Enhancement of Platelet Deposition by Cross-Linked Hemoglobin in a Rat Carotid Endarterectomy Model
Background Purified human cross-linked hemoglobin, which is now being used in clinical trials, increases mean arterial pressure through binding of nitric oxide (NO). We postulated that binding of NO by cross-linked hemoglobin (ααHb) could also increase platelet deposition at sites of subintimal injury.
Methods and Results Male Sprague-Dawley rats were infused with ααHb (0.88 g/kg, n=8) or with the NO synthase inhibitor NG-monomethyl-l-arginine (L-NMMA, 30 mg/kg, n=7) before undergoing microsurgical carotid endarterectomy. 111In-labeled platelets were infused after endarterectomy, and platelet deposition was measured 20 minutes later. In control endarterectomized rats (n=8), mean platelet deposition was 7.7±0.7×106/mm2. Platelet deposition was significantly increased above controls in rats that received ααHb (13.2±0.9×106/mm2, P=.0004) and in rats infused with L-NMMA (13.9±1.0×106/mm2, P=.0002). The increase was prevented by infusion of l-arginine (150 mg/kg) immediately after ααHb or L-NMMA. To determine whether aspirin (ASA) blocked the increased deposition induced by ααHb, rats received oral ASA (10 mg/kg) 18 hours before endarterectomy. Platelet deposition in animals receiving ASA alone was 6.4±0.9×106/mm2 (n=8). This was significantly increased to 10.8±0.8×106/mm2 (P=.002) for the ASA-treated group that received ααHb at the time of endarterectomy (n=8). The prolonged bleeding times induced by ASA were unaffected by the infusion of ααHb.
Conclusions These data suggest that in a rat endarterectomy model, ααHb increases platelet deposition at sites of subintimal injury by binding NO. Increased deposition induced by ααHb can be prevented by administration of l-arginine but not by pretreatment with aspirin.
Several different preparations of purified human hemoglobin, cross-linked to prevent renal damage, are currently undergoing clinical trials as an oxygen-carrying agent.1 In addition to binding oxygen, hemoglobin is known to have a very high affinity for NO.2 3 Infusion of hemoglobin into animals increases blood pressure,4 5 6 7 8 at least in part because of binding of NO to hemoglobin, which also results in inhibition of NO-mediated vasodilation.
In addition to moderating blood pressure, NO inhibits platelet function through increasing levels of platelet cGMP.9 10 11 12 13 Several studies suggest that this interaction can be modulated by hemoglobin. Incubation of hemoglobin with monolayers of bovine endothelial cells promotes platelet adhesion,9 and platelets obtained from rabbits that have been infused with hemoglobin demonstrate an increased response to agonists ex vivo, as measured by aggregometry.13 In vivo studies have not
been done to determine whether hemoglobin, by binding to NO, can promote platelet deposition at the site of endothelial injury. We therefore investigated the effects of ααHb on platelet deposition in rats undergoing microsurgical carotid endarterectomy.
The NO synthase inhibitor L-NMMA was obtained from Calbiochem-Novabiochem Corp and was administered intravenously in 0.5 mL normal saline. L-Arg (Sigma Chemical Co) was administered intravenously in 1.5 mL normal saline. Generic ASA powder (Mallinckrodt Chemical Works) was dissolved at a concentration of 10 mg/mL in polyethylene glycol (average molecular mass, 4 kD; Sigma Chemical Co).
Cell-free human hemoglobin, lot 91196, was produced at Letterman Army Institute of Research from outdated red blood cells by methods previously described.14 Final concentration of the hemoglobin solution was 7.0 g/dL. The hemoglobin was cross-linked between α-chains with bis(3,5-dibromosalicyl)fumarate, preserved in Ringer’s acetate, divided into aliquots, and frozen at −80°C until the day of use.
The endotoxin concentration, as determined by a kinetic turbidometric assay using limulus amebocyte lysate,14 was 0.25 endotoxin units/mL. Two samples of this preparation of hemoglobin (ααHb; molecular mass, 64 kD) passed the rabbit pyrogenicity test as administered at the Center for Biologics Evaluation and Research (Food and Drug Administration). The solution had no bacterial contamination, as shown by a sterility test conducted over a period of 1 week. The concentration of methemoglobin was 2.6%, and the pH was 7.35. The P50 was 25 mm Hg.
Plasma hemoglobin was measured with the Radiometer OSM 3 Hemoximeter. Platelets were enumerated in the Baker Hematology Series Cell Counter System 9000.
Male Sprague-Dawley rats (mean weight, 434 g) were allowed ad libitum access to water and standard rat chow until the time of anesthesia. Animal procedures were approved by the Walter Reed Army Institute of Research Laboratory Animal Care and Use Committee and were conducted in accordance with the Guide for the Care and Use of Laboratory Animals, NIH publication 86-23, revised 1985.
Preparation of Donor Platelets
Homologous platelets for each experiment were obtained from a separate donor rat, which was first anesthetized with pentobarbital (50 mg/kg IP). A femoral artery catheter was then inserted, and the donor rat was exsanguinated. Platelets from 12 mL whole blood (anticoagulated with 2 mL acid citrate dextrose) were labeled with 100 μCi 111In oxine (111In, MPI Pharmacy Inc) by a modification of the labeling technique described by Bauman and Landry.15 Briefly, platelet-rich plasma was centrifuged at 300g for 10 minutes, the supernatant was decanted, and the platelets were resuspended in acid citrate dextrose anticoagulant. 111In was then incubated with the platelet suspension for 30 minutes at 22°C. The platelets were then washed and resuspended in 2 mL native platelet-poor plasma. Mean labeling efficiency was 84±7% (mean±SD).
The animal model, which has been described previously,16 is a modification of the microsurgical carotid endarterectomy model developed by Sasaki et al.17 The experimental rats were initially anesthetized with pentobarbital at a dose of 50 mg/kg IP, with subsequent doses (20 mg/kg IP) used as necessary to maintain adequate anesthesia. Surgical procedures were performed under ×5 to ×20 magnification with a Zeiss OPMI 6 series operating microscope (Carl Zeiss, Inc, Medical Products Division).
Fig 1⇓ outlines the experimental protocol. The experimental rat was secured, and a catheter was placed in the femoral vein. Normal saline was infused through this line at a maintenance rate of 2.75 mL · kg−1 · h−1 with a model 355 Orion syringe pump (Orion Research Inc). A femoral artery catheter was then placed, and arterial pressure was monitored through a transducer with a Datascope 2002A hemodynamic monitor. The initial MAP was recorded, and then blood (2 mL) was withdrawn from the arterial line for measurement of the platelet count and plasma hemoglobin level. The blood was replaced by an equal volume of normal saline. An initial bleeding time was then determined by methods described below. A preinfusion MAP was recorded, and group-specific reagents (L-NMMA or ααHb) were then administered intravenously. The dose of ααHb administered (1.25 mL/100 g body wt, or 0.88 g ααHb/kg) was equivalent to 10% of the total hemoglobin in the rat. The ααHb was infused through the venous catheter during 5 minutes, with simultaneous withdrawal of an equal amount of blood from the arterial catheter to avoid volume changes. Postinfusion MAPs were recorded 2 minutes after the completion of each infusion.
The trachea and both common carotid arteries were exposed through a midline cervical incision, and a tracheostomy tube was placed. The right (experimental side) and left (sham-operated side) common carotid arteries were isolated and clamped with low-pressure Kleinert microvascular clamps. Two-millimeter longitudinal arteriotomies were made in both the experimental and sham-operated arteries. Stay sutures of 10-0 nylon were used to expose the lumen of the experimental carotid artery, and an endarterectomy of the back wall was performed by placement of two transverse and parallel score marks in the intima 0.5 mm apart with the point of a 27-gauge needle and removal of the intervening intima with microforceps, exposing the vessel media. Both arteriotomies were then closed with running 10-0 nylon sutures.
111In-labeled platelets prepared from the donor rat were given intravenously (1.5 mL volume) after arteriotomy closures. MAP was recorded 2 minutes after the platelet infusion. Platelets were allowed to circulate for 10 minutes before removal of the arterial clamps and exposure to the carotid endarterectomy site for 20 minutes until termination of the experiment. A bleeding time was obtained at the beginning of platelet circulation time, and 2 mL of blood was then withdrawn from the arterial catheter for repeat determination of the platelet count and plasma hemoglobin level.
After the 20-minute circulation time, a final MAP was recorded. The animal was hemodiluted by withdrawal of blood from the arterial catheter and injection of saline into the venous catheter and was perfusion-fixed with the injection into the arterial line of 20 mL 4% formaldehyde/1% glutaraldehyde. Three-millimeter sections of both carotid arteries that contained the suture lines were excised and placed in fixative. Occlusive thrombi were not present in the carotid arteries of any of the experimental animals.
After excision of the distal 1 mm of the tail, a cotton-tipped applicator was gently applied to the wound every 30 seconds for the first 5 minutes and every 10 seconds thereafter until cessation of oozing. The time at which blood was no longer evident on the swab was taken as the bleeding time.
111In-Labeled Platelet Counting
The radioactivity of the experimental and sham carotid specimens and of blood samples was determined in a TM Analytic Gamma Trac 1193 gamma well-counter programmed for 111In photon peak.
Endarterectomy Area Measurement
After measurement of radioactivity, the experimental carotid artery was opened longitudinally through the arteriotomy suture line and mounted on a silicon rubber block with minutien pins. The specimen was postfixed with osmium tetroxide, critical point–dried with carbon dioxide, and coated with gold palladium. The artery was then examined with scanning electron microscopy at ×25 to ×50 magnification, and an electron photomicrograph that contained the entire endarterectomized segment was generated. This photomicrograph was analyzed with computer image analysis software (Optimus Bioscan Inc), and the surface area of the endarterectomy was determined.
Primary Outcome Measure
The effect of the experimental treatments on platelet deposition was measured as the number of adherent 111In-labeled platelets per square millimeter of the carotid endarterectomy. The radioactivity of whole blood (cpm/μL) and platelet count were measured just before euthanasia, and the specific activity of circulating platelets (cpm/platelet) was calculated. The 111In-labeled platelet activity per square millimeter of the endarterectomy site was calculated by subtracting the activity of the sham specimen from the activity of the experimental specimen and dividing the net radioactivity by the area (in square millimeters) of the endarterectomy sites. Platelet deposition, defined as total platelets per square millimeter on the endarterectomy site was computed as platelet deposition (platelets/mm2)=cpm/mm2×platelets/cpm.
The experimental protocol was performed in two series. Series 1 (n=40) consisted of six experimental groups: (1) control (n=8); (2) infusion of L-NMMA alone (30 mg/kg, n=7); (3) infusion of L-NMMA followed by L-Arg (150 mg/kg, n=7); (4) infusion of ααHb alone (0.88 g/kg, n=7); (5) infusion of ααHb followed by L-Arg (n=7); and (6) infusion of L-Arg alone (n=4).
Series 2 (n=16) included two additional experimental groups: (1) oral administration of ASA alone (10 mg/kg, n=8) and (2) oral administration of ASA followed by intravenous ααHb (0.88 g/kg). ASA was administered orally through a gastric tube the evening before the experiment to both donor and experimental animals, and ααHb was administered before the endarterectomy as in series 1. In each series, to control for possible time effects, block randomization was used to assign animals to experimental groups. For series 1, each block contained 6 rats (1 from each experimental group), and for series 2, each block contained 2 rats (1 from each of the two experimental groups).
A total of 59 rats underwent carotid endarterectomy. Three rats died during the surgical procedure and were excluded from the analysis. One rat developed copious pulmonary secretions and respiratory distress after platelet administration, and 1 rat became hypotensive after ααHb and L-Arg infusion. The third rat, a recipient of L-NMMA and L-Arg, developed hypotension and died after the carotid arteries were unclamped.
Summary results for continuous outcomes are reported as mean±SEM. ANOVA for independent groups was used to determine the overall statistical significance of differences in mean platelet deposition.18 19 Analysis was done on transformed (square root and rank) and untransformed data with similar results. The unpaired t test was used for individual comparisons in mean platelet deposition between two groups by use of the ANOVA mean square error for untransformed data. Overall differences between the initial and final bleeding times (logarithmically transformed) among groups were assessed by repeated measures ANOVA. Paired t tests were used for mean changes in bleeding time for a single experimental group (by repeated measures mean square error). All reported probability values are two-sided; no adjustment was made for multiple testing. A value of P<.05 was considered to be significant. Statistical calculations were carried out with Minitab statistical software.20
Effect of L-NMMA and ααHb on Subintimal Platelet Deposition
We hypothesized that in animals infused with ααHb, platelet deposition could be increased because of binding of NO to circulating free hemoglobin. To determine whether increased platelet deposition could be detected in this animal model if NO were inhibited, one group was infused with L-NMMA, a known inhibitor of NO synthase. In control rats, mean platelet deposition at the endarterectomy site was 7.7±0.7×106/mm2 (Fig 2⇓), and in the group receiving L-NMMA, platelet deposition was significantly increased by 81% to 13.9±1.0×106/mm2 (P=.0002). In the group infused with ααHb, platelet deposition was 13.2±0.9×106/mm2, an increase of 71% above controls (P=.0004).
L-Arg, a substrate for NO synthase, was infused immediately after L-NMMA or ααHb to determine whether sufficient NO could be generated to prevent increased platelet deposition. In one group that received L-Arg alone, platelet deposition was less than that of the control group, although the reduction was not significant (6.1±1.1×106/mm2 versus 7.7±0.7×106/mm2, P=.42). In the group receiving ααHb followed by L-Arg, platelet deposition was 4.8±0.9×106/mm2, a value less than that of the control group (P=.028). Platelet deposition was 7.5±0.6×106/mm2 in the group infused with L-NMMA and L-Arg and was not different from the control group (P=.88). Thus, infusion of L-Arg into groups that had received ααHb or L-NMMA successfully blocked the increase in deposition, resulting in platelet deposition that was either the same as or less than that of the control group.
Additional studies were performed to determine whether ASA blocked the increased deposition noted with ααHb. Two groups received ASA 18 hours before carotid endarterectomy; one group also received ααHb just before the surgery. In the ASA-treated group that also received ααHb, platelet deposition was 10.8±0.8×106/mm2. This was significantly higher than the value of 6.4±0.9×106/mm2 for the group receiving ASA alone (P=.002) and for the control group (7.7±0.7×106/mm2, P=.016). There was no difference in platelet deposition between the control group and the rats receiving ASA alone (7.7±0.7×106/mm2 versus 6.4±0.9×106/mm2, P=.28).
Bleeding Times and Platelet Counts in Experimental Groups
Comparison of initial and final bleeding times within groups showed that for the group receiving L-NMMA, the final bleeding time was decreased by 10% from that of the initial bleeding time (673±19 versus 604±26 seconds, P=.012). In all other groups, including those receiving ASA alone and ASA and ααHb, the final bleeding time did not differ significantly from the initial bleeding time (Table 1⇓).
A comparison of bleeding times among the groups showed that the initial bleeding time in the two groups receiving ASA was significantly higher than the initial bleeding time for the control group (P<.01). For the group receiving ASA followed by ααHb at the time of endarterectomy, the initial and final bleeding times were not different from the group receiving ASA alone (P>.5).
Although no significant differences in platelet counts were obtained between the beginning and end of the experimental protocol in any of the groups, platelet counts tended to be lower at the end of the experiment in those groups that received ααHb (Table 1⇑).
Hemodynamic Effects of L-NMMA and of ααHb
In control rats, the MAP ranged from 88 to 102 mm Hg throughout the experiment (Table 2⇓). In the group receiving ααHb and L-Arg, the MAP increased from 94±5 to 118±6 mm Hg after infusion of ααHb (P=.0042); after the infusion of L-Arg, the MAP was not significantly different from the baseline value (P=.079). Similarly, for rats that received both L-NMMA and L-Arg, MAP was significantly increased after infusion of L-NMMA (P=.0012), but shortly after the infusion of L-Arg, it was reduced to values that were not different from baseline (P=.206). Pretreatment of rats with ASA before the infusion of ααHb did not affect the hemodynamic response.
In the ααHb-treated group, the MAP measured after the infusion of 111In-labeled platelets was not different from that measured at the termination of the experiment. In all other groups, the MAP decreased during this time. However, the MAP did not appear to influence platelet deposition. There was no correlation between platelet deposition and the MAP measured immediately after the infusion of 111In-labeled platelets (r=−.172, P=.22) or between platelet deposition and the absolute decrease in MAP that occurred during the time between the infusion of the 111In platelets and the termination of the experiment (r=.096, P=.48).
Plasma Hemoglobin Levels
In the three groups of rats receiving ααHb (n=22), the mean preinfusion plasma level of hemoglobin was 86±18 mg/dL. When measured at the end of the experiment (ie, 100 minutes after infusion of ααHb), the mean plasma hemoglobin level was 618±49 mg/dL (P<.0001). The mean plasma hemoglobin level for rats (n=14) not receiving ααHb was 71±16 mg/dL at the beginning of the experiment and 157±89 mg/dL at the end; the increase was not statistically significant (P=.143).
The present study demonstrates that in the rat microsurgical model, the infusion of the NO synthase inhibitor L-NMMA or ααHb before endarterectomy significantly increases platelet deposition. Administration of L-Arg after L-NMMA or ααHb abrogated this response, strengthening the association of platelet deposition with NO availability and indicating a possible role for L-Arg in modifying the ααHb-associated increase in platelet deposition.
Platelet deposition in ASA-treated rats infused with ααHb was increased above that of control rats or rats receiving ASA alone, indicating that ASA does not block the proaggregatory effects of ααHb. These findings are consistent with recent studies by Broekman et al,21 who demonstrated that NO-mediated inhibition of platelet reactivity was ASA independent.
ASA caused a significant increase in bleeding time, which was sustained even with the infusion of ααHb. Thus, the bleeding time did not accurately reflect the increased platelet deposition at the site of the endarterectomy. These data are consistent with other reports demonstrating that the bleeding time does not predict the hemostatic function of platelets in animal models22 or in patients subjected to surgery.23 24 25
MAPs paralleled those expected with changes in NO availability, with significant increases occurring shortly after infusion of L-NMMA or ααHb, followed by restoration to preinfusion values after the administration of L-Arg. Infusion of L-Arg alone caused no significant change in MAP. The effect of ααHb on MAP in this model was similar to that found in conscious rats.4
Rats received ααHb at a dose that was ≈10% of the whole blood hemoglobin; by the end of the experiment, the plasma values were equivalent to 5% of the total hemoglobin concentration. The relatively short half-life is consistent with that found in other studies.5 In control rats, plasma hemoglobin levels remained at ≤1% of the total hemoglobin throughout the experiment. Similar plasma hemoglobin values have been reported in control rats in other studies; this may be due to an increased fragility of rat erythrocytes, which results in lysis either in vivo or ex vivo.26
The role of NO in regulating platelet function has been suggested in several in vivo and in vitro studies. In vitro studies have confirmed that NO can partially inhibit platelet aggregation27 or platelet adhesion to endothelial cells.9 These effects can be reversed by the addition of hemoglobin.2 9 Hemoglobin does not appear to have a direct effect on platelet function in vitro.27 28
In one study, platelet deposition, as assessed by electron microscopy, was increased on injured endothelium of rabbits infused with L-NMMA.29 Yao et al30 found that infusion of L-NMMA into dogs with a mechanical injury and stenosis of either a coronary artery or femoral artery resulted in recurrent platelet aggregation as assessed by cyclic flow variation; the aggregation was prevented by infusion of L-Arg. Groves et al31 performed bilateral carotid angioplasty on anesthetized pigs and compared aggregation of 111In-labeled platelets between controls and animals receiving the NO donor SIN-1. Compared with controls, SIN-1 significantly reduced platelet aggregation at sites of both superficial and deep arterial injury. The use of SIN-1 was further accompanied by a significant increase in platelet cGMP.
It is unknown whether the increased platelet deposition associated with the infusion of ααHb is a dose-dependent phenomenon, since only one concentration of ααHb, which was calculated to be equivalent to a one-unit transfusion in a human recipient, was used in this study. The lowest doses at which NO will be completely inhibited have not been established. It is also unknown whether the effect of hemoglobin on platelet deposition is a property of the hemoglobin preparations that are now in clinical trials. The significance of these findings therefore remains to be investigated in human recipients of commercial products.
Selected Abbreviations and Acronyms
|ααHb||=||purified human cross-linked hemoglobin|
|ASA||=||acetylsalicylic acid (aspirin)|
|MAP||=||mean arterial pressure|
The authors gratefully acknowledge the assistance and support provided to this project by the following individuals at the Walter Reed Army Institute of Research: Edward Asafoadjei, William Mondy, and Marcos Gomez, Department of Ultrastructural Pathology; Stephen Van Albert, Division of Biometrics; and William H. Baker, LTC, DVM, Division of Pathology.
The opinions expressed herein are solely those of the authors and do not reflect official views or policies of the Department of the Army or the Department of Defense.
- Received July 26, 1995.
- Revision received November 10, 1995.
- Accepted November 15, 1995.
- Copyright © 1996 by American Heart Association
Hamilton I, Schultz SC, Cole F, Burhop K, Malcolm D. Characterization of diaspirin cross-linked hemoglobin’s pressor response. Crit Care Med. 1992;20:S106. Abstract.
Hess JR, MacDonald VW, Brinkley WW. Systemic and pulmonary hypertension after resuscitation with cell-free hemoglobin. J Appl Physiol. 1993;74:1769-1778.
Bauman JM, Landry AJ. Practical platelet labeling with indium-111-oxine. J Nucl Med Technol. 1986;14:15-17.
Armitage P, Berry G. Statistical Methods in Medical Research. 2nd ed. Boston, Mass: Blackwell Scientific Publications; 1987:90-92.
Glantz SA, Slinker BK. Primer of Applied Regression and Analysis of Variance. New York, NY: McGraw-Hill Co; 1990:272-313.
Minitab Inc. Minitab Reference Manual, Release 9 (for Windows). State College, Pa: 1993.
Broekman MJ, Eiroa AM, Marcus AJ. Inhibition of human platelet reactivity by endothelium-derived relaxing factor from human umbilical vein endothelial cells in suspension: blockade of aggregation and secretion by an aspirin-insensitive mechanism. Blood. 1991;78:1033-1040.
De Caterina R, Lanza M, Manca G, Buti-Strata G, Maffei S, Salvatore L. Bleeding time and bleeding: an analysis of the relationship of the bleeding time test with parameters of surgical bleeding. Blood. 1994;84:3363-3370.
Lind S. The bleeding time does not predict surgical bleeding. Blood. 1991;77:2547-2552.
Yao S-K, Ober JC, Krishnaswami A, Ferguson JJ, Anderson V, Golino P, Buja LM, Willerson JT. Endogenous nitric oxide protects against platelet aggregation and cyclic flow variations in stenosed and endothelium-injured arteries. Circulation. 1992;86:1302-1309.
Groves PH, Lewis MJ, Cheadle HA, Penny WJ. SIN-1 reduces platelet adhesion and platelet thrombus formation in a porcine model of balloon angioplasty. Circulation. 1993;87:590-597.