Synthesis of New GABAA Receptor Modulator with Pyrazolo[1,5-a]quinazoline (PQ) Scaffold

We previously published a series of 8-methoxypirazolo[1,5-a]quinazolines (PQs) and their 4,5-dihydro derivatives (4,5(H)PQ) bearing the (hetero)arylalkylester group at position 3 as ligands at the γ-aminobutyric type A (GABAA) subtype receptor. Continuing the study in this field, we report here the design and synthesis of 3-(hetero)arylpyrazolo[1,5-a]quinazoline and 3-(hetero)aroylpyrazolo[1,5-a]quinazoline 8-methoxy substituted as interesting analogs of the above (hetero)arylalkylester, in which the shortening or the removal of the linker between the 3-(hetero)aryl ring and the PQ was performed. Only compounds that are able to inhibit radioligand binding by more than 80% at 10 μM have been selected for electrophysiological studies on recombinant α1β2γ2L GABAA receptors. Some compounds show a promising profile. For example, compounds 6a and 6b are able to modulate the GABAAR in an opposite manner, since 6b enhances and 6a reduces the variation of the chlorine current, suggesting that they act as a partial agonist and an inverse partial agonist, respectively. The most potent derivative was 3-(4-methoxyphenylcarbonyl)-8-methoxy-4,5-dihydropyrazolo[1,5-a] quinazoline 11d, which reaches a maximal activity at 1 μM (+54%), and it enhances the chlorine current at ≥0.01 μM. Finally, compound 6g, acting as a null modulator at α1β2γ2L, shows the ability to antagonize the full agonist diazepam and the potentiation of CGS 9895 on the new α+/β− ‘non-traditional’ benzodiazepine site.


Introduction
γ-aminobutyric type A (GABA A ) receptors Type A (GABA A R) and Type B (GABA B R) are the principal inhibitory neurotransmitter receptors in the mammalian brain, and are the targets of many important drugs on the market for a variety of neurological conditions, including epilepsy, anxiety, spasticity, pain, and psychiatric illness [1]. The neurotransmitter GABA produces slow (sub-second) synaptic response by interaction with the GABA B Rs, belonging to the G-protein-coupled receptors (GPCR) category, or exerts fast (<10 ms) and powerful synaptic inhibition by acting on the GABA A Rs, which is a Cys-loop pentameric ligand gated ion channel (LGIC). This last family of receptors also includes the nicotinic acetylcholine receptors (nAChRs), the 5-hydroxytriptamine type 3 receptors (5-HT 3 Rs), and the glycine receptors, all of which show a similar topology, namely five homologous ligands at the GABAA subtype receptor, we considered the shortening or the removal of the linker between the 3-(hetero)aryl ring and the PQ core as an interesting chemical modification of the previously reported ester derivatives. Thus, the synthesis of 8-methoxy 3-(hetero)arylpyrazolo[1, 5a]quinazoline and 3-(hetero)aroylpyrazolo[1,5-a]quinazoline was planned. These compounds are most likely more metabolically stable then ester derivatives [18], and are structurally strictly related to compounds that have been claimed to be GABAA subtype receptor ligands such as Indiplon, Ocinaplon [4,19,20], and compounds of type I and II [21][22][23] (Figure 1, Panel A). All of these compounds belong to the pyrazolopyrimidines, triazolophtalazine, or pyrazolobenzotriazine family, whose core is easily recovered in our new synthesized PQ compounds (Figure 1, Panel B).  All new compounds were evaluated for their ability to displace [3H]flumazenil(Ro-151788) from its specific binding to Bz receptors in bovine membrane samples. Moreover, chlorine current on recombinant GABAA receptors of the α1β2γ2L type (expressed in frog oocytes of the Xenopus laevis species) was limited to those compounds that are able to inhibit radioligand binding by more than 80% at 10 μM. Additionally, for the most interesting compounds, the modulation on the new α+/β− 'non-traditional' benzodiazepine site was evaluated. All new compounds were evaluated for their ability to displace [3H]flumazenil(Ro-151788) from its specific binding to Bz receptors in bovine membrane samples. Moreover, chlorine current on recombinant GABA A receptors of the α1β2γ2L type (expressed in frog oocytes of the Xenopus laevis species) was limited to those compounds that are able to inhibit radioligand binding by more than 80% at 10 µM. Additionally, for the most interesting compounds, the modulation on the new α+/β− 'non-traditional' benzodiazepine site was evaluated.
The aromatization of the quinazoline ring was obtained for compounds 3a-e by treatment with LiAlH 4 in anhydrous THF, and further oxidation in the same vessel reaction (the corresponding 4,5-dihydropyrazolo[1,5-a]quinazolines were not isolated) to directly give the aromatic compounds The aromatization of the quinazoline ring was obtained for compounds 3a-e by treatment with LiAlH4 in anhydrous THF, and further oxidation in the same vessel reaction (the corresponding 4,5dihydropyrazolo[1,5-a]quinazolines were not isolated) to directly give the aromatic compounds 5ae. Instead, compounds 3f-h were transformed into the corresponding 5-chloroderivatives (type 4 compounds), and the next reduction with HCOONH4 and 10% Pd/C yielded the final desired compounds 6a-c, without showing the corresponding dihydroderivatives.
In the second approach, modifications of position 3 of the pyrazolo[1,5-a]quinazoline scaffold were performed, depending on the final designed product (schemes 2, 3). For the introduction of the (hetero)aryl ring, the Suzuki coupling reaction was carried out (Scheme 2). The key intermediate 8 was obtained from 7, the 8-methoxypyrazolo[1,5-a]quinazolin-5(4H)-one, by treatment with LiAlH4/THF, and the next dehydrogenation with 10% Pd/C and toluene; compound 7, in turn, was achieved by decarboxylation of ethyl 8-methoxy-5-oxo-4,5-dihydropyrazolo[1,5-a] quinazoline 3carboxylate [18]. Halogenation of position 3 of the PQ core, by bromine or N-iodosuccinimide (NIS), gave 9a,b, which were used as counterparts with the suitable boronic acid (2-MeO-phenyl-, 3-furan-, and 1-Boc-2-pyrrolboronic acid) in the Suzuki coupling reaction, yielding compounds 5f-h. The best yields were obtained starting from compound 9b. The deprotection of 3-(1-Boc-2-pyrrolyl)-8methoxypyrazolo[1, 5-a] For the introduction of a keto group at position 3, still working on the tricyclic system, the 8methoxypyazolo[1,5-a]quinazoline-3-carboxylic acid 10 [18] was used as the starting material. The 3-(4-methoxyphenylcarbonyl)-8-methoxypyrazolo[1,5-a]quinazoline 6d was obtained with Eaton's reagent and anisole, while the other compounds were obtained through the 3-acylchloride intermediate. The attempt to obtain the 3-carbonylderivatives by Suzuki coupling reaction failed, since for example, when using 2-MeO-phenylboronic acid as the counterpart, unexpectedly, only compound 5f was recovered, evidencing CO elimination and Ar-Ar coupling (Scheme 3). Thus, the For the introduction of a keto group at position 3, still working on the tricyclic system, the 8-methoxypyazolo[1,5-a]quinazoline-3-carboxylic acid 10 [18] was used as the starting material. The 3-(4-methoxyphenylcarbonyl)-8-methoxypyrazolo[1,5-a]quinazoline 6d was obtained with Eaton's reagent and anisole, while the other compounds were obtained through the 3-acylchloride intermediate. The attempt to obtain the 3-carbonylderivatives by Suzuki coupling reaction failed, since for example, when using 2-MeO-phenylboronic acid as the counterpart, unexpectedly, only compound 5f was recovered, evidencing CO elimination and Ar-Ar coupling (Scheme 3). Thus, the Friedel-Craft reaction was applied and, after treatment of the 8-methoxypyazolo[1,5-a]quinazoline-3-carboxylic acid 9 with SOCl 2 or Cl 3 CCN and PPh 3 to give the 3-acylchloride intermediate, the appropriate (hetero)aryl was added, and the final products 6e-g were obtained. Compound 6g was also obtained by alkylation of 6f with DMF, K 2 CO 3 , and methyl iodide, but in low yield. Friedel-Craft reaction was applied and, after treatment of the 8-methoxypyazolo[1,5-a]quinazoline-3-carboxylic acid 9 with SOCl2 or Cl3CCN and PPh3 to give the 3-acylchloride intermediate, the appropriate (hetero)aryl was added, and the final products 6e-g were obtained. Compound 6g was also obtained by alkylation of 6f with DMF, K2CO3, and methyl iodide, but in low yield. Finally, type 6 compounds were transformed into the corresponding 4,5-dihydroderivatives by treatment with NaBH3CN in AcOH, obtaining compounds 11a, b, and d-g, as shown in Scheme 4. When 3-(hetero)aryl substituted compounds (type 5) were reacted in the same conditions, the corresponding 4,5-dihydroderivatives were evidenced only in the reaction mixture by TLC, and were not isolated as pure compounds, because after the normal work up, they were converted again into the starting material. Finally, type 6 compounds were transformed into the corresponding 4,5-dihydroderivatives by treatment with NaBH 3 CN in AcOH, obtaining compounds 11a, b, and d-g, as shown in Scheme 4. When 3-(hetero)aryl substituted compounds (type 5) were reacted in the same conditions, the corresponding 4,5-dihydroderivatives were evidenced only in the reaction mixture by TLC, and were not isolated as pure compounds, because after the normal work up, they were converted again into the starting material.

Biological Evaluation
All of the new compounds were previously evaluated for their ability to displace [ 3 H]flumazenil (Ro-151788) from its specific binding to Bz receptors in bovine membrane samples. Only compounds that were able to inhibit radioligand binding by more than 80% at 10 μM (5d-f, 6a-d, 6f,g, 11a,b,d) were further selected for electrophysiological studies on recombinant α1β2γ2L GABAA receptors.
Recombinant α1β2γ2L GABAA receptors were expressed in Xenopus laevis oocytes, and the

Biological Evaluation
All of the new compounds were previously evaluated for their ability to displace [ 3 H]flumazenil (Ro-151788) from its specific binding to Bz receptors in bovine membrane samples. Only compounds that were able to inhibit radioligand binding by more than 80% at 10 µM (5d-f, 6a-d, 6f,g, 11a,b,d) were further selected for electrophysiological studies on recombinant α1β2γ2L GABA A receptors.

Biological Evaluation
All of the new compounds were previously evaluated for their ability to displace [ 3 H]flumazenil (Ro-151788) from its specific binding to Bz receptors in bovine membrane samples. Only compounds that were able to inhibit radioligand binding by more than 80% at 10 μM (5d-f, 6a-d, 6f,g, 11a,b,d) were further selected for electrophysiological studies on recombinant α1β2γ2L GABAA receptors.
Recombinant α1β2γ2L GABAA receptors were expressed in Xenopus laevis oocytes, and the effects of compounds tested at 1 to 100 μM were assessed on the modulation of GABAA receptor function ( Figure 2). The three sections of Figure  As evident in Figure 2 (section A), the first group of compounds (5d-f) was not able to modulate the GABAA function through acting as a null modulator or antagonist. Among compounds 6a-d,f,g bearing an aroyl moiety at position 3 ( Figure 2, section B), the 3-(2-methoxyphenylcarbonyl) -8methoxypyrazolo[1,5-a]quinazoline 6a slightly, but not significantly, inhibited the GABAA receptor function at 100 μM. On the contrary, the 3-(thien-2-yl-carbonyl)-8-methoxypyrazolo[1,5a]quinazoline 6b, slightly but significantly enhanced the GABAA receptor function until 100 μM, As evident in Figure 2 (section A), the first group of compounds (5d-f) was not able to modulate the GABA A function through acting as a null modulator or antagonist. Among compounds 6a-d,f,g bearing an aroyl moiety at position 3 ( Figure 2, section B), the 3-(2-methoxyphenylcarbonyl) -8-methoxypyrazolo[1,5-a]quinazoline 6a slightly, but not significantly, inhibited the GABA A receptor function at 100 µM. On the contrary, the 3-(thien-2-yl-carbonyl)-8-methoxypyrazolo[1,5-a]quinazoline 6b, slightly but significantly enhanced the GABA A receptor function until 100 µM, showing an EC 50 = 11.97 µM and acting as a partial agonist (E max + 42% at 100) For the 4,5-dihydroderivatives 11a,b,d, which are reported in the third part of Figure 2 (section C), compound 11d stands out as enhancing the GABA A receptor function and reaching a maximum of activity at 1 µM (Emax + 54%), but maintaining a slightly positive variation in chloride current even at ≥0.01 µM.
In order to confirm whether those compounds that are null modulators act at the benzodiazepine binding site, 6g as a representative sample was evaluated for its ability to antagonize the full agonist diazepam. Figure 3 clearly demonstrated that the null modulator 6g is able to completely abolish the potentiation of the GABA A receptor function induced by 1 µ diazepam at 10 µM. Recently, other low-affinity sites for benzodiazepine, in addition to the high-affinity benzodiazepine binding site, have been discovered [9,17,26]. Among them, the site located in the extracellular domain at the α+/β− interface was demonstrated to be the target for the pyrazoloquinoline CGS 9895 (2-pmethoxyphenylpyrazolo[4,3-c]quinolin-3-(5H)-one, as shown in Figure 4), which was already identified as a null modulator (antagonist) at the high-affinity benzodiazepine site, and acts as a positive allosteric modulator [27]. Therefore, compound 6g was further tested for its ability to antagonize the potentiation of the GABAA receptor induced by CGS 9895. Recently, other low-affinity sites for benzodiazepine, in addition to the high-affinity benzodiazepine binding site, have been discovered [9,17,26]. Among them, the site located in the extracellular domain at the α+/β− interface was demonstrated to be the target for the pyrazoloquinoline CGS 9895 (2-p-methoxyphenylpyrazolo[4,3-c]quinolin-3-(5H)-one, as shown in Figure 4), which was already identified as a null modulator (antagonist) at the high-affinity benzodiazepine site, and acts as a positive allosteric modulator [27]. Therefore, compound 6g was further tested for its ability to antagonize the potentiation of the GABA A receptor induced by CGS 9895.
binding site, have been discovered [9,17,26]. Among them, the site located in the extracellular domain at the α+/β− interface was demonstrated to be the target for the pyrazoloquinoline CGS 9895 (2-pmethoxyphenylpyrazolo [4,3-c]quinolin-3-(5H)-one, as shown in Figure 4), which was already identified as a null modulator (antagonist) at the high-affinity benzodiazepine site, and acts as a positive allosteric modulator [27]. Therefore, compound 6g was further tested for its ability to antagonize the potentiation of the GABAA receptor induced by CGS 9895. The data obtained in GABAA receptors devoid of the γ subunit (α1β2) indicates that compound 6g reduces the potentiation of the GABAA receptor induced by CGS 9895 (10 μM) (about −58% at both 10 μM and 30 μM), suggesting that this compound binds the α+/β− low-affinity site, too. This interesting result seems to suggest that our compounds acting as null modulators or antagonists could be selective α+/β− low-affinity benzodiazepine binding site ligands. This hypothesis was The data obtained in GABA A receptors devoid of the γ subunit (α 1 β 2 ) indicates that compound 6g reduces the potentiation of the GABA A receptor induced by CGS 9895 (10 µM) (about −58% at both 10 µM and 30 µM), suggesting that this compound binds the α+/β− low-affinity site, too. This interesting result seems to suggest that our compounds acting as null modulators or antagonists could be selective α+/β− low-affinity benzodiazepine binding site ligands. This hypothesis was preliminarily confirmed by compounds 5d and 5e, which in the same test completely blocked or strongly reduced the potentiation of the GABA A receptor function induced by CGS 9895, respectively (data not shown). In accordance with Sieghart et al. (2012) [8], we believe that drugs acting at the "non-canonical" α+/β− low-affinity binding site might display potential clinical relevance. It has been proposed that the compounds acting at this site could be beneficial for long-term epilepsy treatment. In fact, they will be able to interact with a broader variety of GABA A receptors a subtypes such as the δ, ε, and π subunit-containing GABA A receptors.
Results from electrophysiological studies on recombinant α1β2γ2L GABA A receptors showed that the 3-aroyl derivatives are able to slightly induce a change in the agonist-evoked current. This evidence led us to speculate that the aroyl moiety at position 3 of the PQ scaffold could be responsible for this effect. The more active compounds 6a and 6b respectively bear the 2-methoxybenzoyl and thien-2-yl-carbonyl groups at position 3, which are both able to engage hydrogen bond interaction with receptor proteins. Also, it is noteworthy that compound 6b retains the same pyrazolo[1,5-a]pyrimidine substructure as indiplon (Figure 1), which is already known as a partial agonist. In particular, the carbonyl moiety (CO) and the 2-methoxy (2-OMe) substituent on the phenyl ring for 6a and the sulfur atom of the thienyl ring for 6b could provoke a change of conformation favorable for the effect; this is different from the 3-aryl substituted PQ, which is devoid of activity.

Material and Methods
Melting points were determined with a Gallenkamp apparatus and were uncorrected. Silica gel plates (Merk F 254 ) and silica gel 60 (Merk 70-230 mesh) were used for analytical and column chromatography, respectively. The structures of all the compounds were supported by their IR spectra (KBr pellets in nujol mulls, Perkin-Elmer 1420 spectrophotometer) and 1 H-NMR data (measured with a Bruker 400 MHz). Chemical shifts were expressed in δ ppm, using DMSO-d 6 or CDCl 3 as solvent.
The chemical and physical data of new compounds are shown in Table S1; all of the microanalyses were performed with a Perkin-Elmer 260 analyzer for C, H, and N. Mass spectra (m/z) were recorded on an Electrospray ionisation time-of-flight mass spectrometry ESI-TOF mass spectrometer (Bruker Micro TOF), and reported mass values are within the error limits of ±5 ppm mass units. Microanalyses indicated by the symbols of the elements or functions were performed with a Perkin-Elmer 260 elemental analyzer for C, H, and N, and they were within ±0.4% of the theoretical values.

General Procedure for the Synthesis of 5a-e
LiAlH 4 /THF solution (2.8 mmol) was added to a suspension of 3a [25] and 3b-e (1.0 mmol) in anhydrous THF and the mixture refluxed. The reaction was monitored with TLC (toluene/ethyl acetate/methanol 8:2:1.5 v/v/v as eluent) until the starting material disappeared and a mixture of compounds was formed: the 4,5-dihydro and the 4,5-dehydro derivatives. Thus, the reaction was maintained at air to permit the complete oxidation to final pyrazolo[1,5-a]quinazoline derivatives. The careful addition of water, the next extraction with ethyl acetate, and the followed evaporation gave the final desired compounds.

General Procedure for the Synthesis of 5f-h
Triphenylphosphine palladium (0) (Tetrakis, 0.035 mmol), 2.4 mL of Na 2 CO 3 solution 2M, and suitable (hetero)arylboronic acid (0.60 mmol) solubilized in 1 mL of absolute ethanol were added to a solution of 3-iodo-or 3-bromo-8-methoxypyrazolo[1,5-a]quinazoline (9a, 9b, see below) (0.2 mmol) in anhydrous THF (6 mL). The suspension was refluxed until the starting material disappeared in TLC; then, it was diluted with ethyl acetate, and water was added. The organic layer was separated and dried over Na 2 SO 4 anhydrous and evaporated to dryness. The final compounds were purified by recrystallization by a suitable solvent. In general, it has been observed that from starting material 9a, the yield of coupling is better than starting from 9b.     The decarboxylation of the ethyl 8-methoxy-5-oxo-4,5-dihydropyrazolo[1,5-a]quinazoline 3-carboxylate [18] (1.0 mmol) was obtained in H 3 PO 4 at fusion condition. After that starting material disappeared in TLC (toluene/ethyl acetate/methanol 8:2:1.5 v/v/v as eluent), the treatment with ice/water gave a precipitate that was isolated and purified by recrystallization with ethanol. Cream crystals, yield 80%; IR ν cm −1 3137, 1697; 1 H-NMR (DMSO-d 6

General Procedure for the Synthesis of 9a,b
A solution of starting material 8 (0.35 mmol) in dichloromethane (5 mL) was supplemented with an excess of bromine (0.8 mL) to obtain compound 9a or N-iodosucinimide (NIS, 1:2) for final product polyethylene tubing (Clay Adams, Parsippany, NJ, USA). Oocytes were impaled at the animal pole with two glass electrodes (0.5 to 10 MΩ) filled with 3 M of KCl, and were clamped at −70 mV with the use of an oocyte clamp (model OC725C; Warner Instruments, Hamden, CT). Currents were measured and analyzed with the pClamp 9.2 software (Molecular Devices, Union City, CA, USA). GABA (Sigma, St. Louis, MO, USA) was dissolved in MBS and applied to the oocytes for 30 s. Oocytes were perfused with test drugs for 30 s either in the absence of the agonists or in its presence at the EC 5-10 (the concentration of agonist that induces a peak current equal to 5-10% of the maximal current elicited by the maximal concentration of the agonist). Compounds were first dissolved in DMSO at a concentration of 10 mM, and then diluted in MBS to the final concentrations. In each experiment, control responses were determined before and 10/15 min after application of the drug.

Statistics
Statistical analysis was performed on normalized data using the Kruskal-Wallis test followed by Dunn's post hoc test or the Mann-Whitney test using GraphPad Prism 7 (Graph Pad Software, Inc., San Diego, CA, USA).

Conclusions
In this paper, new 8-methoxypyrazolo[1,5-a]quinazolines bearing at position 3 of the (hetero)aryl group (type 5 compounds) or (hetero)aroyl moiety (type 6 compounds) and their corresponding 4,5-dihydroderivatives (type 11 compounds) were synthesized and evaluated for their ability to modulate the recombinant α 1 β 2 γ 2L GABA A receptors. Compounds that showed a certain modulation of chlorine current are 3-(hetero)aroylpyrazolo[1,5-a]quinazolines, and 6a and 6b were the most representative compounds. These products modulate the GABA A R in an opposite manner, suggesting that 6b acts as partial agonist and 6a acts as an inverse partial agonist.
Among the 3-(hetero)aroyl derivatives with the 4,5-dihydropyrazolo[1,5-a]quinazoline scaffold, the most interesting compound was 11d, for which an effect on the chlorine current is measurable at ≥0.01 µM. This compound will be the object of further studies.
Finally, we found the profile of the null modulator 6g interesting, since it not only acts as an antagonist blocking the potentiation of the GABAergic function induced by diazepam (so interacting at the high-affinity benzodiazepine site located in the extracellular domain at the α+/β− [2]), but it also works as a positive allosteric modulator at the low-affinity site located in the extracellular domain at the α+/β− interface. These results suggest that this compound could act to both the high-affinity and the low-affinity benzodiazepine site, so we can hypothesize that it will display a much broader action than the classic high-affinity benzodiazepine site ligands.
Moreover, since 6g is the first compound acting via the α+/β− interface of GABA A receptors with the PQ scaffold, it may represent an interesting lead for the discovery of a new class of ligand, which would be useful for the study of this recently discovered low-affinity benzodiazepine site.

Conflicts of Interest:
The authors declare no conflict of interest.