Biochimica et Biophysica Acta, 1052(1990) 351-359
351
Elsevier BBAMCR12695
Inhibition of mural thrombus formation by novel nipecotoylpiperazine antiplatelet agents C h a r l e n e K. O w e n s 1, L a r r y V, M c l n t i r e I a n d A n d r e w L a s s l o 2 1Department of Chemical Engineering (Biomedical Engineering Laboratory) Rice Unioersity, Houston, and 2Department of Medicinal Chemistry, Health Science Center, The University of Tennessee, Memphis, TN (U.S.A.)
(Received 13 November1989)
Key words: Platelet;Collagen;Muralthrombogenesis;Antiplateletagent; Nipecotoylpiperazine
The effectiveness of two dosely related nipecotoylpiperazine derivatives, BPAT-143 and BPAT-II7, as antiplatelet agents was measured by their ability to inhibit the accumulation of human blood platelets on collagen-coated (type 1) glass in a parallel plate flow chamber. Whole human blood, with fluorescently labeled platelets, was perfused through the flow chamber, and epi-fluorescent video microscopy was used to visualize the dynamics of individual platelet adhesion and thrombus formation on the collagen-coated surface. Digital image processing was used to analyze the dynamics of thrombus growth on the surfaces. The collagen-coated surface serves as a model for the damaged blood vessel wall, as collagen is a primary component of the matrix beneath endothelial cells. At a concentration of 50 ttM, BPAT-117 (the considerably more hydrophobic molecule) inhibited platelet accumulation by striking 90 4- 2% ( + S.E.), while it took 2to 4-fold higher concentrations of BPAT-143 to register meaningful to comparable effects (52 + 6% and 80-1-4%, respectively). This further corroborates the substantial impact of hydrophobic features within the matrix of appropriately structured molecules on their ability to alter platelet function.
Introduction The importance of platelets in hemostasis is well known. The platelet's reactivity must be great enough to maintain blood vessel integrity while being focused sufficiently to restrict thrombus formation only to the site of injury. Excessive formation of platelet thrombi may restrict necessary blood flow, or the thrombi may embolize from the vessel wall and be carried downstream to occlude flow in the vasculature. The potential for these events to occur is present in surgery and other instances in which injury to the vessel wall may occur [1,2], in blood exposed to prosthetic devices [3], and in the evolution of atherosclerotic lesions in the blood vessel wall [4,5]. In this study, two nipecotoylpiperazine derivatives (N, N '-bis(1-decylnipecotoyl)piperazine dihydriodide (BPAT-117), and N,N'-bis(1-hexylnipecotoyl)pipera. zine dihydriodide (BPAT-143)), and a reference compound (chlorpromazine) were evaluated for their ability
Correspondence: L.V. Mclntire, Department of Chemical Engineerhag, RiceUniversity,P.O. Box 1892, Houston,Texas 77251, U.S.A.
to inhibit the formation of platelet thrombi on collagen-coated glass. These compounds have previously been evaluated for potency in ADP- and thrombin-stimulated aggregations of platelet rich plasma (PRP) (4-min incubation) in the presence of ethanol (0.095% v/v) [6-9]. The objective of this study was to visualize and quantify platelet adhesion and aggregation on collagen-coated surfaces, using whole blood and physiological flow conditions, to establish whether the antiplatelet compounds were equally effective in a mural thrombus model. Epi-fluorescent videomicroscopy was used to visualize real-time thrombus growth in a parallel plate flow chamber under well characterized fluid flow conditions. Quantification of platelet adhesion was possible by assaying the collagen-coated slide for platelets after the 2-rain flow period, and by digital image processing analysis of the video recording taken throughout the flow period. Rather than dissolving the antiplatelet compounds in an ethanol solution, as they had been in the aggregation studies, they were dissolved in distilled water or saline. This change was necessary since we have shown that bulk ethanol concentrations as low as 0.02%, v/v, inhibit 48 + 8% ( + S.E.) of mural thrombus formation in the parallel plate flow chamber model employed here [10].
0167-4889/90/$03.50 © 1990 ElsevierSciencePublishersB.V. (BiomedicalDivision)
352 Materials and Methods
Blood was drawn from the antecubital vein of healthy non-smoking aspirin-free volunteers into heparin (anticoagulant; heparine sodium, Elkins-Sinn, Cherry Hill, N J) and mepacrine (platelet specific fluorescent probe; quinacrine dihydrochloride, Sigma Chemicals, St. Louis, MO) to yield final bulk concentrations of 10 Units/ml and 10 #M, respectively. Mepacrine has little or no effect on platelet function at this concentration [11]. The blood was stored in polypropylene test-tubes at room temperature. 20 rain before perfusion, it was moved to a water bath at 37 ° C. All blood was used between 30 rain and a maximum of 5 post-venipuncture. Control experiments run at the beginning and end of this time period were identical with respect to platelet accumulation. Before being assembled into the base of the flow chamber, a glass cover slip (Coming Glass Works, Coming, NY; No. 1, 24 × 50 mm) was coated with a solution of fibrillar, type 1, insoluble collagen from bovine achilles tendon (Sigma, C-9879) in glacial acetic acid. The concentration of collagen in this fibrillar suspension was determined, by hydroxyproline assay (12), to be approx. 2125/~g/ml in a 0.552 mol/1 acetic acid solution. The collagen solution was spread over all but the first 1.5 cm of the slide's surface. Care was taken to make the interface between the bare glass portion of the cover slip and the collagen-coated portion of the cover slip as square as possible. The coated slide was set aside in a humid environment for a minimum of 45 min before the supernatant was rinsed away with 10 ml of sterile isotonic saline. The collagen concentration remaining on the slide was determined to be approx. 18.5/~g/cm2. BPAT-143 (F.Wt. C28H54N40212, 732.57) and chlorpromazine hydrochloride (F.Wt. C17H20NEC12S, 355.32; Sigma Cat. No. C-8138, Lot No. 71F-7704) were evaluated at blood concentrations of 100 and 200/tM, and BPAT-117 (F.Wt. C36H70N402I 2, 844.78) was evaluated at a blood concentration of 50/~M. BPAT-143 and BPAT-117 were designed and synthesized in our laboratories and have been discussed in detail elsewhere [6-9]. Chlorpromazine and BPAT-143 solutions were prepared the morning of the experiment by adding the appropriate amount of the compound to be tested to a 1 ml volumetric flask. The flask was filled 4/5 full by adding distilled water, immersed intermittently into hot water maintained on the verge of boiling, and agitated until the solid entered solution. The flask was allowed to cool and distilled water was added to a final volume of 1 ml. Final test media concentrations of 100 ~M and 200/~M were attained by introducing 12 t~l and 24/~1 of the antiplatelet agent solution to 12 ml whole blood, respectively. Due to the relative insolubility of BPAT117, as compared to those of BPAT-143 and chlor-
promazine, the following procedure was employed for the preparation of its solutions: BPAT-117 solutions were prepared the morning of the experiment by adding 15 ml distilled water to 10.56 mg compound and heating in a hot water bath until the solid entered solution. 5 ml of 3.6% NaC1 solution were added to the mixture to produce a final NaC1 concentration of 0.9%. It was necessary to use 0.9% saline solution as a solvent (rather than distilled water) in order to maintain osmotic balance, as addition of this BPAT-117 solution (1,044/~1 to 12 ml whole blood) - yielding a 50/~M final concentration - reduced the whole blood content of the test medium to 92%. The test solutions were maintained at appropriately elevated temperatures, as needed, to assure that the entire quantity of each compound remained solvated. Precisely 5 min prior to perfusion, the appropriate amount of the compound to be tested was injected into the whole blood sample (12.0 ml) and gently mixed. A control in which the corresponding amount of antiplatelet-agent-free vehicle was injected into the blood, was executed in each instance (e.g., each day, for each donor). The flow chamber used has been previously described [11,13]. The flow chamber was assembled and filled with sterile isotonic saline to prevent blood-air contact. A syringe pump (Model 935, Harvard Apparatus, South Natick, MA) was used to draw the blood from the test tube reservoir through the flow chamber, displacing the saline, at a constant wall shear rate of 1000/s. This shear rate corresponds physiologically to a high venous shear rate or a low arteriolar shear rate [14]. A silicon-intensified target video camera (Model C-1012, Hamamatsu, Waltham, MA) was connected to an inverted stage microscope (DIAPHOT-TMD, Nikon, Garden City, NY) and epi-fluorescence illumination was used to visualize thrombus growth during the 2-rain perfusion period. This was possible in whole blood since mepacrine is taken up by the dense granules of platelets and the granules of leukocytes, while any fluorescence from within the erythrocyte is quenched by hemoglobin. Macroscopic measurement of end-point platelet accumulation was performed by scanning the slide and continuously reading the whole field intensity with a photodiode (Model PIN-10DP/SB, United Detector Technology, Hawthorne, CA) connected to the front camera port of the microscope. Signals from this photodiode were sent to a chart recorder via an amplifier (Model 101C, United Detector Technology). It was assumed that the fluorescence intensity was proportional to the platelet density on the surface. However, this assumption may result in small underestimates of the sizes of large thrombi, because some intensity is lost due to reabsorption of emitted light and portions of the thrombi which are out-of-focus [11]. After perfusion, the collagen-coated slide used in
353 each experiment was rinsed, in 20 ml isoton (electrolyte solution; Coulter Diagnostics, Hialeah, FL) with 6 drops zap-oglobin II (stromatolysing reagent; Coulter Diagnostics), to eliminate red blood cells from the surface, and gently crushed into 1 ml of 1% Triton X-100 (cell lysing buffer; Sigma). The sample was then sonicated for 5 s and centrifuged for 10 min at 1300 rpm. An LDH assay (LD-14 PL Kit; Gilford Systems, Oberlin, OH) on this supernatant, and on a control with known platelet concentration, was used to determine the number of platelets adhering to the slide. From the number of platelets on the slide and integration of the intensity distribution on the slide, as seen using the photodiode, an intensity/platelet ratio could be determined. This in turn was used to determine the platelet density profile. Biochemical assays performed on experimental and control samples are explained in greater detail elsewhere
[151. During the experiment, images were recorded with a 0.5 inch color video cassette recorder (Model BR-3100U, JVC, Industrial Audio/Video, Houston, TX). They were individually digitized and analyzed using a digital image processor (Model 327, Perceptive Systems, Houston, TX), in selected cases the output was sent to a photo module (Model 635, Dunn Instruments, San Francisco, CA) so that slides or prints of the final images could be produced. Length, width, area and average grey level intensity of individual growing thrombi were evaluated and combined with information on the intensity and size of a single platelet. Determination of the height of the thrombi and number of platelets which they contained was then possible by normalizing each intensity measurement to that of a single platelet. For calculation of average platelet composition, thrombi were considered to be in the microscope field only if their entire boundary was visible in the frame. The percentage of the total surface area covered by thrombi was calculated, for microscope fields 0.38 mm downstream from the collagen interface, by dividing the total area of the field occupied by thrombi (calculated using digital image processing as described above) by the area of the smallest rectangle enclosing all of these thrombi. Only the area of those thrombi with their entire boundary within the microscope field were used to obtain the cumulative thrombus area. Therefore, this method slightly underestimates the actual percent surface coverage when thrombi grow out of the boundaries of the frame during the experiment. All averaged data was analysed with a two-tailed, non-paired Student's t-test. Platelet densities on slides exposed to treated blood were compared, point by point, to the corresponding densities on slides exposed to untreated blood. This same approach was used to compare the average size of thrombi on treated and untreated slides in frames 0.38 mm downstream from
the interface. Differences were considered to be statistically significant for P values less than 0.05. Results
Platelet density profiles for untreated blood and blood treated with the antiplatelet agents are shown in Figs. 1-3. Each curve represents an average of all platelet density profiles obtained for the corresponding conditions. The error bars shown represent the standard error of the mean (S.E.) of the platelet density for the respective axial position and blood treatment. The control profile is an average of controls performed on the same days, with the same donors, as the corresponding profiles for treated blood. The abscissa represents the distance downstream from the collagen interface. The curves extend horizontally to an axial position of 24 mm (the end of the side) with only slight changes in magnitude from the last position shown. BPAT-143 was solvated in distilled water and tested at 100 /xM and 200 /~M bulk concentrations. It is observed that 100 /~M BPAT-143 inhibits 52 + 6% (+ S.E.) of the total platelet accumulation on the slide, in the first 3.8 mm downstream from the collagen interface, and a peak in platelet density is formed 0.4 mm downstream from the interface (Fig. 1). The platelet density profile is significantly different from the control only in the first 0.25 mm downstream from the collagen interface (P < 0.05). Two hundred micromolar BPAT-143 inhibits 80 + 4% ( + S.E.) of the accumula500
c~o g: ~
400'
o0
300"
W
O
200"
100"
II"w'~ 0
z |
0
1 Axial
i
2 Position,
|
3 mm
Fig. 1. P]atelct density profiles for untreated blood, n = 6 (11); for blood incubated with bulk concentrations of 100 pM, n = 6 (A); and 200 /~M, n = 6 (@) BPAT-143. Blood samples were incubated with BPAT-143 for 5 min, then perfused over collagen-coated glass for 2 min at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. T h e error bars represent the standard error of the mean.
354 200 180 160 -
~
140 -
°0
120-
Ioo_~
80-
~
60
o
a.
40 20 0 0
2
Axial
Position
3
4
(mm)
Fig. 2. Platelet density profiles for untreated blood, n = 7 (11); and for blood incubated with a bulk concentration of 50 # M BPAT-117, n = 9 ([3). Blood samples were incubated with BPAT-117 for 5 min, then perfused over collagen-coated glass for 2 rain at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. The error bars represent the standard error of the mean. Standard errors for 50 # M BPAT-117 are smaller than the plot symbol.
tion and prevents the formation of a maximum in platelet density near the interface (Fig. 1). The profile for treatment with 200 # M BPAT-143 is significantly different ( P < 0.02) from the control for axial positions less than 2.5 mm downstream from the interface. 400"
At a concentration of 50 /xM, BPAT-117 inhibits a striking 90 + 2% ( + S.E.) of platelet accumulation, in the first 3.8 mm downstream from the collagen interface, relative to controls in which the blood was treated with the corresponding amount of antiplatelet-agent-free NaC1 solution. Platelet density on the slide, Fig. 2, is low and uniform over the entire length of the slide, and is significantly different ( P < 0.01) from control platelet densities at all positions. Since the whole blood content of the test media was reduced more in the BPAT-117 studies (92% vs. 99.9% or 99.8% in the other experiments reported), additional experiments were performed to ensure that the comparison of results from blood treated with BPAT-117 with those from blood treated with BPAT-143 or chlorpromazine was valid. Results from experiments performed on the same day, with the same donor, with untreated blood and blood diluted 8% with NaC1 solution show that neither the amount of platelet accumulation on the slides, nor the platelet density profiles along the slides are significantly different for the two cases ( P >> 0.05, data not shown). Experiments were also run in which BPAT-143 was prepared in NaC1 solution, using the same technique as for BPAT-117, so that addition of BPAT-143 to a 100 /~M bulk concentration resulted in a reduction of the whole blood content to 92%. Data from these experiments showed that 100/~M BPAT-143 inhibited 59 + 5% ( -t- S.E., n = 9) of platelet accumulation, relative to controis in which the whole blood content was reduced to 92% with antiplatelet-agent-free saline solution. This result is not significantly different ( P >> 0.05) from that reported above for 100/~M BPAT-143 added from the more concentrated solution (52 + 6%, S.E.). The platelet
('4
E ::L
500
300
o o o
~
,,,I e~
200
E 2 "
Q.
400"
I O0
300"
Q. 200"
0
|
0
!
2
Axial
ii
8 A
i
01
i
4
Position,
S
6
8
a.
1oo-
mm
Fig. 3. Platelet density profiles for untreated blood, n = 5 (ll); for blood incubated with bulk concentrations of 100/AM, n = 7 ( e ) ; and 200 /LM, n = 2 (a) chlorproma~ine. Blood samples were incubated with chlorpromazine for 5 rain, then perfused over collagen-coated glass for 2 rain at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. The error bars represent the standard error of the mean. The standard errors for 100 # M and 200 /~M chlorpromazine are smaller than the data symbols.
0
i
-
-
-
i
20
•
•
•
i
40
•
•
-
i
•
60
-
•
i
80
.
•
-
i
100
,
-
•
120
Seconds
Fig. 4. Average platelet composition of thrombi for untreated blood (ra), blood incubated (5 rain) with 100/~M BPAT-143 ( o ) , and with 200/JM BPAT-ld3 (zx). Error bars represent the standard error of the mean. Evaluation was performed on images taken 0.38 rnm downstream from the collagen interface.
355 respectively. These histograms show the kinetics of thrombus growth and the overall distribution of platelet sizes which contributed to the averages shown in Figs. 4
400
300
E o
70t
r,-
I-
15 s), Fig. 5. Histograms of thrombus size are shown in Figs. 6 and 7 for blood treated with BPAT-143 and BPAT-117,
50
40
o ~.
30
20 10 0 60
1
t 11
• 20-90 []
351
Elsevier BBAMCR12695
Inhibition of mural thrombus formation by novel nipecotoylpiperazine antiplatelet agents C h a r l e n e K. O w e n s 1, L a r r y V, M c l n t i r e I a n d A n d r e w L a s s l o 2 1Department of Chemical Engineering (Biomedical Engineering Laboratory) Rice Unioersity, Houston, and 2Department of Medicinal Chemistry, Health Science Center, The University of Tennessee, Memphis, TN (U.S.A.)
(Received 13 November1989)
Key words: Platelet;Collagen;Muralthrombogenesis;Antiplateletagent; Nipecotoylpiperazine
The effectiveness of two dosely related nipecotoylpiperazine derivatives, BPAT-143 and BPAT-II7, as antiplatelet agents was measured by their ability to inhibit the accumulation of human blood platelets on collagen-coated (type 1) glass in a parallel plate flow chamber. Whole human blood, with fluorescently labeled platelets, was perfused through the flow chamber, and epi-fluorescent video microscopy was used to visualize the dynamics of individual platelet adhesion and thrombus formation on the collagen-coated surface. Digital image processing was used to analyze the dynamics of thrombus growth on the surfaces. The collagen-coated surface serves as a model for the damaged blood vessel wall, as collagen is a primary component of the matrix beneath endothelial cells. At a concentration of 50 ttM, BPAT-117 (the considerably more hydrophobic molecule) inhibited platelet accumulation by striking 90 4- 2% ( + S.E.), while it took 2to 4-fold higher concentrations of BPAT-143 to register meaningful to comparable effects (52 + 6% and 80-1-4%, respectively). This further corroborates the substantial impact of hydrophobic features within the matrix of appropriately structured molecules on their ability to alter platelet function.
Introduction The importance of platelets in hemostasis is well known. The platelet's reactivity must be great enough to maintain blood vessel integrity while being focused sufficiently to restrict thrombus formation only to the site of injury. Excessive formation of platelet thrombi may restrict necessary blood flow, or the thrombi may embolize from the vessel wall and be carried downstream to occlude flow in the vasculature. The potential for these events to occur is present in surgery and other instances in which injury to the vessel wall may occur [1,2], in blood exposed to prosthetic devices [3], and in the evolution of atherosclerotic lesions in the blood vessel wall [4,5]. In this study, two nipecotoylpiperazine derivatives (N, N '-bis(1-decylnipecotoyl)piperazine dihydriodide (BPAT-117), and N,N'-bis(1-hexylnipecotoyl)pipera. zine dihydriodide (BPAT-143)), and a reference compound (chlorpromazine) were evaluated for their ability
Correspondence: L.V. Mclntire, Department of Chemical Engineerhag, RiceUniversity,P.O. Box 1892, Houston,Texas 77251, U.S.A.
to inhibit the formation of platelet thrombi on collagen-coated glass. These compounds have previously been evaluated for potency in ADP- and thrombin-stimulated aggregations of platelet rich plasma (PRP) (4-min incubation) in the presence of ethanol (0.095% v/v) [6-9]. The objective of this study was to visualize and quantify platelet adhesion and aggregation on collagen-coated surfaces, using whole blood and physiological flow conditions, to establish whether the antiplatelet compounds were equally effective in a mural thrombus model. Epi-fluorescent videomicroscopy was used to visualize real-time thrombus growth in a parallel plate flow chamber under well characterized fluid flow conditions. Quantification of platelet adhesion was possible by assaying the collagen-coated slide for platelets after the 2-rain flow period, and by digital image processing analysis of the video recording taken throughout the flow period. Rather than dissolving the antiplatelet compounds in an ethanol solution, as they had been in the aggregation studies, they were dissolved in distilled water or saline. This change was necessary since we have shown that bulk ethanol concentrations as low as 0.02%, v/v, inhibit 48 + 8% ( + S.E.) of mural thrombus formation in the parallel plate flow chamber model employed here [10].
0167-4889/90/$03.50 © 1990 ElsevierSciencePublishersB.V. (BiomedicalDivision)
352 Materials and Methods
Blood was drawn from the antecubital vein of healthy non-smoking aspirin-free volunteers into heparin (anticoagulant; heparine sodium, Elkins-Sinn, Cherry Hill, N J) and mepacrine (platelet specific fluorescent probe; quinacrine dihydrochloride, Sigma Chemicals, St. Louis, MO) to yield final bulk concentrations of 10 Units/ml and 10 #M, respectively. Mepacrine has little or no effect on platelet function at this concentration [11]. The blood was stored in polypropylene test-tubes at room temperature. 20 rain before perfusion, it was moved to a water bath at 37 ° C. All blood was used between 30 rain and a maximum of 5 post-venipuncture. Control experiments run at the beginning and end of this time period were identical with respect to platelet accumulation. Before being assembled into the base of the flow chamber, a glass cover slip (Coming Glass Works, Coming, NY; No. 1, 24 × 50 mm) was coated with a solution of fibrillar, type 1, insoluble collagen from bovine achilles tendon (Sigma, C-9879) in glacial acetic acid. The concentration of collagen in this fibrillar suspension was determined, by hydroxyproline assay (12), to be approx. 2125/~g/ml in a 0.552 mol/1 acetic acid solution. The collagen solution was spread over all but the first 1.5 cm of the slide's surface. Care was taken to make the interface between the bare glass portion of the cover slip and the collagen-coated portion of the cover slip as square as possible. The coated slide was set aside in a humid environment for a minimum of 45 min before the supernatant was rinsed away with 10 ml of sterile isotonic saline. The collagen concentration remaining on the slide was determined to be approx. 18.5/~g/cm2. BPAT-143 (F.Wt. C28H54N40212, 732.57) and chlorpromazine hydrochloride (F.Wt. C17H20NEC12S, 355.32; Sigma Cat. No. C-8138, Lot No. 71F-7704) were evaluated at blood concentrations of 100 and 200/tM, and BPAT-117 (F.Wt. C36H70N402I 2, 844.78) was evaluated at a blood concentration of 50/~M. BPAT-143 and BPAT-117 were designed and synthesized in our laboratories and have been discussed in detail elsewhere [6-9]. Chlorpromazine and BPAT-143 solutions were prepared the morning of the experiment by adding the appropriate amount of the compound to be tested to a 1 ml volumetric flask. The flask was filled 4/5 full by adding distilled water, immersed intermittently into hot water maintained on the verge of boiling, and agitated until the solid entered solution. The flask was allowed to cool and distilled water was added to a final volume of 1 ml. Final test media concentrations of 100 ~M and 200/~M were attained by introducing 12 t~l and 24/~1 of the antiplatelet agent solution to 12 ml whole blood, respectively. Due to the relative insolubility of BPAT117, as compared to those of BPAT-143 and chlor-
promazine, the following procedure was employed for the preparation of its solutions: BPAT-117 solutions were prepared the morning of the experiment by adding 15 ml distilled water to 10.56 mg compound and heating in a hot water bath until the solid entered solution. 5 ml of 3.6% NaC1 solution were added to the mixture to produce a final NaC1 concentration of 0.9%. It was necessary to use 0.9% saline solution as a solvent (rather than distilled water) in order to maintain osmotic balance, as addition of this BPAT-117 solution (1,044/~1 to 12 ml whole blood) - yielding a 50/~M final concentration - reduced the whole blood content of the test medium to 92%. The test solutions were maintained at appropriately elevated temperatures, as needed, to assure that the entire quantity of each compound remained solvated. Precisely 5 min prior to perfusion, the appropriate amount of the compound to be tested was injected into the whole blood sample (12.0 ml) and gently mixed. A control in which the corresponding amount of antiplatelet-agent-free vehicle was injected into the blood, was executed in each instance (e.g., each day, for each donor). The flow chamber used has been previously described [11,13]. The flow chamber was assembled and filled with sterile isotonic saline to prevent blood-air contact. A syringe pump (Model 935, Harvard Apparatus, South Natick, MA) was used to draw the blood from the test tube reservoir through the flow chamber, displacing the saline, at a constant wall shear rate of 1000/s. This shear rate corresponds physiologically to a high venous shear rate or a low arteriolar shear rate [14]. A silicon-intensified target video camera (Model C-1012, Hamamatsu, Waltham, MA) was connected to an inverted stage microscope (DIAPHOT-TMD, Nikon, Garden City, NY) and epi-fluorescence illumination was used to visualize thrombus growth during the 2-rain perfusion period. This was possible in whole blood since mepacrine is taken up by the dense granules of platelets and the granules of leukocytes, while any fluorescence from within the erythrocyte is quenched by hemoglobin. Macroscopic measurement of end-point platelet accumulation was performed by scanning the slide and continuously reading the whole field intensity with a photodiode (Model PIN-10DP/SB, United Detector Technology, Hawthorne, CA) connected to the front camera port of the microscope. Signals from this photodiode were sent to a chart recorder via an amplifier (Model 101C, United Detector Technology). It was assumed that the fluorescence intensity was proportional to the platelet density on the surface. However, this assumption may result in small underestimates of the sizes of large thrombi, because some intensity is lost due to reabsorption of emitted light and portions of the thrombi which are out-of-focus [11]. After perfusion, the collagen-coated slide used in
353 each experiment was rinsed, in 20 ml isoton (electrolyte solution; Coulter Diagnostics, Hialeah, FL) with 6 drops zap-oglobin II (stromatolysing reagent; Coulter Diagnostics), to eliminate red blood cells from the surface, and gently crushed into 1 ml of 1% Triton X-100 (cell lysing buffer; Sigma). The sample was then sonicated for 5 s and centrifuged for 10 min at 1300 rpm. An LDH assay (LD-14 PL Kit; Gilford Systems, Oberlin, OH) on this supernatant, and on a control with known platelet concentration, was used to determine the number of platelets adhering to the slide. From the number of platelets on the slide and integration of the intensity distribution on the slide, as seen using the photodiode, an intensity/platelet ratio could be determined. This in turn was used to determine the platelet density profile. Biochemical assays performed on experimental and control samples are explained in greater detail elsewhere
[151. During the experiment, images were recorded with a 0.5 inch color video cassette recorder (Model BR-3100U, JVC, Industrial Audio/Video, Houston, TX). They were individually digitized and analyzed using a digital image processor (Model 327, Perceptive Systems, Houston, TX), in selected cases the output was sent to a photo module (Model 635, Dunn Instruments, San Francisco, CA) so that slides or prints of the final images could be produced. Length, width, area and average grey level intensity of individual growing thrombi were evaluated and combined with information on the intensity and size of a single platelet. Determination of the height of the thrombi and number of platelets which they contained was then possible by normalizing each intensity measurement to that of a single platelet. For calculation of average platelet composition, thrombi were considered to be in the microscope field only if their entire boundary was visible in the frame. The percentage of the total surface area covered by thrombi was calculated, for microscope fields 0.38 mm downstream from the collagen interface, by dividing the total area of the field occupied by thrombi (calculated using digital image processing as described above) by the area of the smallest rectangle enclosing all of these thrombi. Only the area of those thrombi with their entire boundary within the microscope field were used to obtain the cumulative thrombus area. Therefore, this method slightly underestimates the actual percent surface coverage when thrombi grow out of the boundaries of the frame during the experiment. All averaged data was analysed with a two-tailed, non-paired Student's t-test. Platelet densities on slides exposed to treated blood were compared, point by point, to the corresponding densities on slides exposed to untreated blood. This same approach was used to compare the average size of thrombi on treated and untreated slides in frames 0.38 mm downstream from
the interface. Differences were considered to be statistically significant for P values less than 0.05. Results
Platelet density profiles for untreated blood and blood treated with the antiplatelet agents are shown in Figs. 1-3. Each curve represents an average of all platelet density profiles obtained for the corresponding conditions. The error bars shown represent the standard error of the mean (S.E.) of the platelet density for the respective axial position and blood treatment. The control profile is an average of controls performed on the same days, with the same donors, as the corresponding profiles for treated blood. The abscissa represents the distance downstream from the collagen interface. The curves extend horizontally to an axial position of 24 mm (the end of the side) with only slight changes in magnitude from the last position shown. BPAT-143 was solvated in distilled water and tested at 100 /xM and 200 /~M bulk concentrations. It is observed that 100 /~M BPAT-143 inhibits 52 + 6% (+ S.E.) of the total platelet accumulation on the slide, in the first 3.8 mm downstream from the collagen interface, and a peak in platelet density is formed 0.4 mm downstream from the interface (Fig. 1). The platelet density profile is significantly different from the control only in the first 0.25 mm downstream from the collagen interface (P < 0.05). Two hundred micromolar BPAT-143 inhibits 80 + 4% ( + S.E.) of the accumula500
c~o g: ~
400'
o0
300"
W
O
200"
100"
II"w'~ 0
z |
0
1 Axial
i
2 Position,
|
3 mm
Fig. 1. P]atelct density profiles for untreated blood, n = 6 (11); for blood incubated with bulk concentrations of 100 pM, n = 6 (A); and 200 /~M, n = 6 (@) BPAT-143. Blood samples were incubated with BPAT-143 for 5 min, then perfused over collagen-coated glass for 2 min at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. T h e error bars represent the standard error of the mean.
354 200 180 160 -
~
140 -
°0
120-
Ioo_~
80-
~
60
o
a.
40 20 0 0
2
Axial
Position
3
4
(mm)
Fig. 2. Platelet density profiles for untreated blood, n = 7 (11); and for blood incubated with a bulk concentration of 50 # M BPAT-117, n = 9 ([3). Blood samples were incubated with BPAT-117 for 5 min, then perfused over collagen-coated glass for 2 rain at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. The error bars represent the standard error of the mean. Standard errors for 50 # M BPAT-117 are smaller than the plot symbol.
tion and prevents the formation of a maximum in platelet density near the interface (Fig. 1). The profile for treatment with 200 # M BPAT-143 is significantly different ( P < 0.02) from the control for axial positions less than 2.5 mm downstream from the interface. 400"
At a concentration of 50 /xM, BPAT-117 inhibits a striking 90 + 2% ( + S.E.) of platelet accumulation, in the first 3.8 mm downstream from the collagen interface, relative to controls in which the blood was treated with the corresponding amount of antiplatelet-agent-free NaC1 solution. Platelet density on the slide, Fig. 2, is low and uniform over the entire length of the slide, and is significantly different ( P < 0.01) from control platelet densities at all positions. Since the whole blood content of the test media was reduced more in the BPAT-117 studies (92% vs. 99.9% or 99.8% in the other experiments reported), additional experiments were performed to ensure that the comparison of results from blood treated with BPAT-117 with those from blood treated with BPAT-143 or chlorpromazine was valid. Results from experiments performed on the same day, with the same donor, with untreated blood and blood diluted 8% with NaC1 solution show that neither the amount of platelet accumulation on the slides, nor the platelet density profiles along the slides are significantly different for the two cases ( P >> 0.05, data not shown). Experiments were also run in which BPAT-143 was prepared in NaC1 solution, using the same technique as for BPAT-117, so that addition of BPAT-143 to a 100 /~M bulk concentration resulted in a reduction of the whole blood content to 92%. Data from these experiments showed that 100/~M BPAT-143 inhibited 59 + 5% ( -t- S.E., n = 9) of platelet accumulation, relative to controis in which the whole blood content was reduced to 92% with antiplatelet-agent-free saline solution. This result is not significantly different ( P >> 0.05) from that reported above for 100/~M BPAT-143 added from the more concentrated solution (52 + 6%, S.E.). The platelet
('4
E ::L
500
300
o o o
~
,,,I e~
200
E 2 "
Q.
400"
I O0
300"
Q. 200"
0
|
0
!
2
Axial
ii
8 A
i
01
i
4
Position,
S
6
8
a.
1oo-
mm
Fig. 3. Platelet density profiles for untreated blood, n = 5 (ll); for blood incubated with bulk concentrations of 100/AM, n = 7 ( e ) ; and 200 /LM, n = 2 (a) chlorproma~ine. Blood samples were incubated with chlorpromazine for 5 rain, then perfused over collagen-coated glass for 2 rain at a shear rate of 1000/s. The abscissa represents axial distance downstream from the collagen interface. Each profile is the average of ' n ' experiments. The error bars represent the standard error of the mean. The standard errors for 100 # M and 200 /~M chlorpromazine are smaller than the data symbols.
0
i
-
-
-
i
20
•
•
•
i
40
•
•
-
i
•
60
-
•
i
80
.
•
-
i
100
,
-
•
120
Seconds
Fig. 4. Average platelet composition of thrombi for untreated blood (ra), blood incubated (5 rain) with 100/~M BPAT-143 ( o ) , and with 200/JM BPAT-ld3 (zx). Error bars represent the standard error of the mean. Evaluation was performed on images taken 0.38 rnm downstream from the collagen interface.
355 respectively. These histograms show the kinetics of thrombus growth and the overall distribution of platelet sizes which contributed to the averages shown in Figs. 4
400
300
E o
70t
r,-
I-
15 s), Fig. 5. Histograms of thrombus size are shown in Figs. 6 and 7 for blood treated with BPAT-143 and BPAT-117,
50
40
o ~.
30
20 10 0 60
1
t 11
• 20-90 []
