Materials and Methods

Cell Culture.

COS-7 cells (SV40-transformed African green monkey kidney cells; American Type Culture Collection, Bethesda, MD) were grown at 37°C in a 5% CO₂, humidified atmosphere on standard plastic cultureware in Dulbecco’s modified Eagle’s medium (DMEM; GIBCO BRL Life Technologies, Gaithersburg, MD) supplemented with 10% fetal calf serum (GIBCO BRL) and 100 U/ml penicillin and 100 μg/ml streptomycin (GIBCO BRL) to obtain complete DMEM. When the cells were confluent, subcultures were made using 0.25% trypsin (Sigma Chemical Co., St. Louis, MO) at a ratio of 1:7 into 12-well plates, 2 days before transfection. The cells were transiently transfected with the cDNA encoding either the rat NET (in the pEUK-C1 vector), human NET (in the pEUK-C1 vector), or bovine NET (in the pSG5 vector) or the pEUK-C1 vector (without insert) using Lipofectamine reagent (GIBCO BRL). Cloning of the transporter cDNAs into the pSG5 and pEUK-C1 eukaryotic expression vectors has been described previously (Lingen et al., 1994; Brüss et al., 1997). The transfection mixture containing 200 ng of cDNA and 2 μl of Lipofectamine in 500 μl of Opti-MEM (GIBCO BRL) was added to each well and incubated at 37°C in a 5% CO₂, humidified atmosphere for 7 h. Complete DMEM supplemented with an additional 10% fetal calf serum was then added to each well (500 μl). The transfection solution was discarded 17 h later and replaced with complete DMEM. Experiments were performed 24 h later.

Buffer for Incubation Experiments.

The Krebs/HEPES buffer used in the experiments (unless otherwise indicated) contained 125 mM NaCl, 4.8 mM KCl, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 1.3 mM CaCl₂, 25 mM HEPES, 5.55 mMd-(+)-glucose, 1.02 mM ascorbic acid, 10 μM U-0521 [to inhibit catechol-O-methyltransferase (S-adenosyl-l-methionine:catechol-O-methyltransferase; EC 2.1.1.6)], and 100 μM pargyline [to inhibit monoamine oxidase (amine:oxygen oxidoreductase [deaminating] [flavin containing]; EC1.4.3.4)], and the pH was adjusted to 7.4 with NaOH (which added a further 12.5 mM Na⁺). In experiments on the kinetics of the sodium dependence of the NETs, the NaCl concentration in the Krebs/HEPES buffer was 10, 20, 40, 80, or 160 mM, with 150, 140, 120, 80, or 0 mM LiCl added to maintain the same concentration of Cl⁻ in each solution, and the pH was adjusted to 7.4 with Tris.

Norepinephrine Uptake Assays.

DMEM was removed from the cells, and they were washed twice with 1 ml of Krebs/HEPES buffer containing 0.1% BSA at 37°C. The same buffer, where necessary containing 1 μM nisoxetine or other drugs or ionic changes as indicated, was then added to each well (1 ml) for 15 min at 37°C. The cells were then incubated for exactly 2 min at 37°C with solutions of the same composition but with 10 nM [³H]norepinephrine added to determine initial rates of norepinephrine uptake into the cells. The incubation solution was then rapidly removed, and the cells were immediately washed three times with 2 ml of ice-cold Krebs/HEPES buffer to terminate uptake and remove extracellular amine. The cells were lysed by incubation with 1 ml of 0.1% Triton X-100 in 10 mM Tris · HCl, pH 7.5, for 60 min at 37°C. Determination of the protein content by the Lowry method (Lowry et al., 1951) was carried out on 100 μl of the lysate, and the³H content of 800 μl of the lysate was determined by the addition of 2 ml of Starscint scintillation medium (Packard, Melbourne, Australia) and liquid scintillation counting.

Drugs and Solutions.

Citalopram hydrobromide was obtained from Lundbeck (Copenhagen-Valby, Denmark). Cocaine hydrochloride was purchased from Drug Houses of Australia (Sydney, Australia). Desipramine hydrochloride, (−)-epinephrine bitartrate, imipramine hydrochloride, (−)-norepinephrine bitartrate, and pargyline hydrochloride were obtained from Sigma Chemical Co. U-0521 (3′,4′-dihydroxy-2-methylpropiophenone) was obtained from Upjohn Pty. Ltd. (Kalamazoo, MI). Fluoxetine hydrochloride and nisoxetine hydrochloride were purchased from Lilly Research Laboratories (Indianapolis, IN). GBR 12909 (1-[2-[bis(4-fluorophenyl)methoxy]ethyl]-4-(3-phenylpropyl)piperazine dihydrochloride) and MPP⁺ iodide were purchased from Research Biochemicals International (Natick, MA). (+)-Oxaprotiline hydrochloride was obtained from Ciba-Geigy Ltd. (Basel, Switzerland). Paroxetine hydrochloride was obtained from SmithKline Beecham Pharmaceuticals (Worthing, West Sussex, UK). The radiolabeled norepinephrine used in the experiments was [ring-2,5,6-³H](−)-norepinephrine (specific activity, 2100 Bq/pmol; NEN Life Science Products, Boston, MA). Stock solutions of norepinephrine, epinephrine, and dopamine (10 mM) were prepared in 10 mM HCl, and stock solutions of the other drugs were prepared in water as 10 mM solutions for all drugs except U-0521 and GBR 12909, which were prepared as 1 mM and 100 μM solutions, respectively. All dilutions were prepared on the day of the experiment in Krebs/HEPES buffer.

Calculation of Results.

The results of liquid scintillation counting and protein determinations were used to calculate the [³H]norepinephrine uptake into the cells, expressed as fmol/mg protein. In each experiment, the mean result from duplicate wells for each treatment was used. Specific uptake was calculated as the difference between uptake of [³H]norepinephrine in the absence (total uptake) and presence (nonspecific uptake) of 1 μM nisoxetine for each plate. The n values shown represent the number of different experiments on separate plates for each treatment.K_m values of norepinephrine and of Na⁺ for norepinephrine uptake by the NETs were calculated from nonlinear regression analysis of the data for each individual experiment according to a hyperbolic model. IC₅₀ values for inhibition of [³H]norepinephrine uptake by the drugs used in the study were calculated from nonlinear regression analysis of percent inhibition of specific uptake versus log drug concentration data for each individual experiment according to a sigmoidal model and were used to calculate the K_i value for each drug in each experiment (Cheng and Prusoff, 1973). pK_m and pK_i values were calculated as the negative log of the corresponding K_mand K_i values expressed in molar concentrations. Arithmetic mean ± S.E.M. or geometric mean values and 95% confidence limits were calculated as indicated inResults. The significance of differences between treatment groups (e.g., inhibitors or species) was determined by the Student’s or paired t test or by one-factor repeated measures ANOVA (because data for each treatment group or species were included in each experiment) followed by Tukey-Kramer post hoc t tests as indicated in Results. All data were analyzed using Prism 2 software (GraphPAD Software, San Diego, CA).

Results

Functional Properties of Rat NET.

When COS-7 cells were transfected with the human NET cDNA or the rat NET cDNA that we recently cloned from PC12 cells (Brüss et al., 1997), there was marked uptake of [³H]norepinephrine (969 ± 95 fmol/mg protein, n = 3, and 1159 ± 146 fmol/mg protein, n = 3, respectively) that was inhibited in both cases by 98% in the presence of 1 μM concentration of the NET inhibitor nisoxetine (P < .001). This confirmed the functional expression of the NETs in our system. On the other hand, control transfection of COS-7 cells with the vector pEUK-C1 and subsequent incubation with [³H]norepinephrine resulted in only a very small amount of norepinephrine accumulation (10.9 ± 1.71 fmol/mg protein, n= 3) that was not significantly affected by 1 μM nisoxetine.

In another series of experiments, the effects of several NET- and SERT-selective inhibitors (at a concentration of 100 nM) on norepinephrine uptake in cells transfected with the rat NET cDNA (2492 ± 224 fmol/mg protein, n = 3) were determined to establish whether the recently cloned rat NET displayed an inhibitor profile typical of a NET. The selective NET inhibitors nisoxetine and (+)-oxaprotiline caused almost complete inhibition of norepinephrine uptake (93 and 92% inhibition, respectively,n = 3, P < .001). With the selective SERT inhibitors, citalopram had no effect, and in accordance with previous findings (Shank et al., 1987), paroxetine caused a small inhibitory effect on norepinephrine uptake by the rat NET (33% inhibition, n = 3, P < .05). In a further series of experiments, the selective DAT inhibitor GBR 12909 (100 nM) did not significantly affect norepinephrine uptake (23% decrease compared with uptake in the control without GBR 12909 of 2661 ± 345 fmol/mg protein, n = 4,P > .05). These data confirm that the transporter that we cloned from rat PC12 cells is a typical NET.

Pharmacological Comparison of Rat, Human, and Bovine NETs.

Sequence alignment of the rat, human, and bovine NETs (Fig.1) showed that there are five amino acid exchanges that lead to changes in the local charge distribution in the rat sequence compared with the species variants. To investigate the potential effects of these amino acid exchanges, a comparison was made of the kinetics of norepinephrine uptake by the rat, human, and bovine NETs by incubating COS-7 cells expressing each of the NETs with 10 nM or 0.3, 1, 3, or 10 μM [³H]norepinephrine. Uptake of [³H]norepinephrine by the NETs was saturable (Fig. 2A), and the results of the kinetic analyses are shown in Table1. The Hill coefficients for all three NETs were not significantly different from 1 (P > .05, Student’s t test). The pK_mvalue of norepinephrine for the human NET was significantly greater than that for the rat NET (Table 1), indicating a lower affinity of norepinephrine for the rat NET than for the human NET. TheV_max values for norepinephrine transport were 188 ± 22.5, 95.4 ± 8.62, and 141 ± 17.7 pmol/mg protein/min (n = 4) for the rat, human, and bovine NETs, respectively. However, the meaning of differences in these V_max values is unclear because they are dependent not only on the transporter but also on the vector and transfection efficiency.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

A diagrammatic representation of the topology of the rat NET (Brüss et al., 1997). ●, the 26 amino acid residues that are divergent in rat NET but are conserved in the human and bovine NETs. Of the 13 nonconservative exchanges included in these 26 exchanges, the 5 that predict striking alterations in local charge distribution are shown at positions 7 (rat, Lys; human and bovine, Asn), 62 (rat, Glu; human and bovine, Lys), 198 (rat, Asp; human and bovine, Asn), 375 (rat, Lys; human and bovine, Asn), and 612 (rat, Arg; human and bovine, Gln). In addition, the exchange at position 198 results in loss of one N-glycosylation site in rat NET compared with human and bovine NET (N-glycosylation site; carbohydrate group).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Saturation of specific uptake of [³H]norepinephrine uptake with increasing concentrations of [³H]norepinephrine (A) or increasing Na⁺ion concentrations (LiCl replacing NaCl) with a constant [³H]norepinephrine concentration of 10 nM (B) in COS-7 cells expressing the rat (●), human (▴), or bovine (▪) NET. Values are arithmetic mean ± S.E.M. (n = 4) of the rate of specific uptake of [³H]norepinephrine calculated as the difference between uptake in the absence and the presence of 1 μM nisoxetine divided by the incubation time of 2 min. Curves were obtained by nonlinear regression analysis according to a hyperbolic model.

K_i values for inhibition of norepinephrine uptake by the three NETs were determined for the three NET substrates dopamine (0.1, 0.3, 1, and 3 μM), MPP⁺ (0.1, 0.3, 1, and 3 μM), and (−)-epinephrine (1, 3, 10, and 30 μM) and for the four inhibitors nisoxetine (1, 3, 10, and 30 nM), desipramine (1, 3, 10, and 30 nM), imipramine (10, 30, 100, and 300 nM), and cocaine (60, 200, 600, and 2000 nM). There were no significant differences between the pK_i values of dopamine for the three NETs, but the pK_i values for MPP⁺ were greater for human NET than for rat NET and the pK_i values for epinephrine were greater for human and bovine NETs than for rat NET (Table2). It can be concluded that, as found above for norepinephrine, MPP⁺ and epinephrine, but not dopamine, have lower affinities for rat NET than for human NET, and epinephrine also has a lower affinity for rat NET than for bovine NET. For the inhibitors tested, the pK_i value of cocaine was greater for human and bovine NETs than for rat NET (Table 2), but there were no significant differences between the three NETs for the pK_i values of nisoxetine, desipramine, or imipramine (Table 2). Hence, cocaine has a lower affinity for the rat NET than for the human or bovine NETs, and there were no species differences for the other inhibitors tested. As shown in Fig.3, there is a marked correlation between the pK_i values for inhibition of rat NET by the compounds and data obtained previously for inhibition of uptake of norepinephrine in rat PC12 cells (Bönisch and Harder, 1986), indicating that the properties of the rat NET were not changed when expressed in COS-7 cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Correlation between pK_ivalues of substrates and inhibitors for inhibition of [³H]norepinephrine uptake in COS-7 cells expressing the rat NET and corresponding values determined in rat PC12 cells (Bönisch and Harder, 1986). The correlation coefficient (r) obtained by linear regression analyses of the data was 0.982 (P < .01, df = 4) with a slope of 0.90 ± 0.09, which is not significantly different from 1 (Student’s t test). The dotted lines represent the 95% confidence limits of the line of best fit.

The pharmacological properties of the human, bovine, and rat NETs were further compared by determining the dependence of their activities on Na⁺ ion concentration. When increasing concentrations of Na⁺ from 10 to 160 mM were used to stimulate the activity of the NETs at a constant nonsaturating concentration of norepinephrine (10 nM), the rate of norepinephrine uptake increased and the effect was saturable (Fig. 2B). Analysis of the data showed that the Hill slopes were close to 1 (Table3), indicating a stoichiometry of 1:1 for transport of norepinephrine and Na⁺ by all three NETs. The pK_m values of Na⁺ for stimulation of norepinephrine uptake (Table 3) were significantly lower for bovine NET than for rat NET, whereas no differences were found for the corresponding values between rat and human NETs (Table 3).

Discussion

One aim of the study was to characterize in greater detail the functional properties of the rat NET we recently cloned from rat PC12 cells (Brüss et al., 1997). In transiently transfected COS-7 cells, [³H]norepinephrine uptake by rat NET was inhibited by selective NET inhibitors and not to any marked extent by selective SERT or DAT inhibitors; theK_m value of norepinephrine uptake and the K_i values of NET substrates and inhibitors were correlated with previous values in rat PC12 cells (Fig.3), and the K_m of Na⁺ for stimulation of norepinephrine uptake was comparable with that obtained previously in rat PC12 cells (Friedrich and Bönisch, 1986). Hence, the typical properties of the PC12 cell rat NET are not changed when this transporter is expressed in non-neuronal cells such as COS-7 cells.

The main aim of the study was to examine species differences in the pharmacological properties between the rat NET and the previously cloned human and bovine NETs. In a study on SERT, tricyclic antidepressants were less potent, amphetamine was more potent, and nontricyclic inhibitors showed no differences in potency for rat compared with human SERT (Barker et al., 1994). Furthermore, for DATs expressed in COS-7 cells, binding of a cocaine analog and MPP⁺ uptake decreased in the order human > rat > bovine (Lee et al., 1996). These results suggest that small interspecies differences in amino acid sequences of the monoamine transporters are sufficient to result in significant functional differences. The rat NET shows 93 and 91% amino acid identity to its human and bovine counterparts, respectively (Brüss et al., 1997). Thus, we examined whether these small differences in amino acid sequence might be sufficient to cause differences in their functional and pharmacological properties.

The rat, human, and bovine NETs expressed in COS-7 cells showed, as expected, Na⁺-dependent and saturable norepinephrine uptake that was inhibited by NET substrates and inhibitors. The K_m value of norepinephrine and the K_i values of the substrates and inhibitors were comparable with those obtained in previous studies with the human (Pacholczyk et al., 1991; Pifl et al., 1996) and bovine (Lingen et al., 1994) NETs, and there were highly significant correlations between the values obtained for rat NET in this study and those obtained previously in rat PC12 cells (Bönisch and Harder, 1986; Fig. 3) and perfused lungs (Bryan-Lluka and O’Donnell, 1992; Paczkowski et al., 1996;r = 0.888, P < .01).

Experiments with a range of Na⁺ concentrations showed that transport of norepinephrine by the three NETs is clearly Na⁺-dependent and that the rat NET exhibits a higher affinity for Na⁺ than the bovine NET. The Hill coefficient of about 1 for the Na⁺concentration dependence indicates that a single Na⁺ ion is involved in transport by each of the three NETs. These results are in accordance with previous studies on the rat NET in PC12 cells (Friedrich and Bönisch, 1986) and on the heterologously expressed human NET (Gu et al., 1994).

The interspecies comparison of the pharmacological properties of the three NETs showed that the K_m orK_i values of three of the four substrates tested (norepinephrine, epinephrine, and MPP⁺) were greater, and hence their affinities were less, for the rat NET than the human NET, to a small but significant extent, but the K_i value of dopamine showed no species differences. TheK_i values of the NET inhibitors nisoxetine, desipramine, and imipramine did not show any species differences, but the K_i value of cocaine was greater for the rat NET than for the human and bovine NETs, indicating a higher affinity of cocaine for the latter two NETs. It should be noted that the experimental design of the present study, in which strictly parallel experiments included each of the three NETs, provided a very sensitive experimental paradigm for the detection of small differences in the affinities of the compounds for the transporters. This was further aided by the high level of reproducibility (with small variances) of the pK_m and pK_i values presented in Tables 1 to 3.

The slightly higher affinities of the substrates tested (except dopamine) for human NET than for rat NET contrasts with comparative data for SERT where substrate affinities generally did not differ between rat and human, except in the case of amphetamine, where rat SERT had the higher affinity (Barker et al., 1994). For inhibitors, there also were differences between the results obtained in the present study for NET, where cocaine was the only inhibitor to show any species differences in affinities (higher affinity for human and bovine than for rat NETs), and the previous study with SERT, where tricyclic antidepressants had markedly higher affinities for human than for rat SERT and there was no difference for cocaine (Barker et al., 1994). In the present study, cocaine exhibited a significantly lower affinity for the rat NET than for the bovine and human NETs. A similar species difference in the affinity of cocaine was also shown previously for the DAT, where the affinity of cocaine was lower for rat DAT than for human DAT (Giros et al., 1992). These differing species-dependent effects for cocaine compared with other inhibitors are compatible with the circumstantial evidence from NET/DAT chimera studies (Giros et al., 1994; Buck and Amara, 1995) and site-directed mutagenesis studies on the DAT (Kitayama et al., 1992, 1993) that the binding sites on the monoamine transporters for cocaine and its analogs are different from those for other inhibitors, such as the tricyclic antidepressants.

For species homologs of receptors, relatively conservative mutations close to the ligand-binding site have been shown to have marked effects on ligand affinity, such as rat and human neurokinin NK₁ receptors (Pradier et al., 1995) and human and rat 5-hydroxytryptamine_1B receptors (Oksenberg et al., 1992). The three NETs compared in this study can be regarded as naturally occurring variants of a common functional motif. Alignment of the amino acid sequences of the rat, human, and bovine NETs showed that of the amino acids that are divergent in rat NET but are conserved between the human and bovine NETs (Fig. 1), five of them in particular predict striking alterations in local charge distribution and occur at positions that are conserved among the other cloned NETs (Fig. 1). In addition, one of these exchanges results in loss of one putative N-glycosylation site in rat NET compared with the human and bovine NETs (Fig. 1). Interestingly, the mouse NET (Fritz et al., 1998) is identical with the rat NET at this position but identical with the human and bovine transporters at the other four positions. It is not possible to speculate from the data obtained in the present study, even in combination with previous reports on chimeras of NET and DAT (Buck and Amara, 1994, 1995; Giros et al., 1994), as to the particular amino acid exchanges that may contribute to the small, but significant, species differences found in the affinities of substrates, cocaine, and Na⁺ for the NETs.

In conclusion, we confirmed that the recently cloned rat NET (Brüss et al., 1997) shows the typical functional properties of a NET. The fact that the affinities for some substrates, cocaine and Na⁺ exhibited small, but significant, interspecies differences among the rat, human, and bovine NETs suggests that ligand recognition, the translocation process, and Na⁺ dependence are influenced differentially by just a few amino acid exchanges in the primary sequences of the transporters. Site-directed mutagenesis studies should clarify whether the amino acid differences between the NETs are responsible for the observed small functional differences in the pharmacological properties among the rat, human, and bovine NETs. On the other hand, the lack of any major differences in the pharmacological properties of the rat, human, and bovine NETs in this study suggest that data obtained in previous studies on rat tissues and bovine cells should reflect, in all but the most quantitative analyses, the properties of the human NET.