Thrombus formation and platelet aggregation at the site of vascular injury locally trigger the liberation of high concentrations of serotonin (5-HT) from platelets, and the released 5-HT induces vasoconstriction, platelet aggregation, increase of vascular permeability, and cell proliferation (1-4). In patients with peripheral vascular disease (PVD), high plasma levels of 5-HT were clinically observed (5,6), which may suggest release of significantly high concentrations of 5-HT at the site of vascular injury in this disease. Moreover, factors such as age, atherosclerosis, and hypertension are known to augment 5-HT-induced vasoconstriction (7-10), and they are frequently associated with PVD. The collateral vessels in these patients are also exquisitely sensitive to the vasoconstrictive effects of 5-HT (11-15). Therefore 5-HT could be one of the relevant factors that are closely related to the pathophysiology of PVD.
These pathophysiologic effects of 5-HT are all mediated by the 5-HT2A receptor (16-18). Although ketanserin substantially attenuates these 5-HT-dependent phenomena (11-17) and has been shown to have potential effects in the treatment of patients with PVD in several studies (19,20), it was shown that this agent has an adverse activity of prolongation of an electrocardiogram QT interval (19,20), which is a serious side effect in clinical use.
AT-1015, N-[2-[4-(5H-dibenzo[a,d]cyclohepten-5-ylidene)piperidino]ethyl]-1-formyl-4-piperidinecarboxamide monohydrochloride monohydrate (Fig. 1), is a novel 5-HT2 receptor antagonist, and preliminary experiments have revealed that AT-1015 has high affinity for 5-HT2A, 5-HT2B, and 5-HT2C receptors and does not affect the QT interval in an animal model. In this study, we demonstrated a blocking effect of AT-1015 on 5-HT2A receptor-mediated platelet aggregation and vasoconstriction in various in vitro and ex vivo experiments in rats. Moreover, we evaluated its inhibitory effect on laurate-induced peripheral vascular lesions (PVLs) in rats in comparison with ketanserin, sarpogrelate, and antiplatelet agents such as cilostazol, beraprost, and aspirin.
Male Wistar rats were obtained from Japan SLC, Inc. (Shizuoka, Japan) and maintained under specific pathogen-free conditions. The procedures involving the care and use of animals in this study was approved by Institutional Animal Care and Use Committee of Pharmaceutical Research Laboratories of Ajinomoto Co. before the study.
In vitro human platelet aggregation
Human blood from normal healthy volunteers was collected by venipuncture into a 3.8% trisodium citrate solution (9:1 vol/vol). Platelet-rich plasma (PRP) and platelet-poor plasma (PPP) were prepared by centrifugation at 1,100 rev/min for 10 min and 2,700 rev/min for 15 min, respectively. Platelet number of PRP was adjusted to 250,000 cell/μl by dilution with PPP. Platelet aggregation was measured with an aggregometer (NBS HEMA TRACER 801; Niko Bioscience Inc., Tokyo, Japan) under constant stirring of 225 μl of PRP at 1,000 rev/min at 37°C. Two microliters of test agent was added 3 min before the addition of 25 μl of agonist. Aggregation was performed at pH 7.8. The IC50 value (concentration required to inhibit 50% of platelet aggregation) of the test agent was calculated by linear regression of the concentration-response curve.
Rat platelet aggregation
Blood was collected into a 3.8% trisodium citrate solution (9:1 vol/vol). PRP and PPP were prepared by centrifugation at 1,300 rev/min for 10 min and at 3,000 rev/min for 15 min, respectively. For the platelet aggregation study, PRP was adjusted to 500,000 cell/μl. The platelet aggregation test was performed in the same method as already described. For the ex vivo experiments, PRP and PPP samples were prepared after oral administration of the test agents. The ID50 value (dose required to inhibit 50% of platelet aggregation) of the test agent was calculated by linear regression of the concentration-response curve.
Vascular contraction in rat aorta
Rats were killed by cervical dislocation, and the thoracic aortae were quickly removed and placed in the O2-saturated (95% O2/5% CO2) modified Tyrode's solution (in mM: NaCl, 158.3; KCl, 4.0; CaCl2, 2.0; NaH2PO4, 0.42; MgCl2, 1.05; glucose, 5.0; and NaHCO8, 10). Deendothelialized vascular rings were prepared and mounted in organ baths containing the modified Tyrode's solution bubbled with 95% O2/5% CO2 at 37°C, and connected to the force transducer (T7-8-240; Orientec, Tokyo, Japan) to measure isometric force. The ring was stretched and maintained at 2 g and allowed to equilibrate for 2 h. After confirming no relaxation response to acetylcholine (ACH; 10−6M), the ring was incubated with an appropriate concentration of test agent for 2 h. Responses to incremental concentrations of 5-HT in the presence of the agent were measured, and the pA2 values were determined according to the Schild equation (21), with pKB values being calculated from the equation of (antagonist)/(dose ratio − 1).
In the experiments using other agonists, a control contractile response curve for each agonist [KCl, norepinephrine (NE), U-46619, and prostaglandin F2α (PGF2α)] was obtained for each tissue. After washing 3 times, the tissue was then incubated with AT-1015 for 2 h, and the contractile response for each agonist was estimated again in the presence of At-1015. In an experiment of ex vivo 5-HT-induced vascular contraction, thoracic aorta rings were prepared after oral administration of a test agent, and responses to incremental concentrations of 5-HT were measured. The maximal tension attained at 3 × 10−7M NE was considered as 100%.
Laurate-induced PVL model in rats
Laurate-induced PVL was induced according to the previously described method (22) with minor modifications. A rat after overnight fasting was anesthetized with pentobarbital (40 mg/kg, i.p.), and the right femoral artery was freed from surrounding tissue. For inducing PVL, 0.15 ml of laurate (10 mg/ml in physiologic salt solution) was injected into the distal side of the artery. Test agents were orally administered 0.5 h (beraprost), 1 h (ketanserin, sarpogrelate, and aspirin), and 2 h (AT-1015 and cilostazol) before the injection of laurate. These agents were administered once a day on days 1-10 after the injection of laurate, except for beraprost, which was administered twice a day. The treated hindlimb was macroscopically examined on days 3 and 10 after the laurate injection. The progress of the lesion was assessed using a 5-point graded scoring system as follows. Grade 0, normal appearance; grade 1, affected region was limited to the nail parts; grade 2, affected region was limited to the toes; grade 3, necrosis of toes; grade 4, falling off of toes. Each toe was assessed and scored, and the sum of the scores for the five toes was used as the lesion index. If the lesion developed in the sole of the foot, a further 5 points were added.
Agents and chemicals
AT-1015 and sarpogrelate was synthesized at Ajinomoto Co. (Kawasaki, Japan) and Hodogaya Contract Laboratory Co. (Tsukuba, Japan), respectively. Ketanserin was obtained from Research Biochemicals International (Natick, MA, U.S.A.). Cilostazol and beraprost were purified from tablets of Pletaal (Otsuka Pharmaceutical Co., Osaka, Japan) and Dorner (Yamanouchi Pharmaceutical Co., Tokyo, Japan), respectively.
5-HT, U-46619, arachidonic acid (AA), PGF2α, epinephrine (EPI) and thrombin were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Adenosine 5′-diphosphate (ADP) was obtained from MC Medical Co. (Tokyo, Japan). Collagen reagent Horm was obtained from Nycomend Arzneimittel GMBH (Munchen, Germany). ACH, NE, and aspirin were obtained from Wako Pure Chemical Co. (Osaka, Japan). Laurate was obtained from Tokyo Chemical Industry Co. (Tokyo, Japan).
The results are expressed as the mean ± SEM. In the ex vivo study, the AT-1015-treated group was compared with the control group using one-way analysis of variance (ANOVA), followed by Dunnett's (two-tailed) post hoc test. For the laurate-induced model, the test agent-treated group was compared with a control group by a nonparametric Kruskal-Wallis test, followed by Dunnett's test. In all comparisons, the difference was considered to be statistically significant at p < 0.05.
Inhibition of in vitro human and rat platelet aggregation
Inhibitory effects of AT-1015 on human and rat in vitro platelet aggregation are summarized in Table 1. AT-1015 inhibited 5-HT-enhanced human platelet aggregation induced by collagen or ADP in a concentration-dependent manner, and IC50 values (0.22 and 0.13 μM, respectively) were almost equivalent to ketanserin (0.16 and 0.37 μM, respectively) and ∼100 times more potent than sarpogrelate (IC50, 22.0 and >100 μM, respectively). The inhibitory activities of AT-1015 were also observed in rat platelet aggregation, and IC50 values were 0.19 and 0.19 μM, respectively, which were almost equivalent to ketanserin (0.21 and 0.059 μM, respectively), and ∼100 times more potent than sarpogrelate (15.1 and 31.4 μM, respectively).
AT-1015 moderately inhibited collagen-induced human platelet aggregation (IC50, 43.8 μM), and second wave of human platelet aggregation induced by epinephrine (EPI; IC50, 3.4 μM), whereas AT-1015 did not significantly affect the aggregation induced by ADP, thrombin, U-46619, or AA, and the first wave of aggregation induced by EPI.
Inhibitory effect on ex vivo platelet aggregation in rats
AT-1015 (0.1-3 mg/kg, p.o.) inhibited rat ex vivo platelet aggregation induced by 5-HT and collagen in a dose-dependent manner, and significant inhibition was observed from 0.3 mg/kg (Fig. 2). Ketanserin (0.1-10 mg/kg) and sarpogrelate (30-300 mg/kg) also inhibited in a dose-dependent manner (data not shown). The ID50 values of AT-1015, ketanserin, and sarpogrelate were 0.16, 0.13, and 68.8 mg/kg, respectively. The oral antiplatelet effect of AT-1015 depending on 5-HT2A antagonism was approximately the same as ketanserin and ∼400 times more potent than sarpogrelate.
Inhibitory effects on agonist-induced in vitro vascular contraction of rat aorta
As shown in Fig. 3, AT-1015 slightly reduced maximal contraction induced by 5-HT and caused a rightward shift of the concentration-response curve (pKB value, 9.5), whereas ketanserin and sarpogrelate showed competitive antagonism without affecting the maximal contractile response (pA2 value, 9.3 and 8.7, respectively). AT-1015 had no effects on the contraction induced by KCl (50 mM), NE (3 × 10−7M), PGF2α (1 × 10−6M), or U-46619 (3 × 10−6M), even at 10−6M (data not shown).
Inhibitory effects on 5-HT-induced ex vivo vascular contraction of rat aorta
After oral administration of test agent, AT-1015 significantly inhibited the ex vivo vascular contraction induced by 5-HT. AT-1015 at doses of 1 and 3 mg/kg produced a rightward shift of the concentration-response curve. A significant reductions in the maximal contractile response were also observed by 35.5% (1 mg/kg) and 54.6% (3 mg/kg), respectively (Fig. 4A). However, ketanserin (10 mg/kg, p.o.) and sarpogrelate (300 mg/kg, p.o.) did not significantly inhibit the vascular contraction (Fig. 4B).
Laurate-induced PVL model in rats
When laurate was injected into the femoral artery (day 0), the limb turned pale. One day after the injection (day 1), fingers changed to dark red or violet color, and edema of the paw was observed. Then they turned black, and the whole paw became progressively black. We evaluated the grade of the disease on day 3 and day 10. The changes of the lesion score are shown in Table 2. In the control group (vehicle), the scores on day 3 and day 10 were 7.3-10.0 and 14.5-17.8, respectively (Table 2). AT-1015 delayed progression of the lesion and suppressed appearance of the disease, and significant effects were obtained from 1 mg/kg on both day 3 and day 10 (Table 2). Ketanserin significantly suppressed and delayed the progression of the disease at 10 mg/kg on day 10 (Table 2). Sarpogrelate was not effective, even at 300 mg/kg (Table 2). Therefore it was estimated that the effect of AT-1015 was 10 times more potent than ketanserin and ≥300 times more potent than sarpogrelate. Aspirin at doses of 30 and 100 mg/kg did not produce an apparent inhibitory effect. Cilostazol was not statistically effective at 100 mg/kg. It showed a significant effect at a dose of 300 mg/kg (Table 2). Beraprost (0.1, 0.3, and 3 mg/kg) significantly suppressed and delayed the progression of the disease. However, the efficacy was moderately decreased in a dose-dependent manner (Table 2).
In this study, we first evaluated the inhibitory effects of AT-1015 on 5-HT2A-mediated platelet aggregations and vascular contractions. AT-1015 potently inhibited 5-HT-enhanced rat and human platelet aggregations. These enhancements are reported to be mediated by 5-HT2A receptor on platelets (23), and the inhibitory effect of AT-1015 was thought to be dependent on its 5-HT2A antagonistic activity. AT-1015 moderately inhibited collagen-induced platelet aggregation and the second wave of platelet aggregation induced by EPI in humans. These inhibitions are thought to be due to 5-HT2A receptor antagonism, because other 5-HT2A antagonists, ketanserin and sarpogrelate, are reported to inhibit these aggregations (24,25). We confirmed that ketanserin moderately inhibited the second wave of human platelet aggregation induced by EPI in our study. AT-1015 did not affect aggregation induced by ADP, AA, thrombin, or U-46619. These results suggest that antiplatelet effect of AT-1015 depends on its 5-HT2A antagonism. In this study, inhibitory activities of ketanserin and sarpogrelate induced by 5-HT plus collagen were smaller than previous results (24,25). A possible explanation of the discrepancy is that the experimental conditions for platelet aggregation are different among these studies. It is reported that several experimental conditions affect the platelet aggregation. Such conditions are the reaction pH, reaction temperature, agonist concentration, and so on (26). Especially, it is reported that reaction pH affects the inhibitory activity of ketanserin (27). As in our experiments, inhibitory activities were compared at a reaction pH maintained below 7.8; inhibitory activity of ketanserin may be increased at constant pH conditions such as 7.4. It is not clear whether the reaction pH affects the inhibitory activity of AT-1015.
In the vascular contraction study, AT-1015 competitively inhibited contractile response to 5-HT with 10-25% decreases in the maximal response, whereas ketanserin and sarpogrelate showed competitive antagonism without reducing maximal response. This unique pharmacologic profile also is reported as to other 5-HT2 receptor antagonists, LY215840 and ritanserin (28,29). In these reports, several hypothetical mechanisms have been proposed to explain insurmountable antagonism. They include an action of the agent on multiple receptors, a slow dissociation of the agent from a receptor, and an allosteric modification of receptors (30,31).
AT-1015 did not inhibit the contractions induced by KCl, NE, PGF2α, or U-46619 at concentrations ≤10−6M. These results suggest that AT-1015 works as a specific 5-HT2A receptor antagonist on vascular contraction.
In ex vivo vascular contraction, significant inhibition against 5-HT-induced vascular contraction was also observed with insurmountable antagonism. AT-1015 was effective from 1 mg/kg in this experiment; on the other hand, no significant inhibitory activity was observed even at 10 mg/kg of ketanserin and 300 mg/kg of sarpogrelate. Although significant inhibitions of ketanserin and sarpogrelate were observed in ex vivo platelet aggregation, the significant inhibitory activity was not observed in ex vivo vascular contraction. These results suggest a slow dissociation of AT-1015 from the binding site on the artery, compared with ketanserin and sarpogrelate.
PVLs induced by injection of laurate into the femoral artery in rat are known to be a model for severe peripheral vascular disease (22). The injected laurate damages vascular endothelium, and then platelet aggregates are formed over the damaged vascular wall, occlusive thrombi, and concurrently vasoactive substances are released from aggregating platelets, which further enhance local ischemia by causing vasoconstriction (32). Participation of platelets in the generation and extension of the disease has been demonstrated in thrombocytopenic rats (22). Although aspirin failed to prevent the progression of the disease in this model, vasodilators such as potassium (K+) channel openers prevented the disease in this model without inhibiting platelet aggregation (33). These results suggest that both platelet aggregation and especially vasoconstriction induced by vasoactive substances released by platelets are relevant to the progression of the disease in this model. Our results confirmed the lack of effectiveness of aspirin, and effectiveness of cilostazol and beraprost, which had vasodilative and antiplatelet effects, as previously reported (34).
Yoshino et al. (33) reported that a K+-channel opener significantly improved the lesion induced by laurate and a Ca2+ channel antagonist, nilvadipine, was not effective. It is known that K+ channel openers inhibit contraction of collateral vessels more effectively than do Ca2+ channel antagonists, and they speculated that collateral vessels are probably more sensitive to the K+ channel opener than the Ca2+ channel antagonist, and collateral circulation is important for attenuating the progress of this disease (33,35). Some investigator demonstrated that the limb collateral arterial tree was strikingly sensitive to 5-HT in various species (11-15). Therefore it may be one of the mechanisms for attenuating the lesion that 5-HT2A receptor antagonists inhibit 5-HT-induced contraction of the collateral.
In our study, AT-1015 and ketanserin prevented the progression of the lesion, but the effect of sarpogrelate resulted in only a tendency to suppress the progression of the lesion without significance. In another report, a suppressive effect of sarpogrelate on the progression of the lesion in the same model was previously demonstrated (36). A possible explanation of the discrepancy is that the experimental conditions for laurate treatment differed between the two studies. Differences in the laurate treatment may lead to differences in the severity of the disease, and this may be a cause of the different efficacies of the agents. In accordance with these explanations, the potent antagonistic action of AT-1015 may be reflected also in this model in comparison with sarpogrelate.
The anti-5-HT2A activity of AT-1015 on platelet aggregation was almost the same as that of ketanserin in the ex vivo study, but the preventive effect of AT-1015 was 10 times more potent than that of ketanserin. Although the reason for this difference could not be clarified in our study, some possible explanations are possible, as follows. In the vascular contraction studies, AT-1015 reduced the maximal contraction in a dose-dependent manner, which is different from that of competitive antagonists such as ketanserin, and the inhibitory effect on 5-HT was stronger than ketanserin. This may suggest that the inhibition of vasoconstriction induced by 5-HT is important in this disease model and that AT-1015 efficiently inhibits the stimulation of the artery by 5-HT that is locally released in high concentration from platelets. The second possibility is that the duration of the 5-HT2A receptor antagonistic effect of AT-1015 at the site of vascular lesions may be longer than that of ketanserin, as observed in our ex vivo vascular contraction study. A third possibility is that AT-1015 may exert its inhibitory effect through mechanisms other than 5-HT2A receptor antagonism.
In summary, we have clarified that AT-1015, a newly developed 5-HT2 receptor antagonist, selectively and strongly inhibited 5-HT2A receptor-mediated platelet aggregation and vasoconstriction. We also demonstrated that AT-1015 ameliorated PVLs in the laurate-induced rat model and that its effect was more potent than other 5-HT2A receptor antagonists and anti-platelet agents. These findings indicated that AT-1015 could be developed as a useful agent for the therapy for PVD in humans.
1. Cohen ML, Fullar RW, Wiley KS. Evidence for 5-HT2
receptors mediating contraction in vascular smooth muscle. J Pharmacol Exp Ther
2. De Clerck F, Xhonneux B, Leysen JE, Janssen PAJ. Evidence for functional 5-HT2
receptor sites on human blood platelets. Biochem Pharmacol
3. De Clerck F, Van Gorp L, Beetens J, Reneman RS. Platelet-mediated vascular permeability in the rat: a predominant role for 5-hydroxytryptamine. Thromb Res
4. Nemecek GM, Coughlin SR, Handley DA, Moskowitz MA. Stimulation of aortic smooth muscle cell mitogenesis by serotonin
. Proc Natl Acad Sci U S A
5. Barradas MA, Gill DS, Fonseca VA, Mikhailidis DP, Dandona P. Intraplatelet serotonin
in patients with diabetes mellitus and peripheral vascular disease
. Eur J Clin Invest
6. Turpie AGG, de Bore AC, Genton E. Platelet consumption in cardiovascular disease. Semin Thromb Hemost
7. Vanhoutte PM. Aging and vascular responsiveness. J Cardiovasc Pharmacol
8. Heistad DD, Armstrong ML, Marcus ML, Piegors DJ, Mark AL. Augmented responses to vasoconstrictor stimuli in hypercholesterolemic and atherosclerotic monkeys. Circ Res
9. Henry PD, Yokoyama M. Supersensitivity of atherosclerotic rabbit aorta to ergonovine: mediation by a serotonergic mechanism. J Clin Invest
10. Van Neuten JM, Janssen WJ. Augmentation of vasoconstrictor response to serotonin
by acute and chronic factors: inhibition by ketanserin. J Hypertens
11. Schaub RG, Meyers KM, Sande RD. Serotonin
as a factor in depression of collateral blood flow following experimental arterial thrombosis. J Lab Clin Med
12. Verheyen A, Vlaminckx E, Lauwers F, Van Den Broeck C, Wouters L. Serotonin
-induced blood flow changes in the rat hindlegs after unilateral ligation of the femoral artery: inhibition by the S2
receptor antagonist ketanserin. Arch Int Pharmacodyn Ther
13. Janssen PA. Pharmacology of potent and selective S2
-serotonergic antagonists. J Cardiovasc Pharmacol
14. Nevelsteen A, De Clerck F, Loots W, De Gryse A. Restoration of post-thrombotic peripheral collateral circulation in the cat by ketanserin, a selective 5-HT2
receptor antagonist. Arch Int Pharmacodyn Ther
15. Orlandi C, Blackshear JL, Hollenberg NK. Specific increase in sensitivity to serotonin
of the canine hindlimb collateral arterial tree via the 5-hydroxytryptamine-2 receptor. Microvasc Res
16. Wiernsperger N, Rapin JR. 5-HT2
receptors in cerebral and peripheral vascular pathology: a review. J Vasc Med Biol
17. Holmsen H. Platelet activation and serotonin
. In: Vanhoutte PM, ed. Serotonin and the cardiovascular system.
New York: Raven Press, 1985:75-86.
18. Golino P, Ashton JH, Buja LM, et al. Local platelet activation causes vasoconstriction
of large epicardial canine coronary arteries in vivo: thromboxane A2
are possible mediators. Circulation
19. Frishman WH, Huberfeld S, Okin S, Wang YH, Kumar A, Shareef B. Serotonin
antagonism in cardiovascular and noncardiovascular disease. J Clin Pharmacol
20. Brogden RN, Sorkin EM. Ketanserin: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in hypertension and peripheral vascular disease
21. Arunlakshana O, Schild HO. Some quantitative uses of drug antagonist. Br J Pharmacol
22. Ashida S, Ishihara M, Ogawa H, Abiko Y. Protective effect of ticlopidine on experimentally induced peripheral arterial occlusive disease in rats. Thromb Res
23. Li N, Wallen NH, Ladjevardi M, Hjemdahl P. Effects of serotonin
on platelet activation in whole blood. Blood Coagul Fibrinol
24. Hara H, Osakabe M, Kitajima A, Tamao Y, Kikumoto R. MCI-9042, a new antiplatelet agent is a selective S2
-serotonergic receptor antagonist. Thromb Haemost
25. De Clerck F, Xhonneux B. Effects of ketanserin, a selective 5-HT2
serotonergic antagonist, on the secondary recruitment of human platelets in vitro. Agents Actions
26. Han P, Ardli NG. The influence of pH, temperature, and calcium on platelet aggregation
: maintenance of environmental pH and platelet function for in vitro studies in plasma stored at 37°C. Br J Haematol
27. De Clerck F, Xhonneux B, Tollenaere JP, Janssen PAJ. Dependence of the antagonism at human platelet 5-HT2
receptors by ketanserin on the reaction pH. Thromb Res
28. Cohen ML, Robertson DW, Bloomquist WE, Wilson HC. LY215840, a potent 5-hydroxytryptamine (5-HT)2
receptor antagonist, blocks vascular and platelet 5-HT2
receptors and delays occlusion in a rabbit model of thrombosis. J Pharmacol Exp Ther
29. de Chaffoy de Courcelles D, Leysen JE, Roevens P, Van Belle H. The serotonin
receptor: a receptor coupling model to explain insurmountable antagonist effect. Drug Dev Res
30. Wienen W, Mauz AB, Van Meel JC, Entzeroth M. Different types of receptor interaction of peptide and nonpeptide angiotensin II antagonists revealed by receptor binding and functional studies. Mol Pharmacol
31. Panek RL, Lu GH, Overhiser RW, Major TC, Hodges JC, Taylor DG. Functional studies but not receptor binding can distinguish surmountable from insurmountable AT1
antagonism. J Pharmacol Exp Ther
32. Shore PA, Aipers HS. Platelet damage and release of amine by microgram quantities of certain fatty acids. Fed Proc
33. Yoshino T, Ohta H, Jinno Y, Torii Y, Ogawa N, Izawa T, Okada Y. Protective effect of the K+
channel opener KRN4884 on peripheral occlusive arterial disease in rats. Gen Pharmacol
34. Murai T, Muraoka K, Saga K, et al. Effect of beraprost on peripheral circulation insufficiency in rats and rabbits. Arzneimittelforschung
35. Randall MD, Griffith TM. Effects of BRL 38227, sodium nitroprusside and verapamil on collateral perfusion following acute arterial occlusion in the rabbit isolated ear. Br J Pharmacol
36. Hara H, Shimada A, Kitajima A, Tamao Y. Effect of (±)-2-(dimethylamino)-1-[[o
-(m-methoxy-phenethyl)phenoxy]methyl]ethyl hydrogen succinate on experimental models of peripheral obstructive disease. Arzneimittelforschung