immediately prepared, washed free of blood in K-H buffer and incubated. These are referred to as “non-perfused” slices. 14C Arachidonic Acid Metabolisa Arachidonic acid (Sigma) to yiell$ a final buffer concentration of 70 J.IM was combined with 0.008 rJCi of C arachidonic acid (New England Nuclear) and evaporated to dryness under 106% After solubilization with 0.2 ml of 0.1 M Na CO3 and neutralization with 0.2 Ni)* m of 0.1 M citric acid, the arachidonate was dilu ised in 2.0 ml of phosphate buffer at pH 7.4. Liver slices weighing approximately 175 mg were added. Following a 20 minute incubation at 37OC, the reaction was stopped with 4.0 ml of cold ethanol and the samples were homogenized with a Tekmar tissue grinder at full speed for 15 seconds. Samples were solvent extracted and chromatographed on Whatman LKS linear K silica gel thin layer chromatography plates in a solvent system composed of the or anic phase of ethyl acetate : acetic acid : iso-octane : Areas on the plate corres,$~n~c$i:,no,W,n,;s water (110:20:30:100, v/v Pv/v). prostanoid standards were scraped, eluted, and the associated DPM values were used to quantitated by liquid scintillation spectrfiicopy. C w ‘ch was associated with each calculate the percent of the recovered prostanoid fraction. All values are corrected for b C arachidonic acid impurity as assessed by parallel TLC of the labeled material. Experimental Protocol. Six different groups of livers were studied. Two groups of livers were not perfused. Of these groups, fresh slicq were removed from livers of control animals and incubated either directly with 4C arachidonic acid or pretreated in vitro for 10 minutes with 10 &ml endotoxin prior to the arachidonic acid incubation. In four other groups of livers, all were removed from the animals and perfused with K-H buffer for a total of 120 minutes. These livers were subjected to either vehicle, endotoxin in vivo (i.e., pretreatment of the intact animal with 1 mg/kg endotoxin one houxto removal and perfusion of the liver), endotoxin in vitro, or endotoxin in vitro plus meclofenamate at the concentrations and times as described in Tables 1 through 5.

364

Statistical

analysis

Values between groups for each eicosanoid were compared using Student% Y-test for unpaired data. A p value of less than 0.05 was accepted as connoting a significant difference between the means. RESULTS Comparisons of 1’C prostanoid and hydroxy-fatty acid production between fresh liver slices and slices taken from livers perfused with K-H buffer for a total of 2 hours arc s wn in Table 1. In slices. from non-perfused livers, 98.07 t 0.37 percent of the 5% C label was still associated with arachidonic acid after 20 minutes of incubation. In contrast, slices taken from perfused livers metabolized significantly more 14C AA into more polar compounds. Of the products formed, the fraction containing PCA2 and most of the lipoxygenase products (HETE’s and HPETE’s) represented the largest percentage of product. PCD and thromboxane B were also produced in significant quantities while PGE2, i GF2e and L-ketowere synthesized in smaller quantities. When comparing the ratio of P&F perf&&i divided by nonperfused value for each nostanoid, all compounds demonstrated a 10 to 20 fold increase in production of C labeled products by the slices from the perfused livers except for PGE2 The production of this prostanoid increased by 80 fold.

Percent of Rawvered

'@C Activity K-HLinr,nomin

!Ezzz bl0)

92.07t0.37

Sipificmce

c4

71.*2.29

vo.02

0.17to.03

3.0320.:2

pqo.02

0.29+06

2.2)?0.73

PO.02

O-OW.Of

3.17~O.m

p'O.01

O-20-

5.2%1.23

PO.02

o.r2$.09

6.9r+2.24

FO.05

0.6W.N

2.3Ot2.19

p'O.02

365

Table 2 compares I4 C eicosanoid production between slices prepared from perfused livers treated in vitro with either vehicle or 10 ug/ml endotoxin (E coli, 0127-B8). The magnitude and pattern of product formation are almost identical between the two groups of liver slices. Thus, 30 minutes of exposure to lOAg/ml endotoxin in vitro had no effect on hepatic prostanoid production from exogenously administered AA.

effectof Liver Slices. Al bufferfor120m~

mUmh)

w

Endotox&l Admhbtraon

V’ vu8

WIW priuwd

on

“C

in ??fbw-bralt,

pmtadd 8pratw

pmduction from hrochidmntc with Krobo-Hamekbl4ite

by

for 30 mimaoa

of Reco*rred

Percent Prostatwid

Arachidonic

&id

*)C Activity

K-H Perfumed c4nvol Live? b=l)

K-H Pafuad

71.46tt.29

7l.U+62

Si@ifiuMn

NS

x03+0.112

2.7720.76

NS

PGF-

2.63tO.73

2.42t0.71

NS

PCE2

3.17+0.1)8

3.17t1.03

NS

4.2521.23

4.07t1.33

NS

6.95t2.24

7.14t2.61

NS

8.30t2.49

8.6722.41

NS

6-Keto

PGFlalw

Thromboun

02

PCD2

PC+

HHT, 5-HETE.

IbHETE

$;fltare

the

moon t SEM for n obmrvltiau

in each group.

NS = rx)

significant

To compare in vitro versus in vivo effects of endotoxin on arachidonic acid metabolism, rabbits were given mr dose of E coli en&toxin (1 mg/k i.v.) The pfixktion one hour prior to sacrifice and remova M the live~$erfusion. of prostanoids by liver slices from these animals is shown in Table 3. When compared to values from K-H perfused control liver slices, there are no significant differences.

366

of flacwerod “C Activity

Percmt PrortuPid

Aruhidmic

acid

K-H Pafamed

K-HPUftWd

r)r YYLirn

plum Ln*iro h=8)

7 I A6f3.29

70.09f9.u

Sighificmce

NS

6-Kcto PGF ,aph

3.0k0.82

2.93f0.95

NS

PGF mph

2.6RO.73

2.My0.95

NS

PCL2

3.17~0.8a

2.8hO.91

NS

1.25?1.23

4.27+1.18

NS

PGD2

6.9k2.24

7.68:2.17

NS

PGA2, HHT, 5-HETE.

1.3Of2.19

9.38t2.76

NS

Thromhoun

e2

I2-HETE

To confirm enzymatic arachidonic acid metabolism in the perfused livers, livers were perfused with buffer containing 2 ~&/ml of sodkm meclefe@amatt. In addition, these slices received exposure to 10 &ml ertdototi for 36 mimrtes of the 126 minute perfusion period. These results are shown in Table 4. Mkclofenamate significaritly inhibited the perfusion induced increase in the

367

Inhibition of ‘bC Prortanoid Biosynthesis Cleclofenamate (m4o) Was added to the pbrfuutc

in Hepatic bt a

Percent K-H

6-Keto

Sodium

I

Meclofenamate.

‘bC Activity K-H Pcrfwed

Significmce

Plus Ewbtoxin

acid

7 .8?&.62

96.99t0.91

p 0.02

PCF,a,pha

2.77~0.76

0.26f0.08

pco.01

2.4220.71

0.33~0.01

p 0.02

3.17t1.03

O.36t0.1

b.O7?1.33

O.b3$

7.1922.61

0.7b$28

pco.05

8.6722.bl

0.89fO.37

p~o.02

PGF2alpha

PCE2

Thromboxane

82

PGD2

PCA2,

by

prior to the two hour (10 u&n1 x 120 ml/minute) between 10 and

of Recovered

Perfused

Plus Endotoxin

Arachidonic

Tissue

concentration of t&g/ml

perfusion period. Livers then received endotoxin in 60 minutu of perfusion.

HHT,

5-HETE.

I

II

ps 0.02

pco.05

I2-HETE

All values UC the mean t SEM for n observations

in each group.

A final comparison is made in Tabie 5 between twa pups of fresh liver The first pup of slices * slices directly inahtad with composed of slices which were feeshiy prepred from it&t iher lncubrBwl with 10 &ml endetoxin for 10 minu@s and Wn inct&M with 3 k AA fat 20

slices.

czampuwofP~~~from pcrfurdLivsrsIkesIieu&fqv~aEnbf3pxin. cndotoxin for IO mhwtea and than inabated

RadMabeM Aradtibnic Acid Between NanLiver rlkes were incubated with 10 &ml C arachidordc add as previously detcribed.

with

Pncent

Activity Significance

acid

91.0720.37

92.9gt3.03

PGF ,a,ph

0.17~0.03

0.51$0.23

NS

029~0.06

OAOtO. 18

NS

O.Ol?O.O I

0.52to.25

NS

0.3a~o.oa

0.7gt0.39

NS

0.02~0.09

1.17:0.65

NS

0.65t0.11

3.63t1.33

NS

PGF2&il8

PCE2

Tkomboxux

B2

PCD2

PCA2,

‘k

Non-perfused Plus End&win ~vitron=lO)

Arachidmic

6-Kcto

of Recovered

Nca-perfustd Control Liver (n.10)

Prostanoid

HHT.

5-HETE ,

NS

I2-HETE

All values we difference.

the me811 t SEM for n observations

in each group.

NS : no significant

DISCUSSION The isolated, perfused rabbit liver has been characterized as a stable model for investigating endogenous prostanoid production and associated blliogical changes (4, 5). Three hours of perfusion with Krebs-Henseleit-bicarbonate buffer There is neither results in no hemodynamic changes in the preparation. cytoplasmic or lysosomal enzyme release into the venous effluent, nor any increase in the rate of endogenous prostanoid production over the course of the experimental period. In contrast, the data from the present study suggest that the perfusion protocol has a significant effect on prostanoid production from exoKenously administered arachidonic acid. There was a 30% increase in 14C arachidonic acid metabolism in liver slices taken from perfused livers when compared to values obtained from fresh slices. Since the patterns of metabolite

369

production in the two groups of liver slices were very similar, it does not appear that the perfusion effect was via an action on cycle-oxygenase or enzymes further within the cascade. Rather, two hours of perfusion appears to significantly affect the ability of hepatic tissue to esterify the exogenously administered arachidonic acid into membrane phospholipids. Freshly prepared tissue slices from nonperfused livers were able to protect arachidonic acid from metabolism in this way. In these fresh tissues, over 98% of the arachidonate was unavailable to cycle-oxygenase or lipoxygenase enzyme systems. The fact that 30% more free arachidonic acid was available in the tissues from perfused livers was confirmed sodium meclofenamate to the perfusate. Under these conditions, 97% 2 tahdedilng C label remained as AA in tissue taken from K-H perfused livers. These findings clearly demonstrate the physiologic differences between arachidonic acid metabolism arising from endogenous versus exogenous substrate availability in very similar models. It is interesting to note that Morita and Murota (6) have suggested the presence of a heat-labile enzyme or co-factor which is present in This factor facilitates the incorporation of the cytosol of hepatocytes. arachidonic acid into various phospholipids. The loss of this type of material from the liver via the non-recirculating perfusion technique may explain the relative inability of tissue slices from perfused livers to efficiently esterify exogenous arachidonic acid. The effect of bacterial endotoxin on 14C arachidonic acid metabolism was investigated in liver slices obtained from both perfused and non-perfused Tissue from perfused livers subjected to a total dose of 36 mg of sources. endotoxin in vitro demonstrated no significant differences in either the pattern or magnitude of AA metabolism when compared to perfused controls. In addition, endotoxin added to freshly prepared slices from non-perfused livers had no significant effect on the metabolism of exogenously administered arachidonic acid. This lack of an endotoxin effect in vitro has been reported for several other tissues (7, 8). It seems clear from these types of studies that gram negative bacterial endotoxins do not directly result in arachidonic acid ,release from sequestered storage areas. It also appears that endotoxin does not directly affect cycle-oxygenase or lipoxygenase activity in liver tissue. Several investigators (7, 8) have also reported that the in vivo administration of endotoxin results in enhanced prostanoid production in vitro from tissues removed from the endotoxic animals. Others report a depressioncyclo-oxygenase activity in renal medulla in vitro following the in vivo administration of endotoxin (9). Results from the present study suggest that the administration of 1 mg/kg of endotoxin to rabbits 60 minutes prior to sacrifice has no effect on the metabolism of exogenous arachidonic acid in tissues secured from perfused livers. In contrast, Feuerstein and Ramwell (8) reported that the administration of endotoxin to rats in vivo resulted in a greater rate of endogenous prostanoid production by lung strips vitro, although the in vitro administration of endotoxin to control lung strips had no effect on prostanoidduction from endogenous substrate. Once again, these type of data demonstrate the differences in models based upon the source of the Endotoxin treatment in vitro does not appear to affect arachidonic substrate. The acid metabolism from either endogenous or exogenous substrate. administration of endotoxin in vivo appears to enhance the subsequent rate of endogenous arachidonic metin vitro but does not affect the rate of Thus, the effect of arachidonic metabolism in vitro from exogenous AA. endotoxin on stimulating the arachidonic acid cascade rests at the level of In this regard, Conde et al. (IO) have endogenous substrate availability. demonstrated that the in vivo administration of endotoxin to rats results in the release of arachidonic acidm mitochondrial phospholipids. This occurred at a 370

time when there was no parallel release of aithht stearic or llnoleic acids. The precise mechanism by which endotoxin increases endqpous arachidonate availability in vivo is unclear, but preliminary evidence (unpublished data) suggests the involvement of the complement system. ACKNOWLEDGEMENTS I wish to thank Mrs. Sultana Lulias for her excellent technical assistance during the course of this project. This work was supported by a grant from the National Institutes of Health, GM 28023. REFERENCES 1.

Demling RH, Smith M, Gunther R, Flynn JT, Gee MH. Pulmonary injury and prostaglandin production during endotoxemia in conscious sheep. Amer J Physiol 240:H348-H353, 1981.

2.

lipoxygenase inhibitors Flynn JT, Click AJ. Effects of arachidonate survival following endotoxemia in rabbits. Fed Proc 4 1: 1135, 1982.

3.

Flynn JT. Lack of a direct stimulatory effect of endotoxin on hepatic prostanoid synthesis. Abstracts, 4th Int Prostaglandin Conference, P Ramwell, B Samuelsson and R Paoletti, Eds. Raven Press, Washington, p. 36, 1979.

4.

Flynn JT. Influence of the arachidonic acid cascade on the in vitro hepatic response to hypoxia. Prostaglandins 17:39-52, 1979.

5.

Flynn JT. Effect of 2,4 dinitrophenol on the rate of hepatic prostaglandin production. Adv Shock Res 5:149-162, 1981.

6.

Morita I, Murota S. Prostaglandin-synthesizing Biochem 90:441-449, 1978.

7.

Herman AC, Vane JR. Endotoxin and production isolated rat jejunum. Influence of indomethacin. 213:328-329, 1975.

8.

Feuerstein N, Ramwell PW. In vivo and in vitro effects of endotoxin prostaglandin release from rat lung. Br J Pharmac 73:511-516, 1981.

9.

Bhattacherjee P, Phylactos A. Depression of prostaglandin synthetase activity in kidney medulla by Shigella endotoxin injected intravenously. Biochem Pharmacol 27:807-808, 1978.

10.

Conde C, Garcia-Barren0 P, Suarez A. Arachidonate release from rat liver mitochondria in endotoxin shock. FEBS Letters 112:89-91, 1980.

371

system in rat liver.

on

Eur J

of prostaglandins by the Arch Int Pharmacodyn on