Introduction

Top Abstract Introduction Methods Results Discussion References

d-Fenfluramine, an anorectic drug commonly prescribed to treat obesity in Europe, can increase extracellular 5-HT in the hypothalamus (Schwartz et al., 1989), cortex (Series et al., 1994) and striatum (Kreiss et al., 1993) of rodents. With repeated doses it can also evoke prolonged depletion of 5-HT in the brain regions rich in serotonergic nerve endings (Harvey and McMaster, 1975; Clineschmidt et al., 1976, 1978). It has been suggested that 5-HT release without compensatory replenishment of the storage vesicles is largely responsible for the observed decrease in 5-HT tissue concentrations (see Rowland and Carlton, 1986, for review). d-FEN can act as a 5-HT re-uptake inhibitor (Wong et al., 1975; McTavish and Heel, 1992) and releaser (Fuller et al., 1988), and its major metabolite d-NF has been reported to act as a 5-HT releaser (Garattini, 1985).

Recently, information about drug concentrations at their target site (extracellular space) has been determined by brain microdialysis (Ståhle, 1991; Allen et al., 1994; Alonso et al., 1995; Clausing et al., 1995). Such data are useful for several purposes. First, they can help to compare results of in vitro studies (Kramer et al., 1994) with the in vivo situation. Second, such data can facilitate the understanding of the temporal features of drug-induced neurotransmitter and behavioral changes in a dose-response-context (Segal and Kuczenski, 1994), in particular when extracellular drug levels, neurochemical and behavioral changes are determined simultaneously (Bradberry et al., 1993; Clausing et al., 1996). Third, the simultaneous determination of drug levels, neurochemical and behavioral changes enables one to elucidate to what extent individual differences in drug response are caused by pharmacodynamics (Hooks et al., 1992) versus pharmacokinetic mechanisms (Clausing et al., 1996). Finally, the evaluation of relative drug potencies in an in vivo situation ultimately requires knowledge of the drug concentration at the target site (Bradberry et al., 1993).

To further the understanding of the central effects of d-FEN, this study provides data on the levels of FEN and NF in microdialysate collected from the frontal cortex after either subcutaneous or intraperitoneal administration of different doses of d-FEN. For the purpose of comparison, plasma and brain tissue levels were also determined as well as brain 5-HT concentrations 4 days after dosing. The frontal cortex was chosen for these microdialysis studies because it is one of the areas more sensitive to both methylenedioxymethamphetamine and d-FEN (Slikker et al., 1988; Stewart et al., 1997).

	Methods

Top Abstract Introduction Methods Results Discussion References

Animal housing and experimental design. All procedures involving animal care were approved by the National Center for Toxicological Research Institutional Animal Care and Use Committee. Male Sprague-Dawley rats (Crl:COBS CD [SD] BR), 4 to 5 months old, from the National Center for Toxicological Research breeding colony were housed individually, starting on the day of implantation of the microdialysis guide cannula. They were kept in acrylic cages (45 × 22 × 20 cm) on wood shaving bedding. d-FEN was purchased from Research Biochemicals International (Natick, MA) and the d-NF, which was used as a reference standard for HPLC analysis, was a gift from S.A. Lorens (Loyola University, Maywood, IL). All but one experiment was carried out at a standardized room temperature of 23 ± 1°C. To collect brain microdialysates the animals were transferred into the microdialysis bowl 4 hr before administration of d-FEN. A single subcutaneous injection of 2 mg/kg d-FEN was administered to determine what the FEN levels in the microdialysis were over time from a dose which produces minimal to no 5-HT depletion. Three subcutaneous doses of 5 mg/kg d-FEN spaced 2 hr apart were used to enable a direct comparison of FEN-dialysate levels with a previously published similar study with d-amphetamine using the same dosing paradigm (Clausing et al., 1995). Plasma and whole-brain FEN and NF levels were determined in animals (not used for microdialysis) after either 1 × 5 mg/kg s.c. d-FEN or 3 × 5 mg/kg s.c. d-FEN, so that levels in these tissue could be compared with the microdialysis data gathered from 3 × 5 mg/kg s.c. d-FEN. Finally, the time course of FEN and NF dialysate levels was monitored for 8 hr after a single i.p. injection of 10 mg/kg d-FEN, a dose typically used in neurotransmitter microdialysis studies (Schwartz et al., 1989; Hirano et al., 1995; Viana et al., 1996). This latter experiment was conducted at two different ambient temperatures (23 ± 0.5°C vs. 27 ± 0.5°C). During this experiment rectal temperatures were recorded every 60 min, as described previously by Bowyer et al. (1994), and when body temperature exceeded 41.0°C ice was placed in the microdialysis bowel to cool the animals to below 40.0°C. The 5-HT tissue concentrations of frontal cortex were determined 4 days after dosing. This cooling affected the time versus temperature profile; to better visualize the time course of hyperthermia produced in a 27°C environment an additional group profile of nonimplanted rats receiving 10 mg/kg i.p. d-FEN, whose temperature did not exceed 41.0°C, is shown in figure 2B. Also, nonimplanted rats were dosed with 10 mg/kg i.p. d-FEN at either 23°C or 27°C, and were sacrificed either 1 or 8 hr after dosing to determine plasma and brain concentrations of FEN and NF at these time points. An overview of the total experimental design is presented in table 1.

Brain microdialysis. Brain microdialysis was carried out in a manner as described by Ungerstedt (1984) and modified by this laboratory. CMA microdialysis equipment (Carnegie Medicine, Stockholm, Sweden) was used and CMA/12 guide cannulae were implanted with the coordinates AP 3.5 mm, LAT 1.3 mm, DV 3.6 mm relative to bregma (Paxinos and Watson, 1986). The artificial cerebrospinal fluid had the following composition: 145 mM NaCl, 1.5 mM KCl, 1.5 mM MgCl₂·6H₂O, 1.25 mM CaCl₂·2H₂O, 10 mM glucose, 1.5 mM K₂HPO₄, adjusted to pH 7.0 with HCl. After surgery the animals were given a 5- to 7-day recovery period. On the morning of the experiment the animal was hand-held and the CMA/12 dialysis probe (2-mm probe tip) was carefully inserted through the guide cannula into the frontal cortex. Four hours after probe insertion the first d-FEN dose was administered, and (after allowing for the dead volume in the tubing) microdialysis samples were collected every 20 min at a flow rate of 1 µl/min. After the microdialysis experiment, rats were sacrificed for histological verification of the probe position. In vitro probe recovery was measured to assess the functionality of the individual probes and to exclude nonfunctional probes from the experiment. Percent in vitro probe recovery for FEN and NF was determined as [(concentration in collected sample × 100)/concentration in standard solution] at 23°C and 1 µl/min flow rate. The estimated average in vitro probe recoveries for FEN and NF of 21 probes used were 23 ± 1.6% for FEN and 25 ± 1.7% for NF.

Collection of plasma and brain samples. For the collection of plasma and brain samples, rats were decapitated and trunk blood was collected into glass tubes containing heparin. Blood samples were centrifuged under refrigeration for 10 min at 18,000 × g. Brains were dissected quickly on a chilled glass plate. Frontal cortex, hypothalamus, caudate/putamen and substantia nigra were collected into separate tubes. Some extra samples of untreated animals were collected for running blanks and standards during the HPLC analysis. All samples were stored at 150°C until analysis.

HPLC quantitation of FEN and NF. A detailed description of the HPLC method can be found elsewhere (Clausing et al., 1997). Because this method does not distinguish between l- and d-enantiomers we refer to FEN and NF levels in this paper rather than d-FEN and d-NF levels, although it is reasonable to assume these enantiomers do not rapidly racemize after administration of d-FEN. FEN and NF levels were determined by fluorescent detection after derivatization with dansyl chloride, removal of excess dansyl chloride with a strong anion-exchange resin and separation on a Supelcosil LC-18 column, running a step gradient with 50% KH₂PO₄ (0.05 M, pH 5.5) + 50% acetonitrile (mobile phase A) versus 25% KH₂PO₄ (0.05 M, pH 5.5) + 75% acetonitrile (mobile phase B). Microdialysate samples were derivatized directly, whereas plasma and brain samples were extracted into ethyl acetate under basic conditions and reconstituted in KH₂PO₄ buffer (adjusted to pH 2.8) before derivatization. The quantitation limits of this method were 500 fmol in microdialysate, 5 pmol in extracted plasma and 25 pmol in extracted brain tissue. Fluoxetine from Ely Lilly (Indianapolis, IN) was used as internal standard for the analysis of brain tissue and plasma samples.

HPLC quantitation of 5-HT. 5-HT levels in brain tissue were analyzed by electrochemical detection. Each sample was weighed and diluted with a measured volume (20 × w/v) of 0.2 M perchloric acid containing 100 ng/ml 3,4-dihydroxybenzylamine (Sigma, St. Louis, MO) as internal standard. After sonication and centrifugation the supernatant was removed and injected directly onto the HPLC/electrochemical detection system. Samples were analyzed under the following conditions: An isocratic M-6000A pump (Waters Associates, Milford, MA) at a 1.0 ml/min flow rate, a Rheodyne 7125 injector (Rheodyne Inc., Cotati, CA.), a Supelcosil LC18 3 µm analytical column (7.5 cm × 4.6 mm, Supelco, Bellefonte, PA), a BAS-LC4B amperometric detector, a BAS-LC-17 oxidative flow cell and an LCI-100 integrator (Perkin Elmer Corp., Cupertino, CA) were used. The mobile phase consisted of 92% KH₂PO₄ buffer (0.07 M, pH 3.0) and 8% methanol, containing 1 mM Na-1-heptanesulfonic acid and 0.2 mM Na₂-ethylenediaminetetraacetic acid per liter unless otherwise indicated.

Statistics. Data are presented as arithmetic mean ± standard error of the mean (S.E.M.) unless otherwise indicated. Student's t test was used for two-sample comparisons. Multiple groups were analyzed by one-way ANOVA with post hoc Fisher's least significant difference test, if significant main effects were observed. Linear correlation and regression analysis was used to examine the relationship between hyperthermia and 5-HT tissue or drug microdialysis concentrations. Areas under the time-concentration curve (AUC) were calculated applying the trapezoidal rule with the computer program of Tallarida and Murray (1986).

	Results

Top Abstract Introduction Methods Results Discussion References

Levels of FEN and NF in microdialysate. Microdialysate levels of FEN and NF after systemic administration of d-FEN are depicted in figs. 1 and 2. After 2 mg/kg d-FEN s.c. the peak concentration was 0.24 ± 0.07 µM between 40 and 60 min and declined to below detectable levels 4 hr after dosing (fig. 1). However, the difference between the 0- to 20-min fraction and the third fraction did not quite reach statistical significance (P > .05). Administration of 3 × 5 mg/kg d-FEN, spaced 2 hr apart, caused a continuous build-up of FEN throughout the dosing. Although peak levels of 0.33 ± 0.04 µM were reached within 40 to 60 min after the first dose, after multiple doses of 5 mg/kg FEN the time-to-peak level was greater than 80 min with peaks of 0.68 ± 0.04 µM after the second dose and 1.20 ± .07 µM after the third dose. One-way ANOVA revealed significant differences among the FEN concentrations of the different microdialysate fractions (F_20,126 = 60.74, P < .001). The subsequent Neuman-Keuls test indicates that, compared with the first fraction (0-20 min), levels of the third (40-60 min) fraction and all following fractions were significantly higher (P < .05), whereas there was no difference between the third and sixth fraction. FEN levels were still at plateau levels 3 hr after the last dose. Peak levels as well as the time course (fig. 2) were almost identical when d-FEN (1 × 10 mg/kg i.p.) was administered at two different ambient temperatures. Peak FEN levels of 1.63 ± 0.28 µM and 1.54 ± 0.20 µM were reached between 40 and 60 min at 23°C and 27°C room temperature, respectively. One-way ANOVA indicated significant differences in FEN concentration between individual microdialysis fractions at 23°C (F_23,120 = 6.18, P < .0001) and 27°C (F_23,120 = 8.85, P < .0001). A subsequent Newman-Keuls test revealed that the FEN concentration in the 0- to 20-min microdialysate fraction was significantly lower (P < .01) than in the subsequent six fractions, whereas the subsequent six fractions did not significantly (P > .05) differ from each other. Table 2 summarizes the analysis of these data calculated as the area under the concentration versus time curve (AUC). No significant differences were observed. These data indicate that after a single systemic administration of d-FEN microdialysis FEN levels reach their maximum 40 and 60 min after dosing irrespective of the dose given.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   FEN (open symbols) and NF (filled symbols) concentrations in microdialysate collected from the frontal cortex (not corrected for probe recovery) after either one dose of 2 mg/kg s.c. d-FEN or three doses of 5 mg/kg s.c. d-FEN. Results are the mean ± S.E.M. for five (2 mg/kg group) or six rats (3 × 5 mg/kg group).

View larger version (48K): [in this window] [in a new window]   Fig. 2.   FEN and NF concentrations in microdialysate collected from the frontal cortex (A) or body temperature (B) after a single dose of 10 mg/kg i.p. d-FEN. Results are the mean ± S.E.M. for five rats per group for the microdialysis rats. (a) The mean temperature at this time point dropped abruptly because four rats were placed on ice because their body temperature exceeded 41.0°C. A separate group of nonmicrodialysis rats are shown (filled triangles) to show the true profile of the temperature increases in a 27°C environment after 10 mg/kg i.p. d-FEN.

Levels of FEN and NF in plasma and brain tissue. Results for plasma and tissue concentrations of FEN and NF are summarized in table 3. After 1 × 10 mg/kg d-FEN i.p., FEN concentrations were significantly lower in all substrates 8 hr after dosing as compared with 1 hr after dosing. In contrast, NF concentrations 8 hr after dosing were significantly higher than 1 hr after dosing. As compared with a single dose, three s.c. doses of 5 mg/kg d-FEN spaced 2 hr apart resulted in significantly higher levels of both FEN and NF in all the substrates investigated. Regional differences were observed for both FEN and NF. After 1 × 10 mg/kg i.p. d-FEN, concentrations were significantly lower in hypothalamus as compared with striatum and/or frontal cortex 1 hr after dosing. At 8 hr after dosing, concentrations of FEN and NF in the frontal cortex and striatum tended to be higher than in hypothalamus and substantia nigra. This must be considered with caution, however, because the routes of administration were different, and the pharmacokinetics of fenfluramine has been shown to be complex after multiple administrations at varying doses (Caccia et al., 1992).

Tissue concentrations of 5-HT and 5-HIAA in the frontal cortex. Compared with controls, rats receiving 1 × 10 mg/kg d-FEN i.p. had significantly lower concentrations of 5-HT and 5-HIAA 4 days after dosing (fig. 3). After administration of d-FEN at a 23°C room temperature, 5-HT and 5-HIAA concentrations were 36% and 61% of the corresponding controls, respectively. When rats were kept at 27°C ambient temperature for 8 hr after dosing, levels dropped to 13% and 17% of the corresponding controls for 5-HT and 5-HIAA, respectively, which were significantly lower than those levels in rats dosed with d-FEN at 23°C (fig. 3). There was a significant negative correlation (r = 0.900, P = .003) for rats dosed with d-FEN between either peak body temperature or increase in temperature above the animals' base-line temperatures just before dosing during the session and the 5-HT concentration in the frontal cortex 4 days after dosing (fig. 4).

View larger version (44K): [in this window] [in a new window]   Fig. 3.   Tissue concentrations of 5-HT (A) and 5-HIAA (B) in the frontal cortex of rats 4 days after the administration of a single i.p. dose of 10 mg/kg d-FEN hydrochloride either at 23°C or at 27°C room temperature. Results are the mean ± S.E.M. of four or five rats per group. a = significantly different (P < .05) from controls at the same room temperature; b = significantly different (P < .05) from FEN-dosed rats at 23°C.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   The relationship between the increase in body temperature from base line (A) or peak body temperature (B) and 5-HT concentrations in the frontal cortex of rats dosed with 1 × 10 mg/kg d-FEN i.p. at an ambient temperature of either 23°C or 27°C is shown. Linear regressions are depicted in both panels A and B along with the 95% confidence intervals. Serotonin levels were determined in the frontal cortex of rats sacrificed 4 days after FEN exposure.

Drug levels and hyperthermia. Although based on the limited number of five animals dosed with d-FEN at elevated ambient temperature, there was a clear tendency for a positive correlation between peak drug concentrations in microdialysate and peak increase of body temperature the base line (fig. 5). This was clearest for the combined micromolar concentration of NF and FEN (r = 0.858, P = .063). The correlation between FEN levels and peak body temperature increase was r = 0.801 (P = .104), whereas for NF this relationship was virtually nonexistent (r = 0.286, P = .641). On average the combined peak drug levels preceded the peak body temperature increase by 80 ± 20 min. There was also a good correlation between the 5-HT levels in the frontal cortex (4 days after FEN exposure) compared with either the increase in body temperature above the base-line temperature (fig. 5A) or peak body temperature (fig. 5B) for rats dosed in the 23°C and 27°C environments.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Relationship between peak combined (FEN + NF) drug concentrations in microdialysate and peak hyperthermia of rats dosed with 1 × 10 mg/kg d-FEN i.p. at 27°C ambient temperature. Linear regression of maximum concentration (µM) of FEN + NF in the microdialysate, which preceded the peak body temperature over base line by 80 ± 20 min.

	Discussion

Top Abstract Introduction Methods Results Discussion References

This study demonstrates that after a single s.c. or i.p. administration, like amphetamine levels in the striatum (Clausing et al., 1995), peak microdialysate levels of FEN in the frontal cortex are reached at approximately 40 to 60 min after dosing ("Results," figs.1 and 2A). However, after multiple doses of 5 mg/kg s.c. FEN the time-to-peak microdialysis level is extended to 80 min or more. We have also observed that depletion of 5-HT and 5-HIAA (figs. 3 and 4) subsequent to d-FEN administration is increased at an elevated ambient temperature (27°C), but this potentiation is not caused by FEN and NF pharmacokinetics (fig. 2). This indicates that elevated temperature does not appreciably alter the pharmacokinetics of FEN. However, there is a strong tendency for a positive correlation between total drug concentration (FEN + NF) in microdialysate and drug-induced hyperthermia in the 27°C environment (fig. 5). Considering the results of Stewart et al. (1997), which show that at elevated environmental temperatures a dose of 2.5 mg/kg d-FEN is the threshold for depleting brain 5-HT levels whereas 5 mg/kg produces long-term depletions (greater than 1 week), figures1 and 2, it appears that, after systemic administration of d-FEN, FEN levels in microdialysate need to exceed 0.25 µM for appreciable depletions of 5-HT to occur. An 0.25 µM FEN level in microdialysate would correspond to an approximate extracellular level of between 1.25 µM (calculated from the probe in vitro recovery efficiency shown under "Methods") and 2.5 µM (calculated by the actual in vivo recovery efficiency reported previously for amphetamine; Clausing et al., 1995). According to table 3 plasma concentrations expected to produce neurotoxicity would be slightly less than 1 µM, whereas actual brain concentrations would have to approach 50 µM.

When the existing differences in experimental design are taken into account, our data for plasma and brain tissue concentrations of FEN and NF are in good agreement with the relatively few studies that investigated the pharmacokinetics of the d-isomer of FEN (Jori et al., 1978; Caccia et al., 1992; Anelli et al., 1995). In contrast to our earlier study with d-amphetamine (Clausing et al., 1995), a much larger build-up of drug concentration in microdialysate as well as tissue occurred after multiple doses of 5 mg/kg d-FEN, spaced 2 hr apart. This confirms earlier observations by Blundell et al. (1975) who described a more rapid clearance of amphetamine as compared with FEN and a longer excretion half-life of d-FEN as compared with d-amphetamine. Although the rate of clearance of FEN is much lower than of amphetamine, the ratio of microdialysate levels to plasma levels is similar for the two compounds with FEN ratios being between 0.3 and 0.75 at doses ranging from 2 to 10 mg/kg d-FEN compared with 0.5 to 0.7 for doses of amphetamine between 1.0 and 15 mg/kg (Clausing et al., 1995).

Significant regional variation was observed for FEN and NF concentrations. Although regional differences are known for the 5-HT depleting effect of d-FEN, to our knowledge regional variation for drug levels has not yet been described. The cortex, a region particularly sensitive to d-FEN-induced reductions of 5-HT (Steranka and Sanders-Bush, 1979; Kleven et al., 1988), turned out to also be a region of high FEN and NF levels (present study). Accordingly, the hypothalamus was described by these same authors as less sensitive to 5-HT depletion and displayed the lowest FEN and NF concentrations in our investigation. Further studies will be necessary to see if regionally there is a positive correlation in 5-HT depletions with FEN levels and 5-HT innervation.

The interpretation and causes of the well-established 5-HT-depleting properties of d-FEN are controversial. Long-term decreases in tissue concentrations of 5-HT (as well as vesicular 5-HT and 5-HT re-uptake sites) are used as an important indication of toxicity to the serotonergic system (Kleven and Seiden, 1989; Appel et al., 1990). Others argue that such decreases could also be explained as functional down-regulation, i.e., reduced density of the 5-HT transporter per nerve ending (Gobbi et al., 1992, 1996) instead of neurodegeneration. Moreover, several indices, used by other investigators, of serotonergic neurotoxicity used by some investigators (e.g., astrogliosis, neuronal sprouting and persisting behavioral effects) failed to occur after FEN administration (Lorens et al., 1990; Baumgarten et al., 1992). In contrast, Westphalen and Dodd (1995), with a novel technique which supposedly can reflect the quantity of nerve endings able to form synaptosomes, reported that the loss of 5-HT reuptake transporter in FEN-treated rats was similar to the "neurotoxic model" (i.e., loss of nerve terminals) rather than to the "transporter ablation model" with otherwise intact serotonergic nerve terminals.

Our observation that increased indole depletion is not caused by altered pharmacokinetics after hyperthermia is an important finding because pharmacokinetics can play an important role in FEN depletion of 5-HT (Anelli et al., 1995). We have reported previously that amphetamine-induced hyperthermia potentiates dopaminergic terminal degeneration, not by altering the pharmacokinetics (Bowyer et al., 1994; Clausing et al., 1995) but probably by increasing endogenous reactive intermediates responsible for terminal degeneration (Bowyer and Holson, 1995). Stewart et al. (1997) have shown that, even at an elevated environmental temperature of 27°C when hyperthermia occurs and 5-HT depletions are more pronounced, doses of between 2.5 and 5 mg/kg are necessary for 5-HT depletions. With the 0.25 µM microdialysate levels seen after a 2 mg/kg dose of d-FEN (fig.1), we can extrapolate the approximate threshold extracellular brain levels of FEN necessary for 5-HT depletions. With the in vitro recovery of FEN of 23% an extracellular level of about 1.25 µM FEN would be present after 2 mg/kg d-FEN; whereas if the true in vivo recovery is closer to 10% which we have seen previously for d-amphetamine (Clausing et al., 1995), the extracellular level necessary for 5-HT depletion would be greater than 2.5 µM. These would correspond to plasma levels slightly less than 1 µM and brain levels approaching 50 µM.

In conclusion, hyperthermia increases the indole depletion produced by d-FEN but not via pharmacokinetics although pharmacokinetics can influence the occurrence of hyperthermia. The ratio of FEN levels in microdialysate to plasma is similar to that observed with d-amphetamine, although, as expected, the clearance of FEN from microdialysate is much slower than that of amphetamine. Furthermore, the data on brain pharmacokinetics obtained in these studies indicate that the extracellular levels of FEN necessary for 5-HT depletions is between 1.25 and 2.5 µM, whereas plasma and brain levels would have to approach 1 µM and 50 µM, respectively. These data may aid in determining the doses of d-FEN that are necessary to produce neurotoxicity/neuroregulatory changes in the brain.