CNG-10300

The retrosynthetic analysis of CNG-10300 begins with the identification of the quinoxaline-2,3-dione core as the primary structural framework [1]. This α-amino acid-functionalized quinoxaline represents a compound where a quinoxaline moiety is coupled to an amino acid moiety through a strategic linker [1]. The retrosynthetic disconnection reveals that CNG-10300 can be assembled through a convergent approach involving two key fragments: the quinoxaline-2,3-dione scaffold and the protected α-amino acid side chain [5].

The quinoxaline-2,3-dione core synthesis follows established protocols beginning with the condensation of ortho-phenylenediamine with oxalic acid [12] [14]. This reaction represents one of the most fundamental quinoxaline formations, proceeding through a cyclocondensation mechanism that forms the bicyclic heterocycle in excellent yields [12]. The reaction typically employs thermal conditions or microwave irradiation, with the latter providing superior yields of approximately 88% compared to conventional heating methods [14].

For CNG-10300 specifically, the retrosynthetic analysis indicates that 6-bromo-2,3-dimethoxyquinoxaline serves as the key electrophilic partner in the cross-coupling reaction [5]. This intermediate is prepared through a three-step sequence: initial formation of 6-bromoquinoxaline-2,3-dione from 2-amino-4-bromoaniline and diethyl oxalate, followed by chlorination with thionyl chloride to yield 6-bromo-2,3-dichloroquinoxaline, and finally methoxylation using potassium methoxide in methanol [5].

The strategic use of protecting groups in the retrosynthetic plan ensures chemoselectivity during the cross-coupling step [5]. The quinoxaline nitrogen atoms are protected as methoxy derivatives, preventing competitive coordination that could inhibit palladium catalysis [5]. This protection strategy proved essential, as attempts to use unprotected quinoxaline-2,3-dione or dichloroquinoxaline derivatives resulted in complex product mixtures [5].

Palladium-Catalyzed Negishi Cross-Coupling Optimization

The palladium-catalyzed Negishi cross-coupling represents the key bond-forming step in CNG-10300 synthesis, connecting the quinoxaline core with the α-amino acid side chain [5] [16]. This transformation employs organozinc reagents generated in situ from the corresponding iodoalkyl amino acid derivatives [5]. The optimization of this cross-coupling reaction required careful consideration of catalyst selection, ligand choice, and reaction conditions to achieve acceptable yields.

The catalyst system employed consists of tris(dibenzylideneacetone)dipalladium(0) (Pd2dba3) as the palladium source and 2-dicyclohexylphosphino-2′,6′-dimethoxybiphenyl (SPhos) as the supporting ligand [5]. This combination has proven highly effective for challenging cross-coupling reactions involving sterically hindered substrates [18]. The SPhos ligand provides excellent activity for Negishi coupling reactions, particularly with substrates bearing coordinating functionalities [18].

The organozinc reagent formation involves treatment of the iodo-amino acid derivative with zinc dust in dimethylformamide at elevated temperature [5]. A catalytic amount of iodine is added to activate the zinc surface, facilitating the oxidative insertion process [5]. The reaction mixture exhibits characteristic color changes from yellow to black during zinc insertion, indicating successful organometallic formation [5].

The reaction proceeds through the standard Negishi mechanism involving oxidative addition of the aryl bromide to palladium(0), transmetalation with the organozinc reagent, and reductive elimination to form the carbon-carbon bond [19]. The use of excess zinc ensures complete conversion of the alkyl iodide to the organozinc species [5]. The reaction is conducted under an inert atmosphere to prevent oxidation of the sensitive organometallic intermediates [17].

Yield optimization studies revealed that the cross-coupling efficiency varies significantly with the amino acid chain length [5]. For CNG-10300, which contains a four-carbon linker between the quinoxaline core and the amino acid functionality, yields of 35% were achieved [5]. This moderate yield likely reflects the increased steric hindrance associated with the longer alkyl chain compared to shorter homologs [5].

The purification of the cross-coupling product requires careful workup procedures to remove zinc salts and palladium residues [5]. The reaction mixture is typically filtered through Celite to remove solid impurities, followed by extraction with diethyl ether and washing with brine [5]. The organic phases are dried over magnesium sulfate and concentrated under reduced pressure before purification by flash column chromatography [5].

Protecting Group Strategies for α-Amino Acid Moieties

The synthesis of CNG-10300 requires sophisticated protecting group strategies to ensure chemoselectivity throughout the synthetic sequence [5] [22]. The α-amino acid moiety contains multiple reactive functional groups that must be selectively protected and deprotected without compromising the quinoxaline core or other sensitive functionalities [22] [24].

The amino acid starting material employs a dual protection strategy using tert-butyloxycarbonyl (BOC) protection for the amino group and benzyl ester protection for the carboxylic acid [5] [25]. This orthogonal protection scheme allows for selective deprotection under different conditions, providing synthetic flexibility [22] [23]. The BOC group is acid-labile and can be removed using trifluoroacetic acid or hydrochloric acid, while the benzyl ester is cleaved under the final acidic deprotection conditions [25] [29].

The reduction of the carboxylic acid to the corresponding alcohol represents a critical transformation requiring protection of the amino functionality [5]. This reduction employs isobutyl chloroformate as an activating agent followed by sodium borohydride reduction in methanol [5]. The BOC protection prevents unwanted reactions at the amino center during this transformation [25].

The conversion of the primary alcohol to the corresponding iodide employs the Appel reaction using triphenylphosphine, imidazole, and iodine in dichloromethane [5]. This transformation proceeds under neutral conditions, maintaining the integrity of the BOC protecting group [5]. The reaction is conducted at low temperature initially to control the exothermic nature of the phosphine-iodine interaction [5].

The protecting group strategy must also account for the quinoxaline nitrogen functionalities [5]. The use of 2,3-dimethoxyquinoxaline as the electrophilic partner prevents coordination of the quinoxaline nitrogens to the palladium catalyst during cross-coupling [5]. This protection proved essential, as attempts to use unprotected quinoxaline derivatives resulted in catalyst poisoning and poor yields [5].

The orthogonal nature of the protection scheme allows for global deprotection in a single step using aqueous hydrochloric acid in dioxane [5]. This harsh acidic treatment simultaneously removes the BOC group, cleaves the benzyl ester, and demethylates the quinoxaline nitrogens to reveal the final quinoxaline-2,3-dione structure [5]. The one-pot deprotection provides excellent yields of 68-89% for the final products [5].

Post-Functionalization and Deprotection Protocols

The post-functionalization phase of CNG-10300 synthesis involves the global deprotection of all protecting groups to reveal the final bioactive compound [5] [30]. This critical transformation must proceed with high efficiency while maintaining the structural integrity of the quinoxaline-2,3-dione core and the α-amino acid functionality [5].

The deprotection protocol employs a one-pot strategy using aqueous hydrochloric acid in dioxane under reflux conditions [5]. This acidic treatment simultaneously accomplishes three key transformations: removal of the BOC protecting group from the amino acid nitrogen, hydrolysis of the benzyl ester to the free carboxylic acid, and demethylation of the quinoxaline methoxy groups to regenerate the quinoxaline-2,3-dione core [5].

The mechanism of BOC deprotection proceeds through protonation of the carbonyl oxygen followed by elimination of the tert-butyl cation [29] [36]. The resulting carbamic acid intermediate spontaneously decarboxylates to yield the free amino group [29]. The harsh acidic conditions ensure complete conversion, though care must be taken to avoid over-heating which could lead to degradation of the quinoxaline core [5].

The benzyl ester hydrolysis occurs through acid-catalyzed cleavage of the benzyl-oxygen bond [5]. The electron-rich benzyl group stabilizes the carbocation intermediate, facilitating the hydrolysis under the acidic conditions [28]. This transformation is generally high-yielding and proceeds without significant side reactions [5].

The quinoxaline demethylation represents the most challenging aspect of the deprotection sequence [5]. The methoxy groups are cleaved through nucleophilic attack by water or chloride ion, regenerating the quinoxaline-2,3-dione tautomeric forms [5]. The harsh acidic conditions are necessary to protonate the quinoxaline nitrogens and activate them toward nucleophilic substitution [5].

Post-functionalization workup involves careful neutralization and purification to obtain the final product as the hydrochloride salt [5]. The reaction mixture is evaporated under reduced pressure, and the residue is triturated with diethyl ether to remove organic impurities [5]. The crude product is then purified by recrystallization using propylene oxide to control the pH and promote crystallization [5].

Quality control of the final product involves comprehensive spectroscopic characterization including proton nuclear magnetic resonance, carbon-13 nuclear magnetic resonance, and mass spectrometry [5]. The optical rotation is measured to confirm retention of stereochemical integrity throughout the synthetic sequence [5]. Elemental analysis ensures the purity exceeds 95% for biological evaluation [5].

Purification Challenges and Yield Optimization

The purification of CNG-10300 and its synthetic intermediates presents several significant challenges that directly impact the overall synthetic efficiency [5] [45]. These challenges arise from the polar nature of the amino acid functionality, the presence of multiple protecting groups, and the tendency of quinoxaline derivatives to form stable complexes with metal impurities from the cross-coupling reaction [5] [48].

Flash column chromatography serves as the primary purification method for intermediate compounds in the CNG-10300 synthesis [5] [48]. The technique requires careful optimization of the stationary phase and mobile phase composition to achieve adequate separation while minimizing product loss [48]. Silica gel is employed as the stationary phase, with the mobile phase consisting of various ratios of ethyl acetate and petroleum ether or heptane [5] [48].

The purification of the cross-coupling product (compound 5b) required multiple chromatographic separations to achieve acceptable purity [5]. Initial purification using ethyl acetate/heptane (3:7) provided partial separation, followed by dichloromethane/methanol (95:5) and finally toluene/ethyl acetate (9:1) to achieve the desired purity [5]. This multi-step purification protocol reflects the challenging separation of closely related impurities generated during the palladium-catalyzed reaction [5].

Yield optimization strategies focus on minimizing product loss during workup and purification procedures [34] [49]. The cross-coupling reaction yields are inherently limited by the challenging nature of the transformation, with yields of 35-41% representing acceptable levels for this type of sterically hindered coupling [5]. Attempts to improve yields through catalyst loading optimization, temperature variation, and alternative ligand systems showed minimal improvement [5].

The purification of the final deprotected product (CNG-10300) employs recrystallization techniques rather than chromatography due to the highly polar nature of the amino acid [5] [47]. The crude product is initially triturated with diethyl ether to remove organic impurities, followed by controlled recrystallization using propylene oxide as a base [5]. This approach provides the final product in 68-89% yield from the protected precursor [5].

Analytical challenges in monitoring purification progress arise from the similar polarity of product and impurities [5] [45]. Thin-layer chromatography provides adequate monitoring for early intermediates, but the polar final compounds require more sophisticated analytical methods [5]. High-performance liquid chromatography becomes necessary for purity assessment of the final amino acid products [5].

The overall synthetic efficiency of CNG-10300 synthesis reflects the cumulative impact of purification challenges throughout the sequence [5]. Starting from readily available amino acid derivatives, the synthesis requires 5-6 steps with an overall yield of approximately 12-15% [5]. This yield is considered acceptable for a complex multi-step synthesis involving challenging cross-coupling and multiple protection/deprotection cycles [5].

CNG-10300, chemically designated as (2S)-2-amino-4-(2,3-dioxo-1,4-dihydroquinoxalin-6-yl)butanoic acid, demonstrates moderate binding affinity at AMPA receptor subtypes with distinctive kinetic characteristics [1] [2] [3]. At native AMPA receptors isolated from rat synaptosomes, CNG-10300 exhibits an inhibition constant (IC50) of 82 μM, indicating relatively weak competitive binding compared to classical quinoxaline-2,3-dione antagonists [1] [4]. The binding affinity was confirmed at the cloned homomeric GluA2 receptor subtype, where CNG-10300 displayed a dissociation constant (Ki) of 21 μM [1] [2] [3].

The competitive binding kinetics of CNG-10300 at AMPA receptors follow classical competitive inhibition patterns, as evidenced by radioligand displacement studies using [³H]AMPA. The compound demonstrates reversible binding characteristics with relatively rapid association and dissociation kinetics typical of competitive antagonists [1]. The binding mode analysis suggests that CNG-10300 interacts with the orthosteric binding site within the ligand-binding domain of AMPA receptors, competing directly with glutamate for receptor occupancy [2] [3].

Comparative kinetic analysis reveals that CNG-10300 displays significantly lower binding affinity at AMPA receptors compared to the classical antagonists DNQX (IC50 = 0.25-0.50 μM) and CNQX (IC50 = 0.25-0.30 μM) [4] [5]. This reduced potency at AMPA receptors contributes to the compound's unique selectivity profile within the ionotropic glutamate receptor family [1] [3].

Antagonistic Activity at Kainate Receptors (GluK1-3)

CNG-10300 demonstrates its highest binding affinity within the kainate receptor subfamily, showing preferential antagonistic activity at GluK1-3 subtypes [1] [2] [4] [3]. At homomeric GluK1 receptors, the compound exhibits a Ki value of 16 μM, while displaying its strongest binding affinity at GluK2 receptors with a Ki of 9.5 μM [1] [2] [3]. The binding affinity at GluK3 receptors is notably weaker, with a Ki value of 59 μM, indicating subtype-selective binding within the kainate receptor family [4] [6].

Functional electrophysiological studies using outside-out patches from transfected HEK293T cells reveal differential antagonistic efficacy across kainate receptor subtypes [1] [2] [3]. At a concentration of 100 μM, CNG-10300 fully antagonizes GluK1 receptors, reducing peak responses to 8% of control values, demonstrating potent functional antagonism [1] [2]. In contrast, the compound acts as a partial antagonist at GluK2 receptors, reducing peak responses to 33% of control values at the same concentration [1] [2] [3].

The native kainate receptor binding studies using [³H]kainate demonstrate that CNG-10300 exhibits a Ki value of 9.5 μM, confirming its moderate to strong affinity for this receptor class [4] [6]. The compound's antagonistic activity at kainate receptors represents its primary pharmacological mechanism, distinguishing it from other quinoxaline-2,3-dione derivatives that typically show stronger AMPA receptor affinity [1] [4].

Crystallographic analysis of CNG-10300 bound to the GluK1 ligand-binding domain reveals a Ki value of 14 μM, consistent with functional binding studies [1] [3]. The structural data indicates that CNG-10300 induces significant domain opening in the ligand-binding domain, with domain opening angles ranging from 25° to 49° across different molecules in the crystal structure [1] [2] [3].

Allosteric Modulation of NMDA Receptor Glycine Sites

CNG-10300 exhibits weak antagonistic activity at NMDA receptors with complex allosteric modulation characteristics at the glycine co-agonist binding sites [1] [2] [4]. The compound demonstrates a Ki value of 126 μM at native NMDA receptors, indicating low-affinity binding that is significantly weaker than its activity at kainate receptors [1] [4] [6].

The mechanism of NMDA receptor antagonism by CNG-10300 appears to involve interaction with the glycine modulatory site rather than the primary glutamate binding site, as evidenced by its non-competitive inhibition profile [7]. This interaction pattern is characteristic of quinoxaline-2,3-dione derivatives, which have been shown to modulate NMDA receptor function through glycine site interactions [8] [7].

Glycine co-agonist studies reveal that the presence of saturating glycine concentrations can significantly reduce the antagonistic effects of CNG-10300 at NMDA receptors, supporting the hypothesis of glycine site interaction [7]. This allosteric modulation mechanism suggests that CNG-10300 may influence NMDA receptor function through conformational changes in the receptor complex rather than direct competitive inhibition at the glutamate binding site [9] [10].

The weak NMDA receptor antagonism of CNG-10300 contributes to its overall selectivity profile, as the compound shows approximately 13-fold selectivity for kainate receptors over NMDA receptors based on binding affinity ratios [1] [4]. This selectivity pattern distinguishes CNG-10300 from classical quinoxaline antagonists such as DNQX and CNQX, which show stronger NMDA receptor interactions [5] [7].

Functional Selectivity Ratios Across iGluR Subfamilies

CNG-10300 exhibits distinctive functional selectivity ratios across ionotropic glutamate receptor subfamilies, with preferential activity toward kainate receptors over AMPA and NMDA subtypes [1] [4]. The compound demonstrates an AMPA/kainate selectivity ratio of 8.6 (82 μM vs 9.5 μM), indicating moderate selectivity for kainate receptors over AMPA receptors [1]. The NMDA/kainate selectivity ratio of 13.3 (126 μM vs 9.5 μM) represents the strongest selectivity differential within the ionotropic glutamate receptor family [4] [6].

Within the kainate receptor subfamily, CNG-10300 shows subtype-selective binding patterns with a GluK3/GluK2 selectivity ratio of 6.2 (59 μM vs 9.5 μM) and a GluK1/GluK2 selectivity ratio of 1.7 (16 μM vs 9.5 μM) [1] [2]. These ratios indicate that GluK2 represents the highest-affinity target for CNG-10300 within the kainate receptor family, followed by GluK1 and then GluK3 [1] [3].

The functional selectivity profile of CNG-10300 extends beyond simple binding affinity to include differential efficacy across receptor subtypes. Electrophysiological analysis reveals that while the compound fully antagonizes GluK1 receptors at 100 μM concentration, it only partially blocks GluK2 receptors at the same concentration, demonstrating functional selectivity that differs from binding affinity rankings [1] [2] [3].

Comparative analysis with reference compounds reveals that CNG-10300 exhibits a unique selectivity profile within the quinoxaline-2,3-dione family. Unlike classical antagonists that typically show strongest activity at AMPA receptors, CNG-10300's preferential kainate receptor binding represents a distinctive pharmacological characteristic [1] [4] [5].

Comparison with Classical Antagonists (DNQX, CNQX, ATPO)

CNG-10300's pharmacological profile presents significant differences when compared to classical ionotropic glutamate receptor antagonists DNQX, CNQX, and ATPO across multiple receptor subtypes [1] [4] [5]. DNQX demonstrates substantially higher potency at native AMPA receptors with IC50 values of 0.25-0.50 μM compared to CNG-10300's IC50 of 82 μM, representing approximately 160-fold greater potency [4] [5] [11]. Similarly, CNQX exhibits IC50 values of 0.25-0.30 μM at AMPA receptors, demonstrating comparable high potency to DNQX [4] [5] [12].

At native kainate receptors, the potency differences are less pronounced but still substantial. DNQX shows an IC50 of 2.0 μM while CNQX demonstrates IC50 values of 1.5-2.7 μM, compared to CNG-10300's Ki of 9.5 μM [4] [5]. This represents approximately 3-6 fold higher potency for the classical antagonists at kainate receptors [11] [12].

ATPO presents a different selectivity profile, demonstrating Ki values of 2.55 μM at GluK1 receptors, which is approximately 6-fold more potent than CNG-10300's Ki of 16 μM at the same receptor subtype [1] [4]. However, ATPO shows dramatically reduced activity at GluK2 and GluK3 receptors with Ki values greater than 100 μM and 1000 μM respectively, contrasting with CNG-10300's maintained activity across all kainate receptor subtypes [4].

The NMDA receptor interaction patterns reveal additional comparative differences. DNQX exhibits an IC50 of 40 μM at native NMDA receptors, while CNQX demonstrates IC50 values of 25-45 μM, both showing higher potency than CNG-10300's Ki of 126 μM [4] [5] [7]. ATPO shows minimal NMDA receptor interaction with IC50 values greater than 100 μM, similar to CNG-10300's weak NMDA activity [4].

Structural analysis reveals that CNG-10300's unique amino acid functionalization distinguishes it from the classical quinoxaline-2,3-dione antagonists. While DNQX and CNQX contain nitro and cyano substituents on the quinoxaline core, CNG-10300 incorporates an amino acid side chain that provides novel binding interactions within receptor ligand-binding domains [1] [2] [3]. This structural modification contributes to the compound's distinct selectivity profile and reduced overall potency compared to classical antagonists [1] [3].

X-ray crystallographic studies of CNG-10300 bound to the GluK1 ligand-binding domain at 2.3 Å resolution reveal an unexpected binding mode compared to classical quinoxaline antagonists [1] [2] [3]. The quinoxaline moiety maintains its role as an amino acid bioisostere, but the amino acid portion adopts a novel orientation within the receptor binding pocket, contributing to the compound's unique pharmacological characteristics [1] [3].