- Email: [email protected]
Effects of cocaine on leakage of creatine kinase from skeletal muscle: In vitro and in vivo studies in mice
Life Sciences,Vol.57, No. 17,pp. 1569-1578. 1995 Copyright 0 1995Ekvier Science.Inc. Printedin theUSA.All rights resewed
Pergamon
0@24-3205/95 $9.50 + .oO
0024-3205(95)021323
EFFECTS OF COCAINE ON LEAKAGE OF CREATINE KINASE FROM SKELETAL MUSCLE: IN VITRO AND IN WO STUDIES IN MICE Gayle A. Brazeau, Anne McArdle and Malcolm J. Jackson Department of Pharmaceutics, University of Florida, J. Hillis Miller Health Science Center, Gainesville, Florida 326 10 and Department of Medicine, University of Liverpool, Liverpool England L693BX (Received in final form August 16, 1995)
Summarv The mechanism of cocaine-induced rhabdomyolysis and/or muscle damage has not been elucidated. To determine if cocaine has a direct effect on muscle, isolated soleus and EDL muscles were incubated in the presence of 1 mM and 0.2 mM cocaine using a pulse and continuous exposure protocol. The release of creatine kinase from the isolated EDL muscle was statistically significant only when muscles were exposed to 1 mM cocaine for a period of 30 minutes. These findings suggest that cocaine-induced creatine kinase release could be mediated by a direct action on the fibers. It is also possible, however, that cocaine-induced muscle damage and creatine kinase release may be mediated via an indirect effect, It is possible that cocaine’s vasoconstrictor effects could lead to muscle damage via an ischemiareperfusion injury leading to free radical formation and lipid peroxidation. This study, therefore, also investigated the possibility that cocaine-induced cytosolic enzyme release may To test this be mediated via the formation of free radicals leading to lipid peroxidation. a free radical scavenger, and hypothesis, muscle total glutathione levels, muscle thiobarbituric acid reactive substances (TBARS), a measurement of lipid peroxidation, were examined following an acute IV cocaine dose in mice. Sedentary BalbC mice were injected with cocaine (40 mg/kg) or normal saline via the tail vein. Creatine kinase levels in serum and total glutathione and TBARS in liver and muscle were determined at 4, 8, and 24 hrs. Serum creatine kinase levels were significantly elevated 5-fold, while TBARS were elevated 100% in the gastrocnemius muscle of cocaine-treated animals at 4 hrs compared to normal saline controls. However, serum creatine kinase levels, total glutathione and TBARS in the gastrocnemius muscle were not statistically different at 8 or 12 hrs; or in the liver and suggests that lipid anterior tibialis muscle at 4, 8, or 24 hrs. The present findings peroxidation may be occurring in skeletal muscle after a single IV cocaine dose in mice.
ZGzyWords: cocaine, skeletal muscle, intravenous, muscle damage, rhabdomyolysis, lipid peroxidation
Gayle A. Brazeau, Department of Pharmaceutics, University of Florida, J. Hillis Miller Health Science Center, Box 100494, Gainesville, Florida 32610, Phone (904) 392-3223, FAX (904) 846-0298.
1570
ESfects of Cocaine on Skeletal Muscle
Vol. 57,No. 17.1995
The increasing number of clinical reports describing cocaine-induced rhabdomyolysis leading to acute renal failure and the publicized deaths of athletes who used cocaine for its reported ergogenic properties has highlighted our lack of knowledge on the mechanisms responsible for cocaine’s toxicity on cardiac and skeletal muscle (l-7). The potential cardiotoxic mechanisms of cocaine have been extensively studied (8), however very little is known about the underlying mechanism(s) responsible for cocaine toxicity on skeletal muscle (9). Pagala and coworkers recently demonstrated an increased release of creatine kinase from rat soleus in vitro, but not from the extensor digitorum longus muscle (BDL) when incubated in the presence of 0.5-1.0 mM cocaine over a 4-hour period (9). These investigators suggested that cocaine may have a direct toxic effect on muscle tissue that could lead to rhabdomyolysis (9). To confirm if cocaine has a direct toxic effect on skeletal muscle, we investigated whether the release of creatine kinase from the isolated mouse soleus (SOL) and extensor digitorum longus (BDL) muscle was enhanced in the presence of cocaine. It is possible that cocaine-induced muscle damage could also be mediated via effects on other muscle associated tissues (e.g., vasculature). The short systemic half-life of cocaine compared to $lie development of toxic effects (which can be delayed relative to the time of drug administration) could suggest that these effects are caused by cocaine metabolites or secondary to events in other tissues. Cocaine-induced muscle damage, as measured by the release of cytosolic enzymes, could be mediated through its vasoconstrictor effects on muscle vasculature. It is well known that cocaine induces vasoconstriction in a variety of tissues by its ability to block catecholamine re-uptake at presynaptic vessels (10). Cocaine-induced vasoconstriction has been implicated as a possible mechanism in myocardial infarction, cerebrovascular accidents, obstetrical complications, and rhabdomyolysis (2, 7, 10, 11). Recent studies have suggested that cocaine’s toxic effects on the heart and brain seems to be mediated via vasoconstriction and decreased blood flow (12-14). It can be hypothesized that cocaine-induced muscle damage is the result of spasm and vasoconstriction of muscle associated vasculature. This decreased blood flow could cause ischemia with the resulting reperfusion leading to free radical formation. Therefore, we investigated whether an acute intravenous cocaine dose causes increased serum creatine kinase levels in mice and whether this elevation is associated with free radical formation and lipid peroxidation as measured by total glutathione levels and the formation of lipid peroxides through TBARS, respectively. Materials and Methods In vitro muscle exneriments. Male 2-month old BalbC mice were humanely sacrificed via cervical dislocation and the SOL and EDL muscles were carefully isolated and dissected free from adjoining muscles as previously described (15). The muscles were attached to small glass holders and placed into plastic tubes containing 4 ml of a mammalian Ringers solution gassed with 02/CO2 (95:5) as described previously (15). The tubes were mounted in a thermostat controlled water bath maintained at 37’C. At 30 minute intervals over the 3-hour incubation, the incubation solution (with or without cocaine) in each tube was replaced with fresh 37’C solution using a peristaltic pump. In these studies, one muscle from each animal was incubated in the presence of the cocaine and the contralateral muscle served as the control. Two based upon using either experiments,
concentrations of cocaine were utilized in these the work of Pagala and coworkers. The isolated a pulse or a continuous experimental protocol. the muscles were pre-incubated in the mammalian
experiments (1 mM and 0.2 mM) muscles were exposed to cocaine In both the pulse and continuous Ringers solution alone during the
Vol. 57, No. 17, 1995
Effects of Cocaine on SkeletalMuscle
1571
first 30 minutes of the incubation. In the pulse experiments, muscles were incubated in the absence or presence of cocaine between 30-60 minutes only. For the remaining incubation period (60-180 minutes), the muscles were incubated in the mammalian Ringers solution and the media changed every 30 minutes. In contrast during the continuous exposure, the muscles were incubated in the absence or presence of cocaine between 30-180 minutes, with fresh cocaine solution added every 30 minutes. Creatine kinase release from the isolated muscles is expressed as the cumulative sum during the experimental period. In vivo cocaine emeriments. Conscious Female BalbC mice were injected with normal saline (0.1 ml/l0 g) or cocaine (40 m@g) in normal saline (0.1 ml/l0 g) via the tail vein. The injection was administered over S-10 seconds. The animals did not struggle during the injection procedure nor did they experience any tremors or seizures at this dose. These injections were performed by one of the investigators (MJJ) who has extensive experience with this technique. At 4, 8, and 24 hours, the animals were humanely sacrificed by a pentobarbital overdose, an aortic blood sample taken and tissues harvested. The tibialis anterior muscle, gastrocnemius muscle, and a portion of the liver was harvested and frozen immediately in liquid nitrogen. SamDIe analvsis. Creatine kinase activity in the incubation solutions and serum samples was measured using an enzyme-linked assay as described previously (15). Muscle and liver homogenates (20% w/v) were prepared in 1% sulphosalicylic acid. Total muscle glutathione was quantified by the method of Jackson and coworkers (16) developed from the work of Anderson (17). A portion of the homogenate was set-aside for protein determination using the Lowry method (18). Thiobarbituric acid reactive substances (TBARS) in the muscles were measured according to the method of Phoenix and coworkers (19) with the exception that the tissues were homogenized in 1% sulphosalicylic acid. Data analvsis. Data are presented as mean and standard error Statistical differences between cocaine treated muscles and control muscles were analyzed using an unpaired Student’s T-test. In the in vivo studies, between normal saline and cocaine-treated at the three time points were unpaired Student’s T-test.
of the mean (SEM). in the in vitro studies statistical differences determined using an
Results In vitro creatine kinase release from isolated mouse muscles. The cumulative release of creatine kinase from isolated mouse muscles exposed to a pulse 1 mM cocaine dose is shown in Figure 1 (Panel A). Creatine kinase release was significantly higher in EDL muscles exposed to a pulse 1 mM cocaine dose compared to untreated control muscles. There was no statistical difference in the cumulative creatine kinase release from the SOL muscles between a pulse cocaine exposure and control untreated muscles. To determine if muscle damage could be further exacerbated by continuous exposure to EDL and SOL muscles were incubated in fresh cocaine solution at 30 minute intervals (Figure 1, Panel B). The cumulative release of creatine kinase from continuously cocaine exposed EDL or SOL muscles was not statistically different from the untreated control muscles. Furthermore, in the continuous cocaine exposure protocol, cumulative creatine kinase release was less than that seen in the pulse exposure protocol. The reason for the reduced levels of creatine kinase release seen during the continuous exposure protocol is not known. Preliminary studies showed that the presence of cocaine in the bath did not interfere with the 1 mM cocaine,
Eiffects of Cocaine on SkeletalMuscle
1572
Vol. 57, No. 17,lWS
A
0
30 60
90
6 Continuous
120
150
180
30
60
90
120
150
180
Time (mii)
Fig. 1 Cumulative creatine kinase release from isolated EDL or SOL muscles incubated with 1 mM cocaine (m) as a 30-minute pulse exposure (Panel A) or continuous exposure (Panel B). Control muscles (0) were incubated in the Mammalian Ringers. Values are presented as mean and SEM (n = 4-6) with p< 0.05 (*). determination of creatine kinase activity (data not shown). Alternatively, it can be proposed that cocaine itself or an unknown compound generated or released by cocaine subsequently blocks the creatine kinase release. A pulse or continuous exposure to 0.2 uM cocaine did not cause any statistically significant increase in the cumulative release of creatine kinase from either the isolated EDL or SOL muscle (data not shown). Intravenous cocaine studies. There were no fatalities using 40 mg/kg dose and dosing route, however increasing the dose to 50 mg/kg resulted in approximately 50% mortality (data not shown). The animals receiving this dose and dosing route did not exhibit any tremors or signs of muscle contraction following injection or during the experimental period. This dose of cocaine (40 mg/kg) resulted in a significant 5-fold increase in serum creatine kinase levels compared to normal saline controls suggesting that muscle damage may be occurring with this treatment (Figure 2). Serum creatine kinase levels were also higher at 8 hours however, these were not statistically significant compared to saline controls. There was no statistically significant difference between the two groups at 24 hours.
Vol. 57, No. 17, 1995
Effects of Cocaine on SkeletalMuscle
lx3
There was no statistical difference in glutathione levels between cocaine-treated and normal saline-treated animals in the gastrocnemius (Figure 3) and tibialis anterior muscle (data not shown). This could be a function of the large intra-day variability of glutathione levels in these tissues. Muscle glutathione levels were found to vary approximately 2533% in the control normal saline treated animals. In the liver, glutathione levels in cocaine treated animals were lower compared to normal saline at 4 and 8 hours, but these values did not reach statistical significance (Figure 4). Liver glutathione levels in normal saline treated animals were extremely variable at each of the time points and throughout the study period. Consistent with the increased serum creatine kinase levels at 4 hours, TBARS absorbance values were significantly higher in the gastrocnemius muscles from cocaine-treated compared to normal saline-treated mice (Figure 5). At 8 and 24 hours, these values were not different between the two groups. Likewise, the TBARS absorbance values from the liver samples were not different between the two groups at each of the three time points (data not shown). Due to the small size of the anterior tibialis muscle, TBARS could not be quantified in this muscle. Discussion The findings from the in vitro studies suggest that cocaine-induced creatine kinase release, often used as an indirect measure of muscle damage, may be mediated by a direct drug effect on the muscle. Creatine kinase release was significantly enhanced when mice EDL
1400
1
4
a
24
Time (hrs)
Fig. 2 Serum creatine kinase activities in normal saline (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n = 5) with p< 0.05 (*).
1574
Vol. 57, No. 17, 1995
Effects of Cocaine on Skeletal Muscle
0.6
0.5 F P ;
0.4
% % l 0.3 g z 0.2 1 S G 0.1
0.0
24
Time (hrs)
Fig 3.
Total glutathione content of the gastrocnemius muscle from normal saline- (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n =
Liver
110 100
i
4
8
24
Time (hm)
Fig. 4
Total glutathione content of the liver Corn normal saline- (open bars) and cocainetreated (hatched bars) mice. Values are presented as mean and SEM (n = 5).
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17, 1995
0.7
I
0.6
-
0.5
-
0.4
-
5
0.3
-
x’ I; g (3
0.2
-
0.1
I-
G D 9 E
I.575
% 2 8 ._
0.0
8
24
Time (hrs)
Fig. 5
TBARS content of gastrocnemius muscle homogenates from normal saline- (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n = 5) with p< 0.05 (*). muscles were incubated for a short period of time (30 minutes) with a cocaine concentration (e.g., 1 n&I) 2-10x higher than the highest concentration reported in overdose patients (9). Like previous investigators, it was necessary to use this higher dose in mice (versus humans) to demonstrate an effect on muscle because of species differences in the pharmacokinetics, pharmacodynamics or toxicological profile. These findings differ from those reported by Pagala and coworkers who showed significantly higher creatine kinase from only rodent SOL muscles (9). Certainly, species differences might account for the disparate results between the two studies. However, the nature of the experimental design in the two studies (viz., nature of the incubation buffer, duration of the experiment, continuous versus pulse exposure) could also account for the different findings. Furthermore, it is unclear to what extent cocaine may have been degraded or metabolized in the muscle during the 4 hour study period in the studies by Pagala and coworkers (9). Another concern with isolated muscle studies, that has not been addressed in earlier studies, is whether the presence of the drug in the solution interferes in the analysis procedure. Preliminary studies showed that cocaine at the concentrations used in the present investigations did not directly affect the enzymatic determination of creatine kinase using the spectrophotometric kinetic assay (data not shown). Clinical reports of rhabdomyolysis in patients and experimental studies investigating cocaine’s toxic effects on various organ systems, including skeletal muscle, have implicated cocaine-induced vasoconstriction as a possible initiating event in cocaine-induced muscle damage (l-2). It can be postulated that cocaine-induced muscle damage may be the result of vasoconstriction leading to muscle ischemia with the subsequent tissue reperf%sion leading to free-radical formation and lipid peroxidation. The vasoconstrictor effects of cocaine, mediated
1576
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17, 1995
either through the adrenergic or non-adrenergic mechanisms, have been well established in animal models (13-14, 20). Likewise in more recent studies, cocaine metabolites have been shown to have vasoconstrictor properties. Studies have suggested that benzoylecgonine has vasoconstrictor properties and itself may be toxic to tissues (13). In addition, toxicity may be mediated via the formation of cocaine metabolites which generate further free radicals in muscle tissue. It has been suggested that metabolism to norcocaine in the liver is the principle pathway of bioactivation leading to the formation of N-hydroxynorcocaine. This latter metabolite is subsequently converted to the free radical norcocaine nitroxide. Toxicity in tissues may result from the titile recycling between N-hydroxynorcocaine and norocaine nitroxide resulting in the depletion of NADPH and the production of hydrogen peroxide (21). The present study, which uses a lower dose than that reported to cause hepatotoxicity (22), provides evidence that a single acute intravenous dose of cocaine (40 mg/kg via the tail vein) causes increased serum creatine kinase levels at 4 hours. This increase in serum creatine kinase levels corresponds with the presence of lipid peroxidation in the gastrocnemius muscle as measured by increased TBARS. Since cocaine is rapidly metabolized by plasma esterases and hepatic enzymes, the increased serum creatine kinase levels combined with the presence of lipid peroxidation suggests that muscle damage could be caused via the formation of free radicals which damage muscle membranes leading to increased cytosolic enzyme release. At 8 hours, serum creatine kinase levels were still elevated in the cocaine-treated animals relative to normal saline controls, but these values were not statistically different. As in other in vivo studies utilizing creatine kinase as a indicator of muscle damage, the range of the values was large which could account for the lack of statistical significance. However, the rapid decline of serum creatine kinase levels is consistent with the reported short half-life (e.g., < 2 -4 hrs.) of this enzyme in rodent serum (23). Furthermore, it is possible that the peak creatine kinase and TBARS levels and decreased glutathione levels could have occurred earlier than the first sampling time (e.g., 4 hours). The 4-hour initial sampling time was selected based upon other toxicity studies. To verity whether muscle damage is occurring with cocaine dosing, fiture studies will focus on determining total serum creatine kinase levels, creatine kinase isoenzyme profiles (e.g., to determine whether the source is cardiac, muscle or brain) and histological evaluation at various time points. The increased TBARS in the gastrocnemius muscle supports the hypothesis of free radical formation leading to peroxidation of membrane lipids with cocaine administration. It could subsequently be predicted that muscle glutathione levels should be decreased if this hypothesis is correct. In contrast, muscle total glutathione levels were not decreased with cocaine administration. Hepatotoxicity following cocaine administration has been shown to cause decreased liver glutathione levels (22). Muscle and liver glutathione levels were not statistically lower in the cocaine-treated animals compared to the normal saline. The variability of tissue glutathione levels in the normal saline-treated animals over the study period, particularly in the liver, and the lack of statistical significance from the cocaine-treated animals may be a function of daily fluctuations in these levels (24-25). Alternatively, it has been suggested that cocaine-induced hepatotoxicity in mice as measured by changes in glutathione levels may be strain dependent (26). Finally, a single intravenous cocaine dose might increase free radical formation, but other radical scavenger agents and processes (e.g., a-tocopherol, superoxide dismutase) rather than glutathione might be responsible for inactivating these molecules. Decreases in muscle glutathione levels might only occur with repeated cocaine ingestion. It is certainly possible that the effects of cocaine, either direct or indirect, are additive and a f?.mction of repeated doses. Most of the reports of rhabdomyolysis in humans has been associated with those who chronically abuse cocaine.
Vol. 57, No. 17, 1995
Effe&s of Cocaine on Skeletal Muscle
1577
The demonstrated increase in lipid peroxidation as measured by an increase in TBARS would suggest that a likely target for free radicals would be membrane lipids. Similar findings have been reported in cocaine-induced hepatotoxicity (27). However, the present data does not rule out the possibility that other cellular targets (e.g., proteins, DNA) in skeletal muscle might also be affected during acute and chronic cocaine ingestion. Overall, these findings would point to the idea that muscle damage (measured by increased serum creatine kinase levels) following cocaine ingestion is mediated via the formation of free radicals causing lipid peroxidation. It is possible that chronic cocaine ingestion could lead to repeated ischemic events mediated by vasoconstrictor properties of cocaine and its metabolites. Damage to the muscle tissue would result from the subsequent repertusion and the generation of free radicals. If these free radicals are not inactivated by radical scavenging systems in the muscle and associated vasculature, this could result in damage to the muscle fibers. Acknowledgments This work was supported by a Gustavus A. Pfeiffer Faculty Development Research Fellowship from the American Foundation for Pharmaceutical Education. We would like to thank Sue Page for assistance in the assay methodologies. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13. 14. 15. 16. 17. 18.
R.D. FELD, Clin. Chem. News, 3 April (1993) E. A. TURBAT-HERRARA, Ultrastruct. Pathol. 18 127-13 1 (1994). V.A. POGUE and M. NURSE, Amer. J. Med. & 183-186 (1989). B.D. HERZLICH, E.L. ARSURA, M. PAGALA and D. GROB, Ann. Int. Med. 109 335-336 (1988). J.G. ZAMORA-QUEZADA H. DINERMAN, M.J. STADECKER and J. J. KELLY, Ann. Int. Med. &3 564-566 (1988). N. COOPER B. COHN and S. DOHERTY, Newsweek 26, July 7 (1986). A. KETEYIAN and B. SELCRAIG, Sports Illustrated 65 18- 19 (1986). A.A. WELDER P. G RAMMAS and R.B. MELCHERT, Toxicol. Lett. @ 227-238 (1993). M. PAGALA, B. AMALADEVI, D. AZAD, S. PAGALA, B. HERZLICH, T. NAMBA and D. GROB, Life Sci. 52 751-756 (1993). L.L CREGLER and H. MARK, New Eng. J. Med. 315 1495-1500 (1986). T.T. INABA, D. STEWART and W. KALOW, Clin. Pharmacol. Therap. 23 547-552, (1978). B.D. NUNEZ, L. MIAO, Y. WANG, M.M. NUNEZ, F.W. SELLKE, J.N. ROSS, V. SUSULIC, G.Y. CARROZZA, G.Y. PAIK, J.P. CARROZZA and J.P. MORGAN, Circulat. Res. 74 281-290 (1994). J.A. MADDEN and R.H. POWERS, Life Sci. 47 1109-l 114 (1990). C.D. KURTH, C. MONITTO, M. L. ALBUQUERQUE, P. FEUER, E. ANDAY and L. SHAW, J. Pharmacol. Exp. Therap. 587-591 (1993). D.A. JONES, M.J. JACKSON and R.H.T. EDWARDS, Clin. Sci. @ 193-201 (1983). M.J. JACKSON, M.H. BROOKE, K. KAISER and R.H.T. EDWARDS, J Clin. Sci. 80 559-564 (1991). M. ANDERSON, Methods Enzymol. 113_ 548-555 (1985). O.H. LOWRY, N.J. ROSENBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193 265-275 (1951).
1578
19. 20. 21. 22. 23. 24. 25. 26. 27.
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17,195
J. PHOENIX, R.H.T EDWARDS and M.J. JACKSON, Brit. J. Nutrit. &l 245-256 (1990). M.M. KNUEPFER, CA. BRANCH, D.M. WEHNER, Q. GAN and D. HOANG, Can. J. Physiol. Pharmacol. 22 335-343 (1994). M. W., KLOSS, G.M. ROSEN and E. J. RAUCKMAN, Biochem. Pharmacol. 33 169-173 (1984). M.L. THOMPSON, L. SHUSTER and K. SHAW, Biochem. Pharmacol. 28 2389-2395 (1979). S.F. PAGE, Thesis, University of Liverpool (1992). M.J. TUNON, P. GONZALES, P. LOPEZ, G.M. SALIDO, and J.A. MADRID, Arch. Int. Physiol. Biochim. Biophys. 100 83-87 (1992). P.M. BELANGER, M. DESGAGNE and B. Bruguerolle, Drug Met. Dispos. Biol. Fate Chem., 19 241-244 (1991). C.S. BOYER and D.R. PETERSEN, Hepatology 14 1209-1216 (1991). M.W. KLOSS, G.M. ROSEN, E.J. RAUCKMAN, Toxicol. Let. li 65-70, (1983).
Pergamon
0@24-3205/95 $9.50 + .oO
0024-3205(95)021323
EFFECTS OF COCAINE ON LEAKAGE OF CREATINE KINASE FROM SKELETAL MUSCLE: IN VITRO AND IN WO STUDIES IN MICE Gayle A. Brazeau, Anne McArdle and Malcolm J. Jackson Department of Pharmaceutics, University of Florida, J. Hillis Miller Health Science Center, Gainesville, Florida 326 10 and Department of Medicine, University of Liverpool, Liverpool England L693BX (Received in final form August 16, 1995)
Summarv The mechanism of cocaine-induced rhabdomyolysis and/or muscle damage has not been elucidated. To determine if cocaine has a direct effect on muscle, isolated soleus and EDL muscles were incubated in the presence of 1 mM and 0.2 mM cocaine using a pulse and continuous exposure protocol. The release of creatine kinase from the isolated EDL muscle was statistically significant only when muscles were exposed to 1 mM cocaine for a period of 30 minutes. These findings suggest that cocaine-induced creatine kinase release could be mediated by a direct action on the fibers. It is also possible, however, that cocaine-induced muscle damage and creatine kinase release may be mediated via an indirect effect, It is possible that cocaine’s vasoconstrictor effects could lead to muscle damage via an ischemiareperfusion injury leading to free radical formation and lipid peroxidation. This study, therefore, also investigated the possibility that cocaine-induced cytosolic enzyme release may To test this be mediated via the formation of free radicals leading to lipid peroxidation. a free radical scavenger, and hypothesis, muscle total glutathione levels, muscle thiobarbituric acid reactive substances (TBARS), a measurement of lipid peroxidation, were examined following an acute IV cocaine dose in mice. Sedentary BalbC mice were injected with cocaine (40 mg/kg) or normal saline via the tail vein. Creatine kinase levels in serum and total glutathione and TBARS in liver and muscle were determined at 4, 8, and 24 hrs. Serum creatine kinase levels were significantly elevated 5-fold, while TBARS were elevated 100% in the gastrocnemius muscle of cocaine-treated animals at 4 hrs compared to normal saline controls. However, serum creatine kinase levels, total glutathione and TBARS in the gastrocnemius muscle were not statistically different at 8 or 12 hrs; or in the liver and suggests that lipid anterior tibialis muscle at 4, 8, or 24 hrs. The present findings peroxidation may be occurring in skeletal muscle after a single IV cocaine dose in mice.
ZGzyWords: cocaine, skeletal muscle, intravenous, muscle damage, rhabdomyolysis, lipid peroxidation
Gayle A. Brazeau, Department of Pharmaceutics, University of Florida, J. Hillis Miller Health Science Center, Box 100494, Gainesville, Florida 32610, Phone (904) 392-3223, FAX (904) 846-0298.
1570
ESfects of Cocaine on Skeletal Muscle
Vol. 57,No. 17.1995
The increasing number of clinical reports describing cocaine-induced rhabdomyolysis leading to acute renal failure and the publicized deaths of athletes who used cocaine for its reported ergogenic properties has highlighted our lack of knowledge on the mechanisms responsible for cocaine’s toxicity on cardiac and skeletal muscle (l-7). The potential cardiotoxic mechanisms of cocaine have been extensively studied (8), however very little is known about the underlying mechanism(s) responsible for cocaine toxicity on skeletal muscle (9). Pagala and coworkers recently demonstrated an increased release of creatine kinase from rat soleus in vitro, but not from the extensor digitorum longus muscle (BDL) when incubated in the presence of 0.5-1.0 mM cocaine over a 4-hour period (9). These investigators suggested that cocaine may have a direct toxic effect on muscle tissue that could lead to rhabdomyolysis (9). To confirm if cocaine has a direct toxic effect on skeletal muscle, we investigated whether the release of creatine kinase from the isolated mouse soleus (SOL) and extensor digitorum longus (BDL) muscle was enhanced in the presence of cocaine. It is possible that cocaine-induced muscle damage could also be mediated via effects on other muscle associated tissues (e.g., vasculature). The short systemic half-life of cocaine compared to $lie development of toxic effects (which can be delayed relative to the time of drug administration) could suggest that these effects are caused by cocaine metabolites or secondary to events in other tissues. Cocaine-induced muscle damage, as measured by the release of cytosolic enzymes, could be mediated through its vasoconstrictor effects on muscle vasculature. It is well known that cocaine induces vasoconstriction in a variety of tissues by its ability to block catecholamine re-uptake at presynaptic vessels (10). Cocaine-induced vasoconstriction has been implicated as a possible mechanism in myocardial infarction, cerebrovascular accidents, obstetrical complications, and rhabdomyolysis (2, 7, 10, 11). Recent studies have suggested that cocaine’s toxic effects on the heart and brain seems to be mediated via vasoconstriction and decreased blood flow (12-14). It can be hypothesized that cocaine-induced muscle damage is the result of spasm and vasoconstriction of muscle associated vasculature. This decreased blood flow could cause ischemia with the resulting reperfusion leading to free radical formation. Therefore, we investigated whether an acute intravenous cocaine dose causes increased serum creatine kinase levels in mice and whether this elevation is associated with free radical formation and lipid peroxidation as measured by total glutathione levels and the formation of lipid peroxides through TBARS, respectively. Materials and Methods In vitro muscle exneriments. Male 2-month old BalbC mice were humanely sacrificed via cervical dislocation and the SOL and EDL muscles were carefully isolated and dissected free from adjoining muscles as previously described (15). The muscles were attached to small glass holders and placed into plastic tubes containing 4 ml of a mammalian Ringers solution gassed with 02/CO2 (95:5) as described previously (15). The tubes were mounted in a thermostat controlled water bath maintained at 37’C. At 30 minute intervals over the 3-hour incubation, the incubation solution (with or without cocaine) in each tube was replaced with fresh 37’C solution using a peristaltic pump. In these studies, one muscle from each animal was incubated in the presence of the cocaine and the contralateral muscle served as the control. Two based upon using either experiments,
concentrations of cocaine were utilized in these the work of Pagala and coworkers. The isolated a pulse or a continuous experimental protocol. the muscles were pre-incubated in the mammalian
experiments (1 mM and 0.2 mM) muscles were exposed to cocaine In both the pulse and continuous Ringers solution alone during the
Vol. 57, No. 17, 1995
Effects of Cocaine on SkeletalMuscle
1571
first 30 minutes of the incubation. In the pulse experiments, muscles were incubated in the absence or presence of cocaine between 30-60 minutes only. For the remaining incubation period (60-180 minutes), the muscles were incubated in the mammalian Ringers solution and the media changed every 30 minutes. In contrast during the continuous exposure, the muscles were incubated in the absence or presence of cocaine between 30-180 minutes, with fresh cocaine solution added every 30 minutes. Creatine kinase release from the isolated muscles is expressed as the cumulative sum during the experimental period. In vivo cocaine emeriments. Conscious Female BalbC mice were injected with normal saline (0.1 ml/l0 g) or cocaine (40 m@g) in normal saline (0.1 ml/l0 g) via the tail vein. The injection was administered over S-10 seconds. The animals did not struggle during the injection procedure nor did they experience any tremors or seizures at this dose. These injections were performed by one of the investigators (MJJ) who has extensive experience with this technique. At 4, 8, and 24 hours, the animals were humanely sacrificed by a pentobarbital overdose, an aortic blood sample taken and tissues harvested. The tibialis anterior muscle, gastrocnemius muscle, and a portion of the liver was harvested and frozen immediately in liquid nitrogen. SamDIe analvsis. Creatine kinase activity in the incubation solutions and serum samples was measured using an enzyme-linked assay as described previously (15). Muscle and liver homogenates (20% w/v) were prepared in 1% sulphosalicylic acid. Total muscle glutathione was quantified by the method of Jackson and coworkers (16) developed from the work of Anderson (17). A portion of the homogenate was set-aside for protein determination using the Lowry method (18). Thiobarbituric acid reactive substances (TBARS) in the muscles were measured according to the method of Phoenix and coworkers (19) with the exception that the tissues were homogenized in 1% sulphosalicylic acid. Data analvsis. Data are presented as mean and standard error Statistical differences between cocaine treated muscles and control muscles were analyzed using an unpaired Student’s T-test. In the in vivo studies, between normal saline and cocaine-treated at the three time points were unpaired Student’s T-test.
of the mean (SEM). in the in vitro studies statistical differences determined using an
Results In vitro creatine kinase release from isolated mouse muscles. The cumulative release of creatine kinase from isolated mouse muscles exposed to a pulse 1 mM cocaine dose is shown in Figure 1 (Panel A). Creatine kinase release was significantly higher in EDL muscles exposed to a pulse 1 mM cocaine dose compared to untreated control muscles. There was no statistical difference in the cumulative creatine kinase release from the SOL muscles between a pulse cocaine exposure and control untreated muscles. To determine if muscle damage could be further exacerbated by continuous exposure to EDL and SOL muscles were incubated in fresh cocaine solution at 30 minute intervals (Figure 1, Panel B). The cumulative release of creatine kinase from continuously cocaine exposed EDL or SOL muscles was not statistically different from the untreated control muscles. Furthermore, in the continuous cocaine exposure protocol, cumulative creatine kinase release was less than that seen in the pulse exposure protocol. The reason for the reduced levels of creatine kinase release seen during the continuous exposure protocol is not known. Preliminary studies showed that the presence of cocaine in the bath did not interfere with the 1 mM cocaine,
Eiffects of Cocaine on SkeletalMuscle
1572
Vol. 57, No. 17,lWS
A
0
30 60
90
6 Continuous
120
150
180
30
60
90
120
150
180
Time (mii)
Fig. 1 Cumulative creatine kinase release from isolated EDL or SOL muscles incubated with 1 mM cocaine (m) as a 30-minute pulse exposure (Panel A) or continuous exposure (Panel B). Control muscles (0) were incubated in the Mammalian Ringers. Values are presented as mean and SEM (n = 4-6) with p< 0.05 (*). determination of creatine kinase activity (data not shown). Alternatively, it can be proposed that cocaine itself or an unknown compound generated or released by cocaine subsequently blocks the creatine kinase release. A pulse or continuous exposure to 0.2 uM cocaine did not cause any statistically significant increase in the cumulative release of creatine kinase from either the isolated EDL or SOL muscle (data not shown). Intravenous cocaine studies. There were no fatalities using 40 mg/kg dose and dosing route, however increasing the dose to 50 mg/kg resulted in approximately 50% mortality (data not shown). The animals receiving this dose and dosing route did not exhibit any tremors or signs of muscle contraction following injection or during the experimental period. This dose of cocaine (40 mg/kg) resulted in a significant 5-fold increase in serum creatine kinase levels compared to normal saline controls suggesting that muscle damage may be occurring with this treatment (Figure 2). Serum creatine kinase levels were also higher at 8 hours however, these were not statistically significant compared to saline controls. There was no statistically significant difference between the two groups at 24 hours.
Vol. 57, No. 17, 1995
Effects of Cocaine on SkeletalMuscle
lx3
There was no statistical difference in glutathione levels between cocaine-treated and normal saline-treated animals in the gastrocnemius (Figure 3) and tibialis anterior muscle (data not shown). This could be a function of the large intra-day variability of glutathione levels in these tissues. Muscle glutathione levels were found to vary approximately 2533% in the control normal saline treated animals. In the liver, glutathione levels in cocaine treated animals were lower compared to normal saline at 4 and 8 hours, but these values did not reach statistical significance (Figure 4). Liver glutathione levels in normal saline treated animals were extremely variable at each of the time points and throughout the study period. Consistent with the increased serum creatine kinase levels at 4 hours, TBARS absorbance values were significantly higher in the gastrocnemius muscles from cocaine-treated compared to normal saline-treated mice (Figure 5). At 8 and 24 hours, these values were not different between the two groups. Likewise, the TBARS absorbance values from the liver samples were not different between the two groups at each of the three time points (data not shown). Due to the small size of the anterior tibialis muscle, TBARS could not be quantified in this muscle. Discussion The findings from the in vitro studies suggest that cocaine-induced creatine kinase release, often used as an indirect measure of muscle damage, may be mediated by a direct drug effect on the muscle. Creatine kinase release was significantly enhanced when mice EDL
1400
1
4
a
24
Time (hrs)
Fig. 2 Serum creatine kinase activities in normal saline (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n = 5) with p< 0.05 (*).
1574
Vol. 57, No. 17, 1995
Effects of Cocaine on Skeletal Muscle
0.6
0.5 F P ;
0.4
% % l 0.3 g z 0.2 1 S G 0.1
0.0
24
Time (hrs)
Fig 3.
Total glutathione content of the gastrocnemius muscle from normal saline- (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n =
Liver
110 100
i
4
8
24
Time (hm)
Fig. 4
Total glutathione content of the liver Corn normal saline- (open bars) and cocainetreated (hatched bars) mice. Values are presented as mean and SEM (n = 5).
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17, 1995
0.7
I
0.6
-
0.5
-
0.4
-
5
0.3
-
x’ I; g (3
0.2
-
0.1
I-
G D 9 E
I.575
% 2 8 ._
0.0
8
24
Time (hrs)
Fig. 5
TBARS content of gastrocnemius muscle homogenates from normal saline- (open bars) and cocaine-treated (hatched bars) mice. Values are presented as mean and SEM (n = 5) with p< 0.05 (*). muscles were incubated for a short period of time (30 minutes) with a cocaine concentration (e.g., 1 n&I) 2-10x higher than the highest concentration reported in overdose patients (9). Like previous investigators, it was necessary to use this higher dose in mice (versus humans) to demonstrate an effect on muscle because of species differences in the pharmacokinetics, pharmacodynamics or toxicological profile. These findings differ from those reported by Pagala and coworkers who showed significantly higher creatine kinase from only rodent SOL muscles (9). Certainly, species differences might account for the disparate results between the two studies. However, the nature of the experimental design in the two studies (viz., nature of the incubation buffer, duration of the experiment, continuous versus pulse exposure) could also account for the different findings. Furthermore, it is unclear to what extent cocaine may have been degraded or metabolized in the muscle during the 4 hour study period in the studies by Pagala and coworkers (9). Another concern with isolated muscle studies, that has not been addressed in earlier studies, is whether the presence of the drug in the solution interferes in the analysis procedure. Preliminary studies showed that cocaine at the concentrations used in the present investigations did not directly affect the enzymatic determination of creatine kinase using the spectrophotometric kinetic assay (data not shown). Clinical reports of rhabdomyolysis in patients and experimental studies investigating cocaine’s toxic effects on various organ systems, including skeletal muscle, have implicated cocaine-induced vasoconstriction as a possible initiating event in cocaine-induced muscle damage (l-2). It can be postulated that cocaine-induced muscle damage may be the result of vasoconstriction leading to muscle ischemia with the subsequent tissue reperf%sion leading to free-radical formation and lipid peroxidation. The vasoconstrictor effects of cocaine, mediated
1576
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17, 1995
either through the adrenergic or non-adrenergic mechanisms, have been well established in animal models (13-14, 20). Likewise in more recent studies, cocaine metabolites have been shown to have vasoconstrictor properties. Studies have suggested that benzoylecgonine has vasoconstrictor properties and itself may be toxic to tissues (13). In addition, toxicity may be mediated via the formation of cocaine metabolites which generate further free radicals in muscle tissue. It has been suggested that metabolism to norcocaine in the liver is the principle pathway of bioactivation leading to the formation of N-hydroxynorcocaine. This latter metabolite is subsequently converted to the free radical norcocaine nitroxide. Toxicity in tissues may result from the titile recycling between N-hydroxynorcocaine and norocaine nitroxide resulting in the depletion of NADPH and the production of hydrogen peroxide (21). The present study, which uses a lower dose than that reported to cause hepatotoxicity (22), provides evidence that a single acute intravenous dose of cocaine (40 mg/kg via the tail vein) causes increased serum creatine kinase levels at 4 hours. This increase in serum creatine kinase levels corresponds with the presence of lipid peroxidation in the gastrocnemius muscle as measured by increased TBARS. Since cocaine is rapidly metabolized by plasma esterases and hepatic enzymes, the increased serum creatine kinase levels combined with the presence of lipid peroxidation suggests that muscle damage could be caused via the formation of free radicals which damage muscle membranes leading to increased cytosolic enzyme release. At 8 hours, serum creatine kinase levels were still elevated in the cocaine-treated animals relative to normal saline controls, but these values were not statistically different. As in other in vivo studies utilizing creatine kinase as a indicator of muscle damage, the range of the values was large which could account for the lack of statistical significance. However, the rapid decline of serum creatine kinase levels is consistent with the reported short half-life (e.g., < 2 -4 hrs.) of this enzyme in rodent serum (23). Furthermore, it is possible that the peak creatine kinase and TBARS levels and decreased glutathione levels could have occurred earlier than the first sampling time (e.g., 4 hours). The 4-hour initial sampling time was selected based upon other toxicity studies. To verity whether muscle damage is occurring with cocaine dosing, fiture studies will focus on determining total serum creatine kinase levels, creatine kinase isoenzyme profiles (e.g., to determine whether the source is cardiac, muscle or brain) and histological evaluation at various time points. The increased TBARS in the gastrocnemius muscle supports the hypothesis of free radical formation leading to peroxidation of membrane lipids with cocaine administration. It could subsequently be predicted that muscle glutathione levels should be decreased if this hypothesis is correct. In contrast, muscle total glutathione levels were not decreased with cocaine administration. Hepatotoxicity following cocaine administration has been shown to cause decreased liver glutathione levels (22). Muscle and liver glutathione levels were not statistically lower in the cocaine-treated animals compared to the normal saline. The variability of tissue glutathione levels in the normal saline-treated animals over the study period, particularly in the liver, and the lack of statistical significance from the cocaine-treated animals may be a function of daily fluctuations in these levels (24-25). Alternatively, it has been suggested that cocaine-induced hepatotoxicity in mice as measured by changes in glutathione levels may be strain dependent (26). Finally, a single intravenous cocaine dose might increase free radical formation, but other radical scavenger agents and processes (e.g., a-tocopherol, superoxide dismutase) rather than glutathione might be responsible for inactivating these molecules. Decreases in muscle glutathione levels might only occur with repeated cocaine ingestion. It is certainly possible that the effects of cocaine, either direct or indirect, are additive and a f?.mction of repeated doses. Most of the reports of rhabdomyolysis in humans has been associated with those who chronically abuse cocaine.
Vol. 57, No. 17, 1995
Effe&s of Cocaine on Skeletal Muscle
1577
The demonstrated increase in lipid peroxidation as measured by an increase in TBARS would suggest that a likely target for free radicals would be membrane lipids. Similar findings have been reported in cocaine-induced hepatotoxicity (27). However, the present data does not rule out the possibility that other cellular targets (e.g., proteins, DNA) in skeletal muscle might also be affected during acute and chronic cocaine ingestion. Overall, these findings would point to the idea that muscle damage (measured by increased serum creatine kinase levels) following cocaine ingestion is mediated via the formation of free radicals causing lipid peroxidation. It is possible that chronic cocaine ingestion could lead to repeated ischemic events mediated by vasoconstrictor properties of cocaine and its metabolites. Damage to the muscle tissue would result from the subsequent repertusion and the generation of free radicals. If these free radicals are not inactivated by radical scavenging systems in the muscle and associated vasculature, this could result in damage to the muscle fibers. Acknowledgments This work was supported by a Gustavus A. Pfeiffer Faculty Development Research Fellowship from the American Foundation for Pharmaceutical Education. We would like to thank Sue Page for assistance in the assay methodologies. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
13. 14. 15. 16. 17. 18.
R.D. FELD, Clin. Chem. News, 3 April (1993) E. A. TURBAT-HERRARA, Ultrastruct. Pathol. 18 127-13 1 (1994). V.A. POGUE and M. NURSE, Amer. J. Med. & 183-186 (1989). B.D. HERZLICH, E.L. ARSURA, M. PAGALA and D. GROB, Ann. Int. Med. 109 335-336 (1988). J.G. ZAMORA-QUEZADA H. DINERMAN, M.J. STADECKER and J. J. KELLY, Ann. Int. Med. &3 564-566 (1988). N. COOPER B. COHN and S. DOHERTY, Newsweek 26, July 7 (1986). A. KETEYIAN and B. SELCRAIG, Sports Illustrated 65 18- 19 (1986). A.A. WELDER P. G RAMMAS and R.B. MELCHERT, Toxicol. Lett. @ 227-238 (1993). M. PAGALA, B. AMALADEVI, D. AZAD, S. PAGALA, B. HERZLICH, T. NAMBA and D. GROB, Life Sci. 52 751-756 (1993). L.L CREGLER and H. MARK, New Eng. J. Med. 315 1495-1500 (1986). T.T. INABA, D. STEWART and W. KALOW, Clin. Pharmacol. Therap. 23 547-552, (1978). B.D. NUNEZ, L. MIAO, Y. WANG, M.M. NUNEZ, F.W. SELLKE, J.N. ROSS, V. SUSULIC, G.Y. CARROZZA, G.Y. PAIK, J.P. CARROZZA and J.P. MORGAN, Circulat. Res. 74 281-290 (1994). J.A. MADDEN and R.H. POWERS, Life Sci. 47 1109-l 114 (1990). C.D. KURTH, C. MONITTO, M. L. ALBUQUERQUE, P. FEUER, E. ANDAY and L. SHAW, J. Pharmacol. Exp. Therap. 587-591 (1993). D.A. JONES, M.J. JACKSON and R.H.T. EDWARDS, Clin. Sci. @ 193-201 (1983). M.J. JACKSON, M.H. BROOKE, K. KAISER and R.H.T. EDWARDS, J Clin. Sci. 80 559-564 (1991). M. ANDERSON, Methods Enzymol. 113_ 548-555 (1985). O.H. LOWRY, N.J. ROSENBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193 265-275 (1951).
1578
19. 20. 21. 22. 23. 24. 25. 26. 27.
Effects of Cocaine on Skeletal Muscle
Vol. 57, No. 17,195
J. PHOENIX, R.H.T EDWARDS and M.J. JACKSON, Brit. J. Nutrit. &l 245-256 (1990). M.M. KNUEPFER, CA. BRANCH, D.M. WEHNER, Q. GAN and D. HOANG, Can. J. Physiol. Pharmacol. 22 335-343 (1994). M. W., KLOSS, G.M. ROSEN and E. J. RAUCKMAN, Biochem. Pharmacol. 33 169-173 (1984). M.L. THOMPSON, L. SHUSTER and K. SHAW, Biochem. Pharmacol. 28 2389-2395 (1979). S.F. PAGE, Thesis, University of Liverpool (1992). M.J. TUNON, P. GONZALES, P. LOPEZ, G.M. SALIDO, and J.A. MADRID, Arch. Int. Physiol. Biochim. Biophys. 100 83-87 (1992). P.M. BELANGER, M. DESGAGNE and B. Bruguerolle, Drug Met. Dispos. Biol. Fate Chem., 19 241-244 (1991). C.S. BOYER and D.R. PETERSEN, Hepatology 14 1209-1216 (1991). M.W. KLOSS, G.M. ROSEN, E.J. RAUCKMAN, Toxicol. Let. li 65-70, (1983).