Cardiovascular Research 1992;26:75 1-756
75 1
Variable effects of human and canine polymorphonuclear leucocytes on vascular smooth muscle tone J Gonzales, J L Mehta, D L Lawson, W W Nichols, and F A Nicolini Objective: Previous studies have shown variable effects of human and canine polymorphonuclear leucocytes (neutrophils) on vascular tone. The aim of this study was to identify whether these variations in neutrophil function are due to species differences. Methods: Canine and human arterial rings (with and without endothelium) were contracted with the thromboxane A2 analogue U466 19, and then exposed to isolated neutrophils. Results: Human neutrophils caused a significant relaxation of the human mammary arterial rings, and the relaxation was unaffected by the cyclo-oxygenase inhibitor indomethacin, enhanced by superoxide dismutase (SOD), and inhibited by oxyhaemoglobin. The relaxant effect of human neutrophils was also diminished upon pretreatment with NG-monomethyl-1-arginine (L-NMMA), indicating that the vasorelaxant material released by the neutrophils was nitric oxide (NO). Human neutrophils also relaxed canine femoral arterial rings, and the relaxant effect was potentiated by SOD and inhibited by pretreatment with oxyhaemoglobin or L-NMMA, confirming that the vasorelaxation was via release of NO.Canine neutrophils, on the other hand, caused an endothelium dependent contraction of autologous femoral arterial rings. This vasoconstriction was not affected by indomethacin, SOD, oxyhaemoglobin, or L-NMMA. However, treatment of canine neutrophils with the 5-lipoxygenase inhibitor piriprost attenuated ( ~ ~ 0 . 0 their 2 ) contractile effect on vascular rings, suggesting that neutrophil generated 5-lipoxygenase products were probably responsible for smooth muscle contraction. Presence of the leukotriene C4 and D4 receptor antagonist FPL 55,712 totally blocked the contractile effects of canine neutrophils, indicating that femoral arterial ring contraction was mediated by peptido-leukotrienes. Conclusions: The endothelium dependent nature of the canine neutrophil induced contraction suggests that the 5-lipoxygensase product leukotriene A4 is taken up by endothelial cells for conversion to peptido-leukotrienes. Since SOD had no effect and FPL 55,712 totally blocked the vasoconstrictor effects of canine neutrophils, it appears that the vasoconstrictor effects of the latter are mediated primarily through peptido-leukotrienes. In contrast, the vasorelaxation by human neutrophils is mediated through release of NO. Cardiovascular Research 1992;26:75 1-756
P
olymorphonuclear leucocytes (neutrophils) defend the tissues against invasion by pathogens; these cells do so by a variety of mechanisms, among which are release of proteolytic enzymes and free oxygen radicals.' Free oxygen radicals may also injure the adjoining healthy cells because of their tissue damaging effects. The potentially detrimental role of these oxygen radicals has become evident from studies in experimental animals subjected to arterial occlusion and reperfusion, a phenomenon associated with release of large amounts of superoxide radicals.24 Activated neutrophils also release via the 5-lipoxygenase pathway the arachidonate metabolite leukotriene A4. Leukotrienes of C and D series exert potent vasoconstrictor effect^.^ The vascular tissues do not generate significant amounts of leukotrienes, but the endothelial cells can "steal" leukotriene A4 to generate leukotriene C A . ~ Recent studies'-'' indicate that neutrophils generate a vasorelaxant and platelet inhibitory material with a short half life. The biological characteristics of this material, ie, its
degradation by superoxide radicals, prolongation of its half life by superoxide dismutase (SOD), and loss of its effect in the presence of oxyhaemoglobin, all indicate that this material is nitric oxide (NO), similar to that released by the endothelial cells." As such, neutrophils appear to possess both tissue protective (vasorelaxation and platelet aggregation inhibition) and tissue injurious properties. Although, some studies have reported a relaxant effect of human and rat neutrophils on vascular smooth muscle,' other studies suggest primaril a vasoconstrictor effects of rabbit and canine ne~trophi1s.l"~Both resting and ionophore A23187 or f-MLP stimulated cells appear to have qualitatively similar effects. Whether the variations in the effects of neutrophils are species specific or are related to methodological variations in study design is not clear. If indeed these differences are species specific, we postulated that an imbalance between generation of vasoconstrictor and vasorelaxant substances may explain the variability in function of neutrophils from different species. Accordingly,
Department of Medicine, Division of Cardiology, Box 5-277, JHMHC, University of Florida, Gainesville, FL 32610, USA: J Gonzales, J L Mehta, D L Lawson, W W Nichols, F A Nicolini. Correspondence to Dr Mehta.
752
Gonzales, Mehta, Lawson, Nichols, Nicolini
the present study was designed to examine the effects of human and canine neutrophils on smooth muscle tone of various arterial segments.
Methods Preparation of vascular rings Three centimetre long segments of human internal mammary artery were obtained from 12 patients undergoing clinically indicated coronary artery bypass surgery. Femoral artery segments (3-4 cm long) were also obtained from eight anaesthetised dogs. The segments were carefully cleaned of fat and connective tissue and cut into 3 mm long rings, which were kept at 37°C in aerated (95% 0 2 , 5% Cq2) Kreb's buffer of the following composition (mmolelitre- ): NaCl 118, KC1 4.7, CaC12 2.5, KHzP04 1.2, MgClz 1.2, NaHCOj 12.5, NaEDTA 0,001, and dextrose 11.1. The vascular rings were washed, suspended in an organ bath, connected to myograph transducers, and stretched to a preload of 2 g which provided an appropriate length-tension relationship based on several (n=15) experiments. After a 2 h period of equilibration, the vascular rings were exposed to 20 mM KCl to test their viability. In some rings, the endothelium was manually disrupted. The intactness of endothelium (endo+) was examined by addition of lo4 M acetylcholine which caused prompt relaxation. In rings with disrupted endothelium (endo-), acetylcholine failed to induce relaxation of the contracted vascular rings. The buffer in the organ bath was changed every 15 rnin and was continuously aerated with 95% 0 2 and 5% C02. The change in force was recorded on a multichannel recorder (Grass Instruments, Quincy, MA, USA). Preparation of human and canine neutrophils
Peripheral venous blood was obtained from normal healthy subjects who had not consumed any drug in the previous 10 days, and also from the anaesthetised dogs from which the femoral arterial segments were obtained. The blood was collected in heparin (10 units.rn1-I) and neutrophils were isolated by density gradient using monopoly resolving medium.' Red blood cells contaminating the neutrophil rich layer were removed by hypotonic lysis. The neutrophils were suspended in Hanks buffered salt solution (HBSS) and kept in ice until use. Over 95% of cells in the suspension were neutrophils with greater than 95% viability as measured by Trypan Blue exclusion. Time from collection of blood to suspension of neutrophils in the organ bath was about 90 min and it was similar whether blood was collected from human volunteers or dogs.
F g d ) for 10 min prior to their suspension in the organ bath. These concentrations of indomethacin and piriprost decrease cyclo-oxygenase and lipoxygenase activity, respectively, by over 80%. In other experiments, neutrophils were incubated with oxyhaemoglobin ( M) or NGmonomethyl-1-arginine (L-NMMA, 5 X lo-' M) for 10 min prior to their suspension in the organ bath. Indomethacin, piriprost, oxyhaemoglobin, and L-NMMA were washed off before suspension of these neutrophils in the organ bath. In other experiments the arterial rings were treated with piriprost (2 X lo-' M) or the peptido-leukotriene receptor antagonist FPL 55,712 (2 X 10" M) prior to suspension of canine neutrophils. In still other experiments, SOD (200 p g m - ' ) was added in the organ bath at peak contraction following U46,6 19, but before suspension of neutrophils. Sources of chemicals
Monopoly resolving medium (MPRM) used for separation of neutrophils was obtained from Flow Labs, McLean, VA, USA; U46,619 and piriprost were gifts of the Upjohn Co, Kalamazoo, MI, USA; and indomethacin was obtained from Merck, Sharpe and Dohme, West Point, PA, USA. Recombinant human SOD was obtained from Pharmacia, Uppsala, Sweden. FPL 55,712 was a gift of Fison Laboratories, Leicestershire, United Kingdom. L-NMMA was obtained from Dr S Moncada, Wellcome Research Foundation, Beckenham, Kent, United Kingdom. All other reagents were obtained from Sigma Chemical, St Louis, MO, USA. All reagents were prepared fresh on the day of use and kept in ice throughout the experiment. Data analysis
All data from a minimum of four experiments in different human and dog arterial segments are presented as mean (SD). The force generated in vascular rings was divided by the weight of the tissue to obtain data in gamg-l. The effect of neutrophils and acetylcholine (change in force) is expressed as percent contraction or relaxation as compared to U46.619 induced force. The data were analysed by
f 10
9
8
l o 4 lo5 106107 6
7
U46619
(PMNL,ml-')
6
(-l0g.M)
ACh GTN
Study protocols
The vascular rings were washed and then contracted with a thromboxane A2 mimic U46,6 19" in progressively increasing concentrations (lo-'' to lo-' M) to achieve about 80% of maximum tension. After a steady state of contraction, warm (37°C) human or canine neutrophils were suspended in the organ bath in increasing concentration, each suspension followed the previous one at 3-5 rnin intervals in order to observe the maximum effect of each concentration of neutrophils. This protocol was followed in endo+ and endovascular rings. After the addition of neutrophils, acetylcholine was added to the organ bath to determine intactness of the endothelium. In some experiments, the neutrophils were treated with the cyclo-oxygenase inhibitor indomethacin (lo-' M)8 or the M)I6or SOD (200 5-lipoxygenase inhibitor piriprost (2 X
A 10
9
8
A
7
U46619
Figure 1 A representative experiment showing human neurrophil (PMNL) mediated relaxation of a human internal mummary artety ring (top). Acetylcholine (ACh) and glyceryl trinitrate (GTN) caused additional relaxation. There was no spontaneous relaxation in another ring from the same internal mammary artery over a 15-20 rnin period of observation (bottom). ACh and GTN relaxed this ring as well.
Pol.ymorphs and vascular tone
Table I
753
Effect qf various inhibitors on human neutrophil induced internal mummap artep re1a.ration. Values Lire means (SDJ.
Control endo+ rings4 (n=15) Control endo- rings (n=8) Endo+ rings exposed to neutrophils treated with: lndomethacin (n=6) SOU (n= 10) Oxyhaemoglobin (n=7) L-NMMA (n=4)
1.12(0.20) I .20(0.2I )
2(8) 32)
13 I O N
7(11)
17(11) 26(8)$
1.20(0.50) I .00(0.21 ) 1.06(0.12) 1. I2(0.20)
2(2) 3(3) +4(4)f 0
6(6) 8(4) +7(24)f
12W 41 ( 1 2)* +2 1 ( I2)f 4(6)t
of
29(11)
41(14) 0%
36(24) 66( 15)* +30(l2)f 5(6)1
48(14) 71(1S)* +30( I 2 ) t 2S(S)f
41(10)$
80( IS)
8S( 10) 90(7)
99( I ) 47( 10) 85( 10)
"Plus (+) symbol denotes rings with intact endothelium. ACh=acetylcholine; GTN=glyceryl trinitrate: sou=superoxide dismutase: L-NMM~=NG-monomethyl-~-~rgi~iine. $p
75 1
Variable effects of human and canine polymorphonuclear leucocytes on vascular smooth muscle tone J Gonzales, J L Mehta, D L Lawson, W W Nichols, and F A Nicolini Objective: Previous studies have shown variable effects of human and canine polymorphonuclear leucocytes (neutrophils) on vascular tone. The aim of this study was to identify whether these variations in neutrophil function are due to species differences. Methods: Canine and human arterial rings (with and without endothelium) were contracted with the thromboxane A2 analogue U466 19, and then exposed to isolated neutrophils. Results: Human neutrophils caused a significant relaxation of the human mammary arterial rings, and the relaxation was unaffected by the cyclo-oxygenase inhibitor indomethacin, enhanced by superoxide dismutase (SOD), and inhibited by oxyhaemoglobin. The relaxant effect of human neutrophils was also diminished upon pretreatment with NG-monomethyl-1-arginine (L-NMMA), indicating that the vasorelaxant material released by the neutrophils was nitric oxide (NO). Human neutrophils also relaxed canine femoral arterial rings, and the relaxant effect was potentiated by SOD and inhibited by pretreatment with oxyhaemoglobin or L-NMMA, confirming that the vasorelaxation was via release of NO.Canine neutrophils, on the other hand, caused an endothelium dependent contraction of autologous femoral arterial rings. This vasoconstriction was not affected by indomethacin, SOD, oxyhaemoglobin, or L-NMMA. However, treatment of canine neutrophils with the 5-lipoxygenase inhibitor piriprost attenuated ( ~ ~ 0 . 0 their 2 ) contractile effect on vascular rings, suggesting that neutrophil generated 5-lipoxygenase products were probably responsible for smooth muscle contraction. Presence of the leukotriene C4 and D4 receptor antagonist FPL 55,712 totally blocked the contractile effects of canine neutrophils, indicating that femoral arterial ring contraction was mediated by peptido-leukotrienes. Conclusions: The endothelium dependent nature of the canine neutrophil induced contraction suggests that the 5-lipoxygensase product leukotriene A4 is taken up by endothelial cells for conversion to peptido-leukotrienes. Since SOD had no effect and FPL 55,712 totally blocked the vasoconstrictor effects of canine neutrophils, it appears that the vasoconstrictor effects of the latter are mediated primarily through peptido-leukotrienes. In contrast, the vasorelaxation by human neutrophils is mediated through release of NO. Cardiovascular Research 1992;26:75 1-756
P
olymorphonuclear leucocytes (neutrophils) defend the tissues against invasion by pathogens; these cells do so by a variety of mechanisms, among which are release of proteolytic enzymes and free oxygen radicals.' Free oxygen radicals may also injure the adjoining healthy cells because of their tissue damaging effects. The potentially detrimental role of these oxygen radicals has become evident from studies in experimental animals subjected to arterial occlusion and reperfusion, a phenomenon associated with release of large amounts of superoxide radicals.24 Activated neutrophils also release via the 5-lipoxygenase pathway the arachidonate metabolite leukotriene A4. Leukotrienes of C and D series exert potent vasoconstrictor effect^.^ The vascular tissues do not generate significant amounts of leukotrienes, but the endothelial cells can "steal" leukotriene A4 to generate leukotriene C A . ~ Recent studies'-'' indicate that neutrophils generate a vasorelaxant and platelet inhibitory material with a short half life. The biological characteristics of this material, ie, its
degradation by superoxide radicals, prolongation of its half life by superoxide dismutase (SOD), and loss of its effect in the presence of oxyhaemoglobin, all indicate that this material is nitric oxide (NO), similar to that released by the endothelial cells." As such, neutrophils appear to possess both tissue protective (vasorelaxation and platelet aggregation inhibition) and tissue injurious properties. Although, some studies have reported a relaxant effect of human and rat neutrophils on vascular smooth muscle,' other studies suggest primaril a vasoconstrictor effects of rabbit and canine ne~trophi1s.l"~Both resting and ionophore A23187 or f-MLP stimulated cells appear to have qualitatively similar effects. Whether the variations in the effects of neutrophils are species specific or are related to methodological variations in study design is not clear. If indeed these differences are species specific, we postulated that an imbalance between generation of vasoconstrictor and vasorelaxant substances may explain the variability in function of neutrophils from different species. Accordingly,
Department of Medicine, Division of Cardiology, Box 5-277, JHMHC, University of Florida, Gainesville, FL 32610, USA: J Gonzales, J L Mehta, D L Lawson, W W Nichols, F A Nicolini. Correspondence to Dr Mehta.
752
Gonzales, Mehta, Lawson, Nichols, Nicolini
the present study was designed to examine the effects of human and canine neutrophils on smooth muscle tone of various arterial segments.
Methods Preparation of vascular rings Three centimetre long segments of human internal mammary artery were obtained from 12 patients undergoing clinically indicated coronary artery bypass surgery. Femoral artery segments (3-4 cm long) were also obtained from eight anaesthetised dogs. The segments were carefully cleaned of fat and connective tissue and cut into 3 mm long rings, which were kept at 37°C in aerated (95% 0 2 , 5% Cq2) Kreb's buffer of the following composition (mmolelitre- ): NaCl 118, KC1 4.7, CaC12 2.5, KHzP04 1.2, MgClz 1.2, NaHCOj 12.5, NaEDTA 0,001, and dextrose 11.1. The vascular rings were washed, suspended in an organ bath, connected to myograph transducers, and stretched to a preload of 2 g which provided an appropriate length-tension relationship based on several (n=15) experiments. After a 2 h period of equilibration, the vascular rings were exposed to 20 mM KCl to test their viability. In some rings, the endothelium was manually disrupted. The intactness of endothelium (endo+) was examined by addition of lo4 M acetylcholine which caused prompt relaxation. In rings with disrupted endothelium (endo-), acetylcholine failed to induce relaxation of the contracted vascular rings. The buffer in the organ bath was changed every 15 rnin and was continuously aerated with 95% 0 2 and 5% C02. The change in force was recorded on a multichannel recorder (Grass Instruments, Quincy, MA, USA). Preparation of human and canine neutrophils
Peripheral venous blood was obtained from normal healthy subjects who had not consumed any drug in the previous 10 days, and also from the anaesthetised dogs from which the femoral arterial segments were obtained. The blood was collected in heparin (10 units.rn1-I) and neutrophils were isolated by density gradient using monopoly resolving medium.' Red blood cells contaminating the neutrophil rich layer were removed by hypotonic lysis. The neutrophils were suspended in Hanks buffered salt solution (HBSS) and kept in ice until use. Over 95% of cells in the suspension were neutrophils with greater than 95% viability as measured by Trypan Blue exclusion. Time from collection of blood to suspension of neutrophils in the organ bath was about 90 min and it was similar whether blood was collected from human volunteers or dogs.
F g d ) for 10 min prior to their suspension in the organ bath. These concentrations of indomethacin and piriprost decrease cyclo-oxygenase and lipoxygenase activity, respectively, by over 80%. In other experiments, neutrophils were incubated with oxyhaemoglobin ( M) or NGmonomethyl-1-arginine (L-NMMA, 5 X lo-' M) for 10 min prior to their suspension in the organ bath. Indomethacin, piriprost, oxyhaemoglobin, and L-NMMA were washed off before suspension of these neutrophils in the organ bath. In other experiments the arterial rings were treated with piriprost (2 X lo-' M) or the peptido-leukotriene receptor antagonist FPL 55,712 (2 X 10" M) prior to suspension of canine neutrophils. In still other experiments, SOD (200 p g m - ' ) was added in the organ bath at peak contraction following U46,6 19, but before suspension of neutrophils. Sources of chemicals
Monopoly resolving medium (MPRM) used for separation of neutrophils was obtained from Flow Labs, McLean, VA, USA; U46,619 and piriprost were gifts of the Upjohn Co, Kalamazoo, MI, USA; and indomethacin was obtained from Merck, Sharpe and Dohme, West Point, PA, USA. Recombinant human SOD was obtained from Pharmacia, Uppsala, Sweden. FPL 55,712 was a gift of Fison Laboratories, Leicestershire, United Kingdom. L-NMMA was obtained from Dr S Moncada, Wellcome Research Foundation, Beckenham, Kent, United Kingdom. All other reagents were obtained from Sigma Chemical, St Louis, MO, USA. All reagents were prepared fresh on the day of use and kept in ice throughout the experiment. Data analysis
All data from a minimum of four experiments in different human and dog arterial segments are presented as mean (SD). The force generated in vascular rings was divided by the weight of the tissue to obtain data in gamg-l. The effect of neutrophils and acetylcholine (change in force) is expressed as percent contraction or relaxation as compared to U46.619 induced force. The data were analysed by
f 10
9
8
l o 4 lo5 106107 6
7
U46619
(PMNL,ml-')
6
(-l0g.M)
ACh GTN
Study protocols
The vascular rings were washed and then contracted with a thromboxane A2 mimic U46,6 19" in progressively increasing concentrations (lo-'' to lo-' M) to achieve about 80% of maximum tension. After a steady state of contraction, warm (37°C) human or canine neutrophils were suspended in the organ bath in increasing concentration, each suspension followed the previous one at 3-5 rnin intervals in order to observe the maximum effect of each concentration of neutrophils. This protocol was followed in endo+ and endovascular rings. After the addition of neutrophils, acetylcholine was added to the organ bath to determine intactness of the endothelium. In some experiments, the neutrophils were treated with the cyclo-oxygenase inhibitor indomethacin (lo-' M)8 or the M)I6or SOD (200 5-lipoxygenase inhibitor piriprost (2 X
A 10
9
8
A
7
U46619
Figure 1 A representative experiment showing human neurrophil (PMNL) mediated relaxation of a human internal mummary artety ring (top). Acetylcholine (ACh) and glyceryl trinitrate (GTN) caused additional relaxation. There was no spontaneous relaxation in another ring from the same internal mammary artery over a 15-20 rnin period of observation (bottom). ACh and GTN relaxed this ring as well.
Pol.ymorphs and vascular tone
Table I
753
Effect qf various inhibitors on human neutrophil induced internal mummap artep re1a.ration. Values Lire means (SDJ.
Control endo+ rings4 (n=15) Control endo- rings (n=8) Endo+ rings exposed to neutrophils treated with: lndomethacin (n=6) SOU (n= 10) Oxyhaemoglobin (n=7) L-NMMA (n=4)
1.12(0.20) I .20(0.2I )
2(8) 32)
13 I O N
7(11)
17(11) 26(8)$
1.20(0.50) I .00(0.21 ) 1.06(0.12) 1. I2(0.20)
2(2) 3(3) +4(4)f 0
6(6) 8(4) +7(24)f
12W 41 ( 1 2)* +2 1 ( I2)f 4(6)t
of
29(11)
41(14) 0%
36(24) 66( 15)* +30(l2)f 5(6)1
48(14) 71(1S)* +30( I 2 ) t 2S(S)f
41(10)$
80( IS)
8S( 10) 90(7)
99( I ) 47( 10) 85( 10)
"Plus (+) symbol denotes rings with intact endothelium. ACh=acetylcholine; GTN=glyceryl trinitrate: sou=superoxide dismutase: L-NMM~=NG-monomethyl-~-~rgi~iine. $p
