Vascular spasm is a well-known complication during vascular surgery. Topical lidocaine is frequently used to prevent this spasm. However, the effects of lidocaine on the endothelium-dependent antiaggregation are not clear.
The aggregation of platelet-rich plasma (PRP) obtained from healthy volunteers was measured by the turbidimetric technique at 37 degrees C. (1) Cultured porcine aortic endothelial cells were preincubated with lidocaine (3.7 microM to 37 mM), NG-methyl-L-arginine (300 microM), or indomethacin (10 microM) for 30 min. The preincubation medium was exchanged with a medium containing +/- 1 microM bradykinin for 1-min stimulation of endothelial cells. One hundred microliters of the supernatant was then added to PRP (750 micro1) just after stimulation of PRP with collagen (4 micrograms/ml). (2) Authentic nitric oxide (NO) or prostacyclin (PGI2) was applied to collagen-stimulated PRP with or without lidocaine (100 micrograms/ml).
(1) The supernatant from endothelial cells without bradykinin stimulation showed "basal" antiaggregation (13.8 +/- 3.2%; n = 6). Bradykinin enhanced the antiaggregation (100 +/- 0%; n = 6). NG-methyl-L-arginine or indomethacin (antagonists of NO or PGI2) inhibited the bradykinin-evoked antiaggregation (10.3 +/- 2.1% and 13.6 +/- 3.7%, respectively; n = 6). Simultaneous preincubation of both agents completely blocked the effect (-4.2 +/- 2.8%; n = 6). Lidocaine failed to influence basal antiaggregation, but it inhibited bradykinin-stimulated antiaggregation in a concentration-dependent manner (concentration causing 50% inhibition = 108 +/- 41 microM; n = 6). (2) In contrast, lidocaine did not shift the 50% effective concentration of NO (control, 1.3 +/- 0.1 microM vs. lidocaine, 1.6 +/- 0.1 microM) or PGI2 (control, 405 +/- 54 nM vs. lidocaine, 257 +/- 41 nM) for antiaggregation.
Our results suggest that lidocaine has an inhibitory effect on antiaggregation derived from endothelial cells, caused by the inhibition of NO and PGI2 released from endothelial cells.
Methods: The aggregation of platelet-rich plasma (PRP) obtained from healthy volunteers was measured by the turbidimetric technique at 37 degrees Celsius. (1) Cultured porcine aortic endothelial cells were preincubated with lidocaine (3.7 micro Meter to 37 mM), NG-methyl-L-arginine (300 micro Meter), or indomethacin (10 micro Meter) for 30 min. The preincubation medium was exchanged with a medium containing plus/minus 1 micro Meter bradykinin for 1-min stimulation of endothelial cells. One hundred microliters of the supernatant was then added to PRP (750 micro liter) just after stimulation of PRP with collagen (4 micro gram/ml). (2) Authentic nitric oxide (NO) or prostacyclin (PGI2) was applied to collagen-stimulated PRP with or without lidocaine (100 micro gram/ml).
Results: (1) The supernatant from endothelial cells without bradykinin stimulation showed "basal" antiaggregation (13.8 plus/minus 3.2%; n = 6). Bradykinin enhanced the antiaggregation (100 plus/minus 0%; n = 6). NG-methyl-L-arginine or indomethacin (antagonists of NO or PGI2) inhibited the bradykinin-evoked antiaggregation (10.3 plus/minus 2.1% and 13.6 plus/minus 3.7%, respectively; n = 6). Simultaneous preincubation of both agents completely blocked the effect (-4.2 plus/minus 2.8%; n = 6). Lidocaine failed to influence basal antiaggregation, but it inhibited bradykinin-stimulated antiaggregation in a concentration-dependent manner (concentration causing 50% inhibition = 108 plus/minus 41 micro Meter; n = 6). (2) In contrast, lidocaine did not shift the 50% effective concentration of NO (control, 1.3 plus/minus 0.1 micro Meter vs. lidocaine, 1.6 plus/minus 0.1 micro Meter) or PGI2(control, 405 plus/minus 54 nM vs. lidocaine, 257 plus/minus 41 nM) for antiaggregation.
Conclusions: Our results suggest that lidocaine has an inhibitory effect on antiaggregation derived from endothelial cells, caused by the inhibition of NO and PGI2release from endothelial cells.
Key words: Anesthetics, local: lidocaine. Blood, platelets: aggregation. Endothelial cells: collagen; nitric oxide; prostacyclin. Nitric oxide, antagonists: N sup G -methyl-L-arginine. Prostacyclin, inhibitors: indomethacin.
VASCULAR spasm, defined as prolonged contraction of the vascular tissue, is a well-known complication during vascular surgery. The most common and effective method to prevent such complication is the administration of topical vasodilators. Lidocaine, in a concentration used for local anesthesia, is one of the most popular agents frequently used for this purpose during surgical manipulations. [1-4].
Avoidance of endothelial damage is also essential during surgical manipulation because intact endothelial cells around the damaged and anastomosed sites of vascular tissues play an important role in the anti-thrombotic process during reperfusion. [1,5]The commencement of the thrombotic process after vascular injury involves the adhesion of platelets to collagen fibers in the subendothelial matrix through a damaged endothelial lining. [5,6]Stimulation of adhered platelets with collagen leads to platelet aggregation followed by the release of ADP or thromboxane A2. These agents, in turn, cause further aggregation and enhance subsequent blood coagulation cascade. Thus, endothelial cells, known to inhibit platelet aggregation [7-9]and vascular constriction by secreting prostacyclin (PGI2) and nitric oxide (NO), protect the vessel from excessive thrombosis during the first step of a coagulation cascade. Furthermore, stimulation through several mechanisms, including those of the nervous system (e.g., cholinergic stimulation [9,11]), inflammatory chemical mediators (e.g., bradykinin [7-10]), and mechanical stimulation (shear stress [9,12]), enhance the release of NO and PGI2from endothelial cells.
Thus, although lidocaine has a beneficial effect on vascular tone, intravascular thrombosis inside the tissue may progress during reperfusion if the antiaggregatory properties of endothelial cells are inhibited by the agent. Because the effects of lidocaine on these cells have not been yet clarified, we evaluated the influence of the local anesthetic, lidocaine, on antiaggregation derived from basal or receptor stimulated endothelial cells.
Materials and Methods
Porcine aortic endothelial cells were prepared according to the method described previously. In brief, fresh porcine thoracic aortae were obtained from a local slaughterhouse immediately after the animals were killed. After rinsing with sterilized phosphate-buffered saline containing ampicillin (90 micro gram/ml) and kanamycin (90 micro gram/ml), the arteries were incubated with phosphate-buffered saline containing 0.125% (weight in volume) trypsin for 30 min. The endothelial lining was gently scraped using a sterilized cotton rolling pin, and the obtained endothelial cells were suspended in culture medium [1:1 (volume in volume) RPMI 1640 medium-Dulbecco's modified Eagle's medium supplemented with HCO3sup - (2 mg/ml), hydroxyethylpiperazine ethanesulfonic acid (HEPES) (15 mM), ampicillin (90 micro gram/ml), kanamycin (90 micro gram/ml), heparin (10 micro gram/ml), and 10% (volume in volume) fetal bovine serum]. After centrifugation at 200g for 5 min, cells were resuspended in the same culture medium and seeded on collagen-coated plastic flasks (25 cm2). The cells were grown in an incubator at 37 degrees Celsius in a humidified atmosphere of 5% CO2in air. Nonadherent cells were removed by washing with the culture medium within 12 h. Culture medium was replaced every 2 days.
Endothelial cells grown to confluent monolayer in the flask were dispersed with 0.125% (weight in volume) trypsin in phosphate-buffered saline for 5 min at 37 degrees Celsius and subcultured on the other 25-cm2flasks (split ratio 1:3). The subcultured cells (passages 2-4; 2 x 106cells/flask) were then split into a 24-well multiplate (1.8 cm2/well) and used for assay just after confluence.
Trisodium citrate 3.8% was mixed with human blood (1:9 [volume in volume]) from healthy volunteers (n = 15; men, aged 25-31 yr) after obtaining informed consent and the approval of the protocol by the Hiroshima University School of Medicine Committee on Human Research. Platelet-rich plasma (PRP) was obtained by centrifugation of the mixture at 240g for 20 min at room temperature, and kept in a water bath at 37 degrees Celsius. PRP was used for experiments without an adjustment to platelet number (10-30 x 104/micro liter).
Assay of the Effect of Endothelial Cell Function on Platelet Aggregation
The assay procedure is diagrammed in Figure 1. A confluent monolayer of endothelial cells on a 24-well multiplate was preincubated with physiologic salt solution (PSS) in the presence or absence of lidocaine HCl (3.7 micro Meter [1.0 micro gram/ml] to 37 mM [10 mg/ml]) for 30 min before stimulation of endothelial cells with or without bradykinin. The millimolar composition of PSS was NaCl 117, KCl 4.7, CaCl sub 2 1.7, MgSO41.2, KH2PO41.2, D-glucose 11, and HEPES 3 (pH 7.4). In some experiments, cells were preincubated with NG-methyl-L-arginine (L-NMA) (300 micro Meter), a known antagonist of NO, or indomethacin (10 micro Meter), a known inhibitor of PGI2synthesis, instead of lidocaine.
After preincubation of the cells, the incubation buffer was discarded by suction, and the cells were washed with PSS three times. In the next step, the endothelial cells were loaded with or without 1 micro Meter bradykinin-containing PSS (200 micro liter) for 1 min at 37 degrees Celsius. The concentration of lidocaine, L-NMA, or indomethacin in the incubation buffer was reduced to a trace level by these steps to prevent the direct effect of preincubation medium on platelet aggregation. On the other hand, PRP (750 micro liter), in a 1.0-ml siliconized crystal glass chamber, was placed on the stage of an aggregometer, stirred at 1,000 rpm, and stimulated with collagen (4 micro gram/ml) to cause platelet aggregation (37 degrees Celsius). The supernatant from endothelial cells (100 micro liter) was added to the prepared PRP just after stimulation with collagen. Platelet aggregation was measured by the turbidimetric technique, [7,8]with a spectrophotometer equipped for aggregometry (CAF-100, Japan Spectroscopic, Tokyo, Japan), and the change in light transmission of the mixture was recorded. Normal PSS (100 micro liter), instead of the supernatant from endothelial cells, was added to PRP, and the platelet aggregation was recorded as a control. Experimental data were compared with control aggregation values recorded before each experimental recording.
The response of antiaggregation evoked by the supernatant from endothelial cells was expressed as a percentage, assuming control light transmission to be 0% and that obtained before the commencement of aggregation to be 100% (Figure 2).
Effects of Nitric Oxide and Prostacyclin on Platelet Aggregation
NO (100 nM--10 micro Meter) and PGI2(10 nM--10 micro Meter) were added to PRP, and the platelet aggregation was recorded as described above. NO was saturated in deoxygenated PSS by bubbling NO gas as described previously. .
Direct Effect of Lidocaine on Platelets
Although the assay protocol for endothelial cell function (Figure 1) reduced the concentration of lidocaine transferred to PRP, a "trace" level of lidocaine may influence platelet function directly. Therefore, the direct effect of the agent on platelet aggregation and its sensitivity to NO or PGI2were determined. After determining the percentage response of platelet aggregation to lidocaine, the antiaggregatory response to NO or PGI2in the presence of 100 micro gram/ml lidocaine was measured and compared with control (i.e., in the absence of lidocaine).
Data were expressed as means plus/minus SEM. Multiple comparisons were performed by analysis of variance followed by Dunn's test (P < 0.05). Comparisons with control were performed with the paired t test (P < 0.05).
Lidocaine, trypsin, RPMI 1640, Dulbecco's modified Eagle's medium, HEPES, ampicillin, kanamycin, L-arginine, L-NMA, and bradykinin were purchased from Sigma (St. Louis, MO); fetal bovine serum from JRH Bioscience (Lenaxa, KS); collagen from Baxter (Tokyo, Japan); and NO gas from Nippon Sanso (Tochigi, Japan).
(Figure 2(A)) shows the representative recordings of changes in light transmission of PRP in the presence or absence of endothelial cells. The supernatant from endothelial cells without bradykinin stimulation showed a small (13.8 plus/minus 3.2%; n = 6), but significant "basal" antiaggregation compared with control (Figure 2(B)). Bradykinin (1.0 micro Meter) alone had no effect on platelet aggregation, whereas stimulation of endothelial cells by 1.0 micro Meter bradykinin enhanced the antiaggregatory effect of the cells. Platelet aggregation was completely abolished by the addition of the supernatant from bradykinin-stimulated endothelial cells (100%; n = 6) (Figure 2(B)).
Preincubation of endothelial cells with inhibitors of NO or PGI sub 2 for 30 min before the stimulation of cells with bradykinin significantly inhibited antiaggregation (Figure 3(A)). L-NMA (300 micro Meter) inhibited the antiaggregatory effect of bradykinin-stimulated endothelial cells to 10.3 plus/minus 2.1% (n = 6) (Figure 3(B)). Indomethacin (10 micro Meter) also inhibited the antiaggregation to a level similar to that of L-NMA (13.6 plus/minus 3.7%; n = 6) (Figure 3(B)). In contrast, a simultaneous preincubation of both L-NMA and indomethacin completely blocked the antiaggregatory response of endothelial cells (-4.2 plus/minus 2.8%; n = 6) (Figure 3(B)). These results indicated that the antiaggregation evoked by the supernatant from the cells was attributable to NO or PGI2released concomitantly from these cells.
(Figure 4) shows the relation between the antiaggregatory response of endothelial cells and the concentration of lidocaine used during preincubation (n = 6). The "basal" antiaggregatory effect of endothelial cells was not affected by lidocaine up to a concentration of 37 mM (10 mg/ml). In contrast, lidocaine inhibited the antiaggregatory response evoked by bradykinin-stimulated endothelial cells in a concentration-dependent manner. Over a dose of 370 micro Meter (100 micro gram/ml) lidocaine significantly inhibited antiaggregation evoked by bradykinin-stimulated endothelial cells. The percentage response curve was fitted to Hill's equation, and the concentration of lidocaine causing 50% inhibition of bradykinin-stimulated antiaggregation of endothelial cells was determined to be 108 plus/minus 41 micro Meter (29 plus/minus 11 micro gram/ml).
Authentic NO or PGI2showed a concentration-dependent antiaggregation (n = 7) (Figure 5(A and B)). The maximum antiaggregatory effects of NO and PGI2were observed at concentrations of 4 and 3 micro Meter, respectively. The 50% effective concentration of NO obtained from the percentage response curve was 1.3 plus/minus 0.1 micro Meter, whereas that of PGI2was 405 plus/minus 54 nM. Simultaneous application of these agents showed a synergic effect on platelet antiaggregation (n = 4) (Figure 5(C)). The threshold concentrations of NO and PGI2were 400 and 100 nM, respectively (Figure 5(A and B)). However, simultaneous application of a subthreshold concentration of NO (200 nM) and PGI2(100 nM) enhanced antiaggregation to 77.0 plus/minus 7.6%.
In contrast to the inhibition of endothelial cell-dependent antiaggregation by lidocaine, direct application of the agent to PRP evoked antiaggregation in a concentration-dependent manner (n = 3) (Figure 6(A)). Figure 6(B and C) show the relation between antiaggregation of platelets and the concentration of NO or PGI2in the presence or absence of lidocaine. Lidocaine, 100 micro gram/ml, significantly increased the baseline of percentage response curves but not the 50% effective concentrations of NO and PGI2(NO: control 1.3 plus/minus 0.1 micro Meter vs. lidocaine 1.6 plus/minus 0.1 micro Meter; PGI2: control 405 plus/minus 54 nM vs. lidocaine 257 plus/minus 41 nM). These results indicated that the direct effect of lidocaine on NO or PGI2-induced antiaggregation is additive, and that the platelet sensitivity to NO or PGI2is not influenced by lidocaine at this concentration.
Lidocaine is frequently topically applied to prevent traumatic vasospasm during vascular surgery. [1-4]Several groups of investigators have demonstrated a beneficial effect of lidocaine on the prompt resolution of vascular spasm. [1-3]However, it has also been reported that topical lidocaine causes thrombosis in microvessels. In the current experiment, we showed that the supernatant from endothelial cells counteracts platelet aggregation stimulated by collagen, and that the antiaggregatory property is enhanced by the stimulation of endothelial cells with bradykinin, which is one of the inflammatory chemical mediators in damaged tissues. We also demonstrated that lidocaine inhibits the antiaggregation derived from bradykinin-stimulated endothelial cells, in vitro, in a concentration-dependent manner, as it could be speculated from an in vivo study described by Wadstrom and Gerdin. .
Endothelial cells exhibit several antithrombotic properties, including those of NO, [7-9]PGI2, [7,8]heparin-like glycosaminoglycan, [5,15]thrombomodulin, and a tissue-type plasminogen activator. Several such compounds exhibit as cell surface antigens, whereas others are released from the cells to modulate the blood coagulation cascade. The latter "released" compounds can be further classified according to their effect on one or more steps of the blood coagulation cascade. One such compound, plasminogen activator, acts as a fibrinolytic agent on the last step of the cascade. Other compounds, such as NO or PGI2, modulate the first step of thrombosis, that of platelet aggregation. [7-9]In the current experiment, we evaluated the effects of "released" antithrombotic property of endothelial cells at the commencement of collagen-stimulated platelet aggregation using a supernatant from endothelial cells. We showed that the antiaggregatory response of endothelial cells was strongly inhibited by an inhibitor of NO synthesis (L-NMA), or that of PGI2(indomethacin). Simultaneous application of both agents completely blocked antiaggregation. Therefore, it is likely that the effect of the supernatant from endothelial cells on acute platelet aggregation is dependent on the interaction between platelets and NO/PGI2released from endothelial cells. This finding has been consistently observed by other investigators. [7,8].
The former findings are further supported by our results demonstrating the effects of authentic NO or PGI2on antiaggregation. A concomitant application of NO and PGI2had a synergic effect on platelet aggregation, confirming the results of an early observation. We also demonstrated that the simultaneous application of subthreshold concentrations of NO and PGI2strongly enhances the antiaggregation although a single application of one compound failed to influence platelet aggregation. As this result indicates that the potency of antiaggregation cannot be quantified by the respective measurement of NO and PGI2, it points to an advantage of the current study, in which the measurement of the effect of agents concomitantly released from endothelial cells on antiaggregation is performed.
A major and novel finding in the current study is that lidocaine inhibits and antiaggregation, derived from bradykinin-stimulated endothelial cells, in a concentration-dependent manner. This result indicates two possible targets for lidocaine: (1) NO/PGI2release from endothelial cells; (2) or platelet sensitivity to NO/PGI2. To clarify the mechanism of action, we designed the assay protocol to prevent the direct influence of lidocaine on platelets by adding a step of "wash-out" before mixing the supernatant from endothelial cells to PRP. Furthermore, we tested the direct action of lidocaine on NO/PGI2-evoked antiaggregation of platelets. Lidocaine did not modulate the 50% effective concentration of NO or of PGI2for antiaggregation although the agent itself showed a concentration-dependent antiaggregatory effect on platelets. This finding indicates that the sensitivity of platelets to NO/PGI2is not inhibited by lidocaine at the concentration used in the experiment. It is likely, therefore, that lidocaine reduces the release of NO and PGI2from endothelial cells stimulated by bradykinin.
As the relaxation of vascular tissues is another property of NO and PGI2, our finding is supported by several previous reports showing that local anesthetics, including lidocaine, influence the tension of vascular tissues. Johns et al. reported that lidocaine, up to a dose of 1.0 mg/ml, and 0.1 mg/ml bupivacaine constrict vascular tissues in the cremaster muscle in vivo. Under physiologic conditions, endothelial cells in vascular tissues are stimulated by various stimulants (e.g., neural stimuli, inflammatory chemical mediators, and blood flow) to form NO and PGI2, which in turn reduce vascular tone. Therefore, the in vivo vasoconstricting effect of local anesthetics may be partly explained by the inhibition of NO and PGI2release from endothelial cells stimulated by various factors. Johns further confirmed that these agents inhibit endothelium-dependent vascular relaxation evoked by methacholine or A23187 (a Calcium2+ ionophore) in the rat aorta in vitro. However, it is still not known how lidocaine inhibits the release of NO and PGI2from endothelial cells. Several mechanisms may explain the inhibitory effect of lidocaine on the synthesis of NO and PGI2by endothelial cells. NO is enzymatically produced from L-arginine by NO synthase in endothelial cells. The enzyme is classified into two subtypes: constitutive NO synthase and inducible NO synthase. The constitutive type is present in endothelial cells, and requires increased intracellular Calcium2+ concentration to enhance NO formation. In the same manner, phospholipase A2, whose activation is required for PGI sub 2 production, is a Calcium2+ -dependent enzyme. Phospholipase C, a key enzyme for signal transduction as various agonists enhancing NO and PGI2release from endothelial cells (e.g., bradykinin and A23187), stimulates endothelial cells to increase intracellular Calcium2+ concentration through phospholipase C activation. In contrast, lidocaine and other local anesthetics inhibit the activity of these phospholipase(s). [21,22]These findings may explain in part how lidocaine inhibits bradykinin-evoked endothelial function. Lidocaine also acts as an antagonist of calmodulin or other Calcium2+ -regulated proteins. [23-25]This finding may support our results because constitutive NO synthase has the property of a calmodulin-dependent enzyme. [26,27].
In contrast to the inhibitory effect of lidocaine on antiaggregation derived from bradykinin-stimulated endothelial cells, such inhibition was not detected in nonstimulated cells in the current study. Although a basal release of NO and PGI2from endothelial cells was evident from our result showing a significant antiaggregation derived from nonstimulated cells, lidocaine did not inhibit the release even at a dose of 10 mg/ml. However, the effect of basal release of these agents was small (sevenfold less than that from stimulated cells). Thus, whether the contrast poses a specific effect of lidocaine on receptor-mediated stimulation or whether it is attributable to a problem in experimental sensitivity is an unresolved issue in the current study.
In conclusion, our results suggest that lidocaine has an inhibitory effect on antiaggregation derived from cultured porcine aortic endothelial cells.
The authors thank Dr. T. Mizuseki (Hiroshima Prefectural Rehabilitation Centre, Saijo, Japan) for his helpful suggestions. They also thank K. Hori for her assistance in preparing this manuscript and thank Dr. F. G. Issa (Word-Medex, Newtown, NSW, Australia) for his assistance in reading and editing the manuscript.