Synthesis and biological investigations of 3β-aminotropane arylamide derivatives with atypical antipsychotic profile

This work is a continuation of our previous research, concentrating this time on lead structure modification to increase the 5-HT1A receptor affinity and water solubility of designed compounds. Therefore, the compounds synthesised within the present project included structural analogues of 3β-acylamine derivatives of tropane with the introduction of a methyl substituent in the benzyl ring and a 2-quinoline, 3-quinoline, or 6-quinoline moiety. A series of novel 3β-aminotropane derivatives was evaluated for their affinity for 5-HT1A, 5-HT2A, and D2 receptors, which allowed for the identification of compounds 12e, 12i, and 19a as ligands with highest affinity for the tested receptors; they were then subjected to further evaluation in preliminary in vivo studies. Selected compounds 12i and 19a displayed antipsychotic properties in the d-amphetamine-induced and MK-801-induced hyperlocomotor activity test in mice. Moreover, compound 19a showed significant antidepressant-like activity in the forced swim test in mice.


Introduction
The drug files for typical and atypical, including the latest, antipsychotics (aripiprazole, brexpiprazole, cariprazine) on FDA's Accessdata drug data base website contain the information that their mechanism of action is unknown (FDA 2008a(FDA , 2013b(FDA , 2014c(FDA , 2014d. However, in vitro studies show that all antipsychotic drugs bind to D 2 receptors and the dosage correlates with the strength of affinity for these receptors. Dosage has not been shown to correlate with affinity for non-D 2 receptors (Rzewuska 2009). First-generation, or typical, antipsychotics, such as chlorpromazine or haloperidol are antagonists of D 2 receptors, while second-generation, or atypical, antipsychotics (clozapine, olanzapine, risperidone) are described as antagonists of 5-HT 2A /D 2 , the antagonistic effect on 5-HT 2A receptors being greater than that against D 2 receptors (Meltzer 2013;Möller et al. 2015). The three newest antipsychotic drugs listed above are characterised as being partial agonists of D 2 receptors (Citrome 2015;Stahl 2015Stahl , 2016Frankel and Schwartz 2017).
These findings indicate that inhibition of dopaminergic transmission appears to be fundamentally important in the treatment of symptoms of schizophrenia.
The dopaminergic hypothesis also furnishes the best known explanation of the relation between neurochemical factors and the clinical manifestations of schizophrenia. Historically, the hypothesis was made increasingly more precise in three stages (Howes and Kapur 2009). The first concept (I), finally formulated in the 1970's, was called the dopamine receptor hypothesis. The main focus was on dopaminergic hyperactivity, the control of which via blocking dopamine receptors was supposed to provide effective therapy (Creese et al. 1976). Thus, it was a very general idea that did not account for differences between individual dimensions of schizophrenia (e.g., classification of symptoms), risk factors or the relation between different levels of dopamine noted in various areas within the CNS and the clinical expression of symptoms (Snyder 1976). The second concept (II) of 1991 already considered differences in dopaminergic activity between individual regions of the cerebrum, also with regard to various subtypes of the receptor (D 1 vs. D 2 ) (Davis et al. 1991). The main assumption of that concept was frontal dopaminergic hypoactivity resulting in striatal dopaminergic hyperactivity. The negative symptoms of schizophrenia were explained in terms of inadequate dopaminergic transmission in the frontal cortical areas and the positive symptoms were attributed to increased dopaminergic transmission in subcortical nuclei. This concept also had its weaknesses such as the absence of direct evidence at that time for low cortical dopamine levels, ignoring the fact that cortical phenomena are more complex than "hypofrontality" alone, lack of a model linking those abnormalities with clinical phenomena (e.g., the link between dopaminergic hyperactivity and delusions) or failure to include the aetiology of dopaminergic imbalance in schizophrenia (Davis et al. 1991;Tsoi et al. 2008;Howes and Kapur 2009). The third concept (III) (2009) is based on four core assumptions: • risk factors for schizophrenia are linked with one another and result in dysregulation of dopaminergic neurotransmission. Regardless of the cause, it is this dysregulation that is the starting point for the development of psychosis in schizophrenia; • dysregulation of dopaminergic transmission occurs at the level of presynaptic control rather than, as was previously believed, postsynaptic D 2 receptors; • dopaminergic dysregulation may be associated with psychosis rather than schizophrenia, and, for most of its duration, with susceptibility to psychosis. An individual's diagnosis may be related to the type of factors that have produced the psychosis and to socio-cultural conditions; • dopamine dysregulation may alter an individual's assessment of stimuli, perhaps as a result of impaired salience (Howes and Kapur 2009).
The first assumption refers to factors associated with the prenatal period and early childhood, the environment and genetic determinants. It assumes that these are the factors that dopaminergic dysfunction, an abnormality consisting in increased striatal levels of dopamine, is associated with (Haleem 2015;Howes et al. 2017). This assumption changes the approach to antipsychotic therapy and is based on observations indicating that currently available antipsychotics do not treat the underlying abnormalities. Antipsychotics influences the postsynaptic effects of abnormal dopamine release, while the actual problem occurs at an earlier (presynaptic) stage (second assumption). The third and fourth assumptions address psychosis as a salience syndrome. Dopaminergic dysregulation in schizophrenia is viewed here as an extremely significant, but not the only, component contributing to onset of clinical symptoms. Changes in many neuronal and neurotransmitter systems, combined with other biological and environmental factors, lead to dopaminergic hyperactivity in the striatum. It can be said that dopaminergic dysregulation that has reached a certain level of severity, combined with corresponding clinical phenomena, such as delusions and hallucinations, leads to a diagnosis of psychosis and/or schizophrenia. Other neurotransmitters should also be considered besides dopamine (Howes and Kapur 2009). In this regard, significant interaction is hypothesised to take place between dopaminergic and serotonergic pathways. Influence on serotonergic receptors may be the underlying cause of cognitive disturbances and negative symptoms seen in psychoses and mood disorders (Meltzer and Massey 2011;Sumiyoshi et al. 2014). Clozapine, the pioneering atypical antipsychotic with known efficacy against negative symptoms, acts as a 5-HT 2A receptor antagonist (Meltzer and Massey 2011). It demonstrates a much lower affinity for the D 2 receptor compared to classical neuroleptics. Similar properties are exhibited by other atypical antipsychotics, namely olanzapine, risperidone, zotepine, sertindole, quetiapine or ziprasidone, whose discovery was greatly influenced by Meltzer et al.'s hypothesis that compounds of this class should be characterised by a particular 5-HT 2A /D 2 pK i ratio (Meltzer's Index) (Meltzer et al. 1989;Meltzer and Huang 2008;Meltzer and Massey 2011). Cortical 5-HT 2A receptors may play a key role in the development of psychoses via modulating intracortical and cortico-subcortical glutamatergic transmission (Meltzer and Huang 2008). Several new antipsychotic and antidepressant medications (cariprazine, brexpiprazole, quetiapine) reduce the severity of negative symptoms also partly via 5-HT 1A receptors, while producing milder extrapyramidal symptoms (Newman-Tancredi and Albert 2012; Sumiyoshi et al. 2014;Haleem 2015). Taking advantage of action at this receptor may positively influence patients' motivation (Neves et al. 2010).
Our previous publications have described the synthesis and biological activity studies of new derivatives of 3βaminotropane (Słowiński et al. 2011(Słowiński et al. , 2013Stefanowicz et al. 2016). Some of these derivatives demonstrate high activity at the D 2 , 5HT 1A , and 5-HT 2A receptors (Fig. 1).
The aim of our work is to continue the search for new compounds with antipsychotic potential in this group of derivatives. We decided to synthesise and analyse analogues of the above structures containing an additional nitrogen atom in the molecule (Fig. 2). The addition of a nitrogen atom will certainly influence their biological activity and will make it possible to study electron effects on binding affinity for selected molecular targets, while also improving water solubility of salts of these new compounds as salts of the exemplary structures ( Fig. 1) are characterised by very poor solubility.
The additional nitrogen atom is placed in a naphthalene system, producing quinoline and isoquinoline derivatives, or/and in a phenyl ring, producing pyridine derivatives, or in the form of an amine moiety, as a substituent of the phenyl ring (Fig. 2). The linking position for the quinoline and isoquinoline systems remains the same as in 2naphthalene derivatives as this configuration generates much better affinity for the D 2 , 5-HT 1A , and 5-HT 2A receptors than in 1-naphthalene analogues (Słowiński et al. 2011;Zajdel et al. 2012). We are proceeding with synthesis of equatorial isomers (β) only as they possess much better affinity for the receptors of interest than their axial (α) analogues (Słowiński et al. 2011;Stefanowicz et al. 2016).

General remarks
All solvents and raw materials were purchased from commercial sources. Column chromatography was carried out using a Merck Silica gel 60 A (63-200 µm) column as the stationary phase and chloroform:methanol (9:1 v/v) as eluent. Melting points were determined on an Electrothermal 9100 apparatus with open capillary tubes and were uncorrected. IR spectra were obtained using a Shimadzu FTIR-8300 spectrometer. NMR spectra were recorded on a Varian Inova 500 (500 MHz) spectrometer. Chemical shifts (δ) were expressed in parts per million (ppm) relative to tetramethylsilane used as the internal reference. The following abbreviations are used to describe peak patterns when appropriate: s (singlet), bs (broad singlet), d (doublet), dd (double doublet), t (triplet), td (triple doublet), pt (pseudo triplet), 4d (quartet of doublets), m (multiplet), q (quartet), qu (quintet), * (overlapping signals). Coupling constants (J) are in hertz (Hz). ESI-HRMS spectra were obtained on an LCT TOF (Micromass) instrument. Intermediate 8 and 9 (Scheme 1) was obtained following the protocol in Ref. (Słowiński et al. 2011;Dostert et al. 1984). Intermediate 16 and 17 (Scheme 2) was obtained following the protocol in Ref. (Stefanowicz et al. 2016;Dostert et al. 1984) (see Supplementary material). The 1 H NMR spectra of all considered final compounds are presented in Supplementary material. The purity of the tested compounds was determined and was higher than 95% (Fig. 3).
Compound 11c was synthesised with minor modification of general procedure described above.
A solution of 0.003 mol of 10c, 1.1 mL of 10% H 2 SO 4 and 17.2 mL H 2 O was refluxed with stirring for 4 h. The reaction mixture was cooled to room temperature, washed with 20 mL CH 2 Cl 2 to remove dark impurities, then the aqueous phase was alkalised with a saturated NaOH (to a pH of 10-12) and extracted with dichloromethane (3 × 20 mL). The combined organic extracts were dried with magnesium sulphate, filtered, and the solvent was evaporated in vacuo. The crude compound 11c was used in subsequent reactions without further purification. General procedure for synthesis of quinoline amides (12a-w) A solution of suitable quinolinecarboxylic acid (5 mmol), ethyl chloroformate (0.5 mL, 5 mmol) and triethylamine (0.75 mL, 5 mmol) in anhydrous DMF (25 mL) was stirred for 30 min at 0°C. A solution of appropriate amine 10a-f (5 mmol) in anhydrous DMF (15 mL) was added dropwise. The cooling bath was removed and stirring was continued for 24 h. The solvent was evaporated in vacuo and to the residue 10 mL 5% aqueous solution of sodium bicarbonate was added. Next, the aqueous phase was extracted with dichloromethane (3 × 20 mL). The combined organic extracts were dried with magnesium sulphate, filtered, and the solvent was evaporated in vacuo. Final compounds 12ac, 12e-s, and 12v were purified by crystallisation. Compounds 12d, 12t, 12u, and 12w were purified by column chromatography.

HPLC analysis
Dionex system was used. The system consisted of a quaternary pump P580, a UVD detector 340 S, a column thermostat YetStream II Plus (WO Industrial Electronics), all controlled with Chromeleon software (version 6.01). Sample injection was performed through Rheodyne injector valve with a 20 µl sample loop. Chromatographic separations were carried out using the NUCLEODUR C18 Gravity column (Machery-Nagel), 150 × 4.6 mm, 5 µm and guard column NUCLEODUR C18 Gravity 5 µm. Mobile phases consisted of a mixture of 6 mM octane-1-sulphonic acid sodium salt and MeOH (55: 45) adjusted the pH to 3 with acetic acid. The flow rate of the mobile phase was 0.8 ml/min. The temperature in the column was maintained at 30°C. Thanks to the diode array detector, it was possible to record UV spectra of analysed compounds with absorbance maximum at c.a. 236 nm. Detection was carried out at λ = 236 nm.

Radioligand binding assay
All compounds were tested for their affinities for 5-HT 1A , 5-HT 2A , and D 2 receptors according to previously described procedures (Stefanowicz et al. 2016).

In vivo studies
Animals The experiments were performed on male mice (22-26 g, Albino Swiss or CD-1). All animals were kept in an environmentally controlled rooms (ambient temperature 21 ± 2°C; relative humidity 50-60%; 12:12 light-dark cycle, lights on at 8:00) and filtered water were freely available. All the experimental procedures were approved by the I Local Ethics Commission at the Jagiellonian University in Krakow. All the experiments were conducted in the light phase between 09:00 and 14:00 h. Each experimental group consisted of 6-10 animals/dose, and the animals were used only once in each test.

Spontaneous locomotor activity
The locomotor activity was recorded with an Opto M3 multi-channel activity monitor (MultiDevice Software v.1.3, Columbus Instruments). The investigated compounds or vehicle were administered intraperitoneally (i.p.) 60 min before the test running. The mice were individually placed in plastic cages (22 × 12 × 13 cm) for 30 min habituation period, and then the crossings of each channel (ambulation) were measured every 5 for 60 min (in CD-1 mice) and during 1-min or 3-6 min test session for Albino Swiss mice. The cages were cleaned up with 70% ethanol after each mouse.

MK-801-induced hyperactivity
MK-801-induced hyperactivity in mice was recorded according to the method described above. The investigated compounds or vehicle were administered i.p. 30 min., while MK-801 0.2 mg/kg i.p. 15 min before the test running.
Amphetamine-induced hyperactivity d-Amphetamine-induced hyperactivity in mice was recorded according to the method described above. The investigated compounds or vehicle were administered i.p., while amphetamine 2.5 mg/kg subcutaneously (s.c.) 30 min before the test running.

Forced swim test in mice
The experiment was carried out according to the method of Porsolt et al. (1979). Mice (Swiss Albino) were individually placed in a glass cylinder (25 cm high; 10 cm in diameter) containing 10 cm of water maintained at 23-25°C, and were left there for 6 min. A mouse was regarded as immobile when it remained floating on the water, making only small movements to keep its head above it. The total duration of immobility was recorded during the last 4 min of a 6-min test session.

Four-plate test in mice
Test was performed on male Swiss Albino mice. A single mouse was placed gently onto the plate, and each animal was allowed to explore for 15 s. Afterwards, each time a mouse passed from one plate to another, the experimenter electrified the whole floor for 0.5 s (current 0.8 mA), which evoked a visible flight reaction of the animal. If the animal continued running, it received no new shock for the following 3 s. The number of punished crossings was counted for 60 s.

Statistical analysis
The data are presented as the mean ± S.E.M.The obtained data were analysed by one-way analysis of variance (ANOVA) followed by Bonferroni's post-hoc test. p < 0.05 were considered statistically significant.
The next stage was acid catalysed hydrolysis of the amide bond of 10a-f derivatives, giving 8-aryl-8-azabicyclo [3.2.1]oct-3β-yl-amine derivatives (11a-f). Due to the high process yield and purity of the crude products, compounds 11a-f were used in subsequent reactions without further purification. The mixed anhydride method was used in order to obtain the final planned β-quinolineamide derivatives (12a-w).
All except one of the reported synthesis methods for the final compounds proved to be stereospecific. However, compound 12u was obtained as a mixture of isomers (see Scheme 1). This observation is of particular interest in view of earlier our research results of group 3β-acylamine derivatives of tropane. The ratio of 12u isomers in the mixture was confirmed by 1 H NMR and HPLC spectral analysis as described in section Conformational analysis.
The structures of all novel intermediates and final compounds were confirmed by IR, 1 H NMR and 13 C NMR spectroscopy and ESI-HRMS spectrometry. Detailed characterisation data are provided in Material and methods section. For in vivo and in vitro investigations, free bases were converted into the corresponding water-soluble salts.

Conformational analysis
The 1 H and 13 C NMR spectra of the samples 12a-w confirm the assumed structures (see Material and methods section). The signal of the C3H proton (in proton spectra) is particularly interesting as it has the form of a 12-line or 14line multiplet. In order to account for this splitting pattern we need to assume that the C3H proton is axial. For the 12line presentation, we can assume that the multiplet is formed of 3 overlapping quartets, this corresponding to an initial split into a triplet by axial C2H and C4H protons followed by the triplet constituents splitting into quartets as a result of coupling with the three protons of C2H and C4H (equatorial) and NH. We assume that the NH proton coupling constant is the same as (or very similar to) the constants of coupling to the equatorial protons of C2H and C4H. This assumption cannot hold for the 14-line presentation and it can be assumed in this latter case that the signal from the C3H proton is split into a triplet by coupling with the axial C2H and C4H protons, followed by a split of the triplet components into doublets by the NH proton and, finally, followed by the components of the three doublets being split by equatorial C2H and C4H protons. As a result there are 18 theoretical lines, but partial signal overlap simplifies the multiplet to 14 lines, confirming our assumption that the C3H proton is axial, but also leading to the conclusion that the spatial position of the equatorial -NH-CO-R substituent is different in the different compounds analysed (Figs. 4 and 5).
An unexpected effect is seen in the spectra of sample 12u, where there is a marked increase in the number of signals in the 13 C and 1 H NMR spectra, suggesting the presence of a mixture of compounds. Fortunately, the finding of distinct multiplets of the C3H proton (in the proton spectrum) allows the conclusion that there is a mixture of two isomers: one (β) with an axial multiplet of the C3H proton (12 lines) and one (α) with an equatorial multiplet of the C3H proton (4 lines). Apparently, the equatorial C3H proton couples with three protons, namely, the axial C2H and C4H, and NH to produce a pseudoquartet. This requires making the assumption that the coupling constant for C2H and C4H equals 0. Then (according to the Karplus curve), the C3H coupling plane forms an angle of~90 o with the coupling planes of C2H and C4H. Integrals (in 1 H NMR spectra) can be used to calculate that the molar ratio of the β form to the α form is 2:1. This conclusion is corroborated by a good fit of the chemical shifts in 1 H and 13 C NMR spectra with the spectra of similar structures. HPLC studies were conducted following determination of NMR spectra, which revealed that 12u is a mixture of stereoisomers. The synthesis was repeated twice, with HPLC analysis producing very similar results to NMR spectral analysis.
The UV spectra (see supplementary material) are identical and characterised by the same absorbance maximums; i.e., at wavelengths equal to their absorbance maximum, both isomers are detected at the same maximum sensitivity. In this situation, the mass ratio of the isomers can be determined by comparing peak areas.
To check for stereochemical purity, the analysis was repeated for the remaining compounds. This paper contains the results for the compound 12m. The area of the 12m α peak, at t R = 18.283 min., is 2.0553 mAU × min., and the area of the 12m β peak, at t R = 22.747 min., is 300.1300 mAU × min., producing an α:β ratio of 1:146, which corresponds to an admixture of the 12m α isomer of approximately 0.7%.
In summary, all target compounds are equatorial isomers (3β), except for the derivative 12u. To our surprise, the admixture of an axial isomer (3α) was significant (29.4% by HPLC) only in this case. The same results were seen with the re-synthesised compound. We are unable to account for this isomerisation, the less so as the analogues 12t, 12v, and 12w obtained from the same stereochemically pure substrate 10f (NMR) are stereochemically pure equatorial isomers. Work to explain this is under way.

Biological evaluation
Radioligand binding assay for D 2 , 5-HT 1A , and 5-HT 2A receptors As mentioned in the Introduction, ligands with simultaneous affinity for D 2 , 5-HT 1A , and 5-HT 2A receptors seem to be promising compounds for the pharmacotherapy of schizophrenia. In our previous paper, we described the synthesis and biological evaluation of compounds with very good double binding to D 2 and 5-HT 2A receptors; the most potent are shown in Fig. 1. Thus, in the subsequent phase of experimentation, we focused our attention on evaluating the impact of lead structure modification on 5-HT 1A receptor affinity. Therefore, the compounds synthesised within the present project included structural analogues of 3β-acylamine derivatives of tropane with the introduction of a methyl substituent in the benzyl ring and a quinoline moiety. These modifications were designed as a result of previous research, aiming to develop new ligands with enhanced 5-HT 1A binding activity in the investigated group of tropane derivatives.
Compounds 12a-w and 19a-c were tested for their in vitro affinity for the D 2 , 5-HT 1A , and 5-HT 2A receptors using a radioligand binding assay. Competition binding studies were performed according to a previously described procedure in rat brain tissues (Stefanowicz et al. 2016). The results are presented in Table 1.
First, the impact of structure modifications to the quinoline derivatives (12a-w) on D 2 affinity were examined. The influence of the quinoline moiety and its derivatives were analysed. The nitrogen position in the quinolinyl fragment impacted on the affinity for D 2 . We observed the  Table 1 Binding affinities for dopamine D 2 and serotonin 5-HT 1A /5-HT 2A receptors same rank order: 2-quinolinyl > 6-quinolinyl > 3-quinolinyl in the case of all ligands. Therefore, introduction of 2quinoline fragment in ligands was found to be favourable for D 2 binding, with the highest affinity seen for compound 12i. At the same time, the 4-methoxy-quinoline analogues had in general the lowest affinity for this receptor.
Next, by comparing the influence the location of the methyl substitution in the benzyl ring, it was confirmed that p-substituted ligands (12i, 12j, 12k) were considerably more potent than their m-substituted or o-substituted analogues, with compound 12i (p-CH 3 ) displaying a D 2 K i = 7.4 nM. This is in accordance with our previous results (Stefanowicz et al. 2016).
The substitution of a 2-piridylmethyl, 3-piridylmethyl, or 4-piridylmethyl at the in N8 position resulted in a significant loss of activity at the D 2 receptor.
Analysing the K i values for 5-HT 1A receptors, we found that the introduction of quinoline and its derivatives was beneficial in terms of 5-HT 1A receptor binding affinity. Furthermore, the presence of the 4-methoxy-2-quinoline moiety notably ameliorated the affinity for the 5-HT 1A receptor. This enhancing effect was greatest for compound 12h K i = 21.0 nM. The nitrogen position in the quinolinyl fragment also affected binding to 5-HT 1A receptors; these could be ranked in order of their increasing influence as follows: 2-quinolinyl > 3-quinolinyl > 6-quinolinyl.
It is worth mentioning that the replacement of the naphthyl ring with heterocyclic analogues led to the complete loss of 5-HT 2A receptor affinity in the investigated group of ligands. Thus, the presence of a naphthyl moiety is crucial for obtaining ligands in this series with triple binding activity for the D 2 , 5-HT 1A , and 5-HT 2A receptors.
The introduction of an additional nitrogen atom into the molecule as an amine group in the phenyl ring (Fig. 2) in derivatives 19a-c resulted in a marked increase in affinity for all receptors under study. The resulting compounds showed the highest binding affinity for the D 2 , 5-HT 1A , and 5-HT 2A receptors of all derivatives described in this paper. In this respect, the compound 19a (K i [nM] = D 2 = 41.5; 5-HT 1A = 53.0; 5-HT 2A = 46.8) appears to hold the greatest promise. The affinity of the derivative 19a described above and its analogues 19b and 19c was markedly influenced by the position of the -NH 2 moiety in the benzyl system, where the o-isomer was the most active one and the p-isomer was the least active. Of note, unlike the other new structures, these three compounds are naphthalene derivatives. This again seems to lead to the conclusion that the presence of a naphthalene system in these compounds is more beneficial than the presence of a quinoline system in terms of producing a derivative with triple binding affinity for the D 2 , 5-HT 1A , and 5-HT 2A receptors. The salts of compounds 19a-c were also characterised by the best solubility in water among all the derivatives analysed. In summary, the introduction of an additional nitrogen atom into the naphthalene or phenyl ring had an overall adverse effect on the binding affinities of the new compounds compared to the lead structure (compound A) and its derivatives described in our previous publication. This modification had the greatest negative effect on affinity for the 5-HT 2A receptor. Of note, the 2-quinoline derivatives 12e and 12i demonstrated very good binding affinity for the D 2 and 5-HT 1A receptors, being superior in this respect to the 3-and 6-quinoline analogues.
The introduction of a pyridine ring into the molecule had an adverse effect on binding affinity, while the introduction of an amine group as a substituent in the phenyl ring produced very active compounds. Compound 19a is exceptional among the analysed structures as it demonstrates comparable affinities for all three receptor types (D 2 , 5-HT 1A , and 5-HT 2A ), resulting in a very quetiapine-like receptor profile, but with binding affinities 4-fold or 5-fold higher than those of quetiapine.

In vivo studies
General Experiments were carried out on Albino Swiss or CD-1 male mice weighing 22-26 g kept in colony cages in standard laboratory conditions. Experimental groups were chosen randomly and each animal was used only once. The compounds studied were suspended in a 1% solution of Tween 80 (Sigma, St. Louis, MO, USA) and injected intraperitoneally in a volume of 10 ml/kg.

Antipsychotic-like activity
To study the potential antipsychotic activity of selected compounds the d-amphetamine-and MK-801-induced hyperlocomotor activity test in mice were carried out. Compounds 12i (5 mg/kg and 10 mg/kg i.p.), 19a (5 mg/kg and 10 mg/kg i.p.) significantly reduced MK-801-induced hyperlocomotor activity (Table 2). Compound 12e administered at a dose of 5 mg/kg showed a tendency to decrease the MK-801-induced hyperlocomotor activity but the results did not reach a statistically significant level (Table 2). In damphetamine-induced hyperlocomotor activity test, compounds 12i and 12e (at doses of 5 and 10 mg/kg i.p.) significant decreased locomotor hyperactivity in the range of 60-86% vs. respective d-amphetamine group (Table 3). The compound 19a was active in this test only at a dose of 10 mg/kg i.p. (Table 3). The compounds were injected i.p. 30 min. before the test. Values represent the mean ± SEM during last 4-min test session compared to the respective vehicle group (one-way ANOVA is followed by the Bonferroni's post hoc test), N = 6-9, NS-non-significant The investigated compounds were injected i.p. 30 min., while diazepam 60 min. before the test. Values represent the mean ± SEM during 1-min test session compared to the respective vehicle group (one-way ANOVA is followed by the Bonferroni's post hoc test), N = 8-10, NS-non-significant The compounds 19a (10 mg/kg) and 12i (10 mg/kg) significantly decreased spontaneous locomotor activity about 70% since the positive effects observed in hyperlocomotor activity tests may not be specific (Table 4). The compound 12e at the doses used in hyperlocomotor activity tests did not change the spontaneous locomotor activity in mice, thus its antipsychotic-like effect appeared to be specific (Table 4).

Antidepressant-like activity
The potential antidepressant activity of selected compounds in vivo was investigated using the forced swim test in mice. In this test only compound 19a (5 mg/kg i.p.) decreased immobility time about 43% vs. respective control group, showing significant antidepressant-like activity (Table 5).

Anxiolytic-like activity
The potential anxiolytic activity of selected compounds in vivo was investigated using the four-plate test in mice. In this test only compound 12h (1.25 and 2.5 mg/kg i.p.) increased punished crossings in a range of 60% vs. respective control group, showing significant anxiolytic-like activity (Table 6).
Active doses of the investigated compounds had no influence on the spontaneous locomotor activity measured during the time equal to the observation period in the forced swim and the four-plate tests (i.e., from 2-6 min and 1 min 15 s, respectively) (data not shown) thus observed antidepressant-like and/or anxiolytic-like activity of these compounds seems to be specific.

Conclusion
We have described here a series of 26 compounds representing new derivatives of 3β-aminotropane and being analogues of a previously identified compound A, which shows high activity at the D 2 , 5-HT 1A , and 5-HT 2A receptors. Modifications involved the introduction of an additional nitrogen atom, producing quinoline, isoquinoline or pyridine derivatives or derivatives with an amine group as a substituent in the phenyl ring (Fig. 2). Structure-activity relationship studies revealed that these modifications adversely affected the binding affinity of these compounds for the three types of receptors, except for the derivatives 12e and 12i, which demonstrated high binding affinity for the D 2 and 5-HT 1A receptors, and the compound 19a, which showed comparable binding affinities for all three receptor types (D 2 , 5-HT 1A , and 5-HT 2A ), giving it a very quetiapinelike receptor profile, but with 4-fold or 5-fold higher binding affinities than that antipsychotic drug.
Studies of behavioural activity of selected compounds (12d, 12e, 12h, 12i, and 19a) showed that the compounds 12i and 19a exerted a specific antipsychotic-like effect in damphetamine-induced and MK-801-induced hyperlocomotor activity test in mice. Specific antidepressant-like activity (the forced swim test) was displayed only by the compound 19a and a specific anxiolytic-like effect was produced only by 12h (Figs. 3-5).
The beneficial and more comprehensive activity profile of the compound 19a encourages further rational search for new antipsychotics with an affective component in this structural class.