-
Br HeartJ1 1995;73:14-19
Inhibition of superoxide production in humanneutrophils by
combinations of heparin andthrombolytic agents
Klaris Riesenberg, Francisc Schlaeffer, Amos Katz, Rachel
Levy
Infectious DiseaseLaboratory andClinical Biochemistry,Faculty
ofHealthSciences, Ben-GurionUniversity oftheNegev, SorokaMedical
Centre, BeerSheva, IsraelR LevyDepartment ofInternal
Medicine,Faculty ofHealthSciences, Ben-GurionUniversity oftheNegev,
SorokaMedical Centre, BeerSheva, IsraelK RiesenbergF
SchlaefferIntensive Care Unit,Faculty ofHealthSciences,
Ben-GurionUniversity oftheNegev, SorokaMedical Centre, BeerSheva,
IsraelA KatzCorrespondence to:Dr R Levy, InfectiousDiseases
Laboratory,Faculty of Health Sciences,Ben-Gurion University ofthe
Negev, Beer Sheva84105, Israel.Accepted for publication3 May
1994
AbstractObjective-To investigate the effect ofheparin and
thrombolytic agents onsuperoxide generation by human neu-trophils,
as inhibition of superoxide pro-duction may have a role in
reducingischaemia and reperfusion injury.Methods-Neutrophil
superoxide pro-duction stimulated by phorbol myristateacetate
(PMA), opsonised zymosan, orformyl methionyl leucyl
phenylalanine(FMLP) was measured as the superoxidedismutase
inhibitable reduction ofacetyl ferricytochrome c by a
microtitreplate technique.Results-Heparin, at concentrations
of0'5-500 U/ml, caused a gradual inhibitionof superoxide production
stimulated byPMA, opsonised zymosan, or FMLP.Tissue plasminogen
activator was morepotent than heparin in inhibiting super-oxide
production induced by opsonisedzymosan or FMLP, but it did
notaffect the activity stimulated by PMA.Streptokinase or urokinase
had no effecton superoxide production. When heparinwas used in
combination with tissue plas-minogen activator, streptokinase,
orurokinase at their therapeutic concentra-tions there was a
significant inhibition ofsuperoxide generation (70%/,5 30%, and25%,
respectively). The therapeutic con-centrations of tissue
plasminogen activa-tor alone caused a reduction of 40% ofneutrophil
superoxide production. Whentissue plasminogen activator and
strep-tokinase were both added to neutrophils,however, a
synergistic inhibition of 80%was achieved.Conclusions-The
inhibition of superoxide generation by these drug combi-nations may
explain the limited inflam-matory response and reduction
ofreperfusion injury observed in patientsreceiving thrombolytic
treatment.
(Br HeartJ 1995;73:14-19)
Keywords: superoxide production, neutrophils,heparin,
thrombolytic agents.
The fundamental assumption underlying thecurrent thrombolytic
treatment of acutemyocardial infarction is that the early
restora-tion of myocardial blood flow arrests the pro-gression of
myocardial cell death, permittingthe ultimate functional recovery
of reversibly
injured myocardium. Experimental and clini-cal studies have
indicated that reperfusion ofmyocardium immediately after the onset
ofischaemia can reduce infarct size and improvemortality after
acute coronary artery occlu-sion.'-3 Although reperfusion ends
ischaemia,it can also cause further damage to jeopar-dised cells, a
phenomenon which has beentermed reperfusion injury.The precise
pathophysiology of cellular
damage after extreme ischaemia and reperfu-sion has not been
completely established.Several contributing factors have been
sug-gested, including production of toxic oxygenfree radicals,
endothelial cell swelling, anddamage leading to increased capillary
perme-ability, intravascular platelet activation, andfibrin
deposition. Endothelial swelling com-bined with platelet and fibrin
accumulationmay ultimately result in microvascular throm-bosis,
thus preventing reperfusion. Thesesequelae have been termed the no
reflow phe-nomenon.4
Recent work has focused on the role ofneutrophils in the
development of ischaemiaand reperfusion injury. Activated
neutrophilsare a potent source of oxygen derived freeradicals, and
experimental data suggest thatthey may be important in the
pathogenesisof reperfusion injury.5 6 Oxygen radicalsdepress the
contractile function of isolatedpapillary muscles, ventricular
septae, and iso-lated hearts.7 Cardiac tissue exposed to
freeradicals developed swollen mitochondria,endothelial damage, and
abnormal vascularpermeability.8The biochemical basis for the
generation of
superoxide in neutrophils is the enzymaticcomplex NADPH
oxidase.9 10 This enzyme isdormant in resting neutrophils and is
capableof being activated by several stimuli." Theenzyme is a
multicomponent electron trans-port complex which includes a
membranebound b-type cytochrome (flavocytochromeb558)."-" The
cytochrome incorporates theNADPH binding site and both the flavin
andhaeme electron transfer moieties.'4 In addi-tion, superoxide
generating activity dependson the presence of three cytosolic
oxidase pro-teins which have been characterised as 47 kDa(p47), 67
kDa (p67), and ras related GTPbinding proteins.'5-'8We investigated
whether heparin and
thrombolytic agents have an inhibitory effecton superoxide
generation by human neutro-phils. Such an effect may contribute to
theprevention of reperfusion injury.
14
on June 22, 2021 by guest. Protected by copyright.
http://heart.bmj.com
/B
r Heart J: first published as 10.1136/hrt.73.1.14 on 1 January
1995. D
ownloaded from
http://heart.bmj.com/
-
Inhibition of superoxide production in human neutrophils by
combinations of heparin and thrombolytic agents
Materials and methodsREAGENTSThe reagents used were heparin
(LeoPharmaceutical Products, Ballerup, Den-mark); lyophilised
streptokinase (Streptase,Behringwerke, Marburg, Germany);
uro-kinase (Ukidan, Laboratoires Sero Aubonne,Switzerland); and
tissue plasminogen activa-tor produced by recombinant DNA
technol-ogy (Activase, Genentech, San Francisco,Califomia).
15PMA
MEl Opsonised zymosanM FMLP
10
5
n0 0-5 1 5 10 25 50 100 250
Heparin (U/ml)
Figure 1 Effect of heparin on superoxide generation in
stimulated human neutroResults are means (SE) from five different
experiments, each performed in duplicc*P < 0 001 for inhibitory
effect of heparin.
0 15 30 75 150 300 6(
Streptokinase (U/ml)Figure 2 Effect of streptokinase on
superoxide generation in stimulated humanneutrophils. Results are
means (SE) from five different experiments, each
performduplicate.
ISOLATION OF NEUTROPHILSNeutrophils were separated by
Ficoll/Hypaque centrifugation, dextran sedimenta-tion, and
hypotonic lysis of erythrocytes. 19
MEASUREMENT OF SUPEROXIDE ANIONThe production of superoxide
anion (O2- 1) byintact cells was measured as the
superoxidedismutase inhibitable reduction of ferricy-tochrome c by
a microtitre plate technique.20Cells were suspended (5 x 105
cells/well) inHank's balanced salts solution (100,ul) con-taining
150 mmol/l acetyl ferricytochrome c.Neutrophils were stimulated by
the additionof 50 mg/l phorbol myristate acetate (PMA),1 g/l
opsonised zymosan, or 100 nmol/l formylmethionyl leucyl
phenylalanine (FMLP) andthe reduction of acetyl ferricytochrome c
wasfollowed by the change of absorbance at 550nm every five minutes
on a Thermomaxmicroplate reader (Molecular Devices, MelnoPark,
California). The maximal rates of super-oxide generation were
determined using theextinction coefficient E550 = 21 mM- ' cm-
1.
Addition of each of the drugs to a cell freeassay system,
performed as described inour previous study,20 did not affect
superoxidegeneration measured by cytochrome c reduc-tion.
ANALYSIS OF DATAThe differences in means were analysed by
T* * Student's t test. The plots were drawn as leastsquares
regression lines and tested by analysisof variance.
500 ResultsThe effect of both the anticoagulant (heparin)
phils. and the thrombolytic drugs (streptokinase,2te. urokinase,
and tissue plasminogen activator)
on superoxide production in human neu-trophils stimulated by
PMA, opsonisedzymosan, or FMLP is shown in this study.The drugs
were added to the neutrophils
san before the addition of the stimuli.Preincubation with the
drugs for differenttimes, ranging from five minutes to one
hourbefore the reaction, did not change theresults. Heparin, in a
range of 0 5-500 U/ml,caused a gradual inhibition of superoxide
pro-duction stimulated by any of the three stimuli(figure 1). The
inhibition could be detected at25 U/ml and higher. At 100 U/ml
there was asignificant inhibition of superoxide produc-tion (P <
0 001). There was no inhibition ofactivity at the concentration of
heparin inplasma either after a continuous drip (0 1-1U/ml) or
after a bolus intravenous infusion(5-10 U/ml).19
Streptokinase in the range of 15-600 U/ml(figure 2) and
urokinase in the range of10-5000 U/ml (figure 3) had no effect
onsuperoxide production induced by any of thestimuli. The effect of
tissue plasminogen acti-vator in the range of 0-1-100 mg/l on
super-oxide production is shown in figure 4. Therewas a gradual and
significant inhibition of
ed in superoxide generation in neutrophils stimu-lated by
opsonised zymosan or FMLP. The
C'a
.'00
E-0c
~00-000.5
,o
0
C,)
15
Z
COaL)
0to0
-
c
C0
Ca)0-020
a)cn.
~0-020.U)
10
5
15
on June 22, 2021 by guest. Protected by copyright.
http://heart.bmj.com
/B
r Heart J: first published as 10.1136/hrt.73.1.14 on 1 January
1995. D
ownloaded from
http://heart.bmj.com/
-
Riesenberg, Schlaeffer, Katz, Levy
concentration of tissue plasminogen activator(from 0 5 mg/i),
which caused a significantreduction of the activity, was in the
range ofits concentration in plasma during therapeuticinfusions (0
3-3 mg/i).'9 Superoxide produc-tion stimulated by PMA was not
affected bytissue plasminogen activator in this concentra-tion
range. Since the different drugs are often
Fl PMAm OpsorM FMLP
nised zymosan
0 10 25 50 100 250 500 1000 5000Urokinase (U/ml)
Figure 3 Effect of urokinase on superoxide generation in
stimulated human neutrophils.Results are means (SE) from five
different experiments, each performed in duplicate.
15
._
0
0
E
20(D
0
CD*
0 0.1 0.5 1 2 5 10 20
Tissue plasminogen activator (mg/I)
Figure 4 Effect of tissue plasminogen activator on superoxide
generation Xhuman neutrophils. Results are means (SE) from five
different experimentperformed in duplicate. * P < 0 001, ** P
< 0-01 for difference from cont;
50 100
in stimulateds, eachrol value.
used simultaneously, the effect of variouscombinations of these
drugs on the produc-tion of superoxide by neutrophils was
studied.Figure 5 shows the effect of the combinationsof heparin and
streptokinase or heparin andurokinase on superoxide production.
Whenheparin 10 U/ml, a therapeutic dose thatalone did not affect
the generation of super-oxide, was given with streptokinase
(figure5A) or with urokinase (figure 5B), both ofwhich did not
inhibit superoxide production,there was a marked and significant
inhibitionof activity. The maximal inhibition wasachieved at
therapeutic concentrations ofstreptokinase (75 U/ml) or urokinase
(10U/ml).21 A higher concentration of heparin(100 U/ml) was more
efficient in inhibitingthe activity when given together with
strepto-kinase or urokinase.
Figure 6 shows the effect of heparin incombination with tissue
plasminogen activa-tor on superoxide production stimulated
byopsonised zymosan. Heparin in its therapeuticconcentration (10
U/ml) caused a slight andinsignificant inhibition of superoxide
produc-tion stimulated by opsonised zymosan (2-8(0 6) nmol/106
cells/min compared with 3-45nmol/1 06 cells/min in the control). In
thepresence of tissue plasminogen activator (intwo therapeutic
concentrations, 0-1 and 2mg/i) the activity stimulated by
opsonisedzymosan was 3-25 (0-5) and 2 (0 2) nmol/106cells/min,
respectively. When the activity wasmeasured in the presence of both
heparin andtissue plasminogen activator there was anadditional
inhibition of activity (2-2 (0.2) or1-3 (0-1) nmol/106 cells/min by
10 U/mlheparin and 0-1 ,ug/ml tissue plasminogenactivator or 10
U/ml heparin and 2 mg/l tissueplasminogen activator, respectively).
Similarresults were obtained in neutrophils stimu-lated by
FMLP.
Figure 7 shows the effect caused by thecombinations of tissue
plasminogen activatorand streptokinase on superoxide
production.Streptokinase did not affect activity stimu-lated by
opsonised zymosan or FMLP.However, when tissue plasminogen
activatorat concentrations of 0-1 mg/l or 2 mg/I, whichby
themselves caused partial inhibition, wereadded together with
streptokinase a signifi-cant inhibition (P < 0 005) was
achieved.This inhibition was observed at a low andtherapeutic
concentration of streptokinase (75U/ml) and did not change
significantly as thedose of streptokinase was increased.
DiscussionOur results show the direct effect of
threethrombolytic agents and of heparin on thegeneration of
superoxide by human neu-trophils. Heparin caused a gradual,
dosedependent inhibition of superoxide produc-tion by neutrophils
stimulated with any of thethree agents used in the study. The
inhibitionof superoxide production could be observedat 25 U/ml
heparin and higher, with signifi-cant inhibition (P < 0-001) at
a concentrationof 100 U/ml. Heparin is a naturally occurring
15
10
5
cin
(0
0
0
r._
Ec
c0
0
tD
-005-
20a)
0.
Un
o
16
on June 22, 2021 by guest. Protected by copyright.
http://heart.bmj.com
/B
r Heart J: first published as 10.1136/hrt.73.1.14 on 1 January
1995. D
ownloaded from
http://heart.bmj.com/
-
Inhibition ofsuperoxide production in human neutrophils by
combinations of heparin and thrombolytic agents
A OpsonisedzymosanFMLP
_~ ~~~------- ------ //-,f -ItOpsonised-~~~~- ~ ~ ~ ~zymosan
+">> * * heparin 10 U/mI' '~~f---i-FMLP + heparin
10 U/mI-~>>* * Opsonised- * - /b- zymosan +
* heparin 100 U/mIFMLP + heparin100 U/ml
,,50 100 150
Urokinase (U/mi)206 600
B/ Opsonisedzymosan
9\ t - - >,___$~~~~FMLP9
-
Riesenberg, Schlaeffer, Katz, Levy
studies in animals, which have show]radical scavengers are
capable ofinfarct size.2324 In addition, in expmodels of myocardial
infarctioninfarct size could be limited either byneutrophil
activation or by neutroption. The effect of thrombolytic
tregneutrophils is contradictory in theStreptokinase in acute
myocardialis associated with an abrupt reactrophil response.25
However, othehave shown, in accordance with oithat thrombolysis by
streptokinaseplasminogen activator suppresses ractivation and
infiltration, suggestintreatment may even limit the inflresponse
and thus mitigate rinjury.2627 The improved mortathrombolysis and
after fi blockadethat several factors can redu4ischaemic injury.
Thus, the inhibitorneutrophil superoxide production tnation of
thrombolytic and antiagents shown in our study and theof neutrophil
aggregation,28 chemophagocytosis2930 suggest that theseaddition to
their usual mechanismsmay reduce infarct size by inhibitrophil
activity. However, clinical sneeded to confirm the potential
tibenefit of these drug combinations.
4
3
2
- Opsonised zymosan
n that free The mechanism by which heparin or tissuereducing
plasminogen activator inhibits neutrophil
)erimental NADPH oxidase to generate superoxide isin dogs, not
known. Our results suggest that theyinhibiting probably act through
different mechanisms.)hil deple- Studies in whole cells provide
considerableatment on information about signal transduction
path-literature. ways, suggesting that the oxidase may be
acti-infarction vated by various stimuli through several:tive neu-
pathways.3132 When stimulated by PMA or-r studies diacylglycerol,
NADPH oxidase is activatedur results, by protein kinase C. N-formyl
peptide chemo-and tissue attractants stimulate phosphoinositide
hydro-neutrophil lysis and an increase in intracellular freeg that
this calcium ions. Stimuli, such as opsonised par-ammatory ticles,
act through a calcium dependent path-eperfusion way which is
dependent on arachidonate.[lity after Since heparin inhibited the
activity induced! suggests by the three stimuli, it probably
affects intra-ce acute cellular signal transaction pathways that
arery effect of shared by the different agonists, such as pro-by
combi- tein kinase C activity. In contrast, tissue plas-icoagulant
minogen activator inhibited the generation ofinhibition superoxide
induced by FMLP or opsonisedtaxis, and zymosan and not by PMA,
suggesting thatagents, in tissue plasminogen activator interacts
withof action, neutrophil membranes and interferes with theiting
neu- receptors for this agonist. The inhibition of,tudies are
superoxide production by these drugs doesherapeutic not seem to be
due to a scavenging effect as
tissue plasminogen activator did not inhibitPMA stimulating
activity (figure 4) and theaddition of either heparin or tissue
plasmino-gen activator did not affect NADPH oxidaseactivity in a
cell free assay.
Itn conclusion, the significant inhibition ofneutrophil
superoxide by the anticoagulantand thrombolytic agents shown in
this study
Control may suggest that in addition to their effect inreducing
reperfusion in myocardial infarction,they may also prevent
reperfusion injury.
I *
t-PA 0.1 mg/l
\O 2*t-PA 2 mg/I
1
0 100 200 300 400 500 600
4
3
2
FMLP
Control
° t-PA0-1 mg/I
t-PA 2 mg/l
0 100 200 300 400 500 600Streptokinase (U/ml)
Figure 7 Combined effect of tissue plasminogen activatorand
streptokinase on superoxide production stimulated byopsonised
zymosan orFMLP in human neutrophils.Results arefrom a
representative experiment performed intrifilicate; three other
experiments showed similar results.*P < 0-005for difference from
control value.
1 TIMI Study Group. The thrombolysis in myocardialinfarction
(TIMI) trial: phase 1 findings. N Engl Med1985;312:932-6.
2 Guerci AD, Gerstenblith G, Brinker JA, Gottlieb SO,Weiss JL,
Chandra NC, et al. A double-blind, placebo-controlled, randomized
trial of tPA with subsequent ran-domization to elective PTCA for
acute MI. N EnglMed 1988;37:1613-18.
3 International tPA/SK Mortality Trial Study Group. In-hospital
mortality and clinical course of 20,891 patientswith suspected
acute myocardial infarction randomizedbetween tissue plasminogen
activator or streptokinasewith or without heparin. Lancet
1990;336:71-5.
4 Flaherty JT, Weisfeldt ML. Reperfusion injury. Free
RadicBiolMed 1988;5:409-19.
5 Lucchesi BR. Myocardial ischemiia, reperfusion and freeradical
injury. AmJ7 Cardiol 1990;65:14-23.
6 Mehta J, Dinerman J, Mehta P, Saldeen TGP, Lawson D,Donnelly
WH, et al. Neutrophil function in ischemicheart disease.
Circulation 1989;79:549-56.
7 Przyklenk K, Whittaker P, Kloner RA. In vivo infusion ofoxygen
free radical substates causes myocardial systolic,but not diastolic
dysfunction. Am Heart 1990;119:807-15.
8 Otani H, Tanaka H, Inoue T, Umemoto M, Omoto K,Tanaka K, et
al. In vitro study on contribution of oxida-tive metabolism of
isolated rabbit heart mitochondria tomyocardial reperfusion injury.
Circ Res 1984;55:168-75.
9 Malech HL, Gallin JI. Current concepts: immunologyneutrophils
in human diseases. N Engl Med 1987;317:687-94.
10 Babior BM. Oxygen-dependent microbial killing byphagocytes -
II. N Engl Med 1978;298:721-6.
11 Rossi SJ; The 0,-forming NADPH oxidase of thephagocytes
nature: mechanisms and functions. BiochimBiophys Acta
1986;853:65-89.
12 Segal AW, West I, Wientes F, Nugent JH, Chavan AJ,Haley B, et
al. Cytochrome b-245 is a flavocytochromecontaining FAD and the
NADPH-binding site of themicrobicidal oxidase of phagocytes.
Biochem 7 1992;284:781-8.
c
C')C._
Co-0E0C-
00
._
04)
Q'ax
0.
a)Q
18
1
on June 22, 2021 by guest. Protected by copyright.
http://heart.bmj.com
/B
r Heart J: first published as 10.1136/hrt.73.1.14 on 1 January
1995. D
ownloaded from
http://heart.bmj.com/
-
Inhibition of superoxide production in human neutrophils by
combinations of heparin and thrombolytic agents
13 Cross AR, Jones TG, Harper AM, Segal AW. Oxidation-reduction
properties of the cytochrome b found in theplasma-membrane fraction
ofhuman neutrophils: a pos-sible oxidase in the respiratory burst.
Biochem J 1981;194:599-606.
14 Rotrosen D, Yeung CL, Leto TL, Malech HL, KwongCH. Cytochrome
b558: the flavin-binding component ofthe phagocyte NADPH oxidase.
Science 1992;256:1459-62.
15 Numoi H, Rotrosen D, Gallin JI, Malech L. Two forms
ofautosomal chronic granulomatous disease are deficientin distinct
neutrophil cytosol factors. Science1988;242:1298-301.
16 Lomax KJ, Leto TL, Nunoi H, Gallin JI, Malech HL.Recombinant
47-kilodalton cytosol factor restoresNADPH oxidase in chronic
granulomatous disease.Science 1989;245:409-12.
17 Leto TL, Lomax KJ, Volpp BD, Nunoi H, Sechler MG,Nauseef WM,
et al. Cloning of a 67 KDa neutrophilcytosolic oxidase factor and
its similarity to a noncat-alytic region of p60.-" Science
1990;248:727-30.
18 Abo A, Pick E, Hall A, Toty N, Teahan CG, Segal AW.Activation
of the NADPH oxidase involves the smallGTP-binding protein 21racl.
Nature 1991;353:668-70.
19 Levy R, Malech HL, Rotrosen D. Production of myeloidcell
cytosols functionally and immunochemically defi-cient in the 47 KDa
or 67 KDa NADPH oxidase cytoso-lic factors. Biochem Biophys Res
Commun 1990;170:1114-20.
20 Levy R, Rotrosen D, Nagauker 0, Leto TL, Malech HL.Induction
of the respiratory burst in HL-60 cells: corre-lation and protein
expression. J Immunol 1990;145:2595-601.
21 Gilman G, Rall TW, Nies AS, Taylor P, eds. Goodman
andGilman's pharmacological basis of therapeutics. 8th ed.New York:
Pergamon, 1991, 1313-19, 1322-4.
22 Freischlag JA, Colburn MD, Quinones-Baldrich WJ,Moore WS.
Heparin, urokinase, and ancrod alterneutrophil function. J Vasc
Surg 1992;16:565-74.
23 Burke SE, Wright CD, Potoczak RE, Boucher DM,Dodd GD, Taylor
DG, et al. Reduction of caninemyocardial infarct size by CI-959, an
inhibitor of inflam-matory cell activation. J Cardiovasc Pharmacol
1992;20:619-29.
24 Willerson JT, Buja LM. Protection of the myocardiumduring
myocardial infarction: pharmacologic protectionduring thrombolytic
therapy. Am J! Cardiol 1990;65:35-41.
25 Ranjadayalain K, Umachandran V, Davies SW,Syndercombe-Court
D, Gutteridge CN, Timmis AD.Thrombolytic treatment in acute
myocardial infarction:neutrophil activation, peripheral leucocyte
responses,and myocardial injury. Br HeartJt 1991;66:10-4.
26 Muir AL, Bell D. -The role of the neutrophil in
acutemyocardial infarction. Health Bull (Edinb) 1989;47:207-14.
27 Bell D, Jackson M, Nicoll Jn, Miller A, Dawes J, Muir
AL.Inflammatory response, neutrophil activation, and freeradical
production after acute myocardial infarction:effect of thrombolytic
treatment. Br Heart Jf 1990;63:82-7.
28 Goldman G, Welbourn W, Paterson IS, Fracs MB,Klavsner JM,
Lesner K, et al. Ischemia-inducedneutrophil activation and
diapedesis is lipoxygenasedependent. Surgery 1990;107:428-3.3.
29 Cambria RA, Anderson RJ, Dikdan G, Techan
EP,Hemandez-Maldonado JJ, Hobson RW. Leukocyte acti-vation in
ischemia-reperfusion injury of skeletal muscle.JSurgRes
1991;51:13-7.
30 Odeh M. The role of reperfusion-induced injury in
thepathogenesis of the crush syndrome. N Engl J
Med1991;324:1417-22.
31 Parini MI, Tringles MS, Tauber AI. Identification of
dis-tinct activation pathways of the human neutrophilNADPH oxidase.
JImmunol 1986;137:2925-9.
32 Tarsi-Tsuk D, Levy R. Stimulation of oxidative burst inhuman
monocytes by lipoteichoic acid: involvement ofphospholipase A2. J
Immunol 1990;144:2665-70.
19
on June 22, 2021 by guest. Protected by copyright.
http://heart.bmj.com
/B
r Heart J: first published as 10.1136/hrt.73.1.14 on 1 January
1995. D
ownloaded from
http://heart.bmj.com/
