Chemoenzymatic Synthesis and Pharmacological Characterization of Functionalized Aspartate Analogues As Novel Excitatory Amino Acid Transporter Inhibitors.

Aspartate (Asp) derivatives are privileged compounds for investigating the roles governed by excitatory amino acid transporters (EAATs) in glutamatergic neurotransmission. Here, we report the synthesis of various Asp derivatives with (cyclo)alkyloxy and (hetero)aryloxy substituents at C-3. Their pharmacological properties were characterized at the EAAT1-4 subtypes. The l- threo-3-substituted Asp derivatives 13a-e and 13g-k were nonsubstrate inhibitors, exhibiting pan activity at EAAT1-4 with IC50 values ranging from 0.49 to 15 μM. Comparisons between (dl- threo)-19a-c and (dl- erythro)-19a-c Asp analogues confirmed that the threo configuration is crucial for the EAAT1-4 inhibitory activities. Analogues (3b-e) of l-TFB-TBOA (3a) were shown to be potent EAAT1-4 inhibitors, with IC50 values ranging from 5 to 530 nM. Hybridization of the nonselective EAAT inhibitor l-TBOA with EAAT2-selective inhibitor WAY-213613 or EAAT3-preferring inhibitor NBI-59159 yielded compounds 8 and 9, respectively, which were nonselective EAAT inhibitors displaying considerably lower IC50 values at EAAT1-4 (11-140 nM) than those displayed by the respective parent molecules.

Introduction

l-Glutamate (Glu) is the major excitatory neurotransmitter in the mammalian central nervous system (CNS), where it mediates numerous physiological and pathophysiological processes.[1−5] However, accumulation of high levels of extracellular Glu may lead to hyperactivity in the glutamatergic system and neuronal injury.[6] Five subtypes of excitatory amino acid transporters (termed EAAT1–5 in humans) have been identified in glial cells (predominantly EAAT1,2) and neurons (predominantly EAAT3–5), where they are key players in the regulation of glutamatergic transmission.[7] EAAT2 is the major contributor to this, as it is estimated to be responsible for over 90% of total extracellular Glu uptake in the brain, while EAAT1 and EAAT3 are also widely expressed in the CNS. Notably, EAAT4 and EAAT5 are almost specifically located in cerebellum and the retina, respectively. Malfunction of EAATs has been implicated in many neurological disorders, such as Alzheimer’s disease, epilepsy, amyotrophic lateral sclerosis, and Huntington’s disease.[8] However, in contrast to the considerable medicinal chemistry efforts in the fields of ionotropic and metabotropic Glu receptors, the EAATs have received much less attention as putative drug targets.[4]

One of the most important scaffolds for the development of EAAT ligands is the endogenous substrate l-aspartate (1, l-Asp, Figure B). Represented by l-threo-3-benzyloxyaspartate (2, l-TBOA, Figure B), 3-aryloxy substituted Asp analogues were identified as the first class of nontransportable EAAT inhibitors by Shimamoto and co-workers.[9−11] None of these analogues exhibited substantial selectivity or preference toward different EAAT subtypes. Further development of the 3-aryloxy substituted Asp analogues as pharmacological tools in neurobiological research led to the identification of (l-threo)-3-[3-[4-(trifluoromethyl)benzoylamino]benzyloxy]aspartate (3a, l-TFB-TBOA, Figure B), the most potent EAAT inhibitor reported to date.[12−14] Interestingly, elimination of the ether oxygen in l-TBOA (2) yielded the analogue l-threo-3-benzylaspartate (4, l-3-BA, Figure B), which displayed a 10-fold preference as an inhibitor for EAAT3 over EAAT1 and EAAT2.[15] Most recently, [3-(trifluoromethyl)phenyl]sulfonamide-l-aspartate (5, Figure B) was found to be a potent EAAT2 inhibitor exhibiting over 30-fold selectivity for EAAT2 over EAAT1 and EAAT3.[16] In addition, other medicinal chemistry efforts have focused on 4-carboxylate-modified Asp analogues, which resulted in the discovery of N4-[4-(2-bromo-4,5-difluorophenoxy)phenyl]-l-asparagine (6, WAY-213613) as a EAAT2-selective nonsubstrate inhibitor[17−19] and N4-(9H-fluoren-2-yl)-l-asparagine (7, NBI-59159) as an EAAT3-preferring inhibitor (Figure B).[19,20]

Figure 1

EAAT inhibitors derived from the l-Asp scaffold. (A) Structures of chemoenzymatic products and designed hybrid compounds. (B) Structures of previously developed EAAT inhibitors.

The fact that derivatization of the Asp structure has yielded several important classes of EAAT inhibitors (Figure B) encouraged us to further explore the potential of this scaffold in this field (Figure A). Here, we describe the chemoenzymatic asymmetric synthesis of a series of novel Asp derivatives comprising (cyclo)alkyloxy and (hetero)aryloxy substituents at the C-3 position, using an engineered variant of methylaspartate ammonia lyase (MAL-L384A) as the biocatalyst.[21,22] In addition, we report the synthesis of two Asp derivatives by hybridization of the nonselective EAAT inhibitor l-TBOA (2) with the EAAT2-selective inhibitor WAY-213613 (6) or the EAAT3-preferring inhibitor NBI-59159 (7). The pharmacological properties of the newly synthesized Asp derivatives, as well as those of previously synthesized analogues and homologues of l-TFB-TBOA, have been characterized at stable cell lines expressing the human EAAT1–3 and rat EAAT4 and thus the structure–activity relationships of these analogues have been elucidated.

Results and Discussion

Chemoenzymatic Synthesis of 3-Substituted Asp Analogues

The asymmetric synthesis of Asp derivatives with alkyloxy or aryloxy substituents at the C-3 position has proven to be challenging. Shimamoto and co-workers first reported the synthesis of l-TBOA (2) through an elaborate 11-step procedure.[11] A rapid chemoenzymatic methodology for asymmetric synthesis of 2 and its analogues by using MAL-L384A as biocatalyst has been developed in our group.[21,22] To further explore the substrate scope and synthetic usefulness of MAL-L384A, a series of 2-substituted fumarate derivatives with various cycloalkyloxy (12a–f), heteroaryloxy (12g–j), or alkyloxy (12k) substituents were prepared through the addition of appropriate alcohols to dimethyl acetylenedicarboxylate (10), followed by hydrolysis of the methyl esters (Table ). Interestingly, compounds 12a–e and 12g–k were accepted as substrates by MAL-L384A, giving excellent conversions (60–98%) and yielding the corresponding amino acid products in 26–58% isolated yield (Table ). Hence, MAL-L384A has a remarkably broad substrate scope, allowing the addition of ammonia to a wide variety of fumarate derivatives, providing a powerful synthetic tool for the preparation of valuable 3-substituted aspartic acids. The most bulky compound, 2-cyclooctylmethoxyfumarate (12f), was not accepted as substrate by MAL-L384A.

Table 1

Three-Step Chemoenzymatic Synthesis of 3-Substituted Asp Analogues

Conversions were determined by comparing 1H NMR signals of substrates and corresponding products.

Yield of isolated product after cation exchange chromatography.

The diastereomeric excess (de) was determined by 1H NMR.

The purified products were tentatively assigned the threo configuration on the basis of analogy.

The enantiomeric excess (ee) was determined by HPLC on a chiral stationary phase using authentic standards with known l-threo and dl-threo configuration (Supporting Information, Figures S62–S64).

Absolute configuration of products 13a, 13g and 13k were determined unambiguously by comparison of 1H NMR and chiral HPLC data to those of authentic samples with known dl-erythro, dl-threo, and l-threo configuration.

n.d., not determined.

The amino acid product 13a, which is representative for the series of chemoenzymatically prepared 3-cycloalkyloxy substituted aspartic acids, was identified as the desired threo diastereomer (de = 97%, Supporting Information, Figure S1) by comparison of its 1H NMR signals and J-coupling values to those of an authentic standard with known l-threo configuration and chemically synthesized dl-threo and dl-erythro stereoisomers (Scheme ). To determine the absolute configuration of product 13a, HPLC analysis on a chiral stationary phase was conducted by using corresponding reference molecules with known l-threo or dl-threo configuration. This analysis revealed that chemoenzymatically produced 13a was present as a single enantiomer with exclusively the l-threo configuration (ee > 99%, Supporting Information, Figure S62). Similarly, the representative chemoenzymatic products 13g and 13k were also identified as the desired l-threo isomers with excellent de (>98%) and ee (>99%) values (Table , Supporting Information, Figures S2, S3, S63, and S64). Although the relative configurations of products 13b–e and 13h–j were not determined by comparison to authentic standards, we assume the relative configurations to be threo for all enzymatic products on the basis of analogy.

Scheme 1

Synthesis of (dl-threo)- and (dl-erythro)-3-Substituted Asp Analogues

Synthesis of (dl-threo)- and (dl-erythro)-3-Substituted Asp derivatives

To confirm the importance of the relative configuration of 3-substituted Asp derivatives for inhibition of EAATs, representative compounds with dl-threo and dl-erythro configurations were synthesized according to the route given in Scheme . The building block (dl-threo)-16 was prepared based on the Sharpless aminohydroxylation procedure as previously reported.[23] Compound (dl-threo)-17, with the N-protective group transferred from Cbz to Boc, was achieved by hydrogenolysis of (dl-threo)-16 using H2/Pd/C followed by Boc-protection in an excellent yield (2 steps, 83%). To facilitate O-alkylation of 17 with the proper RBr, the strong base NaH was used to deprotonate the hydroxyl group of 17, which provided the desired (dl-threo)-18 with an isolated yield of 11–53%.[24] In addition, a small amount of epimerized product (dl-erythro)-18 was observed and separated by flash column chromatography in 4–11% isolated yield. Subsequently, global deprotection of (dl-threo)-18 was conducted via treatment with trifluoroacetic acid followed by hydrolysis of the methyl esters with LiOH, providing the desired final product (dl-threo)-19 with 34–39% isolated yield over two steps. Following the same procedure, compound (dl-erythro)-19 was obtained in 34–46% yield. Notably, (dl-threo)-19 and (dl-erythro)-19 are not only valuable molecules for exploring the stereochemistry–activity relationship of 3-substituted Asp analogues as EAAT inhibitors but also for determining the stereochemistry of the chemoenzymatically prepared products.

Design and Synthesis of Hybrid Compounds 8 and 9

l-TBOA (2) is a potent nonselective EAAT inhibitor, while WAY-213613 (6) and NBI-59159 (7) are EAAT2-selective and EAAT3-preferring inhibitors, respectively (Figure B).[17−20] We envisioned that hybridization of l-TBOA (2) with WAY-213613 (6) or NBI-59159 (7) would result in a synergistic effect on inhibitory activity at the EAATs. Thus, two hybrid compounds 8 and 9 were designed by integrating 2 with 6 or 7 (Figure A), and their synthesis was achieved according to the route presented in Scheme . The multigram scale chemoenzymatic synthesis of 2 (l-TBOA, de > 98%, ee > 99%), starting from commercially available dimethyl acetylenedicarboxylate 10 and using MAL-L384A as biocatalyst, has been previously reported.[24] Selective monoesterification at the 4-carboxylate of 2, which was accomplished under ambient condition using one equivalent of SOCl2 in dry methanol, delivered intermediate 20 without the need for purification. The 1-carboxylate of compound 20 was subsequently protected by a tert-butyl group via treatment with BF3-Et2O/tBuOAc, providing compound 21 with 87% isolated yield over two steps. Starting from 21, the chiral building block 23 was achieved in two successive reactions including Boc-protection and selective hydrolysis of the 4-methyl ester with LiOH (two steps, 42%). The desired Asp derivatives 8 and 9 were obtained via amidation of 23 with appropriate amine 24 or 25 by using EDCI/HOBT as condensation reagents followed by global deprotection in TFA/DCM with 14–32% isolated yield over two steps.

Scheme 2

Synthesis of Hybrid Compounds 8 and 9

The parent compounds for the blue and black parts of compound 8 are WAY-213613 (6) and l-TBOA (2), respectively, whereas the parent compounds for the blue and black parts of compound 9 are NBI-59159 (7) and l-TBOA (2), respectively.

Pharmacological Characterization of the Asp Derivatives at EAAT1–4

The pharmacological properties of the aspartate derivatives 13a–e, 13g–k, (dl-threo)-19a–c, and (dl-erythro)-19a–c, the hybrid molecules 8 and 9, and selected EAAT reference ligands were determined at HEK293 cell lines stably expressing human EAAT1, EAAT2, and EAAT3 (hEAAT1, hEAAT2, and hEAAT3, respectively) and at a tsA201 cell line stably expressing rat EAAT4 (rEAAT4) in a conventional [3H]-d-Asp uptake assay.[25] The rank order and absolute values of the IC50 values exhibited by the reference EAAT substrates l-Glu, l-Asp, and (l-threo)-3-hydroxyaspartate (l-THA), and the reference EAAT nonsubstrate inhibitors dl-TBOA and WAY-213613 at the four EAATs in the [3H]-d-Asp uptake assay were in good agreement with previously reported pharmacological properties (Table ).[10,11,17,25−27]

Table 2

Pharmacological Properties of EAAT Reference Ligands, 3-Substituted Aspartate Analogues and Hybrid Compounds at hEAAT1, hEAAT2, hEAAT3, and rEAAT4 in the [3H]-d-Aspartate Uptake Assaya

IC50 values are given in μM with pIC50 ± SEM in brackets, and the number of independent experiments (n) are given in superscript behind each pIC50 ± SEM value.

Overall, the 3-substituted Asp derivatives 13a–e and 13g–k were all relatively potent inhibitors of the four EAATs, and even though most of them displayed slightly lower IC50 values at hEAAT1 and hEAAT2 compared to hEAAT3 and rEAAT4, they were essentially nonselective inhibitors (Table ). The Asp derivatives displayed comparable high-nanomolar/low-micromolar IC50 values at the four transporter subtypes, the differences between the highest and lowest IC50 values in the compound series being 5.5-, 4.9-, 13-, and 9.4-fold at hEAAT1, hEAAT2, hEAAT3, and rEAAT4, respectively. Interestingly, the (l-threo)-3-substituted Asp derivatives 13a–e and 13g–k and (dl-threo)-3-substituted Asp derivatives 19a–c were equipotent with dl-TBOA (Table ). On the basis of direct comparisons between racemic Asp analogues (dl-threo)-19a–c and (dl-erythro)-19a–c, it is clear that the (dl-threo)-Asp analogues were significantly more potent inhibitors than the corresponding (dl-erythro)-Asp analogues at all four EAAT subtypes (Table ). Notably, the enantiopure Asp derivatives (l-threo)-13a,g,k only displayed slightly lower IC50 values than their racemic counterparts (dl-threo)-19a–c (Table ). These results are in line with previous observations that l-TBOA and l-TFB-TBOA were more potent inhibitors of EAATs than their corresponding (l-erythro)-diastereomers.[11,28] These structure–activity relationships thus demonstrate that essentially every (cyclo)alkyloxy and (hetero)aryloxy substituent in the 3-position of Asp yields a potent EAAT inhibitor and also confirm that the relative threo configuration of 3-substituted Asp analogues is crucial for their inhibitory activities at EAATs.

The retained EAAT inhibitory activity in the Asp analogues 13a–e and 13g–k regardless of the size of their respective 3-substituent contrasts the substantial gradual reduction in EAAT activity observed upon the introduction of 4-substituents of increasing size into Glu.[29,30] It seems reasonable to ascribe these SAR differences to the substituents in the 3-substituted Asp analogue and the 4-substituted Glu analogue projecting out into different regions of the EAAT substrate binding pocket. Perhaps more interesting, the complete lack of subtype-selectivity or -preference exhibited by all of the analogues in this study (13a–e and 13g–k) differs remarkably from our recent findings for a series of sulfonamido functionalized 3-substituted aspartate analogues that comprises both EAAT1-preferring and EAAT2-selective inhibitors.[16] It seems that even when comparing different 3-substituted Asp analogues, different functionalities (an ether vs a sulfonamide group) in this position of the Asp molecule will result in the substituents being projected out into different binding pocket regions, which again is reflected in markedly different subtype-selectivity profiles across the EAATs.

Whereas Asp (both l-Asp and d-Asp) and its derivative (l-threo)-3-hydroxy Asp (l-THA) are substrates for EAAT1–4, the (dl-threo)-3-benzyloxy Asp (dl-TBOA) analogue is a nonsubstrate inhibitor of all four EAAT subtypes.[10,11,17,25−27] Because the series of Asp analogues in the present study comprised analogues with 3-substituents covering a considerable spectrum in terms of sizes, we decided to test these analogues for both substrate and nonsubstrate inhibitor activity at the hEAAT1–3 cell lines in the fluorescence-based FLIPR Membrane Potential Blue (FMP) assay. This assay measures the membrane potential changes induced in the cell lines by the substrate translocation and the concomitant cotransport of Na+, H+, and the counter-transport of K+ mediated by the EAAT. In a previous study, we have found the assay to be capable of distinguishing substrates from nonsubstrate inhibitors within a series of reference EAAT ligands.[25] None of the 3-substituted Asp analogues (13a–e and 13g–k) were substrates at hEAAT1–3, as they did not evoke a significant increase in the fluorescence levels in the assay when applied on their own (all at concentrations of 300 μM) at hEAAT1-, hEAAT2-, and hEAAT3-HEK293 cell lines (data not shown). In contrast, preincubation and coapplication of the analogues (all at concentrations of 300 μM) together with l-Glu EC80 completely eliminated the Glu-induced response at all three transporters (data not shown), and thus all 3-substituted Asp analogues were nonsubstrate inhibitors of the EAATs. Given that l-THA is a substrate for EAAT1–4, these findings clearly demonstrate that the borderline between when the Asp analogue can be transported by the EAAT or not lies right between the 3-hydroxy substituent (l-THA) and 3-substituents such as cyclopropyl (13a), prop-2-yn (13k), and larger groups in the series of Asp analogues tested here (Figure , Table ).

The IC50 values displayed by the TBOA/WAY-213613 hybrid, compound 8, at hEAAT1, hEAAT2, hEAAT3, and rEAAT4 in the [3H]-d-Asp uptake assay were 190-, 53-, 40-, and 16-fold lower than those of dl-TBOA and 78-, 2.0-, 20-, and 11-fold lower than those of WAY-213613 (6), respectively (Table , Figure ). Thus, introduction of a benzyloxy group into the 3-position of WAY213613 converts this EAAT2-selective inhibitor into a nonselective but much more potent EAAT inhibitor. Hybrid compound 9, derived from TBOA (2) and NBI-59159 (7), displayed IC50 values of 12, 47, 55, and 80 nM and was thus 180-, 40-, 69-, and 29-fold more potent as inhibitor at hEAAT1, hEAAT2, hEAAT3, and rEAAT4, respectively, than dl-TBOA. Because we did not test NBI-59159 (7) in the [3H]-d-Asp uptake assay, we cannot make a direct comparison between the inhibitory potencies displayed by 9 and those of this other parent compound. However, compared to the reported inhibitory data of NBI-59159 (7),[19] the hybrid analogue 9 displayed 83-, 30-, and 1.6-fold increased inhibitory potencies at EAAT1, EAAT2, and EAAT3, respectively (Table , Figure ). Analogously to the observation for the hybrid analogue 8 and its parent compound WAY-213613, introducing a benzyloxy group at the 3-position of NBI-59159 converted this EAAT3-preferring inhibitor into a nonselective but considerably more potent EAAT inhibitor.

Figure 2

Functional properties of the hybrid analogues 8 and 9 as EAAT inhibitors. Representative concentration–inhibition curves for 8, 9, dl-TBOA (parent compound of both 8 and 9), and WAY-213613 (6, parent compound of 8) at hEAAT1-HEK293, hEAAT2-HEK293, hEAAT3-HEK293, and rEAAT4-tsA201 cells in the [3H]-d-Asp uptake assay. Data are from a representative specific experiment and given as mean ± SD values (based on duplicate determinations).

The chemoenzymatic synthesis of l-TFB-TBOA (3a) and its four derivatives 3b–e was reported elsewhere.[24] In agreement with the original study of l-TFB-TBOA,[12] the chemoenzymatically prepared compound 3a was found to be a potent EAAT inhibitor displaying IC50 values of 3.6, 10, 120, and 40 nM at hEAAT1, hEAAT2, hEAAT3, and rEAAT4, respectively, thus displaying some preference for hEAAT1,2 over hEAAT3 (Table ). Whereas substitution of the p-CF3 group in l-TFB-TBOA with o-CF3 (3b) resulted in significantly decreased inhibitory potencies at the EAATs, the m-CF3 analogue (3c) displayed comparable IC50 values as l-TFB-TBOA (3a) at all four transporter subtypes. While extension of the functional group at the C-3 position in l-TFB-TBOA with one carbon (3d) resulted in 5–10-fold increased IC50 values at hEAAT1, hEAAT2, and rEAAT4, the homologue retained the inhibitory potency of l-TFB-TBOA at hEAAT3. Interestingly, extension of the C-3 functional group with two carbons resulted in a homologue (3e) with similar inhibitory properties at the four EAATs as l-TFB-TBOA (Table ).

Conclusion

We have presented the design, synthesis, and pharmacological characterization of an elaborate series of 3-substituted Asp analogues as inhibitors of hEAAT1, hEAAT2, hEAAT3, and rEAAT4. Inspired by the potential of the Asp scaffold for the development of EAAT inhibitors, various new Asp derivatives with (cyclo)alkyloxy and (hetero)aryloxy substituents at the C-3 position were synthesized by using a key stereoselective enzymatic step. Remarkably, all these Asp derivatives were found to be potent nonsubstrate pan inhibitors of the EAATs. The functional properties exhibited by the Asp analogues also provide insight into the relation between ligand structure and EAAT transport. Whereas Asp and (l-threo)-3-hydroxy Asp (l-THA) are substrates of EAATs, it is clear from this series that basically any Asp derivative with a substituent at C-3 larger than a hydroxyl group will be a nonsubstrate inhibitor. Because all of these compounds are Asp derivatives, it is reasonable to assume that they function as competitive inhibitors, like dl-TBOA,[25] and act through the substrate binding site in the transporters. On the other hand, judging from the fact that all 3-substituted Asp analogues comprising a wide range of substituents displayed comparable inhibitory potencies, the 3-substituent either does not contribute substantially to this inhibitory potency by forming interactions with residues in the substrate binding site of the transporter or alternatively all of the 3-substituents in the tested analogues are able to form similar interactions. Finally, our results confirm that the threo configuration of the 3-substituted Asp analogue is crucial for its high inhibitory activities at EAATs.

Unique EAAT inhibitors were developed by hybridization of the nonselective EAAT inhibitor l-TBOA (2) with the EAAT2-selective inhibitor WAY-213613 or the EAAT3-preferring inhibitor NBI-59159 to yield hybrid analogues 8 and 9, respectively. Compounds 8 and 9 displayed significantly higher inhibitory activities at EAATs than both of their respective parent structures, although the hybridization in both cases leads to pan inhibitors of EAAT1–4. Thus, while the additional interactions with the transporter formed by the benzyloxy group at the 3-position in 8/9 in both cases leads to increased inhibitory potency, it also seems to induce a binding conformation of the Asp analogue that dictates a different spatial orientation of the 4-(2-bromo-4,5-difluorophenoxy)phenyl/(9H-fluoren-2-yl) substituent than in WAY-213613/NBI-59159. We propose that this and other hybridization strategies could advance future design and development of more potent EAAT inhibitors.

We recently reported the synthesis and evaluation of l-TFB-TBOA-based inhibitors of the prokaryotic aspartate transporter GltTk with photocontrolled activity, allowing the remote, reversible, and spatiotemporally resolved regulation of transport.[31] Work is in progress to convert the hybrid compounds 8 and 9, which were shown here to be highly potent nonselective EAAT inhibitors, into light-controlled glutamate transporter inhibitors by introducing a photoswitchable azobenzene moiety.

Experimental Section

Chemicals and General Methods

All chemicals were purchased from commercial sources unless otherwise specified. All the solvents were of analytical reagent grade and were used without further purification. 1H NMR, 13C NMR, and 19F NMR analysis were performed on a Bruker Avance III 400/500 MHz spectrometer. High resolution mass spectra (HRMS) were recorded on a LTQ Orbitrap XL. Flash chromatography was performed on a GRACE X2 system with silica gel (200–300 mesh) purchased from Merck. High performance liquid chromatography (HPLC) analysis was performed with a Shimadzu LC-10AT HPLC with a Shimadzu SP-M10A ELSD detector and a Shimadzu SPD-M10A photodiode array detector. Analytical HPLC was performed using a Kinetex C18 column (150 mm × 4.6 mm, 5 μm) with 5–95% MeCN gradient in H2O (0.5% TFA) as mobile phase. The purity of all the target compounds was determined to be >95%. Chiral HPLC analysis for compound 13g was perfomed using Nucleosil Chiral-1 column (250 mm × 4 mm, 5 μm) with 0.5 mM aqueous CuSO4 as mobile phase. Chiral HPLC analysis for compounds 13a and 13k was perfomed using CROWNPAK CR-I (+) column (150 mm × 3 mm, 5 μm) with isocratic MeCN/H2O (98%, v/v, 0.5% TFA) as mobile phase.

Synthesis of Hybrid Compounds 8 and 9

(l-threo)-2-Amino-3-(benzyloxy)-4-[[4-(2-bromo-4,5-difluorophenoxy)phenyl]amino]-4-oxobutanoic Acid (8)

To a stirred solution of 26 (68 mg, 0.1 mmol, see below) in dry DCM (2 mL), in an ice-bath, was added trifluoroacetic acid (2 mL) dropwise. After the complete addition of trifluoroacetic acid, the ice-bath was removed and the reaction was allowed to proceed at room temperature for further 6 h. After completion of the starting material, the solvent was removed in vacuo to provide pure 8 as a trifluoroacetate salt (brown solid, 34 mg, yield 52%). 1H NMR (500 MHz, D2O/DMSO-d6): δ 7.74 (dd, J = 9.8, 8.3 Hz, 1H), 7.43–7.41 (m, 2H), 7.32–7.25 (m, 5H), 7.10–7.02 (m, 1H), 6.89–6.88 (m, 2H), 4.62–4.56 (m, 2H), 4.46 (d, J = 4.0 Hz, 1H), 4.16 (d, J = 3.9 Hz, 1H). 13C NMR (126 MHz, DMSO-d6): δ 168.1, 166.4, 152.4, 149.5 (dd, J = 8.2, 3.8 Hz), 148.5 (dd, J = 374.9, 13.9 Hz), 148.6 (dd, J = 372.3, 13.9 Hz), 136.9, 134.2, 128.2 (2C), 128.0 (2C), 127.9, 122.0, 121.9, 117.8 (2C), 115.3 (d, J = 21.4 Hz), 110.2 (d, J = 20.2 Hz), 108.6 (dd, J = 7.6, 3.8 Hz), 76.8, 72.1, 54.1. 19F NMR (376 MHz, DMSO-d6): δ −73.51 (s), −134.99 (dt, J = 20.6, 10.1 Hz), −141.25 (dt, J = 23.2, 9.0 Hz). HRMS: calcd for C23H20BrF2N2O5 [M + H]+ 521.0518, found 521.0520. HPLC: purity 98%, retention time 9.0 min.

(l-threo)-4-[(9H-Fluoren-2-yl)amino]-2-amino-3-(benzyloxy)-4-oxobutanoic Acid (9)

Hybrid compound 9 was prepared from 27 (100 mg, 0.18 mmol, see below) following a procedure similar to that used for compound 8. The title product was obtained as a white solid (32 mg, 35%). 1H NMR (500 MHz, DMSO-d6): δ 10.35 (s, 1H), 7.99 (s, 1H), 7.84 (t, J = 7.8 Hz, 2H), 7.58 (dd, J = 13.4, 7.9 Hz, 2H), 7.42 (d, J = 7.3 Hz, 2H), 7.38–7.34 (m, 3H), 7.32–7.27 (m, 2H), 4.70 (d, J = 11.7 Hz, 1H), 4.63 (d, J = 4.1 Hz, 1H), 4.59 (d, J = 11.7 Hz, 1H), 4.06 (bs, 1H), 3.92 (s, 2H). 13C NMR (126 MHz, methanol-d4): δ 169.5, 168.4, 145.4, 144.5, 142.4, 140.0, 138.7, 137.7, 137.4, 129.7 (2C), 129.6, 127.9, 127.7, 126.0, 121.0, 120.6, 120.5, 118.6, 107.3, 77.7, 74.6, 56.1, 37.7. HRMS: calcd for C24H23N2O4 [M + H]+ 403.1652, found 403.1653. HPLC: purity 98%, retention time 8.4 min.

Synthesis of Dimethyl 2-Substituted Fumarate Derivatives 11a–k

The synthesis of the dimethyl 2-substituted fumarate derivatives 11a–k was based on a previously published procedure.[22,32] Briefly, to a stirred solution of dimethyl acetylenedicarboxylate (10, 568 mg, 4 mmol) in dichloromethane (30 mL) was added DABCO (45 mg, 0.4 mmol) followed by the appropriate alcohol (5 mmol). The reaction mixture was stirred at room temperature for 12 h. The solvent was removed under reduced pressure to give the crude product as a dark-brown oil. The trans/cis isomers of the products were separated by flash chromatography (silica gel) using 5% EtOAc/petroleum ether (boiling point 40–60 °C) as eluent. The preferred trans-isomers 11a–k (i.e., the fumarate derivatives) were used in further experiments.

Dimethyl 2-(Cyclopropylmethoxy)fumarate (11a)

Clear oil, 230 mg (27% yield). 1H NMR (500 MHz, CDCl3): δ 6.11 (s, 1H), 3.84 (d, J = 7.2 Hz, 2H), 3.72 (s, 3H), 3.63 (s, 3H), 1.16–1.08 (m, 1H), 0.49–0.45 (m, 2H), 0.21–0.18 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 164.5, 163.4, 154.1, 109.6, 78.4, 52.6, 51.4, 10.6, 3.0 (2C).

Dimethyl 2-(Cyclobutylmethoxy)fumarate (11b)

Clear oil, 164 mg (18% yield). 1H NMR (500 MHz, CDCl3): δ 6.11 (s, 1H), 4.00 (d, J = 6.7 Hz, 2H), 3.77 (s, 3H), 3.68 (s, 3H), 2.69–2.61 (m, 1H), 2.04–1.97 (m, 2H), 1.87–1.76 (m, 4H). 13C NMR (126 MHz, CDCl3): δ 164.7, 163.4, 154.5, 108.9, 77.6, 52.7, 51.5, 35.1, 24.5 (2C), 18.3.

Dimethyl 2-(Cyclopentylmethoxy)fumarate (11c)

Clear oil, 136 mg (14% yield). 1H NMR (500 MHz, CDCl3): δ 6.15 (s, 1H), 3.96 (d, J = 6.9 Hz, 2H), 3.82 (s, 3H), 3.74 (s, 3H), 2.34–2.25 (m, 1H), 1.81–1.75 (m, 2H), 1.62–1.53 (m, 4H), 1.35–1.29 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 164.9, 163.7, 154.7, 108.6, 78.0, 52.9, 51.8, 39.8, 29.2 (2C), 25.6 (2C).

Dimethyl 2-(Cyclohexylmethoxy)fumarate (11d)

Clear oil, 130 mg (13% yield). 1H NMR (500 MHz, CDCl3): δ 6.13 (s, 1H), 3.88 (d, J = 6.4 Hz, 2H), 3.82 (s, 3H), 3.74 (s, 3H), 1.84–1.66 (m, 6H), 1.30–1.14 (m, 3H), 1.05–0.97 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 165.0, 163.7, 154.9, 108.1, 79.4, 52.9, 51.8, 38.5, 29.6 (2C), 26.6, 25.9 (2C).

Dimethyl 2-(Cycloheptylmethoxy)fumarate (11e)

Clear oil, 83 mg (8% yield). 1H NMR (500 MHz, CDCl3): δ 6.12 (s, 1H), 3.85 (d, J = 6.6 Hz, 2H), 3.81 (s, 3H), 3.72 (s, 3H), 1.93–1.86 (m, 1H), 1.83–1.77 (m, 2H), 1.69–1.63 (m, 2H), 1.58–1.54 (m, 2H), 1.51–1.39 (m, 4H), 1.27–1.20 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 164.9, 163.6, 154.7, 108.3, 79.2, 52.9, 51.7, 39.9, 30.7 (2C), 28.6 (2C), 26.5 (2C).

Dimethyl 2-(Cyclooctylmethoxy)fumarate (11f)

Clear oil, 56 mg (5% yield). 1H NMR (500 MHz, CDCl3): δ 6.12 (s, 1H), 3.85 (d, J = 6.6 Hz, 2H), 3.82 (s, 3H), 3.73 (s, 3H), 1.97–1.91 (m, 1H), 1.74–1.65 (m, 4H), 1.62–1.46 (m, 8H), 1.36–1.28 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 164.9, 163.7, 154.8, 108.2, 79.6, 52.9, 51.7, 38.2, 29.1 (2C), 27.0 (2C), 26.5, 25.5 (2C).

Dimethyl 2-(Thiophen-3-ylmethoxy)fumarate (11g)

Light-yellow solid, 275 mg (27% yield). 1H NMR (500 MHz, CDCl3): δ 7.35–7.34 (m, 1H), 7.30 (dd, J = 5.0, 3.0 Hz, 1H), 7.16 (dd, J = 5.0, 1.3 Hz, 1H), 6.23 (s, 1H), 5.22 (s, 2H), 3.82 (s, 3H), 3.74 (s, 3H). 13C NMR (126 MHz, CDCl3): δ 164.7, 163.4, 153.6, 137.2, 127.6, 126.3, 124.6, 110.4, 70.2, 53.0, 51.8.

Dimethyl 2-(Thiophen-2-ylmethoxy)fumarate (11h)

Light-yellow solid, 310 mg (38% yield). 1H NMR (500 MHz, CDCl3): δ 7.33 (dd, J = 5.1, 1.2 Hz, 1H), 7.08 (d, J = 3.4 Hz, 1H), 6.98 (dd, J = 5.1, 3.5 Hz, 1H), 6.28 (s, 1H), 5.37 (s, 2H), 3.83 (s, 3H), 3.75 (s, 3H). 13C NMR (126 MHz, CDCl3): δ 164.7, 163.4, 153.0, 138.1, 128.53, 127.4, 126.9, 111.5, 69.2, 53.0, 51.9.

Dimethyl 2-(Furan-3-ylmethoxy)fumarate (11i)

Light-yellow solid, 345 mg (36% yield). 1H NMR (500 MHz, CDCl3): δ 7.46 (dd, J = 1.7, 0.9 Hz, 1H), 7.39 (t, J = 1.7 Hz, 1H), 6.49 (d, J = 1.2 Hz, 1H), 6.24 (s, 1H), 5.10 (s, 2H), 3.82 (s, 3H), 3.74 (s, 3H). 13C NMR (126 MHz, CDCl3): δ 164.7, 163.5, 153.5, 143.6, 141.8, 120.8, 110.6 (2C), 66.6, 53.0, 51.8.

Dimethyl 2-(Furan-2-ylmethoxy)fumarate (11j)

Light-yellow solid, 382 mg (40% yield). 1H NMR (500 MHz, CDCl3): δ 7.42 (dd, J = 1.9, 0.9 Hz, 1H), 6.41 (d, J = 3.3 Hz, 1H), 6.35 (dd, J = 3.3, 1.8 Hz, 1H), 6.29 (s, 1H), 5.17 (s, 2H), 3.83 (s, 3H), 3.73 (s, 3H). 13C NMR (126 MHz, CDCl3): δ 164.6, 163.4, 153.0, 149.7, 143.7, 111.4, 111.3, 110.6, 66.7, 53.0, 51.7.

Dimethyl 2-(Prop-2-yn-1-yloxy)fumarate (11k)

Clear oil, 295 mg (37% yield). 1H NMR (500 MHz, CDCl3): δ 6.35 (s, 1H), 4.90 (d, J = 2.5 Hz, 2H), 3.84 (s, 3H), 3.75 (s, 3H), 2.55 (t, J = 2.5 Hz, 1H). 13C NMR (126 MHz, CDCl3): δ 164.5, 163.2, 152.0, 111.6, 82.9, 77.1, 60.4, 53.1, 52.0.

Synthesis of 2-Substituted Fumaric Acid Derivatives 12a–k

To a stirred solution of dimethyl 2-substituted fumarate 11 (11a, 214 mg, 1.0 mmol; 11b, 150 mg, 0.66 mmol; 11c, 121 mg, 0.50 mmol; 11d, 115 mg, 0.45 mmol; 11e, 70 mg, 0.26 mmol; 11f, 56 mg, 0.20 mmol; 11g, 112 mg, 0.44 mmol; 11h, 87 mg, 0.34 mmol; 11i, 62 mg, 0.26 mmol; 11j, 70 mg, 0.30 mmol; 11k, 198 mg, 1.0 mmol, respectively) in EtOH (2 mL) was added 2 M NaOH (2 mL), and the reaction mixture was heated to reflux for 2 h. After completion of the hydrolysis, the reaction mixture was allowed to cool to room temperature followed by removing the EtOH under vacuum. For compounds 11a–f and 11k, the resulting aqueous layer was acidified with HCl (1 M) until white precipitates appeared (pH 1). The precipitates were filtered and dried under vacuum to provide pure products 12a–f and 12k as white solids. For compounds 12g–j, the resulting aqueous solutions were directly used for the next enzymatic step after adjusting the pH of the solution to pH 9.5 with 1 M HCl.

2-(Cyclopropylmethoxy)fumaric Acid (12a)

White solid, 150 mg (81% yield). 1H NMR (500 MHz, DMSO-d6): δ 6.02 (s, 1H), 3.85 (d, J = 7.1 Hz, 2H), 1.13–1.07 (m, 1H), 0.51–0.47 (m, 2H), 0.25–0.22 (m, 2H). 13C NMR (126 MHz, DMSO-d6): δ 165.4, 164.2, 153.8, 109.8, 76.8, 10.6, 3.0 (2C).

2-(Cyclobutylmethoxy)fumaric Acid (12b)

White solid, 110 mg (84% yield). 1H NMR (500 MHz, methanol-d4): δ 6.07 (s, 1H), 4.02 (d, J = 5.6 Hz, 2H), 2.68–2.62 (m, 1H), 2.04–2.00 (m, 2H), 1.91–1.83 (m, 4H). 13C NMR (126 MHz, methanol-d4): δ 167.9, 165.8, 156.5, 110.0, 78.2, 36.6, 25.5 (2C), 19.2.

2-(Cyclopentylmethoxy)fumaric Acid (12c)

White solid, 77 mg (72% yield). 1H NMR (500 MHz, methanol-d4): δ 6.00 (s, 1H), 3.90 (d, J = 7.0 Hz, 2H), 2.24–2.15 (m, 1H), 1.72–1.66 (m, 2H), 1.57–1.47 (m, 4H), 1.31–1.23 (m, 2H). 13C NMR (126 MHz, methanol-d4): δ 168.0, 165.8, 156.7, 109.3, 78.4, 41.0, 30.1 (2C), 26.4 (2C).

2-(Cyclohexylmethoxy)fumaric Acid (12d)

White solid, 78 mg (76% yield). 1H NMR (500 MHz, DMSO-d6): δ 5.93 (s, 1H), 3.81 (d, J = 6.4 Hz, 2H), 1.74–1.56 (m, 6H), 1.24–1.10 (m, 3H), 1.00–0.92 (m, 2H). 13C NMR (126 MHz, DMSO-d6): δ 165.8, 164.6, 154.8, 108.5, 78.2, 38.3, 29.3 (2C), 26.3, 25.6 (2C).

2-(Cycloheptylmethoxy)fumaric Acid (12e)

White solid, 42 mg (67% yield). 1H NMR (500 MHz, methanol-d4): δ 6.05 (s, 1H), 3.86 (d, J = 6.4 Hz, 2H), 1.88–1.78 (m, 3H), 1.70–1.65 (m, 2H), 1.60–1.43 (m, 6H), 1.29–1.23 (m, 2H). 13C NMR (126 MHz, methanol-d4): δ 168.0, 165.8, 156.5, 109.1, 79.6, 41.1, 31.7 (2C), 29.7 (2C), 27.5 (2C).

2-(Cyclooctylmethoxy)fumaric Acid (12f)

White solid, 20 mg (39% yield). 1H NMR (500 MHz, methanol-d4): δ 6.03 (s, 1H), 3.82 (d, J = 6.7 Hz, 2H), 1.93–1.86 (m, 1H), 1.72–1.64 (m, 4H), 1.59–1.44 (m, 8H), 1.33–1.27 (m, 2H). 13C NMR (126 MHz, methanol-d4): δ 168.0, 165.8, 156.6, 109.0, 80.1, 39.3, 30.2 (2C), 28.0 (2C), 27.7, 26.5 (2C).

Disodium 2-(Thiophen-3-ylmethoxy)fumarate (12g)

1H NMR (500 MHz, DMSO-d6): δ 7.53–7.50 (m, 2H), 7.15 (dd, J = 4.9, 1.3 Hz, 1H), 6.05 (s, 1H), 5.10 (s, 2H). 13C NMR (126 MHz, DMSO-d6): δ 165.3, 164.0, 153.5, 137.7, 127.6, 126.6, 124.2, 110.2, 69.1.

Disodium 2-(Thiophen-2-ylmethoxy)fumarate (12h)

1H NMR (500 MHz, D2O): δ 7.42 (dd, J = 5.0, 1.3 Hz, 1H), 7.08–7.05 (m, 1H), 7.00 (dd, J = 5.1, 3.5 Hz, 1H), 5.91 (s, 1H), 5.14 (s, 2H). 13C NMR (126 MHz, D2O): δ 173.9, 170.9, 152.4, 138.7, 128.6, 127.5, 126.9, 114.4, 66.6.

Disodium 2-(Furan-3-ylmethoxy)fumarate (12i)

1H NMR (500 MHz, D2O): δ 7.50 (dd, J = 1.6, 0.8 Hz, 1H), 7.44 (t, J = 1.9 Hz, 1H), 6.49 (s, 1H), 5.89 (s, 1H), 4.85 (s, 2H). 13C NMR (126 MHz, D2O): δ 174.0, 171.0, 152.7, 143.6, 142.0, 120.8, 113.8, 110.6, 63.8.

Disodium 2-(Furan-2-ylmethoxy)fumarate (12j)

1H NMR (500 MHz, D2O): δ 7.45 (dd, J = 1.9, 0.8 Hz, 1H), 6.38 (dd, J = 3.3, 0.8 Hz, 1H), 6.36 (dd, J = 3.3, 1.9 Hz, 1H), 5.91 (s, 1H), 4.89 (s, 2H). 13C NMR (126 MHz, D2O): δ 173.8, 170.8, 152.2, 150.0, 143.8, 114.6, 111.2, 110.5, 64.2.

2-(Prop-2-yn-1-yloxy)fumaric Acid (12k)

White solid, 142 mg (83% yield). 1H NMR (500 MHz, DMSO-d6): δ 6.14 (s, 1H), 4.82 (d, J = 2.5 Hz, 2H), 3.63 (t, J = 2.5 Hz, 1H). 13C NMR (126 MHz, DMSO-d6): δ 165.1, 163.7, 151.8, 111.7, 79.4, 78.4, 59.1.

Enzymatic Synthesis of 3-Substituted Aspartic Acid Derivatives 13a–k

Reaction mixtures consisted of 2-substituted fumaric acid derivatives 12 (12a, 21 mg, 0.11 mmol; 12b, 40 mg, 0.20 mmol; 12c, 42 mg, 0.20 mmol; 12d, 28 mg, 0.12 mmol; 12e, 23 mg, 0.095 mmol; 12f, 12 mg, 0.047 mmol; 12g, 0.44 mmol; 12h, 0.34 mmol; 12i, 0.26 mmol; 12j, 0.30 mmol; 12k, 9.7 mg, 0.056 mmol, respectively) in buffer (5 M NH3/NH4Cl and 20 mM MgCl2, pH adjusted to pH 9.5). The reaction was started by addition of freshly purified MAL-L384A mutant (0.01 mol %),[21,22] and the reaction mixture was incubated at room temperature for 24 h. After completion of the reaction, the enzyme was inactivated by heating to 70 °C for 10 min. The progress and conversion yield of the enzymatic reaction was monitored by 1H NMR spectroscopy by comparing signals of substrates and corresponding products. For a typical purification procedure, the precipitated enzyme was removed by filtration. The filtrate was acidified with 1 M HCl to pH 1 and loaded onto a column packed with cation-exchange resin (10 g of Dowex 50W X8, 50–100 mesh), which was pretreated with 2 M aqueous ammonia (4 column volumes), 1 M HCl (2 column volumes), and distilled water (4 column volumes). The column was washed with distilled water (2 column volumes), and the desired product was eluted with 2 M aqueous ammonia (4 column volumes). The ninhydrin-positive fractions were collected and lyophilized to yield the product as ammonium salt.

(l-threo)-3-(Cyclopropylmethoxy)aspartate (13a)

White solid, 7 mg (conversion 88%, isolated yield 31%). 1H NMR (500 MHz, D2O): δ 4.38 (d, J = 2.5 Hz, 1H), 3.98 (d, J = 2.5 Hz, 1H), 3.45 (dd, J = 10.7, 7.0 Hz, 1H), 3.28 (dd, J = 10.7, 7.2 Hz, 1H), 1.05–0.97 (m, 1H), 0.53–0.46 (m, 2H), 0.20–0.15 (m, 2H). 13C NMR (126 MHz, D2O): δ 176.1, 171.6, 77.4, 75.7, 56.6, 9.5, 2.6, 2.1. HRMS: calcd for C8H14NO5 [M + H]+ 204.0866, found 204.0861. HPLC: purity 96%, retention time 4.2 min. Comparison of the 1H NMR data of 13a with the 1H NMR data of chemically prepared racemic (dl-threo)-19a and (dl-erythro)-19a showed that the de of product 13a was 97% (Supporting Information, Figure S1). Chiral HPLC analysis: Crownpak CR-I (+) 150 mm × 3 mm. Phase A, ACN + 0.5% TFA; phase B, H2O + 0.5% TFA, A/B = 98:2. Flow rate 0.4 mL/min, column temperature 25 °C, detected by ELSD at 35 °C, tR = 2.8 min, ee > 99% (Supporting Information, Figure S62).

3-(Cyclobutylmethoxy)aspartate (13b)

White solid, 25 mg (conversion 89%, isolated yield 58%). 1H NMR (500 MHz, D2O): δ 4.24 (d, J = 2.4 Hz, 1H), 3.95 (d, J = 2.4 Hz, 1H), 3.60 (dd, J = 9.8, 6.8 Hz, 1H), 3.39 (dd, J = 9.8, 7.3 Hz, 1H), 2.56–2.47 (m, 1H), 2.01–1.95 (m, 2H), 1.86–1.77 (m, 2H), 1.68–1.61 (m, 2H). 13C NMR (126 MHz, D2O): δ 176.4, 171.7, 78.3, 75.6, 56.5, 34.1, 24.5, 24.2, 18.0. HRMS: calcd for C9H16NO5 [M + H]+ 218.1023, found 218.1022. HPLC: purity 98%, retention time 6.5 min.

3-(Cyclopentylmethoxy)aspartate (13c)

White solid, 19 mg (conversion 85%, isolated yield 39%). 1H NMR (500 MHz, D2O): δ 4.28 (d, J = 2.5 Hz, 1H), 3.98 (d, J = 2.5 Hz, 1H), 3.54 (dd, J = 9.4, 7.0 Hz, 1H), 3.25 (dd, J = 9.4, 7.7 Hz, 1H), 2.17–2.11 (m, 1H), 1.70–1.63 (m, 2H), 1.56–1.47 (m, 4H), 1.22–1.13 (m, 2H). 13C NMR (126 MHz, D2O): δ 176.2, 171.5, 78.2, 75.7, 56.4, 38.5, 28.9, 28.7, 24.9 (2C). HRMS: calcd for C10H18NO5 [M + H]+ 232.1180, found 232.1180. HPLC: purity 97%, retention time 9.3 min.

3-(Cyclohexylmethoxy)aspartate (13d)

White solid, 17 mg (conversion 90%, isolated yield 56%). 1H NMR (500 MHz, D2O): δ 4.28 (d, J = 2.6 Hz, 1H), 3.99 (d, J = 2.5 Hz, 1H), 3.48 (dd, J = 9.6, 6.3 Hz, 1H), 3.20 (dd, J = 9.6, 6.9 Hz, 1H), 1.69–1.55 (m, 6H), 1.25–1.11 (m, 3H), 0.91–0.85 (m, 2H). 13C NMR (126 MHz, D2O): δ 175.9, 171.3, 78.1, 76.9, 56.3, 37.0, 29.3, 29.1, 26.1, 25.2 (2C). HRMS: calcd for C11H20NO5 [M + H]+ 246.1336, found 246.1336. HPLC: purity 98%, retention time 7.8 min.

3-(Cycloheptylmethoxy)aspartate (13e)

White solid, 7 mg (conversion 60%, isolated yield 26%). 1H NMR (500 MHz, D2O): δ 4.27 (d, J = 2.1 Hz, 1H), 3.97 (d, J = 2.2 Hz, 1H), 3.45 (dd, J = 9.5, 6.6 Hz, 1H), 3.17 (dd, J = 9.5, 7.2 Hz, 1H), 1.79–1.72 (m, 1H), 1.69–1.58 (m, 4H), 1.55–1.52 (m, 2H), 1.49–1.37 (m, 4H), 1.17–1.09 (m, 2H). 13C NMR (126 MHz, D2O): δ 176.2, 171.4, 78.2, 76.5, 56.4, 38.3, 30.3 (2C), 28.2 (2C), 25.9 (2C). HRMS: calcd for C12H22NO5 [M + H]+ 260.1492, found 260.1494. HPLC: purity 95%, retention time 10.0 min.

(l-threo)-3-(Thiophen-3-ylmethoxy)aspartate (13g)

Light-yellow solid, 47 mg (conversion 98%, isolated yield 47%). 1H NMR (500 MHz, D2O): δ 7.43–7.42 (m, 1H), 7.39–7.38 (m, 1H), 7.11 (dd, J = 4.9, 1.3 Hz, 1H), 4.70 (d, J = 11.9 Hz, 1H), 4.50 (d, J = 11.9 Hz, 1H), 4.29 (d, J = 2.2 Hz, 1H), 3.96 (d, J = 2.2 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.8, 171.4, 137.7, 127.8, 126.6, 124.7, 77.0, 67.2, 56.5. HRMS: calcd for C9H12NO5 [M + H]+ 246.0431, found 246.0430. HPLC: purity 99%, retention time 5.2 min. Comparison of the 1H NMR data of 13g with the 1H NMR data of chemically prepared racemic racemic (dl-threo)-19b and (dl-erythro)-19b showed that the de of product 13g was >98% (Supporting Information, Figure S2). Chiral HPLC conditions: Nucleosil chiral-1 column with 0.5 mM aqueous CuSO4 solution as mobile phase with a flow rate of 1.0 mL/min at 60 °C, UV detection at 254 nm, tR = 7.2 min, ee > 99% (Supporting Information, Figure S63).

3-(Thiophen-2-ylmethoxy)aspartate (13h)

Light-yellow solid, 40 mg (conversion 98%, isolated yield 48%). 1H NMR (500 MHz, D2O): δ 7.46 (dd, J = 5.0, 1.3 Hz, 1H), 7.11 (dd, J = 3.4, 1.1 Hz, 1H), 7.04 (dd, J = 5.0, 3.4 Hz, 1H), 4.88 (d, J = 12.6 Hz, 1H), 4.69 (d, J = 12.6 Hz, 1H), 4.35 (d, J = 2.3 Hz, 1H), 3.99 (d, J = 2.3 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.5, 171.3, 139.0, 128.2, 127.3, 127.0, 76.5, 66.3, 56.4. HRMS: calcd for C9H12NO5 [M + H]+ 246.0431, found 246.0430. HPLC: purity 98%, retention time 5.1 min.

3-(Furan-3-ylmethoxy)aspartate (13i)

Light-yellow solid, 21 mg (conversion 98%, isolated yield 35%). 1H NMR (500 MHz, D2O): δ 7.54–7.53 (m, 1H), 7.49 (t, J = 1.7 Hz, 1H), 6.51–6.48 (m, 1H), 4.59 (d, J = 12.1 Hz, 1H), 4.40 (d, J = 12.1 Hz, 1H), 4.31 (d, J = 2.3 Hz, 1H), 3.98 (d, J = 2.3 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.8, 171.4, 143.8, 141.9, 120.8, 110.5, 76.6, 63.4, 56.5. HRMS: calcd for C9H12NO6 [M + H]+ 230.0659, found 230.0659. HPLC: purity 96%, retention time 4.4 min.

3-(Furan-2-ylmethoxy)aspartate (13j)

Light-yellow solid 26 mg (conversion 98%, isolated yield 39%). 1H NMR (500 MHz, D2O): δ 7.51 (dd, J = 1.9, 0.9 Hz, 1H), 6.47 (d, J = 3.2 Hz, 1H), 6.42 (dd, J = 3.2, 1.9 Hz, 1H), 4.62 (d, J = 13.2 Hz, 1H), 4.50 (d, J = 13.2 Hz, 1H), 4.32 (d, J = 2.4 Hz, 1H), 3.97 (d, J = 2.4 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.4, 171.3, 150.1, 143.8, 111.0, 110.4, 76.6, 63.9, 56.4. HRMS: calcd for C9H12NO6 [M + H]+ 230.0659, found 230.0659. HPLC: purity 98%, retention time 4.4 min.

(l-threo)-3-(Prop-2-yn-1-yloxy)aspartate (13k)

White solid, 5 mg (conversion 83%, isolated yield 47%). 1H NMR (500 MHz, D2O): δ 4.49 (d, J = 2.3 Hz, 1H), 4.30 (dd, J = 16.0, 2.4 Hz, 1H), 4.20 (dd, J = 16.0, 2.4 Hz, 1H), 4.02 (d, J = 2.3 Hz, 1H), 2.83 (t, J = 2.4 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.3, 171.4, 76.8, 76.3, 76.3, 57.7, 56.5. HRMS: calcd for C7H10NO5 [M + H]+ 188.0554, found 188.0550. HPLC: purity 97%, retention time 2.0 min. Comparison of the 1H NMR data of 13k with the 1H NMR data of chemically prepared racemic (dl-threo)-19c and (dl-erythro)-19c showed that the de of product 13k was >98% (Supporting Information, Figure S3). Chiral HPLC analysis: Crownpak CR-I (+) 150 mm × 3 mm. Phase A, ACN + 0.5% TFA; phase B, H2O + 0.5% TFA, A/B = 98:2. Flow rate 0.4 mL/min, column temperature 25 °C, detected by ELSD at 35 °C, tR = 3.2 min, ee > 99% (Supporting Information, Figure S64).

Synthesis of (dl-threo)- and (dl-erythro)-3-Substituted Asp Derivatives

(dl-threo)-Dimethyl 2-[(tert-Butoxycarbonyl)amino]-3-hydroxy Succinate (17)

The chemical synthesis of compound (dl-threo)-16 has been described elsewhere.[23] To a stirred solution of 16 (622 mg, 2 mmol) in THF/MeOH (1:1, 30 mL) was added Pd/C 10% (50.0 mg) under nitrogen atmosphere. The reaction was stirred under H2 atmosphere (balloon) for 2 h at room temperature. After completion of the reaction, the reaction mixture was filtered through Celite and washed with MeOH (5 mL). The filtrate was concentrated under vacuum to provide a colorless oil which was directly used for the next step without purification. To a stirred solution of the colorless oil in dry DCM (20 mL) was added TEA (300 mg, 3 mmol) and Boc2O (654 mg, 3 mmol) under cooling in an ice-bath. After 10 min, the cooling was removed and the reaction mixture was stirred at room temperature for further 24 h. After completion of the reaction, the reaction mixture was diluted with DCM (20 mL) and washed with 0.5 M HCl (50 mL), saturated NaHCO3 solution (50 mL), and brine (50 mL). The organic layer was dried over Na2SO4 and concentrated under vacuum to give crude product 17, which was purified via flash chromatography (EtOAc/petroleum ether, 15%, v/v) to provide pure 17 as a white solid (460 mg, two-step yield 83%). 1H NMR (500 MHz, CDCl3): δ 5.29 (d, J = 9.0 Hz, 1H), 4.78 (d, J = 9.3 Hz, 1H), 4.69 (dd, J = 5.7, 2.0 Hz, 1H), 3.82 (s, 3H), 3.80 (s, 3H), 3.24 (d, J = 5.5 Hz, 1H), 1.42 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 172.4, 170.0, 155.4, 80.4, 71.2, 56.2, 53.2, 53.0, 28.2 (3C). HRMS: calcd for C11H19NO7Na [M + Na]+ 300.1054, found 300.1052.

Dimethyl 2-(Cyclopropylmethoxy)-3-[(tert-butoxycarbonyl)amino]succinate (18a)

To a stirred solution of compound 17 (195 mg, 0.70 mmol) in dry DMF (3 mL) was added bromomethylcyclopropane (191 mg, 1.4 mmol) at −20 °C. After 10 min, NaH (60% in mineral oil, 28 mg, 0.70 mmol) was added to the reaction mixture. The reaction mixture was stirred at −20 °C for 4 h and stirred at 4 °C for further 8 h. After completion of the reaction, the reaction mixture was quenched with cold water, extracted with EtOAc (30 mL × 3), washed with brine (50 mL × 3), and dried over Na2SO4. The solvent was evaporated to provide crude product, which was purified via flash chromatography (EtOAc/petroleum ether, 10%, v/v) to give pure (dl-threo)-18a (26 mg, 11%) and (dl-erythro)-18a (9 mg, 4%) as clear oil.

(dl-threo)-18a

1H NMR (500 MHz, CDCl3): δ 5.33 (d, J = 9.9 Hz, 1H), 4.80 (dd, J = 10.0, 2.4 Hz, 1H), 4.53 (d, J = 2.1 Hz, 1H), 3.77 (s, 3H), 3.74 (s, 3H), 3.50 (dd, J = 10.6, 6.9 Hz, 1H), 3.21 (dd, J = 10.5, 7.0 Hz, 1H), 1.41 (s, 9H), 1.02–0.95 (m, 1H), 0.54–0.46 (m, 2H), 0.19–0.11 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 170.1, 170.0, 155.6, 80.3, 77.6, 76.0, 56.2, 52.8, 52.5, 28.3 (3C), 10.2, 3.4, 2.7. HRMS: calcd for C15H25NO7Na [M + Na]+ 354.1523, found 354.1522.

(dl-erythro)-18a

1H NMR (500 MHz, CDCl3): δ 5.43 (d, J = 8.9 Hz, 1H), 4.85 (dd, J = 8.9, 3.4 Hz, 1H), 4.35 (d, J = 3.4 Hz, 1H), 3.77 (s, 3H), 3.73 (s, 3H), 3.56 (dd, J = 10.7, 6.9 Hz, 1H), 3.34 (dd, J = 10.7, 7.0 Hz, 1H), 1.45 (s, 9H), 1.06–1.00 (m, 1H), 0.56–0.49 (m, 2H), 0.24–0.17 (m, 2H). 13C NMR (126 MHz, CDCl3): δ 170.1, 169.5, 155.3, 80.4, 78.3, 76.2, 55.7, 52.8, 52.4, 28.4 (3C), 10.2, 3.5, 2.7. HRMS: calcd for C15H25NO7Na [M + Na]+ 354.1523, found 354.1523.

Dimethyl 2-(Thiophen-3-ylmethoxy)-3-[(tert-butoxycarbonyl)amino]succinate (18b)

To a stirred solution of compound 17 (197 mg, 0.71 mmol) in dry DMF (3 mL) was added 3-(bromomethyl)thiophene (251 mg, 1.42 mmol) at −20 °C. After 10 min, NaH (60% in mineral oil, 28.4 mg, 0.71 mmol) was added to the reaction mixture. The reaction mixture was stirred at −20 °C for 4 h and stirred at 4 °C for further 8 h. After completion of the reaction, the reaction mixture was quenched with cold water, extracted with EtOAc (30 mL × 3), washed with brine (50 mL × 3), and dried over Na2SO4. The solvent was evaporated to provide crude product, which was purified via flash chromatography (EtOAc/petroleum ether, 10%, v/v) to give (dl-threo)-18b (140 mg, 53%) and (dl-erythro)-18b (27 mg, 10%) as clear oil.

(dl-threo)-18b

1H NMR (500 MHz, CDCl3): δ 7.30 (dd, J = 5.0, 2.9 Hz, 1H), 7.18 (dd, J = 3.0, 1.1 Hz, 1H), 6.98 (dd, J = 4.9, 1.3 Hz, 1H), 5.34 (d, J = 10.0 Hz, 1H), 4.82–4.77 (m, 2H), 4.47 (d, J = 2.3 Hz, 1H), 4.43 (d, J = 12.3 Hz, 1H), 3.77 (s, 3H), 3.65 (s, 3H), 1.42 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 169.9, 169.8, 155.6, 137.7, 127.7, 126.4, 124.2, 80.4, 76.6, 67.8, 56.1, 52.8, 52.6, 28.3 (3C). HRMS: calcd for C16H24NO7S [M + H]+ 374.1268, found 374.1267.

(dl-erythro)-18b

1H NMR (500 MHz, CDCl3): δ 7.30 (dd, J = 4.9, 2.9 Hz, 1H), 7.22 (d, J = 2.8 Hz, 1H), 7.07 (dd, J = 5.0, 1.2 Hz, 1H), 5.33 (d, J = 8.7 Hz, 1H), 4.87 (dd, J = 8.6, 3.0 Hz, 1H), 4.84 (d, J = 12.3 Hz, 1H), 4.57 (d, J = 12.3 Hz, 1H), 4.27 (d, J = 3.1 Hz, 1H), 3.79 (s, 3H), 3.73 (s, 3H), 1.43 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 169.7, 169.2, 155.0, 137.8, 127.3, 126.3, 123.8, 80.3, 77.4, 68.2, 55.6, 52.7, 52.3, 28.3 (3C). HRMS: calcd for C16H24NO7S [M + H]+ 374.1268, found 374.1266.

Dimethyl 2-(Prop-2-yn-1-yloxy)-3-[(tert-butoxycarbonyl)amino]succinate (18c)

To a stirred solution of compound 17 (200 mg, 0.72 mmol) in dry DMF (3 mL) was added propargyl bromide (80% in toluene, 156 μL, 1.44 mmol) at −20 °C. After 10 min, NaH (60% in mineral oil, 29 mg, 0.72 mmol) was added to the reaction mixture. The reaction mixture was stirred at −20 °C for 4 h and stirred at 4 °C for further 8 h. After completion of the reaction, the reaction mixture was quenched with cold water, extracted with EtOAc (30 mL × 3), washed with brine (50 mL × 3), and dried over Na2SO4. The solvent was evaporated to provide crude product, which was purified via flash chromatography (EtOAc/petroleum ether, 10%, v/v) to give (dl-threo)-18c (54 mg, 24%) and (dl-erythro)-18c (25 mg, 11%) as clear oil.

(dl-threo)-18c

1H NMR (500 MHz, CDCl3): δ 5.29 (d, J = 9.8 Hz, 1H), 4.87 (dd, J = 9.9, 2.4 Hz, 1H), 4.78 (d, J = 2.3 Hz, 1H), 4.37 (dd, J = 16.3, 2.4 Hz, 1H), 4.20 (dd, J = 16.2, 2.4 Hz, 1H), 3.78 (s, 3H), 3.77 (s, 3H), 2.45 (t, J = 2.4 Hz, 1H), 1.41 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 169.7, 169.4, 155.5, 80.4, 78.1, 76.0, 75.8, 57.9, 56.0, 52.9, 52.7, 28.3 (3C). HRMS: calcd for C14H21NO7Na [M + Na]+ 338.1210, found 338.1211.

(dl-erythro)-18c

1H NMR (500 MHz, CDCl3): δ 5.44 (d, J = 8.7 Hz, 1H), 4.92 (dd, J = 8.7, 2.9 Hz, 1H), 4.58 (d, J = 2.9 Hz, 1H), 4.43 (dd, J = 16.3, 2.4 Hz, 1H), 4.31 (dd, J = 16.3, 2.4 Hz, 1H), 3.79 (s, 3H), 3.72 (s, 3H), 2.45 (t, J = 2.4 Hz, 1H), 1.45 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 169.4, 169.0, 155.3, 80.4, 78.3, 76.6, 76.2, 58.4, 55.7, 52.9, 52.5, 28.4 (3C). HRMS: calcd for C14H21NO7Na [M + Na]+ 338.1210, found 338.1211.

(dl-threo)-3-(Cyclopropylmethoxy)aspartate [(dl-threo)-19a]

To a stirred solution of (dl-threo)-18a (26 mg, 0.08 mmol) in dry DCM (2 mL) was added trifluoroacetic acid (0.8 mL) dropwise under cooling in an ice-bath. After the complete addition of trifluoroacetic acid, the ice-bath was removed and the reaction was allowed to proceed at room temperature for further 1.5 h. After completion of the starting material, the solvent was removed in vacuo to provide deBoc product quantitatively as a clear oil, which was directly used for the next step without purification.

To a stirred solution of the clear oil in THF/H2O (1:1, each 1 mL) was added LiOH (9.6 mg, 0.40 mmol), and the reaction mixture was stirred at room temperature for 2 h. Volatiles were removed in vacuo, and the residue was washed with EtOAc (1 mL). The aqueous layer was acidified with 1 M HCl (until pH = 1) and loaded onto a column packed with cation-exchange resin (10 g of Dowex 50W X8, 50–100 mesh), which was pretreated with 2 M aqueous ammonia (4 column volumes), 1 M HCl (2 column volumes), and distilled water (4 column volumes). The column was washed with distilled water (2 column volumes), and the product was eluted with 2 M aqueous ammonia (2 column volumes). The ninhydrin-positive fractions were collected and lyophilized to yield the desired product (dl-threo)-19a as ammonium salt (white solid, 7 mg, two-step yield of 37%). 1H NMR (500 MHz, D2O): δ 4.40 (d, J = 2.5 Hz, 1H), 4.00 (d, J = 2.5 Hz, 1H), 3.47 (dd, J = 10.7, 7.0 Hz, 1H), 3.30 (dd, J = 10.7, 7.3 Hz, 1H), 1.06–0.98 (m, 1H), 0.57–0.45 (m, 2H), 0.22–0.16 (m, 2H). 13C NMR (126 MHz, D2O): δ 176.0, 171.5, 77.4, 75.8, 56.4, 9.5, 2.7, 2.1. HRMS: calcd for C8H14NO5 [M + H]+ 204.0866, found 204.0866. HPLC: purity 97%, retention time 4.2 min. Chiral HPLC analysis: Crownpak CR-I (+) 150 × 3 mm. Phase A, ACN + 0.5%TFA; phase B, H2O + 0.5% TFA, A/B = 98:2. Flow rate 0.4 mL/min, column temperature 25 °C, detected by ELSD at 35 °C, tR (d-threo) = 2.3 min, tR (l-threo) = 2.8 min (Supporting Information, Figure S62).

(dl-erythro)-3-(Cyclopropylmethoxy)aspartate [(dl-erythro)-19a]

Compound (dl-erythro)-19a was prepared from (dl-erythro)-18a (9 mg, 0.027 mmol) following a procedure similar to that used for (dl-threo)-19a. The title compound was obtained as a white solid (3 mg, two-step yield of 46%). 1H NMR (500 MHz, D2O): δ 4.29 (d, J = 3.8 Hz, 1H), 4.08 (d, J = 3.8 Hz, 1H), 3.50 (dd, J = 10.6, 7.1 Hz, 1H), 3.37 (dd, J = 10.6, 7.3 Hz, 1H), 1.12–1.04 (m, 1H), 0.58–0.51 (m, 2H), 0.27–0.18 (m, 2H). 13C NMR (126 MHz, D2O/DMSO-d6, 1:1): δ 179.8, 179.0, 99.0, 76.5, 60.2, 11.4, 4.5, 3.9. HRMS: calcd for C8H14NO5 [M + H]+ 204.0866, found 204.0867.

(dl-threo)-3-(Thiophen-3-ylmethoxy)aspartate [(dl-threo)-19b]

Compound (dl-threo)-19b was prepared from (dl-threo)-18b (140 mg, 0.37 mmol) following a procedure similar to that used for (dl-threo)-19a. The title compound was obtained as a white solid (40 mg, two-step yield of 39%). 1H NMR (500 MHz, D2O): δ 7.43 (dd, J = 4.9, 2.9 Hz, 1H), 7.39 (dd, J = 2.9, 1.3 Hz, 1H), 7.12 (dd, J = 4.9, 1.4 Hz, 1H), 4.71 (d, J = 12.1 Hz, 1H), 4.50 (d, J = 11.9 Hz, 1H), 4.30 (d, J = 2.2 Hz, 1H), 3.97 (d, J = 2.3 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.8, 171.4, 137.8, 127.8, 126.6, 124.7, 76.9, 67.2, 56.4. HRMS: calcd for C9H12NO5S [M + H]+ 246.0431, found 246.0430. HPLC: purity 98%, retention time 5.2 min. Chiral HPLC conditions: Nucleosil chiral-1 column with 0.5 mM aqueous CuSO4 solution as mobile phase with a flow rate of 1.0 mL/min at 60 °C, UV detection at 254 nm, tR (l-threo) = 7.3 min, tR (d-threo) = 8.3 min (Supporting Information, Figure S63).

(dl-erythro)-3-(Thiophen-3-ylmethoxy)aspartate [(dl-erythro)-19b]

Compound (dl-erythro)-19b was prepared from (dl-erythro)-18b (27 mg, 0.072 mmol) following a procedure similar to that used for (dl-threo)-19a. The title compound was obtained as a white solid (7 mg, two-step yield of 35%). 1H NMR (500 MHz, D2O): δ 7.47–7.45 (m, 2H), 7.19 (dd, J = 4.8, 1.5 Hz, 1H), 4.74 (d, J = 11.9 Hz, 1H), 4.56 (d, J = 11.9 Hz, 1H), 4.16 (d, J = 4.1 Hz, 1H), 3.93 (d, J = 4.1 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.1, 171.6, 137.8, 127.9, 126.9, 124.9, 78.1, 66.8, 56.6. HRMS: calcd for C9H12NO5S [M + H]+ 246.0431, found 246.0430. HPLC: purity 96%, retention time 4.9 min.

(dl-threo)-3-(Prop-2-yn-1-yloxy)aspartate [(dl-threo)-19c]

Compound (dl-threo)-19c was prepared from (dl-threo)-18c (54 mg, 0.17 mmol) following a procedure similar to that used for (dl-threo)-19a. The title compound was obtained as a white solid (13 mg, two-step yield of 34%). 1H NMR (500 MHz, D2O): δ 4.50 (d, J = 2.4 Hz, 1H), 4.31 (dd, J = 16.0, 2.5 Hz, 1H), 4.21 (dd, J = 16.0, 2.4 Hz, 1H), 4.03 (d, J = 2.4 Hz, 1H), 2.85 (t, J = 2.4 Hz, 1H). 13C NMR (126 MHz, D2O): δ 175.3, 171.4, 78.6, 76.9, 76.4, 57.7, 56.5. HRMS: calcd for C7H10NO5 [M + H]+ 188.0554, found 188.0554. HPLC: purity 95%, retention time 2.0 min. Chiral HPLC analysis: Crownpak CR-I (+) 150 mm × 3 mm. Phase A, ACN + 0.5%TFA; phase B, H2O + 0.5% TFA, A/B = 98:2. Flow rate 0.4 mL/min, column temperature 25 °C, detected by ELSD at 35 °C, tR (d-threo) = 2.4 min, tR (l-threo) = 3.0 min (Supporting Information, Figure S64).

(dl-erythro)-3-(Prop-2-yn-1-yloxy)aspartate [(dl-erythro)-19c]

Compound (dl-erythro)-19c was prepared from (dl-erythro)-18c (25 mg, 0.08 mmol) following a procedure similar to that used for (dl-threo)-19a. The title compound was obtained as a white solid (6 mg, two-step yield of 34%). 1H NMR (500 MHz, D2O): δ 4.37 (dd, J = 16.1, 2.2 Hz, 1H), 4.31 (d, J = 3.8 Hz, 1H), 4.22 (dd, J = 16.1, 2.1 Hz, 1H), 4.01 (d, J = 3.8 Hz, 1H), 2.87 (t, J = 2.4 Hz, 1H). 13C NMR (126 MHz, D2O): δ 174.7, 171.7, 78.2, 78.1 (2C), 57.4, 56.5. HRMS: calcd for C7H10NO5 [M + H]+ 188.0554, found 188.0553.

Synthesis of Precursors for Hybrid Compounds

(l-threo)-1-(tert-Butyl)-4-methyl-2-amino-3-(benzyloxy)succinate (21)

The multigram scale synthesis of l-TBOA (2) was based on a previously published procedure.[24] To a stirred suspension of 2 (1.20 g, 5 mmol) in dry MeOH (20 mL) was added SOCl2 (0.45 mL, 5 mmol) dropwise (in an ice-bath). After 20 min, the cooling system was removed and the reaction mixture was stirred for further 16 h. After completion of the reaction, the solvent was removed to provide crude product 20 as a white solid (1.4 g, 96%). No purification was needed and the crude product 20 was directly used for the next step. To a stirred suspension of compound 20 (1.3 g, 4.5 mmol) in t-BuOAc (20 mL) was added BF3/Et2O (3 mL). The reaction mixture was stirred at room temperature for 16 h. After completion of the reaction, 2 M NaOH (20 mL) was added to the reaction mixture and stirred for another 10 min. The reaction mixture was extracted with EtOAc (30 mL × 3), washed with brine (50 mL × 3), and dried over Na2SO4. The organic solvent was evaporated to provide pure product 21 as a clear oil (1.25 g, 91%). 1H NMR (500 MHz, CDCl3): δ 7.36–7.28 (m, 5H), 4.82 (d, J = 11.3 Hz, 1H), 4.48 (d, J = 3.1 Hz, 1H), 4.45 (d, J = 11.3 Hz, 1H), 3.83 (s, 3H), 3.80 (d, J = 3.1 Hz, 1H), 1.76 (brs, 2H), 1.43 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 171.7, 171.0, 137.1, 128.4 (2C), 128.2 (2C), 128.0, 81.9, 79.3, 73.3, 57.9, 52.2, 28.0 (3C). HRMS: calcd for C16H24NO5 [M + H]+ 310.1649, found 310.1651.

(l-threo)-2-(Benzyloxy)-4-(tert-butoxy)-3-[(tert-butoxycarbonyl)amino]-4-oxobutanoic Acid (23)

To a stirred solution of 21 (1.2 g, 3.9 mmol) in dry DCM (25 mL) was added DIEA (1.3 mL, 8 mmol) and Boc2O (1.7 g, 7.8 mmol) under cooling in an ice-bath. After 10 min, the cooling system was removed and the reaction mixture was stirred at room temperature for further 24 h. After completion of the reaction, the reaction mixture was diluted with DCM (20 mL), and washed with 0.5 M HCl (50 mL), saturated NaHCO3 solution (50 mL), and brine (50 mL). The organic layer was dried over Na2SO4 (s) and concentrated under vacuum to give product 22 as a clear oil (1.35 g, 85%). No purification was needed and compound 22 was directly used for the next step. To a stirred solution of 22 (1.35 g, 3.3 mmol) in THF/H2O (1:1, each 5 mL) was added LiOH (240 mg, 10 mmol) under cooling in an ice-bath. After 10 min, the cooling system was removed and the reaction mixture was stirred at room temperature for further 8 h. After completion of the starting material, THF was removed in vacuo and the resulting aqueous solution was acidified with 1 M HCl to pH = 1. The reaction mixture was extracted with EtOAc (50 mL × 3), washed with brine (50 mL × 3), and dried over Na2SO4. The organic solvent was evaporated to provide crude product, which was purified via flash chromatography (MeOH/DCM 10%, v/v) to give 23 as a clear oil (640 mg, 49%). 1H NMR (500 MHz, DMSO-d6): δ 13.10 (brs, 1H), 7.36–7.29 (m, 5H), 6.45 (d, J = 9.6 Hz, 1H), 4.70 (d, J = 11.6 Hz, 1H), 4.46–4.37 (m, 3H), 1.37 (s, 9H), 1.35 (s, 9H). 13C NMR (126 MHz, DMSO-d6): δ 170.6, 168.3, 155.4, 137.5, 128.1 (2C), 127.7 (2C), 127.6, 81.5, 78.7, 77.4, 72.0, 56.2, 28.1 (3C), 27.5 (3C). HRMS: calcd for C20H30NO7 [M + H]+ 396.2017, found 396.2011.

(l-threo)-tert-Butyl-3-(benzyloxy)-4-[[4-(2-bromo-4,5-difluorophenoxy)phenyl]amino]-2-[(tert-butoxycarbonyl)amino]-4-oxobutanoate (26)

To a stirred solution of 23 (150 mg, 0.38 mmol) in dry DCM (3 mL) was added EDCI (96 mg, 0.5 mmol), HOBT (67 mg, 0.5 mmol), and TEA (50 mg, 0.5 mmol) under cooling in an ice-bath. After 30 min, the cooling system was removed and a solution of amine 24 (100 mg, 0.33 mmol) in DCM (1 mL) was added to the reaction mixture. The reaction mixture was stirred for 12 h at room temperature. After completion of the reaction, the solvent was removed in vacuo. The reaction mixture was extracted with EtOAc (20 mL × 3), washed with brine (50 mL), and dried over Na2SO4. The organic solvent was evaporated to provide crude product, which was purified via flash chromatography (EtOAc/petroleum ether 5%, v/v) to give 26 as a white solid (70 mg, 27%). 1H NMR (500 MHz, CDCl3): δ 8.19 (s, 1H), 7.50–7.46 (m, 3H), 7.41–7.35 (m, 5H), 6.93 (dd, J = 9.0, 3.2 Hz, 2H), 6.73–6.70 (m, 1H), 5.34 (d, J = 9.4 Hz, 1H), 4.83 (d, J = 9.4 Hz, 1H), 4.72 (d, J = 11.2 Hz, 1H), 4.62 (dd, J = 11.3, 3.2 Hz, 1H), 4.51 (s, 1H), 1.48 (s, 9H), 1.33 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 168.9, 167.4, 155.3, 152.7, 149.1 (dd, J = 409.5 Hz, 13.9 Hz), 150.5 (dd, J = 7.6, 3.8 Hz), 147.0 (dd, J = 408.2 Hz, 13.9 Hz), 136.2, 133.5, 129.0 (2C), 128.8, 128.5 (2C), 121.8, 121.7, 121.6, 119.2, 115.6 (d, J = 22.7 Hz), 108.7 (d, J = 21.4 Hz), 107.8 (dd, J = 6.3 Hz, 3.8 Hz), 83.0, 80.8, 80.2, 74.4, 56.0, 28.2 (3C), 28.1 (3C). HRMS: calcd for C32H36BrF2N2O7 [M + H]+ 677.1668, found 677.1675.

(l-threo)-tert-Butyl-4-[(9H-fluoren-2-yl)amino]-3-(benzyloxy)-2-[(tert-butoxycarbonyl) amino]-4-oxobutanoate (27)

Compound 27 was prepared from 23 (100 mg, 0.25 mmol) and (9H)-fluoren-2-amine (25, 58 mg, 0.32 mmol), using EDCI (61 mg, 0.32 mmol), HOBT (43 mg, 0.32 mmol), and TEA (32 mg, 0.32 mmol) and following a procedure similar to that used for 26. The title product was obtained as a white solid (128 mg, 90%). 1H NMR (500 MHz, CDCl3): δ 8.36 (s, 1H), 7.95 (s, 1H), 7.73 (dd, J = 17.7, 7.9 Hz, 2H), 7.55 (d, J = 7.5 Hz, 1H), 7.46–7.37 (m, 6H), 7.34–7.29 (m, 2H), 5.43 (d, J = 9.5 Hz, 1H), 4.91 (dd, J = 9.4, 2.3 Hz, 1H), 4.77 (d, J = 11.1 Hz, 1H), 4.69 (d, J = 11.0 Hz, 1H), 4.57 (d, J = 2.3 Hz, 1H), 3.90 (s, 2H), 1.52 (s, 9H), 1.35 (s, 9H). 13C NMR (126 MHz, CDCl3): δ 169.0, 167.3, 155.3, 144.4, 143.3, 141.3, 138.4, 136.3, 135.8, 129.0 (2C), 128.8, 128.5 (2C), 126.8, 126.5, 125.1, 120.1, 119.6, 118.6, 116.9, 83.0, 80.9, 80.2, 74.3, 56.0, 37.1, 28.2 (3C), 28.1 (3C). HRMS: calcd for C33H39N2O6 [M + H]+ 559.2803, found 559.2808.

Cell Culture

All cell lines were cultured at 37 °C in a humidified 5% CO2 atmosphere. The hEAAT1-, hEAAT2-, and hEAAT3-HEK293 cell lines were maintained in Culture Medium I [Dulbecco’s Modified Eagle Medium Glutamax-I (DMEM) supplemented with 5% dialyzed fetal bovine serum, 100 U/mL penicillin, 100 μg/mL streptomycin, and 1 mg/mL G-418], and the stable rEAAT4-tsA201 cell line was maintained in Culture Medium II (DMEM supplemented with 5% tetracycline-free fetal bovine serum, 100 U/mL penicillin, 100 μg/mL streptomycin, 0.2 mg/mL hygromycin B, and 10 μg/mL blasticidin).

[3H]-d-Asp Uptake Assay

The pharmacological characterization of various reference EAAT ligands and test compounds in the [3H]-d-Asp uptake assay was performed essentially as described previously.[25] The day before the assay, cells were split into poly-d-lysine-coated white 96-well plates (PerkinElmer, Boston, MA) in Culture Medium I (hEAAT1-, hEAAT2-, and hEAAT3-HEK293) or in Culture Medium II supplemented with 1 μg/mL tetracycline (rEAAT4-tsA201). After 16–24 h, the culture medium was aspirated and cells were washed twice with 100 μL of assay buffer (Hank’s Buffered Saline Solution supplemented with 20 mM HEPES, 1 mM CaCl2, and 1 mM MgCl2, pH 7.4). Then 50 μL of assay buffer supplemented with 100 nM [3H]-d-Asp (PerkinElmer, Boston, MA) and various concentrations of test compounds were added to the wells, and the plate was incubated at 37 °C for 4 min. Notably, aspartate uptake is still in the linear range at 4 min after initiation of uptake (for all four cell lines). Nonspecific [3H]-d-Asp uptake/binding in the cells was determined in the presence of 3 mM l-Glu. The assay mixtures were quickly removed from the wells, and the cells were washed with 3 × 100 μL ice-cold assay buffer, after which 150 μL of Microscint-20 scintillation fluid (PerkinElmer, Boston, MA) was added to each well. Then the plate was shaken for at least 1 h and counted in a TopCounter (PerkinElmer, Boston, MA). For each compound, three or four independent assays were performed in duplicate (the specific numbers are stated for each compound at the four different transporters in Table ).

The FLIPR Membrane Potential Blue (FMP) Assay

The substrate or nonsubstrate inhibitor properties of all 3-substituted Asp derivatives were tested at hEAAT1-, hEAAT2-, and hEAAT3-HEK293 cells in the FMP assay essentially as previously described, using l-Glu and l-THA as reference substrates and dl-TBOA as a reference nonsubstrate inhibitor at the EAATs.[25] Briefly, the cells were split into poly-d-lysine-coated black 96-well plates (BD Biosciences, Palo Alto, CA) with a clear bottom. The following day, the culture medium was aspirated, and the cells were washed once with 100 μL assay buffer (140 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl2, 1.2 mM MgCl2, 11 mM HEPES, 10 mM d-glucose, pH 7.4), after which the cells were incubated in 100 μL of assay buffer supplemented with FMP dye (0.5 mg/mL) at 37 °C for 30 min. In the inhibitor experiments, the test compound (300 μM) was added to the assay buffer at this point. The 96-well plate was assayed in a FLEXStation benchtop multimode microplate reader (Molecular Devices) measuring emission [in fluorescence units (FU)] at 565 nm caused by excitation at 525 nm before and up to 90 s after addition of 33.3 μL of assay buffer supplemented with test compound in the substrate experiments (giving rise to an assay concentration of 300 μM of the test compound) or with l-Glu in the inhibitor experiments (giving rise to an assay concentration of l-Glu EC80 for the specific EAAT). For each compound, three independent experiments were done in triplicate.

Data Analysis

Data from the [3H]-d-Asp uptake assay were fitted to the equation %Uptake = 100%Uptake/[1 + ([L]/IC50)nH], and IC50 values for the test compounds were derived from this equation. Concentration–inhibition curves were generated by nonweighted least-squares fits using the program KaleidaGraph 3.08 (Synergy Software, Reading, PA).