*2.2. Recent Syntheses*

Thus, kainic acid raised great interest in synthetics due its activity, wide use in experimental pharmacology, and probability to synthesize analogs and derivatives which could be applicable in medicine to treat schizophrenia and other brain diseases. Taking into account that configurations of its three stereogenic centers are crucial in binding to receptors and functional activities of this compound, the stereoselective synthesis of **2** was directed first to the optically active form of **2**, identical to the natural product. Different synthetic strategies, based on C2–C3 or C3–C4 bond formations (the latter approach was used in majority of syntheses), C4–C5 bond formation and C-*N* bond formation pathways, as well as on the use of the existing pyrrolidine ring and cycloaddition reactions, were applied to synthesize this excitatory amino acid. As a result, about 40 total multi-step syntheses of **2** were reported in literature and reviewed in 2012 [19].

In our review, we discuss only some of syntheses reported later. Most of them are based on novel schemes. We omit listing all the stages of these syntheses here and, instead, pay attention to main ideas and results achieved in comparison with the already known approaches.

Poison et al. [20] elaborated an efficient synthesis of (–)-kainic acid (**2**) using just two of the above-mentioned approaches: application of a pyrrolidine precursor and cycloaddition. A high-pressure Diels-Alder cycloaddition of the obtained from 4-hydroxy-l-prolin (**9**) 3,4-unsaturated pyrrolidine derivative **10** with Danishefsky's diene at high pressure (15 kbar) and room temperature gave bicyclic product **11**. In this case, high pressure provided the gain in reactivity and led, after some additional treatment and 82 h of exposure, to a 96% conversion into **11** that contained *N*-protected trisubstituted pyrrolidine cycle. The subsequent decarboxylation and transformation of its six-membered ring to isopropenyl and carboxymethyl groups gave **12** converted into target product via pyrrolidine derivative **13** with almost total stereocontrol and an approximately 10% yield (Scheme 1). At the same time, the key intermediate product, enone **11**, and related compounds were partly racemic in their first synthesis attempts. However, preliminary purification of the initial Danishevsky's diene by rapid distillation and removal of trace triethylamine from it made it possible to synthesize **11** with excellent yield and high enantiomeric excess.

New approaches, based on the formation of C3–C4 bond and involving the Ireland-Claisen rearrangement of allylic esters, were used by Indian and Japanese groups. Reddy and Chandraseker [21] utilized this rearrangement along with the Sharpless asymmetric epoxidation. The first of these reactions was used to create C3 and C4 *cis* stereocenters, while the Sharpless oxidation allowed the designing of chirality at C2. Authors constructed the target compound 2 from ynone **14**. At the first stages, this compound was converted into alcohol **15** by Noyori reduction and Red-Al

transformation of triple bond into double one. Reaction of **15** with 3-methyl-3-butenoate catalyzed by *N*,*N*- -dicyclohexylcarbodiimide (DCC) and 4-dimethylaminopyridine (DMAP) gave the allylic ester **16**, which was converted into **17** by the Ireland-Claisen rearrangement with lithium bis(trimethylsilyl)amide (LiHMDS) and trimethylsilyl chloride (TMSCl). The subsequent four steps of transformations led to allyl alcohol **18** with removal of p-methoxybenzyloxy (PMBO) protective group. The obtained **18** was epoxidated by asymmetric Sharpless epoxidation using (–)-diethyltratrate (DET), titanium isopropoxide Ti(iOPr)4, and *tert*-butyl hydroperoxide (TBHP) to give **19**. Further conversion provided azide **20**. Reduction of **20** allowed the formation of pyrrolidine ring, and then NH in this ring was protected by reaction with *tert*-butyloxycarbonyl anhydryde ((tBoc)2O) to obtain **21** in several steps. Removal of the protective group and oxidation with Jones reagent (CrO3 in aqueous sulfuric acid) led to kainic acid (**2**) (Scheme 2). Thus, a new strategy for constructing the core system of **2** was realized in this synthesis.

**Scheme 1.** Synthesis of kainic acid by Diels-Alder reaction.

**Scheme 2.** A shortcut scheme of synthesis of kainic acid via chirality transfer through the Ireland-Claisen rearrangement.

A similar strategy, also based on Ireland-Claisen rearrangement, was proposed by Japanese scientists as a unified approach to synthesize not only kainic acid proper, but also bioactive 4-substituted kainoids [22]. A source compound **22** was obtained from l-tartaric acid through four steps. Condensation of 22 with 3-methyl-3-butenoic acid gave ester **23** which, through the Claisen-Ireland rearrangement with LHMDS in the presence of TMSCl, was converted with high diastereoselectivity into carbonic acid **24**. The subsequent reduction of a carboxy group and other transformations gave aminomethyl derivative **25**. The resulting product was cyclized exceptionally easily through the palladium-mediated pyrrolidine-ring formation into the product **26** that had the required stereochemistry for transformation into kainic acid or its derivatives. After protection of the isopropenyl group, the corresponding conversion into **2** was achieved by oxidative cleavage of the vinyl group by ozonolysis followed by the Jones oxidation (Scheme 3).

**Scheme 3.** Synthesis of kainic acid using the Ireland-Claisen rearrangement and palladium catalyzed formation of pyrrolidine ring.

The advantage of the recent synthesis of kainic acid, developed by Japanese scientists [23], unlike most other multi-stage syntheses, consisted in only nine stages. This short total synthesis was carried out on the basis of the Cu-catalyzed Michael addition-cyclization reaction between the bearing chiral auxiliary of camphorsultam-type isonitrile **27** and ester of unsaturated ketoacid **28** (in mixture with an isomeric ester) with Cu(t-ButSal)2 as catalyst (Scheme 4). The additional ester was isomerized into **28** in the conditions of this reaction. The treatment of the obtained adduct **29** by sodium methoxide with the loss of chiral substituent, followed by protection with (Boc)2O in the presence of triethylamine and DMAP, gave unsaturated pyrrolidine **30**. The alkaline hydrolysis of the amide group at the position 2 and the selective reduction of a double bond by boron-organic reagent l-selectride led to the protected analog of kainic acid **31** which was converted into the target compound via intermediates **32**–**34**. This synthesis was carried out with a 16.8% overall yield on chiral isonitrile and was conducted on a 300 mg scale, although usually syntheses of **2** had given smaller amounts of this product.

**Scheme 4.** Short total synthesis of (–)-kainic acid.

A gram-scale synthesis of (–)-kainic acid through six steps and with a 34% overall yield was elaborated using the Pt-catalyzed direct allylic amination [24]. This amination was catalyzed by the combination of dichloro(1,5-cyclooctadiene) platinum (Pt(cod)Cl2) with bis[2-di-phenylphosphino) ether (PPEphos). Such new approach allows catalyzing the direct introduction of amino group into allylic alcohols. The chosen scheme provided the minimum number of stages in this synthesis (Scheme 5). Amination of **35** under microwave heating conditions gave monoallylamine **36**. As a protective group

(Pg), authors preferred to use 2,4-dimethoxybenzyl (DMB) or p-methoxybenzyl (PMB) with almost no loss of enantiopurity. The unsaturated ester **38** was obtained by one-pot oxidation of epoxide **37** with 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) followed by the Wittig reaction. It was found that the transformation of **36** and **38** into diallylamine derivative **39** at a higher concentration inhibited the epimerization, and the desired product was obtained in 95% yield and 98% ee (enantiomeric excess). The following stage of heating in a sealed tube with xylenes and a catalytic amount of *i*Pr2N-Et gave pyrrolidine **40** in 75% yield and 14:1 diastereoselectivity. The Jones oxidation with application of H5IO6 as terminal oxidant gave satisfactory results at conversion of **40** into **2** (Scheme 5). Finally, 1.11 g pure kainic acid was obtained by this method in 34% overall yield. Taking into account its yield and the minimal number of steps, this proved to be the best chemical scheme elaborated to synthesize **2**.

**Scheme 5.** Synthesis of kainic acid using the Pd-catalyzed direct allylic amination.

Kainic and allo-kainic acids (**2** and **3**, respectively) were also synthesized using SmI2, a unique single-electron reducing agent that allows the induction of reductive coupling with the formation of C–C bond under mild conditions [25]. The key initial products for this new synthesis were obtained from D-serine methyl ester hydrochloride using the known procedure and a mixture of E- and Z-isomers of α,β-unsaturated esters, designated as E-14 and Z-14, with a general formula of **41**. The cyclization of **41** gave derivatives of kainic and/or allo-kainic acids (**42** and **43**) with different yields depending on ratio of the reagents (SmI2, hexamethylphosphoramide (HMPA), NiI2 and H2O) and additional ligands such as ethylenediamine, 2,2- -bipyridine, 2,2- -bipyridylamine, triphenylphosphine, etc. Pure compounds were isolated in optimal conditions of this reaction for each of them followed by separation and purification of the products through HPLC. The obtained compounds were converted into kainic acid **2** or allo-kainic acid **3** through a 3-step transformation by conversion ester groups into acids and deprotection of nitrogen in pyrrolidine ring (Scheme 6).

**Scheme 6.** Syntheses of kainic and *allo*-kainic acids using SmI2-induced cyclization.

The formal total synthesis of (–)-kainic acid, recently described by Chinese chemists [26], was based on a new approach to synthesis of the same important intermediate, which was already used for synthesizing kainic acid by different authors [19,20].

In general, there have been several syntheses of kainic acid published in recent years, which confirms the constant attention to this excitatory amino acid and the rapid progress of organic synthesis. These syntheses are shorter and can provide better production of kainic acid as a pharmacological probe for application in experimental pharmacology, although the costs and availability of the synthetic precursors and reagents used were not compared with previously known syntheses. All the studies have discovered a large number of kainoids, a total of 70 [26], including those obtained through recent chemical syntheses.
