A New and Concise Strategy to the Enantioselective Synthesis of ( S )-2-Amino-4-Oxo-4-( Pyridine-2-yl ) Butanoic Acid from Aspartic Acid Evanoel

O a-aminoácido (S)-5 foi sintetizado usando na etapa chave uma reação de substituição nucleofílica quimiosseletiva entre um diéster derivado do ácido L-aspártico e a 2-lítio piridina. O rendimento global (13%, 5 etapas) foi semelhante ao previamente descrito por nosso grupo (12%, 10 etapas) para obtenção do isômero R (o primeiro agonista pleno exógeno de receptores do sub-tipo NMDA) a partir do D-manitol e ao da síntese racêmica relatada por Lovey e Copper (17%, 5 etapas).


Introduction
Around 99% of the essential amino acid L-tryptophan obtained in the diet is metabolized through the kynurenine pathway.Kynurenine (1a) is the first metabolite of tryptophan, being formed by oxidation of the indol ring by enzymes IDO (indoleamine dioxygenase) and TDO (tryptophan-2,3-dioxygenase).This compound can be further transformed by the enzyme KAT (kynurenine aminotransferase) into kynurenic acid, a neuroprotector compound (Figure 1). 1 On the other hand, the enzyme kynurenine 3-hydroxilase oxidizes 1a into 3-hydroxykynurenine (1b), which is further transformed in quinolinic acid, a neurotoxic agent.Quinolinic acid can be alternatively formed from 1a by the action of kyrureninase.Thus, the inhibition of kyrureninase and kynurenine 3-hydroxylase drives the tryptophan metabolism toward kynurenic acid, leading to neuroprotection. 1 It was found that quinolinic acid acts as an agonist at NMDA sub-type of glutamate receptors at the CNS while kynurenic acid acts as an antagonist at the same receptors, suggesting that these pathways could be involved in synaptic plasticity and neurodegeneration. 1 These pathways are also important in the regulation of cell proliferation and present a multitude of potential sites for drug discovery in neuroscience, oncology and visceral pathology. 2ome synthetic analogues of 1a (compounds 2-4, Figure 1) showed to be strong inhibitors of kynureninase, 1 significantly increasing the brain content of kynurenic acid and preventing the induction of seizures.Some years ago compound (R)-5 (Figure 2) was tested on NMDA receptors from rat neurons in culture, using the patch-clamp electrophysiological technique.Whole-cell currents evoked by NMDA (10 mmol L -1 ) were potentiated by the natural co-agonist glycine and also by (R)-5 in a concentration-dependent manner.When compared to glycine in the same cells, (R)-5 showed the same maximal response, but lower potency (50x larger mean effective concentration). 3,4These data suggest that (R)-5 interacts with the glycine site (GlyB) of the NMDA and is the first exogen full agonist of this receptor described in the literature. 3,4Confirming this view, currents induced by co-application of NMDA and (R)-5 were blocked by the selective GlyB antagonist 5,7-dichlorokynurenic acid (5,7-DCKA 1 mmol L -1 ). 4 Compound (R)-5 (unnatural D-configuration), was designed and previously prepared by our group from an enoate derived from D-mannitol, 3 in 12% overall yield after 10 steps. 3However, to obtain its enantiomer using the same strategy, the starting chiral enoate should be prepared from vitamin C, in more steps and lower yield. 5ovey and Copper also synthetized 5 in 5 steps and 17% overall yield, however the a-amino acid was obtained in its racemic form. 6,7he biological importance of (R)-5 as a neuroactive compound makes its synthesis a relevant task.In order to have a shorter enantioselective synthesis for this type of amino acid in both R and S configuration, we decided to use aspartic acid as starting material.We describe in this paper our efforts on the preparation of (S)-5 from more inexpensive L-aspartic acid.The key step proposed in this synthesis is a chemoselective nucleophilic substitution at a C4 methylester group in 6 and 7 by 2-lithiumpyridine, leading to a stable six membered chelated intermediates (Ia,b) which could be transformed into the corresponding ketone after workup (Figure 2).The role of b-amino groups in the control of the reaction course of an ester group toward ArLi species was previously described by our group in the synthesis of (R)-5 from D-mannitol. 3

Results and Discussion
The chemical differentiation between the two carboxyl groups in aspartic acid is a key step to use this chiron, commercially available in L and D configuration, as starting material in organic synthesis. 8The monomethyl ester 6 and the diester 7 (Figure 2) were designed as appropriate intermediates for the syntheses of amino acids type 4 through chemoselective nucleophilic acyl-substitution and were easily prepared from aspartic acid. 8,10e firstly used compound 6 as starting material, avoiding the steps of protection and deprotection at C1 acid group, which are required in the approach using 7.We realize that in basic medium the carboxy group in this intermediate would be transformed into the carboxylate, providing in situ protection of this group toward nucleophiles (Scheme 1).Firstly, we carried out the reaction using n BuLi as a model and 6 in THF at -78 ºC, which led to ketone 8 in 38% yield.The formation of tertiary alcohol was not observed in the crude product.This ketone was transformed to corresponding methyl ester 9 with diazomethane (100%) to prove that the addition had occurred chemoselectively at carbonyl group C1 and the Boc group had been removed leading to amino acid 10.After these encouraging results, we tried the addition of 2-lithiumpyridine to monoester 6 under the same conditions to obtain 11, but unfortunately a complex mixture of products was formed in this case.
Our next goal was to use diester (S)-7 to obtain amino acid (S)-5.This compound was prepared from N-Boc derivative (S)-6 by esterification with tert-butanol in the presence of DCC (Scheme 2). 9The carboxy group in 6 presented low reactivity, requiring the use of 10 equiv. of tert-butanol to consume all starting material.Diester (S)-7 was then allowed to react with 2-lithiumpyridine at -78 o C leading to the product of chemoselective nucleophilic acyl substitution at the methyl ester group (S)-13, after acidic work-up.Also in this case the formation of tertiary alcohol was not observed in the crude product.The intermediate Ib is proposed to explain the exclusive formation of ketone 12.The corresponding amino acid (S)-5 was prepared as a salt by protecting group cleavage using trifluoracetic acid and H 2 O in a 9:1 mixture.

Conclusions
Amino acid (S)-5 was prepared from commercially available aspartic acid in 12% overall yield in 5 steps, improving its stereoselective synthesis.Our approach allows the syntheses of both R and S enantiomers of 5 from the commercially available D and L aspartic acids, respectively.

Scheme 1 .Scheme 2 .
Scheme 1. Attempt to prepare ketone 11 from 6 after optimization of chemoselective addition of n BuLi to 6.