Author for correspondence: Rui Tao, Charles E. Schmidt College of Biomedical Science, Florida Atlantic University, 777 Glades Road, Boca Raton, FL 33431, USA (fax +1 561 297 2221, e-mail firstname.lastname@example.org).
Abstract: Previous studies have demonstrated that serotonin (5-HT) syndromes, particularly for the malignant cases, can be alleviated by ice water mists, cooling blankets and many other external cooling measures. In this study, we tested the hypothesis that external cooling measures reduce the responsivity of 5-HT2A receptors to excessive 5-HT efflux, which may be a possible mechanism underlying the treatment of serotonin syndrome. To test this, rat experiments were carried out in the standard and cool ambient temperature (Tamb) by administration of the 5-HT precursor 5-hydroxy-l-tryptophan combined with the monoamine oxidase inhibitor clorgyline. The first set of experiments was to assess severity of the syndromes by measuring body temperature responses. Consistent with the hypothesis, we found that the syndrome was malignant at the standard Tamb of 22°C but alleviated at 12 or 6°C, these results being similar to those in rats pre-treated with the 5-HT2A receptor antagonist ketanserin. The second set of experiments was to utilize microdialysis to determine the relationship between the syndrome severity and 5-HT levels at the above-mentioned Tamb. We found that excessive 5-HT efflux consisted of primary and secondary components through two distinct mechanisms. Furthermore, the secondary component efflux, which can be ascribed to 5-HT2A receptor activation, was proportionally reduced at the cool Tamb of 12 and 6°C. In conclusion, results of this study support the hypothesis that cooling Tamb reduces the functional activity of 5-HT2A receptors, thus alleviating the malignant syndrome.
By measuring changes in neuromuscular responses and autonomic function, the serotonin (5-hydroxytryptamine; 5-HT) syndromes can be categorized as mild, moderate or severe [1,2]. Alternatively, the syndromes are simply described as benign or else malignant when body-core temperature increases dramatically and is associated with a high risk of death . Under standard laboratory conditions, using rats as a model organism, hyperthermia is an easily measured and reliable marker of the malignant syndrome. As hyperthermia is associated with multiple organ failure and death, monitoring body-core temperature and preventing hyperthermia are of utmost concern in treating the 5-HT syndrome in human beings. In contrast, a reduction in body temperature is a reliable marker for the benign 5-HT syndrome . In general, hyperthermia is recognized as an indicator of a malignant response whereas hypothermia is associated with a benign response to drugs that evoke increases in synaptic levels of 5-HT.
In therapeutic management of the 5-HT syndrome, hyperthermia can be controlled by the 5-HT2A receptor antagonists cyproheptadine and ketanserin [2,4,5]. Besides pharmacologically directed therapies, an alternative approach is to reduce body temperature with ice, mists or cooling blankets [4,6]. Although the effectiveness of external cooling measures for treatment of the 5-HT syndrome has been validated in studies with experimental animals , the neural substrates that carry out interaction between cooling and alleviation are poorly understood.
The purpose of this study was to test the hypothesis that cooling treatments reduce functional activity of 5-HT2A receptors which are responsible for evoking hyperthermia of serotonin syndrome [1,8]. In this study, the syndrome was induced by injection of a 5-HT precursor, 5-hydroxy-l-tryptophan (5-HTP) in rats pre-treated with the monoamine oxidase inhibitor (MAOI) clorgyline. To pinpoint the effect of external cooling on the development of the syndrome, experiments were conducted in the temperature-controlled chamber and thus the whole body of experimental animals was exposed to a given ambient temperature (Tamb). We designed two sets of experiments to test the above-mentioned hypothesis. In the first set, changes in body-core temperature (Tcor) were measured during the syndrome at the standard laboratory Tamb of 22°C and compared with changes in body temperature at cooling Tamb of 12 or 6°C. The involvement of 5-HT2A receptors in the hyperthermia of the malignant syndrome was examined using the 5-HT2A receptor antagonist ketanserin. In the second set of experiments, changes in 5-HT efflux in the preoptic/anterior hypothalamus and prefrontal cortex were determined during the malignant syndrome at Tamb of 22°C as well as for the benign syndrome at Tamb of 12 or 6°C. Ketanserin was also used to test the possible role of 5-HT2A receptors in evoking excessive 5-HT efflux associated with changes in syndrome severity from malignant to benign.
Materials and Methods
Animals. Adult male Sprague–Dawley rats purchased from Charles River Laboratories (Raleigh, NC, USA) were pair-housed with food and water available ad libitum in a temperature- and humidity-controlled room. After 1-week habituation, animals weighing 300–350 g were assigned randomly to either experimental or control groups. Treatment of animals and experimental procedures were in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals and Animal Research Guidelines and approved by the IACUC of Florida Atlantic University. Efforts have been made to reduce the number of animals used and the distress of experimental manipulations.
Drugs. Clorgyline (N-Methyl-N-propargyl-3-(2,4-dichlorophenoxy) propylamine hydrochloride) and 5-hydroxy-l-tryptophan (5-HTP), purchased from Sigma (St Louis, MO, USA), were dissolved in saline (NaCl 0.9%) and injected subcutaneously (s.c.). Ketanserin tartrate obtained from Tocris Bioscience (Ellisville, MO, USA) was dissolved in deionized water for intraperitoneal injection (i.p.).
Induction of the syndrome. Because inhibition of monoamine oxidase is irreversible and long-lasting, clorgyline was given 4 hr before experiments. The dose of clorgyline was 2 mg/kg, as described in previous studies [1,9]. The clorgyline-pre-treated rats were placed individually in a temperature-controlled chamber for habituation at least 2 hr before administration of 5, 15 or 25 mg/kg 5-HTP. As measured by motor and autonomic responses, it had been demonstrated that 15 and 25 mg/kg 5-HTP caused moderate and severe syndromes, respectively, at Tamb of 22°C . Additionally, the syndrome in response to 25 mg/kg (but not 15 mg/kg) 5-HTP was malignant as indicated by hyperthermia and death. In contrast, 5 mg/kg 5-HTP was below the threshold for inducing the syndrome. Thus, the dose range of 5-HTP used in this study allowed us to investigate all degrees of the syndrome. Tamb of 22°C (±1.0) was considered as the standard laboratory condition. To investigate Tamb-dependent responses, cool environmental temperatures were set at 12 and 6°C, or about 10 and 16°C below the standard Tamb, respectively. Temperature-controlled chambers were designed in our laboratory similar to those described in the study  with temperature variation in the range of ±1.0°C.
Measurement of body-core temperature (Tcor). A flexible thermoprobe connected to a digital meter (Traceable®; Fisher Scientific, West Palm Beach, FL, USA) was inserted into the colon to determine changes in body-core temperature. The length of the probe was set at 4.6 cm. After habituation for at least 2 hr, body temperature was taken every 15 min. for a total of 120 min. In this study, hyperthermia was defined as an increase in body temperature of at least 2°C above baseline and hypothermia as a reduction in body temperature of at least 1°C below baseline.
5-HT microdialysis. Guide cannulae were pre-implanted at least 1 week before microdialysis. Briefly, rats were anaesthetised with a combination of xylazine (4 mg/kg, i.p.) and ketamine (80 mg/kg, i.p.) and then placed on a stereotaxic frame (Stoelting Co., Wood Dale, IL, USA). The skull was surgically exposed for implantation of a sterile 22-gauge guide cannula at stereotaxic coordinates for the preoptic/anterior hypothalamus: 1.1 mm posterior to bregma, ±0.9 mm lateral to the midline and 3.0 mm ventral to the skull or for the prefrontal cortex: 3.3 mm anterior to bregma, ±0.7 mm lateral to midline and 2.0 mm ventral to the skull. The guide cannulae were then secured into place with acrylic dental cement and skull screws. One week after the surgery, a hollow probe was inserted through the cannula into the preoptic/anterior hypothalamus or prefrontal cortex for 5-HT microdialysis. The coordinates for the tip of the probe were 1.1 mm posterior to bregma, ±0.9 to midline and 9.0 mm below the skull in the preoptic/anterior hypothalamus or 3.3 mm anterior to bregma, ±0.7 mm lateral to midline and –4.5 mm blow the skull in the prefrontal cortex. After probe insertion, rats were placed in a temperature-controlled chamber and attached to a fluid swivel that allowed animals to move freely (Raturn® system; Bioanalytical System Inc., W. Lafayette, IN, USA). Food and water were available ad libitum. The dialysis probes were perfused overnight with phosphate-buffered Ringer’s solution (140 mM NaCl, 3.0 mM KCl, 1.5 mM CaCl2, 1.0 mM MgCl2, 0.25 mM NaH2PO4, 1.0 mM Na2HPO4, pH 7.4) at a rate of 1.0 μl/min. The chamber was exposed to a room temperature of 22°C for the overnight infusion.
The next day, clorgyline (2 mg/kg, s.c.) was injected 4 hr before experiments. The chamber temperature was then set to 22, 12 or 6°C, depending on the experiment, and animals were housed for habituation to the Tamb for at least another 2 hr. Samples were collected every 15 min. and analysed by high performance liquid chromatography (HPLC) – electrochemical detection (HTEC-500; Eicom, Kyoto, Japan). Separation of 5-HT was achieved on a 150 mm × 1 mm i.d. column packed with a TSK gel ODS-80 TM, 5 μm particle size. The potential on the graphite working electrode was set at +400 mV (relative to the Ag/AgCl reference electrode). The mobile phase was made of 0.1 M phosphate buffer (pH 6.0), 1% methanol, 500 mg/l sodium octanesulfonate, and 50 mg/l EDTA and pumped at a rate of 500 μl/min. The detection limit for 5-HT was 0.05 pg. Drug injection other than clorgyline pre-treatment was made after collection of four baseline samples.
Upon completion of experiments, animals, deeply anaesthetised with pentobarbital (100 mg/kg, i.p.), were infused with 2% Fast Green through the probe for 10 min. and then decapitated. Brains were extracted, frozen and sliced. The dye-tracer location was compared with a rat brain atlas (Paxinos and Watson, 1998). Probes located outside the target boundary were excluded from the data analysis.
Analysis of correlation between secondary 5-HT efflux and 5-HT2A receptor activation. Classically, 5-HT efflux evoked by 5-HTP and clorgyline can be ascribed to increases in 5-HT synthesis and decreases in metabolism, respectively . In theory, enzyme activity for 5-HT synthesis and metabolism should not be altered either by changes in Tamb as rats are homeothermic or by ketanserin pre-treatment because the antagonist has no direct effect on the enzymes. Thus, after blocking 5-HT2A receptors with ketanserin at cool Tamb, 5-HT efflux caused by 5-HTP plus clorgyline may be mainly derived from 5-HT biosynthesis and metabolism, which is defined as ‘primary efflux’ . In this study, the primary efflux was the mean level between 30 and 75 min., in which 5-HT level was at peak. Amounts of ‘total efflux’ were obtained at Tamb of 6, 12 or 22°C in the absence of ketanserin pre-treatment. Amounts of secondary efflux were computed using the equation:
Based in part on our previously obtained data , we suggest that changes in secondary efflux depend on not only amounts of primary efflux but also functional activity of 5-HT2A receptors at a given Tamb. The relationship of secondary and primary efflux to 5-HT2A receptor activity can be expressed in an equation as follows:
We suggest that changes in ratio depend on functional activity of 5-HT2A receptors in response to primary efflux. Specifically, the ratio would be nil or low if 5-HT2A receptors were not sufficiently activated by primary efflux at cool Tamb.
Statistical analysis. Unless otherwise noted, data were expressed as mean ± S.E.M. Changes in body temperature were expressed on the y axis (as °C) plotted against the sampling time on the x axis (in min.). 5-HT efflux expressed as absolute values (pg/sample) was uncorrected from probe recovery. The baseline was calculated from the mean of four sequential samples of 5-HT efflux before 5-HTP injection. For comparison, changes in 5-HT efflux were normalized to relative values expressed as fold increases over respective baselines. Statistical analysis was performed by a two-way (drug treatment × time course) factorial anova. If a significance of drug treatment was found, then further statistical analysis was carried out using the post hoc Scheffe’s test in determining the significance of the respective time points. If appropriate, the unpaired t-test was employed for the data analysis. The p-value was set at 0.05 for statistical significance.
Changes in body temperature (Tcor) in response to 5-HTP at the standard Tamb of 22°C.
The mean basal body temperature was 36.7 ± 0.1°C (n = 19) in clorgyline-pre-treated animals. Fig. 1A shows that 5-HTP produced a time- and dose-dependent increase in body temperature. This conclusion was confirmed by the two-way factorial anova, which revealed a significant main effect of 5-HTP doses (F(3, 15) = 5.222, p = 0.0114) and sampling times (F(7,21) = 12.97, p < 0.0001). The 5-HTP × sampling time interaction indicated that 5-HTP increased body temperature across time more than vehicle injection (F(21,105) = 4.274, p < 0.0001). Specifically, vehicle alone (3 ml/kg, s.c.; 0.9% NaCl) had no effect. The post hoc Scheffe’s test showed that administration of 5 mg/kg 5-HTP caused an insignificant change compared with saline treatment (p > 0.05). There was a tendency for increased body temperature after 15 mg/kg 5-HTP but this was not statistically significant (p > 0.05). In contrast, 25 mg/kg 5-HTP evoked a significant increase (p < 0.0001). The time to maximum increase was ∼60 min. after 5-HTP injection, and the elevation in body temperature was sustained for at least 60 min. In addition, other symptoms including tremor, forepaw treading and hind limb abduction were apparent. All animals died after 2–5 hr at this dose. Thus, the syndrome produced by 25 mg/kg 5-HTP is malignant. This dose was therefore used for investigating the interaction between Tamb and the malignant syndrome in most of the following experiments.
To determine if 5-HT2A receptors mediated hyperthermia in the malignant syndrome, ketanserin (5 mg/kg, i.p.) was administered 15 min. before 25 mg/kg 5-HTP. Control animals were treated with vehicle (H2O, 1 ml/kg, i.p.) followed by 5-HTP or vehicle injection (0.9% NaCl, s.c.) on the same schedule. As shown in fig. 1B, ketanserin pre-treatment antagonized the hyperthermia induced by 5-HTP (p < 0.001, t-test). Instead, body temperature became hypothermic. All animals pre-treated with ketanserin survived 5-HTP injection.
Response at the cool Tamb of 12 and 6°C.
After a 2-hr habituation period, the mean basal body temperature was 36.7 ± 0.1°C at Tamb of 22°C and 36.9 ± 0.1°C (n = 23) at 12°C. Statistical analysis revealed that there was no significant difference in baselines between these two conditions (p > 0.05, unpaired t-test). As shown in fig. 2A, vehicle administration had no effect on body temperature at either environmental Tamb.
Fig. 2B shows the body temperature response to 5-HTP injection in clorgyline-pre-treated rats at Tamb of 12°C (±1.0). Compared with vehicle, 5-HTP produced a dose- and time-dependent decrease in body temperature (hypothermia). Data analysis by the two-way factorial anova revealed a significant difference in doses (F(3, 19) = 4.018, p = 0.0227) and sampling times (F(7, 21) = 6.678, p < 0.0001) as well as a significant interaction of 5-HTP treatment × sampling time (F(21, 133) = 1.997, p = 0.0099). The post hoc Scheffe’s analysis showed a small transient but significant reduction in response to 5 mg/kg 5-HTP (p < 0.05). A large and persistent reduction was found in rats administered 5-HTP at doses of 15 mg/kg (p < 0.05) and 25 mg/kg (p < 0.001). Note that the maximum reduction reached a plateau 30 min. after injection, and it persisted for at least 90 min. during the remaining observation time. A key observation was that animals survived at Tamb of 12°C after 25 mg/kg 5-HTP. In a separate experiment, ketanserin injected 15 min. before 5-HTP failed to block the hypothermic effect of 5-HTP in clorgyline-pre-treated rats at Tamb of 12°C (fig. 2C).
The next experiments were carried out at Tamb of 6°C for two purposes. Firstly, to reaffirm results obtained at the Tamb of 12°C with an even cooler ambient temperature. Secondly, it was of interest to determine if more extreme cooling would have a larger effect on the 5-HTP-evoked hypothermia. After a 2-hr habituation period, the mean basal body temperature at Tamb of 6°C was 37.1°C ± 0.2 (n = 10). Compared with Tamb of 22°C, the difference between the two baselines was not significant (p > 0.05). As shown in fig. 3A, 25 mg/kg 5-HTP produced a marked decrease in body temperature (p < 0.05). The reduction was significantly greater at Tamb of 6°C (p < 0.01) compared with Tamb of 12°C. Interestingly, body temperature progressively declined during the entire 120-min. experimental time, and the maximum reduction was 8.5°C ± 1.2 below the pre-injection level. All animals survived in the Tamb of 6°C after injection of 25 mg/kg 5-HTP. As shown in fig. 3B, ketanserin in the Tamb of 6°C had no effect on hypothermia after 5-HTP (p > 0.05).
5-HT efflux in the preoptic/anterior hypothalamus at Tamb of 22, 12 and 6°C.
To further investigate the relationship between Tamb and the syndrome, we measured extracellular 5-HT in response to 5-HTP injection. The preoptic/anterior hypothalamus was examined because this region is important for thermoregulation . Basal 5-HT efflux in the preoptic/anterior hypothalamus of clorgyline-pre-treated rats at Tamb of 22°C was 0.32 pg/sample (±0.03, n = 36). As shown in fig. 4, injection of 5-HTP elicited a dose-dependent increase in extracellular 5-HT. The two-way factorial anova revealed a significant effect of 5-HTP treatment (F(3, 30) = 9.433; p = 0.0002) and sampling time (F(7, 21) = 7.413; p < 0.0001). In addition, there was a significant interaction between 5-HTP treatment × sampling time (F(21, 210) = 4.473; p < 0.0001). Relative to respective baselines, the increase was 6-, 26- and 97-fold in response to 5, 15 and 25 mg/kg 5-HTP, respectively.
Next, hypothalamic 5-HT was measured at Tamb of 12°C. Basal 5-HT efflux was 0.29 pg/sample (±0.04, n = 25) in animals pre-treated with clorgyline. Injection of 5-HTP produced a time- and dose-dependent increase in 5-HT efflux (fig. 5A–C). At Tamb of 12°C, the maximum increase was 2-, 12- and 86-fold in response to 5, 15 and 25 mg/kg 5-HTP, respectively, in clorgyline-pre-treated rats. There was a strong tendency toward a reduction in 5-HTP-evoked 5-HT efflux at Tamb of 12°C, compared with that at 22°C. However, statistical analysis failed to reach a significant difference between Tamb 22°C and 12°C in response to 5-HTP at doses of 5 mg/kg (F(1, 13) = 2.968, p = 0.1106), 15 mg/kg (F(1, 9) = 0.104, p = 0.7534) or 25 mg/kg (F(1, 16) = 2.031, p = 0.1734). To further explore the effect of environmental temperature, changes in 5-HT efflux in response to 25 mg/kg 5-HTP were next measured at Tamb of 6°C. As shown in fig. 5D, the increase was reduced to 35-times. Compared with the effect at Tamb 22°C, beginning 45 min. after 5-HTP injection, 5-HT efflux was significantly attenuated at the low ambient temperature (F(1, 16) = 5.448, p = 0.0396).
5-HT efflux in the prefrontal cortex at Tamb of 22, 12 and 6°C.
At the Tamb of 22°C, baseline 5-HT efflux in the prefrontal cortex was 0.29 pg/sample (±0.04, n = 29) with a 4-, 30- and 240-fold increase in response to 5, 15 and 25 mg/kg 5-HTP, respectively (fig. 6A–C). At Tamb of 12°C, baseline 5-HT efflux was 0.31 pg/sample (± 0.04, n = 25), not significantly different from that at 22°C (p > 0.05). The increase in 5-HT efflux was 5-, 18- and 129-fold after 5, 15 and 25 mg/kg 5-HTP, respectively. Compared with that at Tamb of 22°C, the efflux at Tamb of 12°C was significantly attenuated in response to 25 mg/kg (F(1, 14) = 20.225, p = 0.0005) but not 5 mg/kg (F(1, 23) = 0.471, p = 0.4996) or 15 mg/kg 5-HTP (F(1,18) = 0.641, p = 0.4337). When Tamb was lowered to 6°C, there was only a 25-fold increase after 25 mg/kg 5-HTP (fig. 6C). The reduction, compared with that at Tamb of 12°C, was significant (F(1, 19) = 12.258, p = 0.0024). Data analysis revealed a significant main effect of Tamb (fig. 6C; F(2, 14) = 38.303, p < 0.0001), a significant effect of sampling time (F(7, 21) = 70.358, p < 0.0001) and a significant interaction of Tamb × sampling time (F(14, 138) = 23.281, p < 0.0001).
Tamb-dependent antagonistic effect of ketanserin.
Fig. 7 shows the Tamb-dependent antagonistic effect of ketanserin (5 mg/kg, i.p.) on 5-HT efflux in the preoptic/anterior hypothalamus evoked by 5-HTP (25 mg/kg, s.c.) in clorgyline-pre-treated animals. Note that neither ketanserin nor vehicle alone had an effect on basal 5-HT at Tamb of 22, 12 or 6°C. As shown in fig. 7A, the increase in 5-HT at Tamb of 22°C was significantly attenuated in rats pre-treated with ketanserin (F(1, 14) = 17.14, p = 0.001). The area under the curve (AUC) analysis revealed that the total reduction in 5-HT efflux was 60%. Fig. 7B shows that the increased efflux at Tamb of 12°C was reduced significantly by ketanserin (fig. 7A; F(1, 12) = 4.643; p = 0.0476), compared with 25 mg/kg alone at the same Tamb. The AUC analysis revealed that the reduction was 50% of the efflux produced by 25 mg/kg 5-HTP. Fig. 7C shows that ketanserin failed to have an antagonistic effect when the test was carried out at Tamb of 6°C (F(1, 9) = 0.511, p = 0.4928).
Fig. 8 shows the Tamb-dependent effect of ketanserin examined in the prefrontal cortex. Pre-treatment with ketanserin blocked 60%, 26% and 0% of the efflux induced by 25 mg/kg 5-HTP at Tamb of 22, 12 and 6°C, respectively. Data analysis shows that the antagonistic effect of ketanserin was significant at Tamb of 22°C (F(1, 12) = 14.230, p = 0.0023). For experiments carried out at Tamb of 12 and 6°C, ketanserin did not significantly attenuate 5-HTP-evoked changes in efflux (fig. 8B; F(1, 19) = 2.088, p = 0.1647; fig. 8C; F(1, 14) = 0.001, p = 0.9823).
Correlation analysis: primary, secondary and ratio of secondary to primary efflux.
Protocols for calculating primary efflux, total efflux, secondary efflux and ratio of secondary to primary efflux have been described in section ‘Materials and Methods’. To minimize the number of animals, we simply re-evaluated the relevant data in figs 4–8 in line with our ‘two-component efflux’ hypothesis . In the preoptic/anterior hypothalamus, basal 5-HT efflux was 0.34 ± 0.05 (n = 5) under the condition of blocking 5-HT2A receptors with ketanserin at Tamb of 6°C in clorgyline-pre-treated rats. Under this specific condition, the excessive 5-HT efflux evoked by 5-HTP is defined as ‘primary efflux’, theoretically relevant to 5-HT synthesis and metabolism. The primary 5-HT efflux after 25 mg/kg 5-HTP was 10.7 ± 3.7 pg/sample (n = 5), approximately 30-fold of the basal efflux. As elucidated in the left column of table 1, 5-HTP injection in the absence of ketanserin produced a Tamb-dependent total efflux at Tamb of 6, 12 and 22°C. The secondary efflux was also Tamb-dependent. In addition, the ratio of secondary to primary efflux, which may reflect the relative strength of 5-HT2A receptor activation at a given Tamb, was determined. Specifically, the ratio was nearly zero at Tamb of 6°C, suggesting that 5-HT2A receptors were not activated although the primary efflux at the preoptic/anterior hypothalamus was already 30-fold above baseline. The ratio of 0.8 at Tamb of 12°C suggested that 5-HT2A receptors were able to be to some extent activated by 30-fold increases in extracellular 5-HT. Under this condition, the amount of 5-HT2A receptor-mediated secondary efflux was nearly equal to that of primary efflux. Lastly, the ratio increased to 2.3 at Tamb of 22°C, suggesting that 5-HT2A receptors were extensively activated by this level of primary efflux. The amount of 5-HT2A receptor-mediated secondary efflux was at least two times greater than those associated with biosynthesis and metabolism.
Table 1. Levels of total and secondary effluxes and ratio of secondary to primary efflux.
Clorgyline-pre-treated rats were administered with 25 mg/kg 5-HTP at different ambient temperatures (Tamb) while 5-HT efflux was measured in the preoptic/anterior hypothalamus (POA) and prefrontal cortex (PFC).
1Total efflux (pg/sample) is mean ± S.E.M. of four samples between 30 and 75 min. in which 5-HT efflux in response to 5-HTP was at peak as illustrated in figs 4–8.
2Secondary efflux (pg/sample) was obtained by subtracting the primary efflux from the total 5-HT efflux as described in section ‘Materials and Methods’.
10.7 ± 3.7 (baseline 0.34 ± 0.05)
9.7 ± 0.6 (baseline 0.40 ± 0.03)
8.3 ± 1.8
19.0 ± 5.8
35.6 ± 11.8
10.6 ± 3.1
18.9 ± 4.3
35.4 ± 5.9
The right column of table 1 shows total efflux, secondary efflux and ratio of secondary-to-primary efflux in the prefrontal cortex. The basal prefrontal cortex 5-HT level was 0.40 ± 0.03 pg/sample (n = 5). The amount of primary efflux was 9.7 ± 0.6 pg/sample (n = 5) after 25 mg/kg 5-HTP in clorgyline-pre-treated rats. Similar to the preoptic/anterior hypothalamus, there were also Tamb-dependent increases in secondary efflux and ratio of secondary to primary efflux. These data agree with the correlation analysis between Tamb and 5-HT2A receptor activation as described in the preoptic/anterior hypothalamus.
The present results indicate that cool Tamb alleviates the severity of serotonin syndrome, as observed that hyperthermia and death evoked by 5-HTP in clorgyline-pre-treated animals at the standard Tamb of 22°C were completely blocked at cool Tamb of 12 and 6°C. Further investigation utilizing microdialysis at cool Tamb revealed that the availability of extracellular 5-HT for evoking the syndrome was significantly reduced, suggesting that the alleviation is due to not only heat loss to the cool environment but also involvement of serotonergic neuronal activity, particularly reduction in functional activity of 5-HT2A receptors. Direct evidence for the latter was obtained by comparing and contrasting changes in 5-HT efflux in animals pre-treated with the 5-HT2A receptor antagonist ketanserin. Altogether, this study suggests that cool Tamb may reduce the responsivity of 5-HT2A receptors, and thereafter 5-HT2A receptor-mediated symptoms of the syndrome are less severe in response to excessive 5-HT efflux which otherwise causes hyperthermia and death at standard Tamb of 22°C.
Another observation under cool Tamb was that some symptoms of the syndrome were still obvious although benign. We therefore suggest that the development of serotonin syndrome undergoes at least two major processes: the syndrome initiation and progression. Evidence supporting the ‘two processes’ hypothesis was obtained in this study by correlating the syndrome severity with changes in 5-HT efflux in rats pre-treated with ketanserin at different Tamb, as observed that there was proportionally reduction in 5-HT efflux at cool Tamb of 12 and 6°C in contrast to that at 22°C. Despite this, remaining 5-HT was still high at cool Tamb. Considering that excessive 5-HT in the CNS for inducing the syndrome consists of primary and secondary effluxes , our data further prompt us to suggest that the syndrome initiation is attributed to the level of primary 5-HT efflux while the syndrome progression is determined mainly by functional activity of 5-HT2A receptors associated with the secondary efflux. In summary, although the syndrome can be initiated but benign at cool Tamb, further deterioration towards the malignant syndrome associated with the 5-HT2A receptor-facilitated secondary efflux is effectively prohibited by cool Tamb, closely resembling the results of 5-HT2A receptor antagonist pre-treatment [11,13], which is likely due to the reduction in functional activity of 5-HT2A receptors under our experimental conditions.
Our animal model of serotonin syndrome employed an approach established in previous studies [1,9]. In this model, rats are pre-treated with an MAOI followed by 5-HTP, which at Tamb of 22°C produces a dose-dependent increase in 5-HT efflux and body temperature. In response to lower doses of 5-HTP, we observe neuromuscular symptoms including tremors, forepaw treading and hind limb abduction similar to those described in earlier studies [1,14]. At the highest 5-HTP dose that we tested, the syndrome becomes malignant as indicated by hyperthermia and death in clorgyline-pre-treated rats. Our observations are generally consistent with previous pre-clinical reports of the effects of combined administration of an MAOI with 5-HTP [9,11] as well as treatment with other drugs such as 3,4-methylenedioxymethamphetamine (MDMA) that evoke large increases in 5-HT efflux [15–17]. This is also consistent with the hypothesis that excessive 5-HT efflux in the CNS may be involved in the initiation of the 5-HT syndrome for some serotonergic drugs [2,11].
In addition to excessive 5-HT efflux, the results support our hypothesis that the malignant syndrome is dependent upon both sufficient activation of 5-HT2A receptors and a warm Tamb. Thus, ketanserin, a 5-HT2A receptor antagonist , prevented the increase in body temperature and lethal effects otherwise evoked by administration of clorgyline followed by the high dose of 5-HTP. However, 5-HT2A receptor antagonists had no effect on neuromuscular symptoms of the syndrome, consistent with previous observations [19,20]. Similarly, increased body temperature and lethality, but not motor symptoms were prevented when animals were housed at cool Tamb of 12 or 6°C instead of the standard laboratory condition of 22°C. At cool temperatures, 5-HTP evoked the opposite response, a substantial decrease in core temperature. Whether core temperature increased or decreased was not well correlated with the extent of drug-evoked 5-HT efflux in the preoptic/anterior hypothalamus or prefrontal cortex. Together with previous evidence, these results are consistent with our hypothesis that development of the malignant syndrome occurs upon sufficient 5-HT2A receptor activation.
Ketanserin pre-treatment in cool Tamb did not affect the decrease in core temperature induced by 5-HT efflux. Remarkably, this observation suggests that the 5-HT2A-related influence on body temperature is completely inhibited in a cool environment. High-dose 5-HTP treatment was benign with no fatalities under cool conditions, presumably secondary to the induced hypothermic response. This last observation is consistent with other reports that hypothermia is protective against brain trauma caused by 5-HT toxicity [21–23]. Thus, for example, in their study of environmental influences on the malignant 5-HT syndrome, Malberg and Seiden (1998) demonstrated hyperthermia and neurotoxic effects of MDMA at Tamb greater than 22°C, with hypothermia and absence of brain damage when rats were housed at less than 22°C.
It is likely that the hypothermic responses to 5-HTP can be ascribed to the role of 5-HT1A receptors in the serotonin syndrome that we and others have observed [1,19,24]. This is based in part on the reduction in core temperature observed in response to 8-OH-DPAT [22,25,26] although recent studies suggest that 5-HT7 receptors are important in this effect [27–29]. Moreover, Nicholas and Seiden (2003) provides indirect evidence that 5-HT-related decreases in core temperature do not involve a change in activity of thermoregulatory neurons in the hypothalamus but represent a hypothermic response, presumably involving excessive 5-HT-related stimulation of peripheral heat loss or inhibition of heat production mechanisms. Our observation that 5-HT efflux in the preoptic/anterior hypothalamus was not well correlated with changes in core temperature is consistent with this conclusion.
Unlike its effect at 22°C, ketanserin had no influence on the 5-HTP-evoked decrease in body temperature at cool Tamb. Apparently, the ‘functional activity’ of 5-HT2A receptors was reduced and thus unresponsive to evoked 5-HT efflux in a cool environment. It is important to note that rats maintained their normal core temperature before 5-HTP administration even when housed for several hours at Tamb of 6°C. Conceivably, rectal temperature did not accurately reflect reductions in temperature of the brain or peripheral vasculature, which might have directly lowered 5-HT2A receptor affinity or sensitivity in these structures. However, as mentioned above, Malberg and Seiden (1998) reported that the hyperthermic response to MDMA at a given dose level was converted to hypothermia when rats were tested at an ambient temperature only 2°C cooler than their standard housing condition. Thus, it seems unlikely that changes in internal body temperature, even in structures close to the body surface, could explain these remarkable differences in thermal response to drug treatment. It is possible that skin surface receptors, namely transient receptor potential proteins which are sensitive to small changes in ambient temperature  activate a neural circuit that influences thermoregulatory processes including responses to 5-HT2A receptor stimulation. Alternatively, cool Tamb induces changes in regulation of some hormones that might interact with functional activity of 5-HT2A receptors. Nevertheless, it is not currently known how ambient temperature affects functional activity of 5-HT2A receptors in the CNS. It is also worth noting the apparent paradox that this 5-HT2A receptor-dependent mechanism for increasing body temperature was inactivated in cool environmental conditions when mechanisms for generating body heat are more important. However, because hyperthermic responses were observed only in response to extreme increases in 5-HT efflux, the physiological relevance of this effect seems unlikely.
Excessive 5-HT efflux in the CNS as opposed to release in the periphery is the probable cause of the motor and autonomic symptoms of the ‘serotonin syndrome’ [2,4]. Our microdialysis experiments were designed to measure, under controlled environmental Tamb, changes in 5-HT efflux in two CNS sites which we suspected might play a role in the malignant syndrome. A major aim was to determine the threshold level at which 5-HT efflux in the CNS is associated with hyperthermia and whether or not this excessive level is always associated with a malignant syndrome. During undisturbed behaviour and even in response to various strong stressors, 5-HT efflux in the rat CNS does not usually exceed two- to three-fold baseline levels according to Rueter and Jacobs . Moreover, SSRIs only produce ∼1- to five-fold increases in 5-HT efflux in rat CNS at doses considered comparable to the therapeutic range for treatment of human depression [32,33]. Consistent with this, we did not observe motor symptoms or any other signs of the serotonin syndrome associated with the five-fold increases in the preoptic/anterior hypothalamus and prefrontal cortex produced by 5 mg/kg 5-HTP. Other studies indicate that even 10-fold increases in 5-HT efflux do not elicit significant neuromuscular symptoms [11,15]. Thus, although the relationship between 5-HT efflux and the syndrome in human beings cannot be determined because of ethical constraints, these microdialysis data suggest that the serotonin syndrome is produced only when 5-HT efflux greatly exceeds physiological levels.
Taking advantage of in vivo microdialysis, we studied the detailed quantitative relationship of extracellular 5-HT to increased body temperature. Previous studies suggested that hyperthermia is correlated with 140- to 1000-fold increases in 5-HT efflux in the preoptic/anterior hypothalamus [9,34]. However, at variance with these reports, we found that an intermediate dose of 5-HTP that produced a 30-fold increase in the preoptic/anterior hypothalamus at Tamb of 22°C was associated with a small, non-significant increase in body temperature. At the highest does of 5-HTP, which elicited an 80-fold increase in 5-HT efflux in the preoptic/anterior hypothalamus, we observed hypothermia instead when administered in a cool Tamb. In summary, 5-HT efflux in the preoptic/anterior hypothalamus is not a reliable marker indicating severity of the serotonin syndrome. Compared with the preoptic/anterior hypothalamus, 5-HTP-evoked increases in 5-HT efflux were much greater in the prefrontal cortex: a 60-fold elevation in response to 15 mg/kg 5-HTP and 250-fold after 25 mg/kg. One possible explanation for this regional difference is that 5-HT2A receptors, which we suggest are responsible for evoking the secondary component of 5-HT efflux, are unevenly distributed in the CNS [35,36]. It is important to note, however, that because our efflux data are correlative in nature, further studies are necessary to firmly establish the location of 5-HT2A receptors necessary for eliciting the malignant syndrome.
Although the present study used rats treated with an MAOI combined with 5-HTP, insights into the mechanisms underlying excessive stimulation of 5-HT efflux should be relevant to understanding causes of the 5-HT syndrome in human beings. When injected separately, 5-HTP and clorgyline, evoke only small, less than 3-fold increases in 5-HT efflux [37,38]. In contrast, combined administration of these two compounds has a synergistic effect on biosynthesis and metabolism resulting in very large initial increases in 5-HT efflux . In a previous study, we defined the initial increase associated with direct effect of these drugs on serotonergic nerve terminals as ‘primary efflux’ . Primary 5-HT efflux might activate all subtypes of 5-HT receptors. However, because the affinity of 5-HT2 receptors is at least 400 times lower than 5-HT1 receptors [39,40], with the relatively small increases in primary efflux produced by lower doses of 5-HTP, 5-HT1 subtypes would be selectively activated and thus produce only motor and other benign symptoms of the serotonin syndrome [19,20].
In contrast, we suggest that secondary efflux is associated with malignant symptoms of the syndrome due to activation of 5-HT2A receptors in a Tamb-dependent manner. As determined by pre-treatment with ketanserin (present data) or cyproheptadine (Zhang et al., 2009) which block 5-HT2A receptors, we suggest that the underlying level of primary efflux evoked by an MAOI combined with 25 mg/kg 5-HTP is about a 30-fold increase above baseline levels in the preoptic/anterior hypothalamus and prefrontal cortex. We further suggest that, in response to primary efflux of this magnitude, 5-HT2A receptors are activated. Subsequently, depending on the responsivity of 5-HT2A receptors, this activates glutamatergic neurons which in turn further stimulate serotonergic neurons to induce ‘secondary efflux’. This hypothesis is consistent with extensive electrophysiological and neurochemical evidence of a feedback loop involving serotonergic projections to the prefrontal cortex that, via stimulation of 5-HT2A receptors, activates glutamatergic neurons which project back to the midbrain raphe nuclei [41–43]. In further support of the role of glutamate in this positive-feedback loop, pre-treatment with NMDA receptor antagonists also blocked hyperthermic responses to 5-HTP [1,34]. Although the physiological significance of this positive feedback loop has not been established, it normally would be restrained by several strong negative feedback mechanisms including 5-HT1 autoreceptors and 5-HT2 receptors that activate GABAergic inputs to serotonergic neurons . We suggest that, at sufficiently high levels of primary efflux and Tamb of 22°C or greater, negative-feedback influences are overwhelmed by the 5-HT2A-dependent positive feedback loop, and the resultant excessive level of secondary efflux causes the malignant syndrome.
The results of the present study demonstrate the importance of 5-HT2A receptor-dependent secondary efflux in producing the malignant syndrome. As shown in table 1, the malignant syndrome was prevented when the ratio of secondary to primary efflux was reduced below two by either pre-treatment with a 5-HT2A receptor antagonist or by carrying out experiments in a cool Tamb. We suggest that higher ratios are indicative of strong 5-HT2A receptor activation leading to a positive-feedback cycle driving the excessive 5-HT efflux thus causing hyperthermia and the serious syndrome. A key observation of this study is that the responsivity of 5-HT2A receptors is attenuated in a cool environment. Thus, in a cool environment, hypothermia and a benign syndrome result from drug-induced increases in 5-HT efflux, an effect that has been linked to stimulation of 5-HT1A receptors in previous reports [19,45]. Further studies are important to determine the molecular mechanisms by which a cool environment attenuates 5-HT2A receptor responsivity and thus ameliorates the malignant serotonin syndrome.