PPBP [4-Phenyl-1-(4-phenylbutyl) Piperidine], a Potent ς-Receptor Ligand, Decreases Brain Injury After Transient Focal Ischemia in Cats
Background and Purpose We tested the hypothesis that administration of 4-phenyl-1-(4-phenylbutyl) piperidine (PPBP), a potent ς-receptor ligand, during transient focal ischemia would affect early postischemic brain injury.
Methods Halothane-anesthetized cats underwent left middle cerebral artery occlusion for 90 minutes followed by 4 hours of reperfusion. Control cats received saline (n=10). Experimental cats (2 groups, n=10 per group) were treated with PPBP at a rate of 0.1 μmol/kg per hour (PPBP-0.1) or administered 1 μmol/kg per hour (PPBP-1) intravenously from 75 minutes after initiation of ischemia and continuing during the 4 hours of reperfusion.
Results As measured by the microsphere method, blood flow to the ipsilateral caudate nucleus was decreased similarly in all groups during ischemia. Blood flow to the ipsilateral inferior temporal cortex was decreased during ischemia in all groups but was higher in cats subsequently treated with PPBP at the highest dose, even before drug administration. There was no difference in blood flow to the ipsilateral caudate nucleus or inferior temporal cortex (area of greatest cortical injury) during reperfusion. Triphenyltetrazolium-determined injury volume of the ipsilateral cerebral hemisphere (control, 29±5%; PPBP-0.1, 17±3%; PPBP-1, 6±1% of ipsilateral hemisphere; mean±SEM) and caudate nucleus (control, 49±5%; PPBP-0.1, 39±6%; PPBP-1, 25±5% of ipsilateral caudate nucleus) was less in cats treated with 1 μmol/kg per hour of PPBP compared with cats treated with saline. Cats treated with 0.1 μmol/kg per hour had a 45% smaller hemispheric injury volume than the control group without differences in intraischemic blood flow. Recovery of somatosensory evoked potential amplitude was greater in cats treated with PPBP-1 compared with control (control, 18±11%; PPBP-0.1, 30±14%; PPBP-1, 54±14% of baseline).
Conclusions These data indicate that ς-receptors may play an important role in the mechanism of acute injury in both the cortex and the caudate nucleus after 90 minutes of transient focal ischemia in the cat. Because PPBP afforded protection when administered at the end of ischemia and during reperfusion, ς-receptors may contribute to the progression of injury in ischemic border regions.
- cerebral blood flow
- middle cerebral artery occlusion
- sigma receptor
- evoked potentials, somatosensory
- triphenyltetrazolium staining
Little is known about the role of ς-receptors in the mechanism of brain injury from ischemia. However, several pharmacological agents that were originally believed to have neuroprotective properties because of effects at the NMDA receptor also have reasonable affinities for ς-receptors. For example, ifenprodil and its derivative SL 82.0715 have been found to decrease infarct volume in cats and rats exposed to permanent focal ischemia.1 These agents appear to act as noncompetitive antagonists in several models of NMDA receptor–mediated activity in vivo and in vitro2 ; they are also potent ς-receptor ligands at 37°C.3 Other ς-receptor ligands have also shown promise as neuroprotective agents. However, each of these agents was only evaluated in a gerbil model of cerebral ischemia,4 5 6 which did not allow for tight control of important physiological variables, or in a rat model of anoxia,7 which may not be relevant to the clinical condition of focal ischemia. Therefore, because neuroprotection afforded by these drugs may have been secondary to effects on important physiological variables (eg, brain temperature) rather than on primary mechanisms that produce injury, further evaluation of their efficacy in the setting of ischemia was warranted.
In the present study we tested the hypothesis that the potent ς-ligand PPBP3 would prevent early evidence of brain injury in a well-characterized model of transient focal ischemia. In addition to assessment of the volume of injury, we also measured regional CBF to determine whether any observed protection by PPBP was related to a more favorable redistribution of perfusion during ischemia or reperfusion.
Materials and Methods
The study was conducted in accordance with National Institutes of Health guidelines for the use of experimental animals, and the protocols were approved by the Animal Care and Use Committee of The Johns Hopkins Medical Institutions.
Male cats weighing 2.3 to 5.4 kg were anesthetized with halothane in oxygen. They were orally intubated and mechanically ventilated to maintain Paco2 at approximately 35 to 40 mm Hg. Anesthesia was maintained with halothane (1.0% to 1.5%) in oxygen-enriched air (Fio2, 0.35 to 0.40). The concentration of anesthetic was not altered during ischemia and reperfusion. The skeletal muscle relaxant pancuronium bromide (0.2 mg/kg IV) was administered as a single dose to prevent movement during electrocautery and muscle artifact during evoked potential monitoring.
Both femoral veins were catheterized for infusion of lactated Ringer’s solution and drugs. Catheters placed in the descending aorta via a femoral artery were used for blood pressure measurement, arterial blood gas sampling, and withdrawal of reference blood samples during injection of radiolabeled microspheres. After left thoracotomy, a catheter was inserted into the left atrium for injection of radiolabeled microspheres. A vascular loop, placed around the descending aorta, was used for blood pressure control during the experimental protocol. The cat was turned prone and its head positioned in a stereotaxic frame approximately 4 cm higher than its heart. A thermistor placed in the right temporal epidural space was used to estimate brain temperature. Epidural temperature was maintained at 38.0±0.5°C with the use of a warming blanket and a heating lamp. The left MCA was exposed by a transorbital approach with the use of microsurgical techniques. To produce focal ischemia, the left MCA was occluded near its origin from the intracranial carotid artery with the use of a microvascular clip for 90 minutes.8 9 At 90 minutes of occlusion, the microvascular clip on the MCA was removed. Reperfusion was then initiated, and it lasted 240 minutes.
Arterial blood pressure was continuously monitored. Arterial pH, Paco2, and Pao2 were measured with a self-calibrating electrode system (Radiometer ABL 3). Hemoglobin and arterial oxygen content were measured with a hemoximeter (model OSM3, Radiometer). Blood glucose was measured with a glucose analyzer (model 2300A, Yellow Springs Instruments). A multichannel signal averager (model CA-1000, Nicolet Biomedical Instruments) was used to measure the SEP with foreleg stimulation, as previously described.9 10 The amplitude to the peak of the first major negative wave was measured from the peak of the preceding positive wave.
Regional CBF was measured with radiolabeled microspheres (15.5±0.5 μm diameter; Du Pont–NEN Products) by the reference withdrawal method.11 Six radioactive isotopes (153Gd, 114mIn, 113Sn, 103Ru, 95Nb, 46Sc) were injected in random sequence into each animal. Approximately 1.5×106 microspheres were injected into the left atrium over a 20-second period, followed by a 5-mL saline flush. Reference blood samples were withdrawn from the aorta at 1.9 mL/min beginning 30 seconds before the injection and continuing for 90 seconds after the saline flush.
After baseline measurements were obtained, experimental cats were assigned randomly to one of three groups. Investigators were blinded to the treatment. Each cat was exposed to 90 minutes of left MCA occlusion and 240 minutes of reperfusion. In each group, treatment with either diluent (saline) or drug was begun at 75 minutes of MCA occlusion, and the infusion was continued throughout reperfusion. In the control group, the saline diluent was infused at a rate of 4 mL/h. In the PPBP-0.1 group, 0.1 μmol/kg per hour (volume, 4 mL/h) of PPBP was infused for 15 minutes of MCA occlusion and 240 minutes of reperfusion. In the PPBP-1 group, 1 μmol/kg per hour (volume, 4 mL/h) of PPBP was infused for 15 minutes of MCA occlusion and 240 minutes of reperfusion. Cats that did not achieve at least 75% reduction in SEP amplitude with MCA occlusion were excluded from the protocol before assignment to a particular treatment group. All variables were measured at baseline, at 30 and 90 minutes of left MCA occlusion, and at 60, 180, and 240 minutes of reperfusion. Cats that did not demonstrate adequate ischemia of the caudate nucleus ipsilateral to the occlusion (blood flow ≤80 mL/min per 100 g) or ipsilateral inferior temporal cortex (blood flow ≤50 mL/min per 100 g) were excluded from further study. This elimination was completed before the investigator had knowledge of the treatment group and resulted in a similar number of cats excluded from each group (control group, 4 cats; PPBP-0.1 group, 4 cats; PPBP-1 group, 5 cats).
In a separate cohort of cats, we also evaluated the effect of PPBP infusion without intervening ischemia. Three cats were treated with saline (4 mL/h), and three cats were treated with PPBP at a rate of 1 μmol/kg per hour (volume, 4 mL/h) for a total duration of 240 minutes. These cats were prepared as those in the ischemia protocol except that they did not undergo surgery for MCA exposure. They were, however, instrumented with a catheter (PE-50) in the superior sagittal sinus for measurement of cerebral venous oxygen content and calculation of CMRO2 (the product of blood flow to the cerebrum and the arterial minus sagittal sinus oxygen content difference). In these cats, all variables were recorded at baseline and at 15, 60, 120, 180, and 240 minutes of saline or PPBP infusion.
At the end of each protocol, cats were killed with intravenous potassium chloride. In cats exposed to ischemia the brain was removed and cut immediately into 12 uniform coronal sections 3 mm thick to estimate brain injury with the TTC (Sigma Chemical Co) technique12 13 as previously described.10 14 After injury volume was estimated, the slices of brain were placed in 10% buffered formalin for 1 to 2 days before sectioning for regional CBF measurement. Ipsilateral and contralateral temporal and parietal lobes of the middle four slices were sectioned into inferior temporal, temporal-parietal, and parietal cortex; ipsilateral and contralateral caudate nucleus, brain stem, and cerebellum also were taken. The arterial microsphere reference samples and weighed tissue specimens were counted in a multichannel autogamma scintillation spectroscopy system, and blood flow was calculated by the reference sample technique.11
Values are expressed as mean±SEM. Statistical comparison to assess changes in measured physiological variables and CBF within groups was performed by repeated-measures ANOVA. Comparison of physiological variables, blood flow, and injury volume between groups was achieved with one-way ANOVA. Post hoc analysis was performed with the Newman-Keuls test. Two-way ANOVA (brain section and treatment group) was used to determine the effect of drug treatment on regional injury volume. Statistical differences were considered significant at P<.05.
In nonischemic cats, there was no difference between groups in any physiological variable, in blood flow to any brain region, or in CMRO2 (Fig 1⇓). In particular, PPBP did not increase CBF in any region.
There were no significant differences in baseline values of any physiological variables among the three experimental groups subjected to ischemia (Table 1⇓). Likewise, arterial blood gas values and hemoglobin and glucose concentrations were similar between groups throughout ischemia and reperfusion. Although treatment with PPBP was associated with decreased MABP compared with baseline and compared with the control group at 180 and 240 minutes of reperfusion, MABP was maintained consistently above 100 mm Hg throughout reperfusion in all groups. Brain temperature was controlled at approximately 38°C in all groups.
Values of blood flow to brain regions ipsilateral to MCA occlusion and to the posterior fossa are given in Table 2⇓. Baseline blood flow was similar between groups in all regions. Reduction in cortical CBF during left MCA occlusion was graded. The greatest reduction occurred in the ipsilateral inferior temporal cortex, and the smallest reduction was seen in the parietal cortex. By chance, within each region there was also some variability in the degree of blood flow reduction between groups. This variation was most evident in the ipsilateral inferior temporal cortex and the ipsilateral temporal parietal cortex. For example, even before drug administration (30 minutes of ischemia) CBF was higher in the PPBP-1 group compared with the control group. On the contrary, CBF to the ipsilateral caudate nucleus was similarly reduced in all groups during MCA occlusion. In addition, CBF to the contralateral hemisphere and caudate nucleus was similar between groups during MCA occlusion (data not shown). During reperfusion, blood flow to the ipsilateral parietal cortex, at 240 minutes of reperfusion, was greater in animals treated with PPBP-1 than animals in the control group. Otherwise there were no differences between groups in regional CBF during reperfusion.
Baseline amplitude (control, 44±6 μV; PPBP-0.1, 44±6 μV; PPBP-1, 37±5 μV) and latency (control, 11.8±0.2 ms; PPBP-0.1, 11.8±0.3 ms; PPBP-1, 12.4±0.2 ms) of the primary cortical SEP were not different among groups. Ipsilateral SEP amplitude decreased to the same extent during left MCA occlusion (control, 4±2% of baseline; PPBP-0.1, 8±3% of baseline; PPBP-1, 10±3% of baseline). Contralateral SEP amplitude was not affected by MCA occlusion. During reperfusion, recovery of SEP amplitude was incomplete in all groups. However, at 240 minutes of reperfusion recovery was greater in cats treated with PPBP-1 compared with the control group (control, 18±11% of baseline; PPBP-0.1, 30±14% of baseline; PPBP-1, 54±14% of baseline; control versus PPBP-1, P=.05 by Dunnett’s one-tailed test). All cats had normal latency of the wave measured over the second cervical vertebra throughout the protocol.
Ipsilateral cerebral hemispheric injury volume was greater in the control group compared with the PPBP-0.1 and PPBP-1 groups when expressed in terms of absolute volume (control, 2685±486 mm3; PPBP-0.1, 1473±286 mm3; PPBP-1, 503±102 mm3) or as a percentage of total ipsilateral hemispheric volume (Fig 2⇓). In addition, injury volume of the caudate nucleus in control cats (143±20 mm3) was greater than that in cats treated with PPBP-1 (73±16 mm3) but not PPBP-0.1 (122±21 mm3) (Fig 2⇓). When injury volume was analyzed separately for each of 12 coronal sections, injury was reduced in the ipsilateral hemisphere and ipsilateral caudate nucleus in the PPBP-1 group (two-way ANOVA) and in the ipsilateral hemisphere in the PPBP-0.1 group (Fig 3⇓).
Statistical significance of injury volume without significant differences in CBF may arise from variability in intraischemic CBF. To test the relationship between these two parameters, we plotted individual volume in the primary MCA territory (inferior temporal cortex) versus CBF in this region at 30 minutes (before PPBP) and 90 minutes (15 minutes after start of PPBP infusion) of occlusion (Fig 4⇓). As expected, injury volume was inversely related to CBF. When CBF was less than 30 mL/min per 100 g, injury area was usually less in both the PPBP-0.1 and PPBP-1 groups than in the control groups at matched levels of CBF.
This study demonstrates that intravenous administration of PPBP beginning at 75 minutes after initiation of a 90-minute focal ischemic insult and continuing throughout 240 minutes of reperfusion substantially reduced the volume of acute brain injury. These results support a role of ς-receptors in modulating focal ischemic injury.
During MCA occlusion CBF was higher in cats that were to subsequently receive the highest dose of PPBP. This finding was clearly not a pharmacological effect because it was even observed before PPBP administration (ie, at 30 minutes of ischemia). This difference occurring by chance in predrug values of CBF was present in cortical regions but not in the caudate nucleus, suggesting greater variability of collateralization between animals in cortical regions compared with subcortical regions. Thus, part of the 81% difference in hemispheric injury volume seen in the PPBP-1 group may be attributable to greater collateral blood flow occurring by chance in this group. However, we believe that the major mechanism of action of PPBP is independent of CBF for several reasons. First, intraischemic CBF to the caudate nucleus was not different between the control and PPBP-1 groups, yet caudate injury volume in the treated group was 49% smaller than in controls. Second, at the lower dose of PPBP there was no difference in intraischemic blood flow in the cerebral cortex, yet hemispheric injury volume was 45% smaller than in controls. Third, to ensure that these effects were not due to an error arising from pooling of data, we plotted the individual injury volume of the MCA territory versus CBF (Fig 4⇑) and found that injury size was generally less with either dose of PPBP than in the controls at equivalent levels of intraischemic CBF. Fourth, the drug did not increase CBF in nonischemic brain regions or in cats not subjected to ischemia. Fifth, only a small fraction of the total dose of PPBP was administered during the last 15 minutes of ischemia, and there was no effect of PPBP on CBF during reperfusion, when most of the drug was administered. Therefore, although we cannot exclude some effect of PPBP on improving intraischemic CBF, the major mechanism of action of the drug appears to be independent of CBF.
A wealth of evidence has implicated ς-receptors in a variety of physiological processes and pathological conditions.15 16 17 18 These include psychosis, antipsychotic actions of neuroleptic drugs, and motor, endocrine, and immune functions. However, the role of ς-receptors in the mechanism of ischemic brain injury has not been completely elucidated. For example, although the ς-receptor ligand ifenprodil has efficacy in a model of permanent focal ischemia in cats, the mechanism of protection was believed to be due to inhibition at the polyamine modulatory site of the NMDA receptor.1 The role of a possible change in cerebral perfusion during drug administration was not assessed. A variety of ς-receptor ligands have also been evaluated in a gerbil model of transient ischemia. In this model, it is very difficult to measure and control many important physiological variables (eg, blood pressure, brain temperature) that may have independent effects on brain injury.
N-Allylnormetazocine [(+)SKF 10,047], a ς-ligand that is also a noncompetitive NMDA receptor antagonist,19 20 is neuroprotective in a gerbil model of transient focal ischemia.21 22 The mechanism of protection for (+)SKF 10,047 may be related to its ability to prevent ischemia-induced increases in intracellular calcium21 or its ability to prevent cortical spreading depression.23 In addition, several drugs in a series of U-50,488H analogues, which are potent ς-receptor ligands, manifested significant efficacy in the gerbil model of transient ischemia, whereas one compound in the series (BD-601) afforded no neuroprotection despite marked affinity for ς-receptors.6 In models of cerebral ischemia, factors that may influence the effects of ς-receptor ligands include differential activities of these drugs at ς-receptor subtypes18 24 and interactions with neurotransmitter systems not directly related to ς-receptors. Furthermore, although most evidence suggests that ς-receptor ligands act in a manner similar to agonists, some reports suggest antagonist actions of ς-compounds.18
These data do not completely exclude the possibility that PPBP is eliciting its protective effects via a mechanism other than that expected from a ς-receptor interaction. For example, other ς-receptor ligands have been found to also be potent direct ligands at a number of different receptors, including the NMDA receptor complex.16 19 20 Because of these effects at other receptors it has been very difficult to establish a definite relationship between ς-receptor affinity and neuroprotection.7 As PPBP is an extremely potent ligand at ς-receptors3 and the binding of [3H]PPBP is inhibited by a number of ς-ligands but not by many other drugs,25 26 PPBP may be an appropriate agent to use to assess the relationship between ς-receptor affinity and neuroprotection.
We found no effect of PPBP on CBF or CMRO2. Contrary to this finding, Altura et al27 have demonstrated cerebral vasoconstriction of ς-ligands tested in vitro. This difference between apparent cerebral vascular effects of ς-ligands may be due to differences intrinsic to each drug. For example, some ς-ligands decrease cerebral metabolism, whereas others increase cerebral metabolism.28 Likewise, depending on the ς-ligand studied, there is great variability in ability to inhibit amine uptake and alter contractile responses of arterial segments studied in vitro.29
Injury volume was estimated with TTC staining. TTC acts as a proton acceptor for mitochondrial oxidative cellular metabolism.30 When compared with light microscopy, TTC tended to overestimate ischemic volumes of hemispheres and cortex and underestimate ischemic volume in caudate for rats exposed to 4 hours of MCA occlusion.31 In another study, TTC staining was found to correlate with histopathology best in cats when the time of ischemia was greater than 2 hours and there was at least 2 hours of reperfusion.32 Bederson et al12 demonstrated that there was good correlation between TTC staining and histopathologic evaluation in rats exposed to 1 to 2 hours or 5 to 6 hours of MCA occlusion but not to 3 hours of MCA occlusion (sample size of only 3 for 3 hours). Nonetheless, since Cole et al13 have demonstrated that the histochemical abnormality revealed by TTC staining may not necessarily represent inevitable infarction when it is used for paradigms of short ischemic periods (3 hours in their study), we cannot be certain that the eventual injury, measured at a later time, would not be similar between groups. Therefore, our results may only indicate that rate of brain injury development is altered by PPBP treatment. However, an alteration in rate of injury may have important clinical implications, as it may increase the window of opportunity for use of other therapeutic agents.
In conclusion, PPBP is effective in decreasing brain injury from transient focal ischemia, even when it is administered after the onset of ischemia. The mechanism of protection does not appear to be primarily related to a more favorable redistribution of blood flow during ischemia or reperfusion. Rather, we speculate that ς-receptors may modulate focal ischemic damage by altering NMDA receptor function.
Selected Abbreviations and Acronyms
|CBF||=||cerebral blood flow|
|CMRO2||=||cerebral oxygen consumption|
|MABP||=||mean arterial blood pressure|
|MCA||=||middle cerebral artery|
|SEP||=||somatosensory evoked potential|
This study was supported by US Public Health Service, National Institutes of Health grant NS-20020 and the Division of Intramural Research of the National Institute on Drug Abuse. The authors thank Toshiki Okada, MD, and Ying Wu for their excellent technical assistance.
Reprint requests to Jeffrey R. Kirsch, MD, Department of Anesthesiology and Critical Care Medicine, 600 N Wolfe St, Blalock 1410, Baltimore, MD 21287-4963.
- Received January 6, 1995.
- Revision received April 14, 1995.
- Accepted May 23, 1995.
- Copyright © 1995 by American Heart Association
Gotti B, Duverger D, Bertin J, Carter C, Dupont R, Frost J, Gaudilliere B, MacKenzie ET, Rousseau J, Scatton B, Wick A. Ifenprodil and SL 82.0715 as cerebral anti-ischemic agents, I: evidence for efficacy in models of focal cerebral ischemia. J Pharmacol Exp Ther. 1988;247:1211-1221.
Carter C, Benavides J, Legendre P, Vincent JD, Noel F, Thuret F, Lloyd KG, Arbilla S, Zivkovic B, MacKenzie ET. Ifenprodil and SL 82.0715 as cerebral anti-ischemic agents, II: evidence for N-methyl-D-aspartate receptor antagonist properties. J Pharmacol Exp Ther. 1988;247:1222-1232.
Clissold DB, Pontecorvo MJ, Jones BE, Abreu ME, Karbon EW, Erickson RH, Natalie KJ Jr, Borosky S, Hartman T, Mansbach RS, Balster RL, Ferkany JW, Enna SJ. NPC 16377, a potent and selective ς-ligand, II: behavioral and neuroprotective profile. J Pharmacol Exp Ther. 1993;265:876-886.
Taylor DP, Yevich JP, Dextraze P, Moon SL, Behling SH, Defnet J, Geissler M. Sigma ligands: a role in neuroprotection? In: Kamenka J-M, Domino EF, eds. Multiple Sigma and PCP Receptor Ligands: Mechanisms for Neuromodulation and Neuroprotection? Ann Arbor, Mich: NPP Books; 1992:767-778.
OBrien MD, Waltz AG. Transorbital approach for occluding the middle cerebral artery without craniectomy. Stroke. 1973;4:201-206.
Matsumiya N, Koehler RC, Traystman RJ. Consistency of cerebral blood flow and evoked potential alterations with reversible focal ischemia in cats. Stroke. 1990;21:908-916.
Matsumiya N, Koehler RC, Kirsch JR, Traystman RJ. Conjugated superoxide dismutase reduces extent of caudate injury after transient focal ischemia in cats. Stroke. 1991;22:1193-1200.
Bederson JB, Pitts LH, Germano SM, Nishimura MC, Davis RL, Bartkowski HM. Evaluation of 2,3,5-triphenyltetrazolium chloride as a stain for detection and quantification of experimental cerebral infarction in rats. Stroke. 1986;17:1304-1308.
Takeshima R, Kirsch JR, Koehler RC, Gomoll AW, Traystman RJ. Monoclonal leukocyte antibody (MoAb 60.3) does not decrease injury following transient focal cerebral ischemia in cats. Stroke. 1992;23:247-252.
Lysko PG, Yue T-L, Gu J-L, Feuerstein G. Neuroprotective mechanism of (+)SKF 10,047 in vitro and in gerbil global brain ischemia. Stroke. 1992;23:1319-1324.
Lysko PG, Gagnon RC, Yue T-L, Gu J-L, Feuerstein G. Neuroprotective effects of SKF 10,047 in cultured rat cerebellar neurons and in gerbil global brain ischemia. Stroke. 1992;23:414-419.
Wu X-Z, Bell JA, Spivak CE, London ED, Su T-P. Electrophysiological and binding studies on intact NCB-20 cells suggest presence of a low affinity Sigma receptor. J Pharmacol Exp Ther. 1991;257:351-359.
Hashimoto K, London ED. Characterization of [3H]4-phenyl-1-(4-phenylbutyl)-piperidine binding in rat brain. Soc Neurosci Abstr. 1993;19:1155. Abstract.
Altura BT, Quirion R, Pert CB, Altura BM. Phencyclidine (‘angel dust’) analogs and sigma opiate benzomorphans cause cerebral arterial spasm. Proc Natl Acad Sci U S A. 1983;80:865-869.
Massamiri T, Duckles SP. Multiple vascular effects of Sigma and PCP ligands: inhibition of amine uptake and contractile responses. J Pharmacol Exp Ther. 1990;253:124-129.
Glenner GG. Tetrazolium salts. In: Little RD, ed. H.J. Conn’s Biological Stains. Baltimore, Md: Williams & Wilkins; 1969:154-162.