An autoradiographic study of dextromethorphan high-affinity binding sites in rat brain: sodium-dependency and colocalization with paroxetine


  • P. Meoni,

    Corresponding author
    1. Department of Pharmacology, The School of Pharmacy, University of London, London WC1N 1AX
      Author for correspondence at present address: Department of Pharmacology, Medical School, University of Birmingham, Vincent Drive, Edgbaston B15 2TT.
    Search for more papers by this author
  • F. C. Tortella,

    1. Neuropharmacology Branch, Walter Reed Army Institute of Research, Washington D.C., U.S.A.
    Search for more papers by this author
  • N. G. Bowery

    1. Department of Pharmacology, Medical School, University of Birmingham, Birmingham B15 2TT
    Search for more papers by this author

Author for correspondence at present address: Department of Pharmacology, Medical School, University of Birmingham, Vincent Drive, Edgbaston B15 2TT.


  • The distribution and some pharmacological properties of centrally located dextromethorphan high-affinity binding sites were investigated by in vitro autoradiography.

  • Sodium chloride (50 mm) induced a 7 to 12 fold increase in dextromethorphan binding to rat brain in all areas tested. The effect of sodium was concentration-dependent with a higher dose (120 mm) exerting a smaller effect on binding.

  • [3H]-dextromethorphan binding in the presence of sodium was inhibited in the presence of the anticonvulsant phenytoin at a concentration of 100 μm, while the σ ligand (+)-3-(-3-hydroxyphenyl)-N-(1-propyl)pipendine ((+)-PPP) had no effect on the binding, suggesting an interaction with the DM2 site.

  • The distribution of the sodium-dependent binding identified in this study correlated significantly with the distribution of the selective 5-HT uptake inhibitor [3H]-paroxetine, and paroxetine and dextromethorphan mutually displaced their binding at concentrations in the low nanomolar range.

  • These data show that dextromethorphan and paroxetine share a sodium-dependent high affinity binding site in rat brain, and suggest that dextromethorphan might interact, in the presence of sodium, with the 5-HT uptake mechanism in rat brain.

British Journal of Pharmacology (1997) 120, 1255–1262; doi:10.1038/sj.bjp.0701043


Dextromethorphan is a dextrorotatory analogue of morphine that acts centrally to elevate the cough threshold in both man (Eddy et al., 1969) and rodents (Benson et al., 1953). Besides this well characterized property, it has also been shown to possess an anticonvulsant (Tortella & Musacchio, 1986) and neuropro-tectant (Steinberg et al., 1991) action in laboratory animals.

Binding sites for dextromethorphan have been described in both the guinea-pig (Craviso & Musacchio, 1980) and rat brain (Klein & Musacchio, 1992), and several centrally acting antitussives such as carbetapentane and caramiphene show high affinity for these sites (Craviso & Musacchio, 1983b). In comparison, dextrorphan, the main active metabolite of dextromethorphan following 0-demethylation by a cytochrome P-450 (Schmid et al., 1985), only displaces dextromethorphan binding from these sites with an IC50 in the micromolar range (Craviso & Musacchio, 1983b).

The functional role of these dextromethorphan high affinity binding sites has yet to be fully elucidated, but the selective displacement of dextromethorphan from these sites by other centrally acting antitussives such as 3-substituted 17-methyl morphinan analogues, and the efficacy of these substances as neuroprotectants and anticonvulsants (Newman et al., 1992) suggest their involvement in the pharmacological actions of dextromethorphan and its derivatives.

Phenytoin and ropizine have been shown to produce a concentration-dependent allosteric increase in the affinity of dextromethorphan to a subpopulation of dextromethorphan high affinity binding sites in guinea-pig (Musacchio et al., 1988) and rat brain homogenates (Klein & Musacchio, 1992). The subsequent observation of a differential effect of phenytoin and other anticonvulsants on dextromethorphan binding in different areas of guinea-pig brain (Canoll et al., 1990) in an autoradiographical study and the computer-assisted modelling of homologous and heterologous competition studies between dextromethorphan and the σ ligand (+)-3-(-3-hydro-xyphenyl)-N-(1-propyl)piperidine ((+)-PPP) (Zhou et al., 1991; Klein & Musacchio, 1992), has led to a model in which dextromethorphan would be binding in guinea-pig and rat brain to two high affinity and one low affinity site. One of the two high affinity sites, termed DM1 by Musacchio et al. (1989), is allosterically enhanced by phenytoin and ropizine (Musacchio et al., 1988; Klein & Musacchio, 1992) and could therefore be involved in the anticonvulsant action of dextromethorphan. The DM1 site has properties in common with the binding site for (+)-PPP, and should therefore be regarded as a DM1/σ1 site (Klein & Musacchio, 1990). By comparison, (+)-PPP only shows a micromolar affinity for the other high affinity binding site (Zhou & Musacchio, 1991; Klein & Musacchio, 1992). DM2, therefore does not represent a σ site, and binding to this site is inhibited (Canoll et al., 1990; Klein & Musacchio, 1992) by phenytoin and ropizine.

SKF 525-A, a wide spectrum inhibitor of cytochrome P-450 (Schenkman et al., 1972), and GBR 12909, a dopamine uptake inhibitor with high affinity for cytochrome P450IID1 (Niznik et al., 1990), have been shown to displace dextromethorphan binding from the DM1/σ1, site (Klein et al., 1991), and similar findings have suggested that this site can be identified with a metabolic site.

Antitussive properties have never been described for σ site ligands, thus the DM2 high affinity site may play a key role in the antitussive action of dextromethorphan. The DM2site is relatively less abundant in guinea-pig than in rat brain, where it represents the majority of the binding (Klein & Musacchio, 1992). Most of the behavioural tests that have highlighted the antitussive, anticonvulsant and neuroprotectant properties of dextromethorphan have been conducted in the rat. This prompted us to characterize in more detail dextromethorphan binding in rat brain, in order to elucidate the identity of the DM2 binding site in this species.

As part of the characterization we have considered the possibility that dextromethorphan might interact with the uptake process for 5-hydroxytryptamine (5-HT). α(—)-Trans-4 - (p-fluorophenyl)-3-[3,4 - methylenedioxyphenoxy-methyl]pi-peridine (paroxetine) is a very potent and selective inhibitor of sodium-dependent 5-HT uptake in vitro (Buus-Lassen, 1978; Habert et al., 1985) and in vivo (Graham & Langer, 1988). We have therefore used it in the present study to examine the possible link between dextromethorphan and the sodium-dependent neuronal 5-HT uptake mechanism.


Dextromethorphan autoradiography

Male Wistar rats weighing 200–300 g were decapitated and the brains were rapidly removed and snap-frozen in liquid nitrogen-cooled isopentane at − 70°C. Brains were stored at − 80°C before being sectioned. Sagittal sections (10 μm) were cut at − 16°C in a cryostat at a level 0.40 mm lateral to the midline and thaw-mounted on gelatine-coated (5% gelatine, 0.05% chromic potassium sulphate) glass slides. Sections were allowed to air-dry before storage at − 20°C until use in binding assays.

Before incubation, sections were thawed at room temperature for 15 to 30 min and pre-incubated in ice-cold Tris-HCl solution (50 mM, pH 7.4) for 30 min. Following pre-incubation, sections were dried under a stream of cool air for 15 to 30 min.

For dextromethorphan binding, incubation was initiated by covering each section with a drop of buffer (150 μl) containing an appropriate concentration of [H]-dextromethorphan (81 Ci mmol−1) in Tris-HCl solution (50 mM, pH 7.4). Non-specific binding was defined as the binding obtained in the presence of unlabelled dextromethorphan 100 μM.

In preliminary studies to assess optimal conditions for autoradiography, the kinetic characteristics of [H]-dex-tromethorphan binding were examined to obtain the best ratio of total to non-specific binding, whilst maintaining the highest possible specific binding.

In these experiments, the level of binding was assessed by liquid scintillation spectrometry: the portion of the slide containing the section was cut out with a diamond pencil, placed in a 20 ml plastic scintillation vial with 10 ml of scintillation fluid (Eco-scint) and counted in a liquid scintillation counter (Rackbeta, Packard).

Association of the ligand with its binding site was studied by incubating the sections for various lengths of time (3 – 96 min) before washing the slides (5 min for three times) in washing buffer at 24 or 4°C after an initial rinse in buffer solution. The dissociation of [H]-dextromethorphan from the slide-mounted sections was examined after incubation with 20 nM [H]-dex-tromethorphan for 60 min at 4°C. The sections were eventually transferred to ice-cooled wash buffer for up to 60 min. These experiments were repeated with three animals, with triplicate determination for each rat.

Saturation experiments were carried out with 5 to 80 nM of [H]-dextromethorphan while mapping studies were performed with a concentration of radiolabelled ligand of 20 nM. After incubation, slides were washed in fresh Tris-HCl solution (50 mM, pH 7.4) containing 100 mM choline chloride and 0.01% v/v Triton X-100 at 4°C (wash buffer), as described by Craviso and Musacchio (1983a). Preliminary studies (data not shown) demonstrated an increase in specific binding from 50 to 70% of total binding when the sections were washed for 5 min in Tris-HCl solution (50 mM, pH 7.4) without or with choline chloride and Triton, respectively.

In saturation and distribution studies, sections were incubated for 1 h in Tris buffer solution (50 mM, pH 7.4) at 4°C and washed for 2.5 min at the same temperature in wash buffer.

Dried, incubated sections were apposed to H-Hyperfilm (Amersham) in slide cassettes with appropriate plastic tritium standards (Amersham) for 7 to 28 days. Films were developed with Kodak D-19 developer for 1 min and fixed for 5 min.

Paroxetine autoradiography

Binding conditions for the study of the distribution of [H]-paroxetine to rat brain were adapted from the work of Chen et al. (1992). Sections were pre-incubated in Tris buffer solution (50 mM at pH 7.4) at room temperature for 20 min. After being dried, the rat tissue was incubated with 1 nM [H]-par-oxetine (15 Ci mmol−1) in Tris HCl solution (containing 300 mM NaCl) for 90 min at room temperature. At the end of the incubation, the sections were quickly dipped in buffer and subsequently washed in the same solution for 30 min. Nonspecific binding was defined as binding in the presence of 100 μM fluoxetine.

Dried sections were then processed as for dextromethorphan autoradiography.

Analyses of autoradiograms

The autoradiograms were analysed by a Quantimet 970 (Cambridge Instruments). The density in each region was recorded by positioning a squared cursor on two areas over each region. Mean values for each nucleus were determined from 4 measurements deriving from two adjacent sections. Variability of measurements is expressed as s.e.mean of the values in fmol mg−1 wet weight of brain tissue. Individual tables were analysed with a two-way ANOVA to determine a possible effect of treatment and interaction between the effect of treatments and the area measured. Data were subsequently analysed with a one-way ANOVA with Tu-key's post-hoc test to assess which treatments significantly affected the binding. In the case of a significant interaction between treatments and nuclei, ANOVA was applied to individual nuclei.


[H]-dextromethorphan and [H]-paroxetine were purchased from New England Nuclear (NEN, Boston, MA, U.S.A.); SKF 525-A (N,N-diethylaminoethyl-2,2-diphenylvalerate hy-drochloride) was obtained through Research Biochemical International (RBI, Natik, MA, U.S.A.). Paroxetine and fluoxetine were a kind gift of Smithkline Beecham (Harlow, Essex, U.K.). GBR 12909 (1-[2-[bis(4-fluorophenyl)methox-y]ethyl]-4-[3-phenylpropyl]piperazine dihydrochloride) was obtained from RBI (Natik, MA, U.S.A.). All the other chemicals in this study were obtained from Sigma Chemical Inc. (St. Louis, MO, U.S.A.).


Dextromethorphan binding kinetics

Experiments conducted to assess the kinetics of dextromethorphan binding to rat brain showed the binding to follow a one-site binding isotherm model with the equilibrium reached after 30 min of incubation at 4°C (Figure 1). Preliminary experiments (data not shown) revealed higher levels of specific binding following incubation at 4°C as compared to room temperature. The influence of washing on the level of radioligand bound to the section was best described by a two-site dissociation model, with the fast dissociating site (t1/2= 1.13 min) having a capacity double the slow-dissociating site (t1/2=5.94 min) (Figure 1). On the basis of these results, binding conditions for successive studies were established as 1 h incubation followed by 2.5 min wash. Under these conditions, specific binding accounted for 70% of total binding when measured with quantitative microdensitometry in distribution studies.

Figure 1.

Effect of different incubation (a) and washing times (b) on [H]-dextromethorphan binding to rat brain. Sections were incubated at 4°C for various lengths of time in the presence of [H]-dextromethorphan 20 nM in TrisHCl 50 mM, pH 7.4, and subsequently washed for 2.5 min at 4°C in the same buffer containing 100 mM choline chloride and 0.01% Triton X-100. The effect of different washing times was tested after an initial incubation of 60 min at 4°C. (▪) Total binding, (▴) non-specific and (▾) specific binding.

Effect of different ions on dextromethorphan binding

Dextromethorphan binding in Tris buffer solution in the absence of any added sodium chloride was very evenly distributed throughout the rat brain. Binding to all nuclei ranged between 10 and 20 fmol mg−1 wet weight tissue with less than a two fold ratio between the lowest (frontal cortex) and the highest (superior colliculi) level of binding (Table 1). Addition of sodium chloride (50 mM) to the incubation solution produced on average a seven fold increase in dextromethorphan binding. This increase in dextromethorphan binding with sodium was significant (ANOVA, P<0.05), with a significant interaction between treatments and nuclei. One-way ANOVA followed by Tukey's post-hoc test applied to individual regions revealed a significant effect of sodium in all areas with the exception of cerebellum, where the effect did not reach statistical significance. The effect of sodium was maximal in the ventral tegmental area, where the binding was increased 12 fold.

Table 1. Effect of sodium on [H]-dextromethorphan binding in discrete regions of rat brain and of calcium and magnesium on the sodium-dependent binding
 ControlsNa± 50 mMaNa± 120 mMaNa+ 50 mM + Ca2+ 2.5mMa,bNa+ 50 mM+ Mg2+ 1.2mMa,b
  1. The tissue sections were incubated with 10nM [H]-dextromethorphan, as described in Methods, in buffer containing the ions indicated above each column. Binding capacity is expressed as fmol mg−1 wet weight of tissue. Each value is the mean of duplicate determinations and represents the mean± s.e.mean of 3–11 rats. aP<0.05 versus binding in the absence of sodium, bP<0.05 versus binding in the presence of sodium 50 mM (ANOVA and Tukey's posthoc test).

Mammillary bodies16.3 ± 2.0132.4± 19.4a138.8 ± 7.7a,b127.7 ± 8.5b79.7 ±10.3b
Ventral tegmental area16.4 ± 3.0198.1± 3.9a157.8 ± 5.7a,b151.8 ± 4.7b114.9 ± 5.5b
Dorsal raphe18.5 ± 2.1191.5± 10.1a142.1 ± 8.4a162.4±11.6b113.9 ± 8.4b
Superior colliculi21.2 ± 3.0152.7± 10.6a114.1 ±3.6a,b93.7 ± 8.3b58.1±4.7b
Central grey area17.0 ± 2.0127.8±10.0a106.3± 3.0a90.4 ±5.2b64.2 ± 2.5b
Dorsal tegmental area19.2 ± 2.6120.1±4.3a102.3± 5.4a,b93.3 ± 3.860.9 ± 3.4b
Anterior hypothalamus14.5 ± 3.1109.2±12.4a104.4 ± 4.1a83.2 ± 3.445.2 ± 3.1b
Pons11.9 ± 2.340.1±6.3a35.1 ± l-9a25.3 ± 6.17.7 ±
Hippocampus (CA3 area)15.3 ± 2.654.7 ± 6.4a42.1±2.2a19.1±4.1b5.3±1.8b
Dentate gyrus12.3 ± 2.642.2 ± 4.6a30.2±1.9a14.9 ± 2.4b1.3 ± 0.7b
Frontal cortex10.9 ± 2.175.8 ± 8.4a68.4 ± 2.9a60.5 ± 6.327.1 ± 3.6b
Occipital cortex16.2 ± 2.162.8 ± 9.4a32.3 ± 2.78a,b12.9 ± 2.9b3.8±1.3b
Mediodorsal thalamic area15.8 ± 2.463.0 ± 9.3a59.8 ± 5.1a35.6 ± 3.27.1±2.3b
Medial pretectal nuclei11.7 ± 1.5137.0±7.5a118.0 ± 3.9a97.2 ± 4.1b65.1±4.0b
Cerebellum19.6 ± 3.042.3±12.519.7 ± 3.25.8±1.8b0.8 ± 0.6b

A rise in the concentration of sodium in the incubation buffer did not induce a further increase in dextromethorphan binding. In fact, binding to the medial mammillary bodies, ventral teg-mental area, dorsal tegmental area, superior colliculi and occipital cortex was significantly reduced when the sodium concentration was increased to 120 mM as compared to binding in the presence of 50 mM sodium. Ca2+ and Mg2+, at the physiological concentration of 2.5 and 1.2 mM, respectively, reduced [H]-dextromethorphan binding in the presence of 50 mM sodium chloride to enhance maximally the sodium-dependent binding. The reduction following Mg2+ was significant in all areas measured, while the same effect after the addition of Ca2+ in the incubation buffer did not reach statistical significance in the dorsal tegmental area, anterior hypothalamus, pons, frontal cortex and mediodorsal thalamic area (Table 1).

Saturation studies showed that the sodium-dependent binding of dextromethorphan to rat brain was saturable with an average Kd value of 14.66 nM (Kd of individual brain areas are presented in Table 2).

Table 2. Autoradiographic saturation analysis of [H]-dextramethorphan binding in discrete areas of rat brain
  1. The tissue sections were incubated with concentrations of [H]-dextromethorphan ranging fron 5 to 80 nM in buffer containing 120 nM sodium chloride. Kd is expressed as nM and Bmax as fmol mg−1 wet weight tissue. Each set of data are from an average of three animals performed in duplicate ± s.e.mean. Data were analysed with GraphPad Prism version 2.01.

Mammillary bodies10.5 ± 4.9286.3 ± 40.1
Ventral tegmental area8.5 ± 5.2257.6 ± 43.5
Dorsal raphe9.7 ± 5.3228.6 ± 36.3
Superior colliculi11.5 ± 5.0204.9 ± 28.0
Central grey area12.2 ± 5.4198.7 ± 28.0
Dorsal tegmental area13.2 ± 5.5212.6 ± 28.9
Anterior hypothalamus11.8 ± 6.0172.9 ± 27.6
Pons13.8 ± 7.087.1 ± 14.8
Hippocampus (CA3 area)13.9 ± 8.177.5 ± 15.1
Dentate gyrus16.5 ± 9.067.2±13.1
Frontal cortex17.3 ± 6.6143.9 ± 19.7
Occipital cortex19.3 ± 9.983.9 ± 16.1
Mediodorsal thalamic area9.3 ± 5.294.9 ± 15.4
Medial pretectal nuclei10.1±5.5219.5 ± 35.4
Cerebellum42.4±18.042.4 ± 20.3

Kd values obtained in different brain areas did not differ significantly (ANOVA test) from each other, and a two-site binding model could not significantly improve the fit to the experimental data.

In the presence of 120 mM NaCl, [H]-dextromethorphan binding followed a distinctive pattern of distribution in rat brain (Figure 2, Table 1). Binding to mammillary nuclei, ventral tegmental area, dorsal raphe, superior colliculi, central grey, anterior hypothalamus and medial pretectal nuclei was higher than 100 fmol mg−1 wet weight of tissue, while the dorsal tegmental area, mediodorsal thalamic nuclei and frontal cortex showed medium levels of binding (50–100 fmol mg−1 wet weight of tissue). Low levels of binding (0–50 fmol mg−1 wet weight of tissue) were detected in the hippocampus, occipital cortex, pons and cerebellum.

Figure 2.

Receptor autoradiographs of [H]-dextromethorphan (a) and [H]-paroxetine binding (b) to parasagittal rat sections. Sections were incubated for 1 h at 4°C with [H]-dextromethorphan or 90 min at 23°C with [H]-paroxetine. Abbreviations: cerebellum (Cer), dorsal raphe (Dr), central gray (Cg), hippocampus (Hip), thalamus (Th), dorsal tegmental area (Dtg), ventral tegmental area (Vtg), medial mammilllary bodies (Mm).

Competition at dextromethorphan binding sites

The σ ligand (+)-PPP, at a concentration of 88 nM, had no effect on dextromethorphan binding (Table 3) in the presence of sodium.

Table 3. Effect of (+)-PPP, phenytoin and paroxetine on sodium-dependent [H]-dextromethorphan binding in discrete regions of rat brain
 Controls(+)-PPP 88 nMPhenytoin 100μMaParoxetine 20 nMa
  1. The tissue sections were incubated with 10 nM [H]-dextromethorphan in the presence of sodium 120 mM, as described in Methods, in buffer containing the displacer indicated above each column. Binding capacity is expressed as fmol mg−1 wet weight of tissue. Each value is the mean of duplicate determinations and represents the mean± s.e.mean of 3–11 rats. aP<0.05 versus controls (ANOVA).

Mammillary bodies136.9 ± 7.9136.4 ± 7.589.9±17.575.8 ± 4.7
Ventral tegmental area154.8 ± 6.4144.3 ± 7.094.7 ± 7.981.9 ± 2.9
Dorsal raphe140.7 ± 8.5126.6 ± 8.1112.1 ± 21.299.6 ± 8.7
Superior colliculi110.8 ± 3.8101.2 ± 5.171.5 ± 4.185.0 ± 3.6
Central grey area103.5 ± 3.2100.6 ± 4.764.1 ± 2.472.0 ± 3.1
Dorsal tegmental area99.9 ± 5.194.4 ± 5.652.5 ± 8.663.3 ± 4.8
Anterior hypothalamus100.2 ± 4.1103.0 ± 4.561.9±10.646.7 ± 4.7
Pons36.4±1.933.2 ± 3.425.4±3.729.3 ± 5.3
Hippocampus (CA3 area)43.6 ± 2.235.7 ± 3.523.±1.927.1 ± 3.2
Dentate gyrus32.1±2.022.4 ± 3.014.7±1.821.5 ± 2.2
Frontal cortex69.0 ± 2.861.2 ± 4.830.9 ± 3.942.6 ± 4.9
Occipital cortex34.2 ± 2.825.8 ± 3.817.2 ± 2.626.8 ± 3.2
Mediodorsal thalamic area60.7 ± 5.050.3 ± 3.248.8 ± 4.135.3 ± 2.5
Medial pretectal nuclei114.0 ± 4.2110.9 ± 4.080.3 ± 3.965.2 ± 5.9
Cerebellum20.3 ± 3.19.8 ± 1.911.9±1.824.8 ± 2.7

In comparison, Na+-dependent dextromethorphan binding was clearly inhibited by the anticonvulsant phenytoin (100 μM). On average, phenytoin reduced dextromethorphan binding to 58.4% of control. This effect was statistically significant with no interaction between the treatments and nuclei (ANOVA). Binding of [H]-dextromethorphan was also reduced by the 5-HT uptake inhibitor paroxetine at a concentration of 20 nM by about 40% (Table 3). Again, the effect was significant with no interaction between treatments and nuclei.

When tested against the sodium-dependent binding of dextromethorphan to rat brain at concentrations of 7.5 and 20.5 nM, respectively, SKF 525-A and GBR 12909 caused a small but statistically significant decrease in dextromethorphan binding to rat brain (Table 4).

Table 4. Effect of cytochrome P-450 inhibitors on [H]-dextromethorphan binding in discrete areas of rat brain
 ControlsSKF 525-A 7.5 nMaGBR 12909 20.5 nMa
  1. Tissue sections were incubated with 10 nM [H]-dextromethorphan in the presence of sodium chloride (120 mM), as described in Methods, in buffer containing the displacer indicated above each column. Binding capacity is expressed as fmol mg−1 wet weight of tissue. Each value is the mean of duplicate determinations and represents the mean± s.e.mean of 3–11 rats. aP<0.05 versus controls (ANOVA).

Mammillary bodies132.8 ± 8.2129.4 ± 8.3130.6 ± 7.6
Ventral tegmental area150.7 ± 6.5145.1±7.1137.4 ± 5.2
Dorsal raphe146.7 ± 6.8131.2 ± 5.2133.6 ± 4.4
Superior colliculi107.1 ± 5.791.1±3.988.3 ± 4.1
Central grey area97.2 ± 3.291.4 ± 5.387.1±2.9
Dorsal tegmental area82.1±8.775.6 ± 4.776.8 ± 5.0
Anterior hypothalamus100.3 ± 7.992.0 ± 2.689.0 ± 3.2
Pons33.4 ± 2.636.4 ± 2.438.2 ± 5.5
Hippocampus (CA3 area)41.4 ± 5.337.3 ± 2.328.5 ± 3.0
Dentate gyrus29.3 ± 4.926.6 ± 2.529.1 ± 5.7
Frontal cortex79.4 ± 5.176.9 ± 2.865.7 ± 5.8
Occipital cortex36.3 ± 5.423.8 ± 3.026.3 ± 3.1
Mediodorsal thalamic area56.2 ± 5.052.8 ± 2.947.1±2.8
Medial pretectal nuclei99.8 ± 5.699.8 ± 3.784.9 ± 6.2
Cerebellum16.9 ± 3.410.9 ± 1.711.0 ± 3.4

Paroxetine binding to rat brain

The binding of [H]-paroxetine to brain sections had a similar pattern of distribution to that of dextromethorphan. [H]-paroxetine binding was high (>90 fmol mg−1 wet weight of tissue) in the mammillary bodies and ventral tegmental area, medium (45–90 fmol mg−1 wet weight of tissue) in the dorsal raphe, superior colliculi and medial pretectal nuclei, and low (<45 fmol mg−1 wet weight of tissue) in all other areas tested (Figure 2 and Table 5).

Table 5. Effect of dextromethorphan on [H]-paroxetine binding in discrete areas of rat brain
NucleusControlsDextromethorphan 20 nMDextromethorphan 40 nMa
  1. The tissue sections were incubated with 1 nM [H]-paroxetine, as described in Methods, in buffer containing dextromethorphan at the concentration indicated above each column. Binding capacity is expressed as fmol mg−1 wet weight of tissue. Each value is the mean of triplicate determinations and represents the mean ± s.e.mean of 3 rats. aP<0.05 versus controls (ANOVA).

Mammillary bodies139.2 ±12.6132.5 ± 15.181.8 ± 8.8
Ventral tegmental area92.5±12.980.0±10.656.0 ± 8.2
Dorsal raphe58.6 ± 8.165.2 ± 8.043.8 ± 6.0
Superior colliculi52.7 ± 5.564.9 ± 5.735.8 ± 4.3
Central grey area29.3 ± 3.343.7 ± 5.727.4 ± 3.6
Dorsal tegmental area43.0 ± 5.740.5 ± 4.528.0 ± 4.53
Anterior hypothalamus40.2 ± 7.328.2 ± 4.418.4 ± 4.1
Pons21.9 ± 4.317.7 ± 3.38.1±2.3
Hippocampus (CA3 area)9.1±2.49.0 ± 2.66.2 ± 2.6
Dentate gyrus4.0±1.53.0 ± 0.93.4 ± 0.6
Frontal cortex12.2 ± 2.110.3 ± 1.616.1±4.4
Occipital cortex4.5 ± 3.17.0 ± 3.13.4±1.1
Mediodorsal thalamic area11.7 ± 3.98.5 ± 2.24.1 ± 1.0
Medial pretectal nuclei52.0 ± 8.546.3 ± 6.239.8 ± 4.4
Cerebellum1.2 ± 0.88.2 ± 3.61.1±0.8

When dextromethorphan was present in the incubation buffer, it reduced the binding of [H]-paroxetine to 64.9% of controls. This effect was statistically significant at 40 nM with no interaction between treatment and nuclei.

The correlation plot between the distributions of dextromethorphan and paroxetine in rat brain is shown in Figure 3. Data points are best described by a straight line (correlation was significant, r = 0.922). The slope of the line differed significantly from zero (P < 0.0001), revealing a non-casual relationship between the distributions. The intercept of the line on the abscissae may suggest the presence of a component of dextromethorphan binding that is not shared by paroxetine.

Figure 3.

Correlation between dextromethorphan and paroxetine binding in rat brain. Corresponding values of [H]-dextromethorphan and [H]-paroxetine binding measured in different brain areas were plotted on the × and Y axis, respectively. Linear regression analysis gave a straight line (r = 0.992, P< 0.0001) whose slope was significantly different from zero. Dashed lines represents 95% confidence intervals.


The present work provides autoradiographical evidence for the existence of a sodium-dependent fraction of dextromethorphan binding to rat brain. This sodium-dependent binding, characterized for the first time in this study, constitutes the majority of dextromethorphan binding to rat brain, and shares most of the pharmacological properties with the DM2 site identified in previous studies (Zhou et al., 1991; Klein & Musacchio, 1992).

In particular, the sodium-dependent binding was unaffected by a concentration of (+)-PPP equal to 4 times the Kd value for the common DM1/σ1 site in the rat (Klein & Musacchio, 1992), and it was only minimally affected (8 and 11% reduction respectively) by concentrations of SKF 525-A and GBR 12909 equal to three times their Kd for the DM1/σ1 site (Klein et al., 1991). Dextromethorphan binding in guinea-pig brain has been shown to be reduced by about 50% by these two cytochrome P450 inhibitors (Klein et al., 1991).

The present study showed for the sodium-dependent binding of dextromethorphan an average Kd of 14.7 nM; that is appreciably lower than the Kd for the DM1/σ1 site in rat brain (47 nM, Klein & Musacchio, 1992). This difference in Kd between the DM1/σ1 and the DM2 site might also be reflected in the higher value of Kd in the cerebellum. The cerebellum was the only area in which dextromethorphan binding was not affected by sodium, and could therefore present relatively high levels of DM1/σ1 binding.

Phenytoin significantly decreased the sodium-dependent binding of dextromethorphan. This anticonvulsant was shown to interact with the voltage-dependent sodium channel in rat brain (Francis & McIntyre-Burnahm, 1992), and its effect on sodium-dependent binding might be either the result of a direct competition with dextromethorphan or an indirect effect following the action on the sodium channel.

The proposed identity of the sodium-dependent dextromethorphan binding with 5-HT uptake and the requirement of external sodium for this process seem to suggest a direct interaction of phenytoin with the dextromethorphan binding site or to a connected allosteric site.

Previous studies on the distribution and characteristics of dextromethorphan binding to guinea-pig and rat brain have all been conducted in sodium phosphate 50 mM (Musacchio et al., 1988; Klein & Musacchio, 1992), and therefore reflect the properties of both the sodium-dependent and non-dependent binding site.

A specific effect of sodium following the addition of sodium chloride to the incubation buffer is supported by the analogous effect on binding exerted by sodium phosphate (data not shown).

This effect of sodium led us to investigate the possible interaction of dextromethorphan with a sodium-dependent uptake mechanism. The particularly high levels of sodium-dependent dextromethorphan binding in the medial mammil-lary, ventral tegmental area and dorsal raphe prompted us to examine the relationship between dextromethorphan binding and the 5-HT uptake mechanism.

The significant colocalization between dextromethorphan binding and the selective 5-HT uptake inhibitor paroxetine, and their mutual displacement strongly suggest an interaction of dextromethorphan at this site. Competition of these two ligands at a metabolic site is very unlikely, since the two cy-tochrome P450 inhibitors did not affect the majority of binding, and this enzyme is implicated in the main metabolic pathway of both drugs.

Analysis of the dextromethorphan saturation curve suggests binding to one site only. This conclusion is contradicted by evidence emerging from the present study such as the residual binding in the absence of sodium, that follows a different pattern of distribution, and the heterogeneous effect of different pharmacological agents in different brain areas.

Binding to the cerebellum for example, was not potentiated by sodium ions, was reduced by (+)-PPP and paroxetine had no effect on the levels of binding. The cerebellum could therefore express particularly high levels of sodium-independent binding, that appears to have the same pharmacological profile as the DM1/σ1 site.

This heterogeneity in the modulation of dextromethorphan binding to rat brain revealed in this study suggests an unequal distribution of the sodium-dependent and non-dependent binding sites, as confirmed by the two different patterns of distribution for [H]-dextromethorphan obtained in the absence and presence of sodium.

The binding properties in different brain areas therefore appeared to be determined by the relative proportion of sodium-dependent binding, which was influenced by paroxetine, and sodium-independent binding which corresponded to the previously identified DM1/σ1 site.

Since σ ligands have never been shown to possess any an-titussive activity, the sodium-dependent site identified in this study as the DM2 site may represent a good candidate for a role in cough modulation, and its possible identity with 5-HT uptake may provide an important clue to the mechanism of action of the antitussive effect of dextromethorphan.

5-HT has been demonstrated to play a key role in the antitussive mechanism of morphine, dihydrocodeine and dextromethorphan (Kamei et al., 1987; 1988), and a further study showed an increase in the release of 5-HT from the nucleus of the solitary tract in the rat following dextromethorphan administration (Kamei et al., 1992). Since this nucleus represents an important relay in the modulation of cough, a local effect of dextromethorphan through DM2 sites could be implicated.

Other evidence for the possible functional effects of the interaction of dextromethorphan with 5-HT uptake have appeared in the literature: Ahtee (1975) demonstrated inhibition of 5-HT uptake in human platelets by nanomolar concentrations of the antitussive agent, and Finnegan et al. (1991) demonstrated a protective effect of dextromethorphan on the acute depletion of 5-HT produced by the administration of p-chloroamphetamine in rats.

An action of dextromethorphan at 5-HT uptake sites in man is also supported by evidence of adverse reactions witnessed by clinicians following the coadministration of dextromethorphan and monoamine oxidase (MAO) inhibitors (Sovner & Wolfe, 1988; Nierenberg & Semprebon, 1993). The symptoms described in these reports correspond to the central nervous system 5-HT syndrome, which is a condition resulting from the interaction of drugs which enhance 5-hydroxy-tryptaminergic tone at receptors in the central nervous system. The same effect following the administration of MAO inhibitors and dextromethorphan has been observed in rabbits (Sinclair, 1976) and ascribed to the inhibition of 5-HT uptake (Sinclair & Lo, 1976).

Further studies are required to establish unequivocally the identity of the sodium-dependent dextromethorphan high affinity binding site with the 5-HT uptake mechanism, but the commonality between a subpopulation of dextromethorphan binding sites and paroxetine binding might lead the way to new possible applications for 5-HT uptake inhibitors and dextromethorphan.