Materials and Methods

Drugs and Chemicals.

(+)-METH, (+)-AMP, and [³H](+)-METH were obtained from the National Institute on Drug Abuse (Rockville, MD). Norephedrine was obtained from Sigma Chemical Co. (St. Louis, MO). The [³H](+)-METH [(+)-[2′,6′-³H(n)](+)-methamphetamine; 23.5 Ci/mmol, 1.01 mCi/ml] was synthesized with radiolabel at positions 2 and 6 of the aromatic ring, which are metabolically stable sites. The synthesis was performed by the Research Triangle Institute (Research Triangle Park, NC) for the National Institute on Drug Abuse. All drug doses and concentrations are expressed as the free base form. All other reagents used in these studies were obtained from Fisher Scientific (Springfield, NJ).

Animals.

Male Sprague-Dawley rats (250–300 g;n = 3 per time point) with a single surgically implanted jugular venous cannula were obtained from Hilltop Laboratory Animals (Scottsdale, PA). The rats were shipped with the cannulas imbedded in the s.c. space between the scapulae. The day after delivery, the rats were anesthetized with ethyl ether, and the cannulas were exposed. Each cannula was flushed after exposure with 200 μl of saline and 50 μl of heparin (50 U) to avoid clotting of the cannulas. Rats were then allowed to recover for 1 week during which the cannulas were flushed with heparinized saline every 2 days. Rats were maintained in an animal care facility, with a 12-h light/dark cycle (7:00 AM–7:00 PM) and a mean temperature of 22°C. All experiments performed in these studies were in accordance with the Guide for the Care and Use of Laboratory Animals, as adopted and promulgated by the National Institutes of Health. The University of Arkansas for Medical Sciences Institutional Animal Care and Use Committee (Little Rock, AR) also approved all animal protocols.

Tissue Distribution Experiments.

Rats were placed in metabolism cages the day before the experiment with free access to water. At the beginning of each experiment, 1.0 mg/kg (+)-METH diluted in saline containing 80 μCi of [³H](+)-METH (as a tracer) was injected as a 15-s bolus dose into the jugular venous cannula. The volume of injection was 1.0 ml/kg. At predetermined time points (2, 5, 10, 20, and 40 min; 1, 2, 3, 4, and 5.5 h) after the injection, rats were anesthetized with ethyl ether. A laparotomy was performed and blood was drawn from the inferior vena cava. The animal was decapitated and the brain, heart, left hepatic lobe, left kidney, and spleen were removed rapidly. For early time points (2–10 min), rats were anesthetized with ethyl ether before drug administration to facilitate removal of blood and tissues on time. Ethyl ether was used because it maintains hemodynamic stability. Three rats were used per time point. Tissues were rinsed with saline, weighed, quick frozen in liquid nitrogen, and stored at −80°C until analysis. All tissues were removed in 3 min or less. The blood was allowed to clot for 1 h at room temperature, after which it was centrifuged to collect serum. The serum was stored at −80°C until analysis.

In a study of (+)-METH and (+)-AMP steady-state tissue distribution after a (+)-METH infusion, three rats were implanted with s.c. osmotic minipumps (Alza Corporation, Palo Alto, CA), which deliver a constant flow rate over 24 h (9.2 μl/h). The rats were anesthetized with ether and, after shaving and ethanol skin preparation, an incision (2 cm) was made across the lower back. The underlying tissues were dissected bluntly to the fascia overlying the back muscles, and a pocket was formed with blunt dissection cephalad from the incision. The loaded pump was inserted into the pocket, and the incision closed. Rats received a total dose of 1.0 mg/kg of (+)-METH diluted in sterile saline containing 80 μCi of [³H](+)-METH over a 20-h period. Blood and tissues were harvested 20 h after pump implantation as described above. Because thet_1/2λZ (+)-METH is 63 min (Rivière et al., 1999), the animals would have achieved steady state at 4–7 t_1/2λZ (i.e., 4–7 h).

Tissue Analysis.

For extraction of (+)-METH and (+)-AMP, tissues were homogenized after addition of 4 volumes (v/v) of ice-cold water with a tissue homogenizer (Tekmar Company, Cincinnati, OH). (+)-METH and (+)-AMP were extracted from the tissue homogenates with a two-step procedure, which consisted of an extraction step with hexane, and a back extraction step with HCl (0.1 N) (Rivière et al., 1999).(+)-METH and (+)-AMP concentrations were then determined with the HPLC procedure described previously (Rivière et al., 1999). (+)-METH and (+)-AMP tissue concentrations were corrected for blood content with the following equation:C_TOTAL = [(C′_TISSUE −(C_B *V_B )) ÷ (1 −V_B )] whereC_TOTAL is the (+)-METH or (+)AMP total tissue concentration, C′_TISSUE is the concentration of (+)-METH or (+)-AMP determined in tissue,C_B is the concentration in blood [(+)-METH and (+)-AMP are equally distributed in blood and serum], and V_B the volume fraction of blood contained in each tissue (Triplett et al., 1985).

Pharmacokinetic Analysis.

The tissue average concentration versus time profiles for (+)-METH and (+)-AMP were analyzed with model-independent pharmacokinetic methods. Area under the serum concentration versus time curve (AUC) values were calculated for the averaged data with the log-trapezoidal rule from time zero to the last experimental data point. AUC values were then extrapolated from the last experimental data point to infinity by calculatingC/λ_Z, where C is the predicted concentration at the last measured time point, and λ_Z is the terminal rate constant. The terminal half-life was estimated with data from the terminal elimination phase for each animal [from 1 to 5.5 h and from 3 to 5.5 h after the injection for (+)-METH and (+)-AMP concentration-time curves, respectively]. In addition, the brain and serum average concentration versus time profiles for (+)-METH were analyzed with model-dependent methods. The nonlinear least-squares fitting routine of the software WinNonlin (Scientific Consulting, Inc., Cary, NC) was used for the model-dependent analysis. Two- and three-compartment models were fit to the averaged concentration versus time data with 1/y or 1/y² weighting, where y is the predicted concentration. The best fit line was selected with visual inspection of the line, analysis of the residuals and a statistical F test for selecting between the alternative equations (Boxenbaum et al., 1974).

Results

General Experimental Strategy.

The 1.0 mg/kg i.v. bolus dose of (+)-METH used in this study was chosen based on our previous study of (+)-METH dose-response relationships (Rivière et al., 1999). This previous study of rat spontaneous motor activity and serum pharmacokinetics of (+)-METH and its metabolite (+)-AMP showed that effects lasted 2 to 3 h and the pharmacokinetics of (+)-METH was not different following a 10-fold range of doses (0.1–1.0 mg/kg) (Rivière et al., 1999). In addition, this 1.0-mg/kg dose was chosen because it is in the range of doses that are reportedly used by humans on a milligram per kilogram basis (Cho, 1990; Beebe and Walley, 1995). Conversely, the dose used for the s.c. infusions (i.e., 1.2 mg/kg/day) was chosen because it does not produce pharmacological effects. This allowed us to compare (+)-METH and (+)-AMP tissue partitioning at steady-state concentrations following s.c. doses with the tissue partitioning after i.v. doses. We were particularly interested in the time course of tissue partitioning during and after the period of pharmacological activity.

To more accurately determine brain tissue concentrations, we first measured whole-brain concentrations and then subtracted the amount of drug in the blood content in the brain based on inferior vena cava venous concentrations. We used this approach in part because it is known that many rapidly acting drugs exhibit a high arterial-to-venous concentration gradient especially during the first few minutes after i.v. administration. For instance, Tucker and Boas (1971) noted a 10:1 peak in arterial versus venous concentrations for lidocaine immediately after i.v. injection in humans, and Gourlay and Benowitz (1997) noted a temporary 2:1 peak for nicotine at the end of a 30-min i.v. infusion in humans. However, in both studies the arterial-to-venous differential decreased quickly over the next 10 min. Because the volume percentage of blood in the brain is so small in the human and the rat (4 and 3%, respectively; Birnbaum et al., 1994) compared with the total volume of brain tissue, temporarily elevated drug concentrations in the arterial blood are unlikely to contribute to the significantly elevated brain/serum ratios found in this study. Indeed, we calculated that if concentrations of blood in the brain were actually 10 times greater than our measured venous concentrations, this would only result in a 10% decrease in our calculated brain concentrations.

Pharmacokinetics of (+)-METH and (+)-AMP in Serum after i.v. Dosing.

The time course of (+)-METH concentrations in serum demonstrated a biexponential decline that was best described by a two-compartment pharmacokinetic model with 1/y² weighting (Fig.1). The fitted curves were used to estimate the distribution half-life values (t_1/2λ1) of (+)-METH in the serum (and brain), to help determine the starting point of the terminal elimination phase, and to check the accuracy of the model-independent pharmacokinetic analysis in serum. The calculated pharmacokinetic values determined by model-independent and model-dependent methods did not vary by >10% (data not shown). Based on visual inspection of the graphs of the concentration-time data for (+)-METH and (+)-AMP, we observed that the 3-h time point appeared to be consistently above the general trend of the exponential decay of the concentration-time data in all tissues. In addition, when we compared this trend with our previous pharmacokinetic studies in serum (Rivière et al., 1999), we did not see aberrant data at this time point to suggest a reason (e.g., physiological) for an increase in concentration at ∼3 h. Thus, we considered the data from the 3-h time point in all tissues to be outliers and did not include them in our pharmacokinetic analysis. We did include this time point in our graphs for completeness (Fig. 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Average concentration versus time profiles for (+)-METH (●) and (+)-AMP (○) in various tissues following a 1.0-mg/kg i.v. dose of (+)-METH in rats (n = 3 rats per time point). These plots are shown in rank order from the largest to smallest (+)-METH AUC (left to right, top to bottom). All values are represented as the mean ± S.D. The solid line represents a linear regression fit to the terminal log concentration versus time data.

The serum (+)-METH pharmacokinetic parameters calculated in this study were determined with model-independent analysis of averaged concentration versus time data from three animals at each time point. Consequently, it was not surprising that the serum pharmacokinetic parameters were somewhat different from our previous study in which serum pharmacokinetic values were determined from multiple samples in individual animals with the same dose (Rivière et al., 1999). Data from the previous study showed theVd_SS was 9.0 l/kg (versus 4.9 l/kg in the current study), the Cl_T was 126 ml/min/kg (versus 76.7 ml/min/kg), and thet_1/2λZ was 63.0 min (versus 54.2 min). Because the analytical methods were the same and many aspects of the experimental protocol were similar between the two studies, we do not think the differences were due to analytical errors. The differences more likely resulted from the use of averaged serum concentration-time data to determine pharmacokinetic parameters in the current study compared with the use of multiple serum samples from individual animals to calculate pharmacokinetic values in the previous study. We observed similar pharmacokinetic differences when we analyzed averaged serum concentrations from groups of animals versus multiple serum samples from individual animals in a previous study of phencyclidine (i.e., the Vd_SS and Cl_T were lower with averaged serum concentration-time data; Valentine et al., 1994; Valentine and Owens, 1996). Although we think multiple samples from an individual animal yield the most accurate results for serum pharmacokinetics, this approach cannot be used when quantifying tissue (i.e., organ) pharmacokinetics.

The time course of (+)-AMP concentrations in serum showed that the highest concentration of (+)-AMP was detected at the 20-min time point after the injection of (+)-METH (Fig. 1). The serumt_1/2λZ of (+)-AMP was 74.9 min, which appeared to be greater than the 54.2 mint_1/2λZ calculated for (+)-METH (Table 1). In our previous serum pharmacokinetic studies of (+)-METH and (+)-AMP in individual animals, the (+)-AMP t_1/2λZ also was found to be longer (and statistically different, P < .05) than the (+)-METH t_1/2λZ following the same dose of (+)-METH used in the current study (Rivière et al., 1999).

Pharmacokinetics of (+)-METH and (+)-AMP in Brain and Other Tissues after i.v. Dosing.

In all tissues except the spleen, the (+)-METH concentrations were highest at the first measured time point (2 min), and the concentration versus time curves showed biexponential declines. In the spleen, the maximum (+)-METH concentration occurred at the 10-min time point (Fig. 1). As in the other tissues, (+)-AMP concentrations increased to a maximum value at 20 min before decreasing. The t_1/2λZ values of (+)-AMP were from 3.8 min longer in the liver to 20.7 min longer in the serum and kidney than the (+)-METH t_1/2λZvalues (Table 1).

Pharmacokinetics of (+)-METH and (+)-AMP during s.c. Infusion.

Because we wanted to determine steady-state brain-to-serum partitioning after a chronic infusion of a nonpharmacologically active (+)-METH dose, we chronically infused rats at a rate of 1.2 mg/kg/day for a total dose of 1.0 mg/kg at the time of sacrifice. Although this was the same total dose as used in the acute i.v. administration, when infused over the 20-h period, it did not produce apparent pharmacological effects.

Based on the t_1/2λZ values in serum for (+)-METH (i.e., 54.2 min) and (+)-AMP (i.e., 74.9 min; Table 1), we estimated that steady-state concentrations would be reached by ∼4–9 h (i.e., 4–7 t_1/2λZ) after starting the s.c. infusion. However, for convenience, we chose to infuse the animals for a 20-h period. The brain:serum (+)-METH concentration ratio at 20 h was similar to that calculated with concentrations determined in the terminal elimination phase starting at 2 h after the i.v. bolus injection (9.3:1 for s.c. infusion versus 8.4:1 for i.v. bolus). We observed the same results for (+)-AMP where the brain/serum ratio at steady state was 7.4:1 versus 7.6:1 for the i.v. bolus. In all other tissues (heart, liver, kidney, and spleen), we also observed that the steady-state tissue/serum (+)-METH or (+)-AMP partitioning values were similar to the values during the terminal elimination phase starting at 2 h. However, the absolute magnitude of the value varied depending on the tissues.

Discussion

In this study, we characterized the concentration history of (+)-METH and its metabolite (+)-AMP in the brain and other tissues following i.v. bolus administration of (+)-METH (Fig. 1). These data showed that (+)-METH distributes rapidly into the brain after i.v. bolus dosing. Indeed, the high (+)-METH concentrations observed in the brain immediately after (+)-METH administration suggested that there is essentially no hindrance to passage of (+)-METH at the blood-brain barrier. In addition, the (+)-METH brain concentrations were ≥7 times the serum concentrations from the first measured time point (2 min) through the remainder of the study. The extremely rapid partitioning of (+)-METH into the brain is consistent with its physicochemical properties (Cho, 1990). It is a small (mol. wt. 149.2), lipid-soluble molecule (octanol/water partition coefficient of 2.3; Brodin et al., 1976) that would be expected to distribute extensively and rapidly into high lipid-content tissues (e.g., brain) that have high blood flow-to-tissue volume ratios. Furthermore, the partitioning could be affected by (+)-METH and (+)-AMP binding to active sites in the central nervous system, and by drug transporters at the blood-brain barrier. For instance, P-glycoprotein is an important efflux pump for some drugs and carrier-mediated transport into the brain via a choline transporter has been reported for other lipophilic drugs such as lidocaine and propranolol (Pardridge et al., 1984; Yamazaki et al., 1994).

An important finding of the current study was that the concentration-time profile in the brain was not readily predictable from the concentration-time profile in the serum, or from the physicochemical properties of (+)-METH (Fig.2). Indeed, if the rate and direction of (+)-METH movement across the blood-brain barrier were only dependent on cerebral blood flow and (+)-METH lipid solubility, the concentration-time curves of (+)-METH in the brain and serum should be parallel. However, we found that the concentration-time courses did not decrease in parallel in the two tissues during the first hour after drug administration. This was primarily because brain concentrations did not decrease as rapidly as serum concentrations for at least 1 h (Fig. 2). This also is demonstrated by the differentt_1/2λ1 values for the brain and serum (11.8 versus 2.3 min, respectively), and by the apparent rise and fall in the brain to serum concentration ratios during the first hour after (+)-METH administration. The brain/serum concentration ratio rose from 6.8:1 at 2 min to a peak of 12.5:1 at 20 min before decreasing to 8.1:1 at 2 h, where it remained constant for the duration of the experiment (Fig. 2, inset). Melega et al. (1995) reported a relatively constant value for the (+)-METH brain/serum ratio of 10:1 during their relatively short sampling period (5 to 60 min versus 2 to 5.5 h in the current study). They did not find an apparent dysequilibrium between the amount of METH in the brain and serum during the first 2 h after (+)-METH administration, as found in the current study.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Average concentration versus time profiles for (+)-METH in the brain (●) and serum (○) (n = 3 rats per time point) following i.v. dosing. The solid line represents a best-fit line determined by model-dependent analysis. The inset shows the brain-to-serum ratios for METH, which allows a better visual representation of the apparent early (first 2 h) dysequilibrium between the brain and serum METH concentrations. All values are represented as the mean ± S.D. The inset shows the average (+)-METH brain-to-serum concentration ratios.

We could not relate the time-dependent partitioning found in the current studies (Fig. 2) to the distribution or eliminationt_1/2 values in the brain or serum because the partitioning and pharmacological effects at a 1.0-mg/kg i.v. (+)-METH dose last well past the distribution phase and continue into the terminal elimination phase. Rivière et al. (1999) report that spontaneous locomotor activity (i.e., distance traveled and number of rearing events) lasts ∼3 h after a 1.0-mg/kg i.v. bolus dose of (+)-METH in rats. As an aid to better understanding the pharmacological significance of a rapid (+)-METH input (i.e., i.v. bolus administration) on the tissue distribution of (+)-METH, we studied the partitioning of (+)-METH after a steady-state infusion of a total dose of 1.0 mg/kg given over a 20-h period. These data showed that the steady-state partitioning at this low, pharmacologically inactive dose was essentially the same as the (+)-METH partitioning (9.3:1 for the s.c. infusion versus 8.4:1 for the i.v. bolus) after the end of the pharmacological effect period following the i.v. bolus dose. Thus, except for the 2-min time point (7:1), (+)-METH brain-to-serum ratio values of >8:1 appeared to be associated with the pharmacological effect period. However, these data do not suggest a direct relationship between locomotor effects and the brain-to-serum ratio because a plot of the locomotor effects (distance traveled or rearing events) versus the brain-to-serum values resulted in a reverse hysteresis loop (data not shown). The results of this analysis suggest the pharmacological effects lag behind the brain-to-serum ratio values. Nevertheless, we wondered if the temporary but detectable increased partitioning (above 8:1) of (+)-METH into the central nervous system during the pharmacological effect period might reflect (+)-METH binding to active sites.

A similar hypothesis has been proposed for nicotine. Russell and Feyerabend (1978) observed an elevated brain/blood ratio for nicotine after i.v. bolus administration in mice that was not observed following i.p. administration based on the data of Stalhandske (1970). The nicotine brain/blood ratio after i.v. bolus administration remained elevated for ∼1 h, after which it decreased to a relatively constant value for the remainder of the study. Russell and Feyerabend (1978) go on to suggest that these same results indicate “that the brain cells bind and retain nicotine against a concentration gradient over and above what is determined by lipid solubility.” The influence of rate of drug administration on time-dependent brain partitioning needs further study to clarify the relationships between drug onset time and factors that affect drug distribution in the brain.

The current studies of distribution of i.v. (+)-METH into other tissues also revealed that the highest concentrations were observed at the first measured time point in the kidney, liver, and heart in concentration rank order (Fig. 1). High early concentrations were not surprising in these tissues because they receive a high percentage of cardiac output compared with their tissue mass in the rat (11, 27, and 4% of total cardiac output, respectively; Ebling et al., 1994). Of the tissues studied, only the spleen showed an early rise in concentrations, which peaked at 10 min after i.v. injection. The percentage of cardiac output going to the spleen (1%; Ebling et al., 1994) is much smaller than that going to the kidney or liver in the rat, so more time would be expected for distribution of (+)-METH into the spleen. Furthermore, (+)-METH concentrations in all of the organs were greater than in the serum. Based on data from the AUC after i.v. dosing, the rank order of (+)-METH tissue accumulation was kidney > spleen > brain > liver > heart > serum. Based on the AUC values for (+)-AMP, the rank order for (+)-AMP was kidney > spleen > liver > brain > heart >serum (Table 1; Fig. 1). We previously reported that renal elimination of (+)-METH is a much greater component of total body clearance in humans than in rats (Rivière et al., 1999). Based on analysis of (+)-METH serum pharmacokinetics in rats (Rivière et al., 1999) and humans (Cook et al., 1993) after i.v. administration, 12.8% of a 1.0-mg/kg dose in rats is eliminated in the urine, whereas 45% of a 0.2-mg/kg i.v. dose in humans is eliminated in the urine. Thus, for both species physiological factors that alter urinary elimination of (+)-METH could be an important factor in the rate of elimination.

The consistently elevated (+)-AMP concentrations in all tissues above serum concentrations indicated that, like (+)-METH, (+)-AMP distributes extensively into these tissues. Analysis of the molar ratio of the (+)-AMP to (+)-METH AUC showed that (+)-AMP accounted for ∼39 to 48% of the (+)-METH AUC (Table 1). This has important implications for the overall pharmacological effects of i.v. (+)-METH because these data suggest that approximately one-third of the molar drug concentration in the tissues is due to (+)-AMP. Because (+)-AMP concentrations increased in tissues at rates similar to the rate of rise of (+)-AMP concentrations in the serum, the (+)-AMP formation rate and not the (+)-AMP elimination rate appears to be the limiting factor in determining (+)-AMP tissue concentrations. In addition, during the terminal elimination phases, tissue/serum partitioning of (+)-METH and (+)-AMP were similar.

In conclusion, (+)-METH distributes very rapidly to all of the tissues in this study except the spleen. Furthermore, the relationship of (+)-METH and (+)AMP brain and serum concentrations is complex. It also appears that time-dependent brain-to-serum partitioning of (+)-METH could be related to the time course of behavioral effects. Finally, based on molar amounts of (+)-AMP and (+)-METH in tissues, it appears that (+)-AMP could make a significant contribution to the pharmacological effects after i.v. administration of (+)-METH. Consequently, these data suggest that medical treatments for (+)-METH abuse must consider the effects of both (+)-METH and (+)-AMP.