Neutrophils contribute to TNF induced myocardial tolerance to ischaemia

J Mol Cell Cardiol

24, 485-495 (1992)

Neutrophils

James Carl

Contribute

to TNF Induced Ischaemia

Myocardial

Tolerance

to

M. Brown, Benjamin 0. Anderson, John E. Repine, Paul F. Shanley, W. White, Michael A. Grosso, Anirban Banerjee, Denis D. Bensard and Alden H. Harken

The Departments of Surgery and Medicine and the Webb- Waring Lung Institute, University of Colarado Health Sciences Center, Denver, Colorado 80262, USA (Received 6 November 1990, accepted in revisedform 16 December 1991) BROWN, B. 0. ANDERSON, J. E. REPINE, P. F. SHANLEY. C. W. WHITE, M. A. GROSSO. A. BANER.JEE,, D. D. BENSARD AND A. H. HARKEN. Neutrophils Contribute to TNF Induced Myocardial Tolerance to Ischaemia. Journal of Molecular and Cellular Cardiology (1992) 24, 485-495. Sublethal endotoxin (ETX) pretreatment of rats induces protection from cardiac ischaemia-reperfusion injury. This protective state is associated with increased endogenous myocardial cataiase activity. Since turnour necrosis factor (TNF) is one mediator of ETX effects, we hypothesized that (TNF) pretreatment of the rat (30pg/kg ip) 36 h prior to cardiac ischaemiareperfusion could induce myocardial protection. We found that TNF administration increased both myorardial tolerance to ischaemia reperfusion injury (modified Langendorff, buffer perfusion, global, normothermir ischaemia) and myocardial catalase activity at 36 h. Moreover, we found that 6 h after TNF administration, myocardial hydrogen peroxide (H,O,, assessed by aminotriazole-H202 inactivation of catalase) and myocardial neutrophil accumulation (assessed by histology) were both increased. When neutrophil function was inhibited either by neutrophil depletion (vinblastine) or by ibuprofen treatments of the rat before TNF, the protection previously apparent at 36 h was blocked. We conclude that TNF can induce myocardial resistance to ischaemia reperfusion injury. This protection is related to prior tissue neutrophil accumulation and concomitant increases in H,O, levels.

J, M.

KEY WORDS:

Cytokines;

Toxic

oxygen

metabolites;

Antioxidant

Introduction Toxic oxygen metabolites (TOM) have been shown to cause reperfusion injury of postischaemic myocardium [I]. Neutrophils have also been shown to contribute to postischaemic myocardial necrosis and dysfunction [2, 31. To date, protection from TOM or neutrophil-mediated injury has for the most part been induced by exogenous therapies such as superoxide dismutase (SOD), cataneutrophil depletion, and lase, allopurinol, anti-neutrophil monoclonal antibodies [1, 3, 41. Endotoxin (ETX) can induce endogenous resistance to numerous insults including haemorrhagic shock and radiation injury [5, (j]. Endotoxin, tumour necrosis factor (TNF) and interleukin-1 (IL-l) can increase in vitro antioxidant systems such as managanous SOD [ 7]. We have previously shown that IL-1 Address Colorado

for correspondence: USA.

James

M. Brown,

Department

enzymes.

pretreatment of the rat (30pg/kg) resulted in protection from ischaemia reperfusion injury [8]. This state of protection was characterized by increased myocardial glucose-6-phosphate dehydrogenase (G-GPDH) activity. Similarly, we have shown that endotoxin pretreatment in sublethal amounts (500pg/kg) was related to in vivo increases in myocardial catalase activity and associated protection from isolated heart ischaemia-reperfusion injury [ 91. Despite these observations, the mechanism by which endotoxin and cytokines induce protection from subsequent ischaemia reperfusion injury remains unknown. In the present investigation, we hypothesized that TNF pretreatment of the rat, like ETX and IL-l would result in a protective state and that this protection was related to earlier myocardial oxidant stress and neutrophi1 accumulation followed by subsequent of Surgery,

Box C-305,

4200 E. 9th Avenue,

Denver,

80262,

OOZZ-2828/92/050485

+ 11 $03.00/O

(c)l 1992 Academic

Prec\

Limitrd

J. M. Brown

486

alteration in myocardial catalase activity. To address this hypothesis we employed a standard Langendorff buffer perfused rat heart model of global ischaemia [ 11, which provides a blood free system for assessing ischaemiareperfusion injury.

Methods Reagents and treatments All chemicals for buffers were from Sigma Chemical Company, St. Louis, unless otherwise specified. Vinblastine (0.75 mg/kg iv), was given to rats 4 days prior to TNF treatment or 4 days prior to isolated heart preparation and resulted in white cell depletions which were similar to published results [IO]. This dose decreased polymorphonuclear cells to 6% of control values (1008 k 251 vs 63 f. 24, WO.05) while it decreased total white cell (7.7 k 1.8 vs 4.1 +0.5), mononuclear (187 + 70 vs 95 k 28), lymphocyte (6404 k 1400 vs 3863 + 448), and platelet counts (993 * 125 vs 634 + 48) to greater than 50% of control values (all P: no difference). Recombinant human tumour necrosis factor alpha (TNF) was used in this study. This TNF, having an activity of 24 x lo6 Ulmg, was found to have less than 50 ng/mg endotoxin contamination by the limulus amebocyte lysate assay. TNF was prepared in phosphate buffered (5 mM, pH = 7.4) saline solution (PBSS) containing 0.5 mg/cc of endotoxin-free bovine serum albumin. TNF was injected intraperitoneally into rats (3Opg/ip). Comparable volumes of PBSS with albumin (vehicle) were injected into control rats without TNF. TNF was

p++-rqts Lquorontlne

-4doYs 1 Vinblostine treotmerj,t

v

or

-I h

0

lbuprofen ?reo tment

TNF

et al. supplied by Abla Greasy, Cetus Corporation, California. Ibuprofen was Emeryville, administered in two sequential doses, the first (30mg/kgiv) being given 1 h prior to TNF administration and the second (15 mg/kg ip) being given 6 h later. Ibuprofen was provided by the Upjohn Corporation. Control experiments were also performed in which vinblastine or ibuprofen were each given 30 h after TNF administration. (For general experimental outline see Fig. 1).

Isolated heart preparation Healthy male Sprague-Dawley rats were fed a standard diet and acclimated for 2 weeks in a quiet quarantine room. Rats were anaesthetized (sodium pentobarbital 60 mg/kg ip) and heparinized (500 U. inferior vena cava). Hearts were rapidly excised, placed in PBSS (5”C), and mounted on a modified Langendorff apparatus and perfused (70 mmHg) with Krebs Henseleit solution (37’) at the aortic root [I]. Hearts were subjected to 20 min of global normothermic ischaemia and 40 min of reperfusion. Ventricular function was assessed using a left intraventricular water filled latex balloon, a pressure transducer, and a Gould recorder (Houston, Texas). Left ventricular end diastolic pressure (LVEDP) was set at 6mmHg by adjusting the volume of the balloon. Developed pressure (DP, the difference between peak systolic pressure and LVEDP), contractility ( + dP/dt) and relaxation rate (-dI’/dt) were recorded continuously. In non-ischaemic control experiments (n = 4); DP, + dP/dt, - dPldt and LVEDP

+5h

+6h

Ibuprofen or vmblostme (controls) t30 h

Horvest heart: Hz02 levels PMN counts by i Histology

“early”

+36h

Isolate and perfuse heort: Function ofter lschoemio enzymes I Antioxidant H202 detoxificotlon “late”

FIGURE 1. General experimental design. After 2 weeks of quarantine rats were treated with TNF or vehicle. effects studied were myocardial neutrophil accumulation or H,O, levels in viva. Late effects studied were isolated perfused heart function after ischaemia-reperfusion, myocardial antioxidant enzyme activities. or myocardial detoxification ability PX vim. Interventions were vinblastine (neutrophil depletion) or ibuprofen.

Early buffer H,O,

TNF:

changed by less than 6% from values over 9Omin of perfusion.

.tfvocardial

H,O,

detoxification ex TNF pretreatment

Protective

their

vivo 36h

initial

after

Myocardial detoxification of hydrogen peroxide was assessed by measuring the aminotriaLole HIOZ-dependent inactivation of catalase. i\minotriazole inactivates catalase in the presence of H,O,. This assay indirectly measured tissue H,O,, based on the assumption that a decrease in myocardial catalase with added aminotriazole is proportional to the amount of H,O, present in the tissue [II]. Howcvrr by 36 h in this study catalase activities between groups were not the same, i.e., the TNF group had more catalase. ‘l’hervfore H,O, levels could not be estimated and the H,O? index was not valid. Instead, I.atalase inactivation was used as a reflection of mvocardial H,O? detoxifying ability. X-amino-l ,2,4-triazole (AMT) was dissolved in PBSS and infused (Harvard Pump) into the isolatc,d heart preparation at 8 mg/min (mean c’oronary llow = 20 cc/min) for 15 min prior to ischacmia and for 15 min after ischaemia. At thr end of reperfusion, hearts were rapidly removt,d from the apparatus and freeze clamped in liquid nitrogen. The hearts were weighed. ground under liquid nitrogen, and homogenized (high speed, 15 s) in 50mM potassium phosphate buffer (pH = 7.8, I I. 1 1n.v EDTA). Supernatants were collected lifter cttntrifugation (20 000 g for 10 min) and were assayed for catalase activity [I]. Myocardial H,O, detoxification for control and TNF pretreated rats was calculated by subtracting the amount of remaining catalase activity from the mean non-AMT treated catalase L alries.

Myocardial

enzyme

assays

Hearts were excised 36 h after treatment, perfused blood-free on the Langendorff apparatus, weighed, and immediately homogenizcd in potassium phosphate buffer (5Om~, pH = 7.8, 0.1 mM EDTA). Aliquots of homogenates were centrifuged (20 000 g for IO min) and supernatants were analyzed for protein, catalase, superoxide dismutase (SOD), glucose-6-phosphate dehydrogenase

Mechanism

487

(G-6PD), glutathione peroxidase, and glutathione reductase activities [12, 1.71. In addition, myocardial catalase activity was measured in hearts from rats that wtre prctreated with TNF 2 h before.

Myocardial

H,O? levels (HpO, index) after TNF Treatment

in viva fi h

AMT was dissolved in 5mM PBSS and administered (300 mg/kg ip) 4.5 h after TNF or vehicle injection. Rats were anaesthetized 90 min after AMT administration, and hearts were rapidly excised, perfused blood-free. and freeze clamped in liquid nitrogen. After hearts were weighed and ground under liquid nitrogen, homogenized, and centrifuged, myocardial catalase activities were assayed. The in vivo myocardial hydrogen peroxide index was calculated as follows: H,O,

index = 1 -

remaining baseline

CAT activity CAT activit)

Baseline catalase (CAT) activity was 907 U/g myocardial tissue (n = 12). The AMT method for in vivo H,O, assessment correlates with increases in oxidized glutathione [a].

Myocardial

neutrophil

accumulation treatment

after

TNF

Animals were anaesthetized and their hearts were excised, 3, 6, 12, and 36h after TNF or vehicle pretreatment. Cross sections were made through the midportion of each heart and the tissues were immediately immersion-fixed in 10% buffered formalin (Fisher Scientific, Fairlawn, New Jersey). The sections were embedded in glycol methacrylate, sectioned (lpm), and stained with naphthol-as-d chloroacetate esterasc polymorphonuclear cell stain. Five random high power (40x) fields (HPF) were taken from right and left ventricles and septum (total 15 HPF). The total number of myocardial neutrophils were counted by a pathologist in a blinded fashion.

Statistical

Data

were

compared

analysi.r

using

non-paired

J. M. Brown

488

Student’s t-tests. Data with more than two using analysis of groups were compared variance (ANOVA) followed by Bonferroni modified t-tests. Nine-five percent confidence limits (WO.05) were considered significant.

et al.

treated groups. Isolated heart ventricular functional including DP, + dPldt, values - dPJdt, and LVEDP were the same (P>O.O5) in all groups after equilibration prior to ischaemia.

Results

Late (36h) ejects of TNF

supernatant protein analysis Lowry (0.45 kO.05 vs 0.51kO.04, in mg/g wet weight), wet to dry ratios (4.78kO.73 vs 5.41 f 0.81), and heart weights (1.26 f 0.8 vs 1.31 f 0.9g of wet weight) were the same (PcO.05) between control and TNF pre-

2 E I--Inn, E a.-

-

Hearts which were subjected to 20 min of ischaemia and 40 min of reperfusion had decreased (WO.05) values for DP when compared to hearts prior to ischaemia (Figs 2 and 3, Table 1). By contrast, hearts from rats pre-

1

l

D

$

75

2 h $

50-

s z %

25-

!! 3 .u EL

0

7

Control

f

TNF 36 h

TNF2h

FIGURE 2. Ventricular developed pressure in isolated rat hearts. Compared to pre-ischaemia values (white bars) which were the same (P>O.O5) across all groups, control hearts (from vehicle injected rats) had decreased (P
600

0

Perfused Ih No lschoemio

TNF 36 h Control Hschoemia-reperfusiond

FIGURE 3 Myocardial hydrogen peroxide detoxification by catalase. Aminotriazole inactivation of catalase in the presence of hydrogen peroxide indicated that hearts which were subjected to ischaemia reperfusion injury had incresed (PcO.05) detoxification by catalase of hydrogen peroxide compared to hearts which were perfused without ischaemia. In contrast, turnour necrosis factor pretreatment of rats resulted in more hydrogen peroxide detoxification by myocardial catalase compared to non-ischaemic or ischaemic control hearts. Values are the mean + s E M of the number shown (*P
TNF:

TABLE

1.

Control TNF 36 h *f-O.05 Units:

Protective

Mechanism

489

Myocardial enzyme activities. Myocardial antioxidant enzyme activities. Treatment of rats with TNF 36 h prior to isolated heart preparation and measurement of antioxidant enzyme activities with TNF resulted in increased (WO.05) myocardial catalase activity (U/mg of supernatant protein). In contrast, superoxide dismutase (SOD), glucose-6-phosphate dehydrogenase (G-GPD), glutathione peroxidase, and glutathione reductasr werr not (P>O.O5) increased by TNF Glutathione peroxidase

Catalazse

SOD

G-6PD

1897 k 44 2910+84*

1589 k 18 1615k30

0.31 f 0.04 0.26 + 0.03

vs control, N_>5 U/mg supernatant

Glutathione wductast:

15.1 f 0.5 14.5 + 0.9

1.16iO.2 0.90 i 0.”

all groups

protein

treated with TNF 36 h before had increased (P~0.05) DP values after ischaemia when compared to hearts from control rats. Hearts from rats which were pretreated with TNF 2 h before had the same (P >0.05) DP values after ischaemia when compared to hearts from control rats. Hearts from rats pretreated with TNF 36h before had increased (WO.05) myocardial catalase activities (Table 1) but did not have increased (P>O.O5) myocardial SOD, G-GPD, glutathione peroxidase, or glutathione reductase activities. Hearts from rats pretreated with TNF 2 h before had the same (P>O.O5) myocardial catalase activities when

compared to hearts from control rats (1929 + 132 vs 1706 + 19 U/mg). Hearts from rats pretreated with vehicle (saline and albumin = control) 36 h before had increased (PcO.05) hydrogen peroxide (H202) detoxification by catalase during isolated heart ischaemia reperfusion when compared to hearts which were isolated and perfused but not subjected to ischaemia. In hearts from rats pretreated with contrast, TNF 36h before had increased (WO.05) hydrogen peroxide detoxification by catalase when compared to untreated hearts subjected to ischaemia reperfusion.

I.“” L

x aI

E

0.75

G ; 0.50 0 ‘0 g 0.25 T2 0.00

0

I

:::::::::: ::::x:::: :::::::::: :::::::x: :::::::l:i :::::::::: ~.S~..\>.> Is$p;x__-. :::::::::: \\%~\.%...

-______

:::::::::: :q:::::::> -~$
TNF6h

Control

FIGURE 4. Myocardial hydrogen peroxide levels and myocardial hydrogen peroxide index was increased (P (0.05) was increased (PCO.05) compared to control (vehicle treated) shown (*P
PMN and heart

accumulation. myocardial values.

Values

Early (tih) polymorphonuclear are the mean

after

TNF treatment. cell accumulation

f s E M of the number

490

J. M. TABLE

2.

Brown

et al.

Myocardial neutrophil accumulation after tumour necrosis factor treatment. Myocardial neutrophil accumulation after TNF. TNF treatment of rats resulted in increased (PO.O5) by 36h N

Control 3 h after TNF 6 h after TNF 12 h after TNF 36 h after TNF

4 4 3 4 4

*P
(6 h) e$‘ects of TNF

neutrophils 5+2 12+ 1* 28 f 4’ 16 f 3* 2*1

vs 3h

or

12h.

with TNF 6 h before had increased myocardial neutrophil accumulation when compared to hearts from control animals. Myocardial neutrophil accumulation was measured in hearts from rats pretreated with TNF 3, 6, 12, and 36h before. Peak neutro-

treated

Hearts from rats pretreated with TNF 6 h before had increased (P
No.

(a) IO01 + I, j

5

5

~

I

I

2250

0

Control

TNF 36 h

TNF 36 h vmblastlne pretreated

TNF 36 h Ibuprofen pretreored

FIGURE 5. Ventricular function after ischaemia and myocardial cat&se activity 36 h after TNF. (Pre-ischaemic values are white bars.) Compared to control heart values (from vehicle injected rats), hearts treated with TNF had increased (P-=0.05) ventricular function after ischaemia and reperfusion in addition to increased (P
TNF: I‘ABl,E

3.

Protective

Mechanism

49 1

Ventricular function after ischaemia reperfusion. Ventricular function after ischarrnia and reperfusion in hearts from rats which were treated with TNF, neutrophil depleted (\‘BI,) prior to TNF, treated with ibuprofen prior to TNF, VBL alone (4 days), ibuproten alonr (36 h), VBL or ibuprofen 30 h after TNF. TNF improved (KO.05) a II ventricular functional parameters after ischaemia reperfusion injury compared to control. Neutrophil depletion and ibuprofen pretreatment of rats prior to TNF prevented (PO.O5) values tar ventricular function after ischaemia reperfusion injury compared to control hearts and givmg vinblastine after TNF or ibuprofen after TNF resulted in the same protective ctkcta as TTNF alone (P>O.O5) N

C:ontrol TNF VBL-4 days/TNF 36 h Ibuprofen/TNF 36 h VBL 4 days Ibuprofen 36 h VBL 30 h after TNF Ibuprofen 30 h after TNF

DP

12 7

5 8 5 3 3 4

*P
dP/dl=

44.7 76.7 61.2 54.0 42.6 43.7 78.0+ 87.2

+ dP/dt

f + + k A +

3.0 2.2* 5.1’ 2.2’ 3.9 4.8 1.1* + 5.8*

mmHg/s

- dP/dt

16.4 i 1.4 3 1 ,3 i 1 .5 *

22.2 f 3.0’

11.2 “0.0

f 0.8 + l.“* i 1.2’ f 0.b’

21.1

k 1.0’

14.1 15.3

15.0 17.0?

?r 3.7 1.5

10.0 * 2.1 1 1 (1 + 1 ,5

28.6

* 1 .H* 1.7*

?0.7 It 0.3* 20.7 + 0.9*

31.oir

1.\-EI)l’

x lo-’

2 E

0.75

4 ;

0.50

n 6 $

0.25

s”

Control

TNF 6 h

TNF6h vlnblostlne pretreoted

TNF6h Ibuprofen pretreated

FIGURE 6. Myocardial hydrogen peroxide concentrations and PMN accumulation in neutrophil depleted and ibuprofen pretreated rat hearts. TNF alone increased myocardial hydrogen peroxide levels and PMN acrumulation (P
J. M. Brown

492

phi1 accumulation was observed at 6h (PO.O5). Histologic assessment of hearts from TNF pretreated and control rats revealed no changes in myocardial morphology. Furthermore, it appeared that the neutrophils were localized to capillaries and with crystalloid perfusion (70mmHg) at the aortic root could be washed out of myocardium such that neutrophil counts were the same as control (data not shown). However, the exact location of the neutrophils in myocardium was not determined in this study.

E$ects

of vinblastine

neutrophil depletion and ibuprofen pretreatment

Hearts from rats pretreated with TNF 36 h before had increased (PcO.05) cardiac function after ischaemia reperfusion when compared to hearts from control animals (Figs 5 and 6, Table 3). By contrast, hearts from rats pretreated with TNF following neutrophil depletion with vinblastine or ibuprofen administration had increased (PcO.05) recovery of function after ischaemia reperfusion when compared to hearts from control animals but had less (PO.O5) myocardial catalase activities when compared to hearts from control animals and had less (PcO.05) myocardial catalase activities when compared to hearts from rats pretreated with TNF. Vinblastine and ibuprofen treatment alone appeared to have no deleterious effects on myocardial function (Table 3). Hearts from rats treated with vinblastine 4 days prior to heart isolation and from rats given ibuprofen 36h before had the same (P>O.O5) preischaemic values for DP, +dPldt, -dPldt and LVEDP (data not shown) and the same (P>O.O5) values for ventricular function after ischaemia reperfusion injury (Table 3). Furthermore, hearts from rats given vinblastine

et al. 30h after TNF pretreatment 6 h prior to ischaemia reperfusion, which would not allow time for neutrophil depletion to occur [4] had the same (P>O.O5) protection from ischaemia reperfusion as hearts from rats given TNF alone. Similarly, hearts from rats given ibuprofen 30 h after TNF had the same (PBO.05) function after ischaemia as hearts from rats treated with TNF alone. TNF treatment of neutrophil depleted rats or ibuprofen treated rats resulted in the same (P > 0.05) myocardial hydrogen peroxide levels 6 h after TNF administration compared to hearts from control rats not given TNF. TNF treatment of neutrophil depleted rats resulted in the same (PBO.05) myocardial neutrophil accumulation 6 h after treatment compared to control rats not given TNF. However, ibuprofen treatment of rats only partially prevented neutrophil accumulation 6 h after TNF administration when compared to control rats not given TNF (PcO.05 vs control or TNF treated).

Discussion In this study we have presented findings at various times after TNF treatment of the intact rat involving the heart in vivo and the isolated buffer perfused heart ex vivo. Our cardiac injury model was one of global ischaemia at normothermia and was a system devoid of cellular components [I]. We found that TNF pretreatment of the rat in sublethal quantities was associated with protection from isolated heart myocardial ischaemia reperfusion injury 36 h later. Concomitantly, myocardial catalase activity was increased. Previous evidence suggests that manganous SOD activity is increased in vitro by endotoxin, TNF, or IL-1 treatment of endothelial cells [o]. Although we found no difference in total myocardial SOD activity, we did not measure SOD isoenzymes or subcellular fraction enzyme activities which could have been increased. No protection or increased catalase activity was noted 2 h after TNF, indicating the process of protection was not related to early more direct effects of TNF but rather appeared to involve an induction over a longer time. These findings extend our previous work in which endotoxin, IL-l, or TNF with IL-l pretreatments have caused protection associated with increased antioxi-

TNF: Protective dant enzymes in both myocardial ischaemia reperfusion and hyperoxic lung injuries [8, 9,

15]* Previous evidence indicates that hydrogen peroxide in rat myocardium is detoxified both by the glutathione pathway and by catalase [ 61. Augmented catalase activity was inferred in this study both by increased myocardial homogenate catalase activity and increased catalase detoxification of hydrogen peroxide by isolated hearts. As opposed to the situation 6 h after TNF when we were interested in myocardial H202 levels; at 36 h after TNF we knew the catalase activities of TNF and control hearts were different. Hence we could not use the H202 index; rather we only could use units of catalase inactivation as an indirect indicator of the relative amounts of catalase-H202 interaction in the isolated heart. Though catalase alone (and not SOD) infused exogenously had afforded protection from ischaemic injury in the isolated rabbit heart [17], the usual pattern in both large animal and small animal experiments has been protection from infusion of SOD alone or catalase in addition to SOD [Z, 3, Z7]. In this regard our results are somewhat surprising. Nonetheless, the pattern of endogenous myocardial catalase elevation coincidentally appearing with a protected state has been observed before [9, 181 and so increased myocardial catalase appeared to be one mechanism by which myocardial tolerance to ischaemia reperfusion injury was promoted. It is not known whether our measures represent stimulated catalase activity or increased de nova catalase molecule synthesis. Knowing that our isolated heart ischaemiareperfusion injury is caused in part by oxygen metabolite[s], we accordingly hypothesized that the protection and increased catalase at 36 h after TNF were a result of an earlier oxidant stimulus. We found that in vivo pretreatment was associated with increased myocardial neutrophil accumulation 6 h later. Based on available data this would reflect increased neutrophil adhesion to endothelium and increased organ neutrophil accumulation without changes in peripheral neutrophil counts at the TNF dosage range used in this study [19, 201. We could not determine from our histologic methods the exact location of appeared the neutrophils though they

Mechanism

493

localized to capillaries and could be washed out of the myocardium. These results are similar to results we found following in vivo IL-1 pretreatment in rats [8]. Furthermore, these findings are similar to our findings in the lung after treatment of the rat with a sublethal dose of endotoxin [2Z] which demonstrated polymorphonuclear cell accumulation at 6 h. In vitro experiments have indicated that TNF can increase neutrophil toxic oxygen metabolite generation [22]. The in vitro effect of TNF on neutrophil accumulation and adhesiveness to endothelial cells which peaks at 4-6 h requires new protein and receptor synthesis [ZO]. Similarly, we found that in vivo peak myocardial neutrophil accumulation occurred concomitantly with increased myocardial H202 levels 6 h following TNF pretreatment. Prior neutrophil depletion prevented both the early increases in myocardial H,O? and the later increases in catalase activity and improvement in function after ischaemia reperfusion injury though there was incomplete reversal of the protection from ischaemia in the vinblastine and TNF treated rats. Because of the decrease in protection caused by vinblastine, however, this might suggest that the protective effects of TNF are mediated in part by early myocardial neutrophi1 accumulation and related oxidant stress. It is possible that vinblastine may have caused its effect by depletion of white cell species other than neutrophils or by depletion of platelets. We noted, however, that there was a 95% reduction of neutrophils but less than 50% reduction in other white cell types. Vinblastine may have had other non-specific in vivo effects in our rats. We noted, however, that vinblastine treatment 30h after TNF administration and 6 h before myocardial ischaemia reperfusion did not alter cardiac function. Although our findings are suggestive of neutrophil-mediated oxidant related process, it is possible that TNF’s effects could have been conveyed via other mediators or cellular protective processes. It follows logically, however, that protection from an oxidant injury (i.e. ischaemia reperfusion) could be induced by a prior oxidant conditioning. Ibuprofen antagonizes the effects of TNF in multiple experimental models by mechanisms which are incompletely understood [19]. Ibuprofen inhibits neutrophil activity in vitro and

494

J. M. Brown

prevents myocardial PMN mediated postischaemic necrosis and dysfunction [5, 191. Ibuprofen blocks the effects of endotoxin and TNF administration on blood pressure, organ blood flow, and acid base changes; and it increases survivial in multiple animal shock models [ 7, 19,]. It is unclear whether these effects result from cycle-oxygenase inhibition or direct effects on neutrophil function. Our results indicate that ibuprofen in similar dosages as the above partially inhibits the TNF-induced effect on myocardial neutrophil accumulation, protection against myocardial ischaemia-reperfusion injury and early myocardial H202 oxidant stress. These data suggest that ibuprofen antagonizes the neutroPhil-related production of oxidant stress and may in part explain ibuprofen’s protective abilities in other settings on in vivo cardiac ischaemia-reperfusion injury. Whether cyclooxygenase inhibition also prevents a neutroPhil-platelet interaction is unclear. Flynn and colleagues [14] found that ibuprofen but not indomethacin protects against myocardial injury in the cat. This would suggest that these mechanisms involve more than simple cycle-oxygenase inhibition.

et al. Using a small amount of a potentially toxic stimulus or substance to induce protective states is an old finding. Reimer et al [24] have demonstrated that tolerance to ischaemia can be induced by prior episodes of ischaemia. Endotoxin pretreatment prevents death following hemorrhagic shock and attenuates injury following radiation [I, 251. Heat shock increases myocardial catalase activity and protects from subsequent ischaemia-reperfusion injury [8]. We previously have demonstrated that 1/50th of a lethal dose of endotoxin given to rats induced increases in myocardial catafrom subsequent lase and protection ischaemia reperfusion-injury [ 91. Furthermore, previous work with interleukin-1 has shown pretreatment induced protection without increases in catalase but with increases in glucose-6-phosphate dehydrogenase [ 81. These changes were likewise associated with early (6 h after treatment) PMN and H202 increases. Though there are differences between above models and the’ specific the characteristics of the protective states, clearly there is a common thread suggesting the presence of an inducible defense mechanism in myocardium and in other tissues.

References 1 BROWN, J. M., TERADA, L. S., GROSSO, M. A., WHITMAN, G. J. R., VELASCO, S. E., PATT, A., HARKEN, A. H., REPINE, J. E. Xanthine oxidase produces hydrogen peroxide which contributes to reperfusion injury of ischemic isolated rat hearts. J Clin Invest 81, 1297-1301 (1988). 2 ENGLER, R., COVELL, J. W. Granulocytes cause reperfusion ventricular dysfunction after 15-minute ischemia in the dog. Circ Res 61, 20-28 (1987). 3 SIMPSON, P. J., TODD, R. F., FANTONE, J. C., MICKELSON, J. K., GRIFFEN, J. D., LUCCHESI, B. R. Reduction of experimental canine myocardial reperfusion injury by monoclonal antibody (anti-MOl, anti-CDllb) that inhibits leukocyte adhesion. J Clin Invest 81, 624-629 (1988). 4 ROMSON, J. L., HOOK, G. B., RIGOT, V. H., SCHORK, M. A., SWANSON, D. P., LUCCHESI, B. R. The effect of ibuprofen on accumulation of Iridium-ill-labeled platelets and leukocytes in experimental myocardial infarctions. Circulation 66, 1002-1011 (1982). 5 NOWOTNY, A. Beneficial effect of endotoxins. London, Plenum Press, ~~213-226 (1983). 6 WONG, G. H., GOEDDEL, D. V. Induction of manganous superoxide dismutase by tumor necrosis factor possible protective mechanism. Science 242, 942-944 (1988). 7 WISE, N. C., COOK, J. A., ELLER, T., HOSHISH, P. V. Ibuprofen improves survival from endotoxic shock in the rat. J Pharmacol Exp Ther 215, 160-164 (1980). 8 BROWN, J. M., WHITE, C. W., TERADA, L. S., GROSSO, M. A., SHANLEY, P. F., MULVIN, D., BANERJEE, A., WHITMAN, G. J. R., HARKEN, A. H., REPINE, J. E. Interleukin-1 pretreatment decreases ischemialreperfusion injury. Proc Nat1 Acad Sci 87, 5026-5030 (1990). 9 BROWN, J. M., GROSSO, M. A., TERADA, L. S., WHITMAN, G. J. R., BANERJEE, A., WHITE, C. W., HARKEN, A. H., REPINE, J. E. Endotoxin pretreatment increases endogenous myocardial catalase activity and decreases ischemia reperfusion injury of isolated rat hearts. Proc Nat1 Acad Sci 86, 2516-2520 (1989). 10 LEMANSKE, R. F., GUTHMAN, D.A., OERTEL, H., BARR, L., KALINER, M. The biologic activity of mast cell granules. The effect of vinblastine-induced neutopenia on rat cutaneous late phase reactions. J Immunol 130(6), 2837-2842 (1983). 11 COHEN, G., SOMERSON, N. L., Catalase aminotriazole method for measuring secretion of hydrogen peroxide by microorganisms. J Bacterial 92(2), 543-546 (1969).

TNF:

Protective

Mechanism

495

12 BEUTLER, E. Red Cell Metabolism. A Manual ofBiochemical Methods. New York, Grune and Stratton. pp. I-160 (1975). 13 CRAPO. J.D., MCCORD, J.M., FRIDOVICH, J. Preparation and assay of superoxide dismutases. Methods Enzymol 53, 382-393 (1978). 14 FLYNN, D. F., BECKER, W. K., VERCELLOTTI, G. M., WEISDORF, D. J., CRADWCK, P.R.. HAMMERSCHMIUT. D. E., LILLEHEI, R. C., JACOB, H. S. Ibuprofen inhibits granulocyte responses to inflammatory mediators: A proposed mechanism for reduction of experimental myocardial infarct size. Inflammation 8. 33-44 (1984). 15 WHITE. C. W.. GHEZZI, P., MCMAHON, S., DINARELLO, C. A., REPINE, J, E. Cytokines increase rat lung antioxidant enzymes late during exposure to hyperoxia. J Appl Physiol 66, 1003-1007 (1989). 16 THAYER, W. S. Role of catalase in metabolism of hydrogen peroxide by the perfused rat heart. FEBS Letters 202, 137-140 (1986). 17 MYERS, C. L.. WEISS, S. J., KIRSH, M. M. SHEPARD, B. M., SHLAFER, M. Effects of supplementing hypothermn crystalloid solution with catalase, superoxide dismutase, allopurinol, or deferoxamine on functional recovery of globally ischemic and reperfused isolated hearts. J Thorac Cardiovasc Surg 91, 281-289 (1986). 18 CURIE. N. R.. KARMAZYN, M., KLOC, M., MAILER, K. Heat shock response is associated with enhanced postischemic ventricular recovery. Circ Res 63, 543-549 (1988). 19 EVANS, D.A., JACOBS, D.O., REVHAUG, A., WILMORE, D. W. The effects of tumor necrosis factor and their selective inhibition by ibuprofen. Ann Surg 209, 312-321 (1989). 20 GAMBLE. J. R., HARLAN, J. M., KLEBANOFF, S. J., VADAR, M. A. Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Ptoc Nat1 Acad Sci 82, 8667-8671 (1985). 21 ANDERSON, B.O., BENSARD, D. D., BROWN, J. M., BANERJEE, A., SHANLEY, P. F., LUFF, J. A., REPINS, ,J. E.. TF.RAI)A. L. S., HARKEN, A. H. FNLP injures endotoxin-primed rat lungs by neutrophil dependent and independent mechanisms. Am J Physiol 260 (Regulatory Integrative Comp. Physiol. 29), R413-R420 (1991) 22 TSUJIMOTO, M., YOKOTA, S., VILCEK, J., WEISSMANN, G. Tumor necrosis factor provokes superoxtde anion generation from neutrophils. Biochem Biophys Res Commun 137(3), 1094~1100 (1986). 23 NISHIJIMA, K., BRESLOW, M. J., MILLER, C. F., TRAYSTMAR, R. J. Effect of naloxone and ibuprofen on organ blood flow during endotoxic shock in pig. Am J Physiol 255, Hl77-H184 (1988). 24 RIF.MER, K. A., MURRY, C. E., YAMASAWA, J.. HILL. M. L., JENNINGS, R. B. Four brief periods of myocardral ischemia cause no cumulative ATP loss or necrosis. Am J Physiol 251(Gl42), Hl306-1315 (1986). 25 Z~EIFACH. B. W.. THOMAS, L. The relationship between the vascular manifestations of shock produced hy endotoxin. trauma, and hemorrhage. J Exp Med 106, 385-401 (1958).