Logo Logo
Switch language to English
Zhao, Xueqing (2002): Development of new GABA uptake inhibitors derived from proline or from pyrrolidin-2-yl acetic acid. Dissertation, LMU München: Fakultät für Chemie und Pharmazie



GABA transporters GAT-1, GAT-2 and GAT-3 are new targets for drug design. The substitution of the nitrogen atoms in Nicopetic acid (11), Guvacine (12) and cis-4- Hydroxynicopetic acid (13) with appropriate bulky lipophilic groups resulted in very potent GABA uptake inhibitors for GAT-1 as well as for GAT-3. Pyrrolidine-2-acetic acid derivatives with the three N-substituents 24a-c (Scheme 54) also showed a highly potent inhibition at GAT-1 and GAT-3, respectively. My intention was to investigate how the potency of pyrrolindine-2-carboxylic acid derivatives and of pyrrolidine-2-acetic acid derivatives at GAT-1 and GAT-3 is influenced by the introduction of a hydroxy group or both of a hydroxy and a (4-methoxy)phenyl group at C-4. For this study, the N-substituents 24a-d were chosen. Thus, the four series of pyrrolidine derivatives 20-23 shown below were designed as potential GABA uptake inhibitors. L-trans-4-hydroxypyrrolidine [(2S,4R)-25] was chosen as a precursor, from which the four key intermediates (2S,4R)-38, (2R,4R)-38, (2S,4R)-86 and (2R,4R)-64 were synthesized. The known compounds (2S,4R)-38 and (2R,4R)-38 were prepared from (2S,4R)-25 according to literature procedures. N-protection of (2S,4R)-25 with Cbz group gave (2S,4R)-58 (90% yield). After a series of reactions  electrolysis (97% yield), O-silyl protection (85%), nucleophilic addition of 1- ethoxy-1-(trimethylsilyloxy)ethene (cis-71 79%; trans-71 9%) and finally O-deprotection (88%) and N-deprotection (89%)  (2R,4R)-64 was obtained in 46% overall yield [from (2S,4R)-25]. After the reaction conditions for the conversion of 70 into 71 have been optimized, the best stereoselectivity [a ratio of cis/trans 97:3; yield cis-71 74%, trans-(2S,4R)-71 1.7%] and a good yield of 88% (cis 79%, trans 9%) were achieved. BF3⋅Et2O appeared to be slightly better for a higher stereoselectivity than TiCl4. (2S,4R)-83 was obtained in 90% yield by protecting the hydroxy group of (2S,4R)-58. An Arndt-Eistert reaction (64% yield) starting from (2S,4R)-83 followed by a simultaneous N,Odeprotection (90% yield) of (2S,4R)-85 led to (2S,4R)-86 in 47% overall yield [from (2S,4R)- 25]. As illustrated in Scheme 58 and 59, (2S,4R)-38 and (2S,4R)-86, (2R,4R)-38 and (2R,4R)-64 were used as starting materials for the synthesis of the N-substituted target compounds (2S,4R)-40a-b, (2S,4S)-40a-b, (2R,4S)-40a-b, (2R,4R)-40a-b, (2S,4R)-89a-d, (2S,4S)-89a-d, (2R,4R)-89a-d and (2R,4S)-89a-d. N-alkylation of these four starting materials with the halides of 24a-d yielded the corresponding tertiary amines. Mitsunobu reactions gave access to the stereoisomers by inversion of the stereocenter at C-4 of the pyrrolidine ring. Finally upon hydrolysis, all the N-substituted pyrrolidine derivatives with a 2-carboxylic acid side chain or a 2-acetic acid side chain were obtained. In the same manners (scheme 58 and 59), the series 96 also were synthesized and finally their hydrogenolysis over Pd-C provided each of four stereoisomers 97. The N-substituted 4-oxopyrrolidine derivatives (2S)-52a-b and (2S)-100a-c (see Scheme 61) were prepared (81-92% yields) via Swern oxidation, and fortunately, the acid-sensitive Nsubstituent 24b was not affected. The N-Cbz-protected 4-oxopyrrolidine derivatives 60 and 108 (Scheme 61) were prepared in good yield (71-78%) via Jones’ oxidation, but in low yield (10-27%) via Swern oxidation. The addition reactions of the organometallic reagents to the N-substituted pyrrolidine derivatives 52a-b and (2S)-100b-c were carried out in two different ways. Depending on the starting material and the employed organometallic reagent, two different results were obtained: Under condition A [(4-MeOC6H4)MgBr at –78 °C in ether] the cis addition product (cis refers to the ester group) was formed as the major diastereomer and a good diastereoselectivity was achieved (cis/trans addition from 79:21 to 89:11; total yields 45- 65%); Under condition B [(4-MeOC6H4)MgBr/CeCl3 at –78 °C in THF], the trans addition product was obtained as a major diastereomer (cis/trans addition from 30:70 to 17:83; total yields 32-48%). In the case of the N-Cbz-protected pyrrolidine derivative (2S)-60, a single diastereomer (2S,4R)-61 was formed in 56% yield under condition B [(4-MeOC6H4)MgBr/CeCl3 in THF at – 60 °C for 4 h]. However, for the homologous (2S)-108 under the same conditions, (2S,4R)- 109 (25% yield) and a side product (5%) resulting from a simultaneous addition to the ester group were formed. The addition of (4-MeOC6H4)MgBr to (2S)-108 (at –78 °C in ether for 4 h), however, led to (2S,4R)-109 (38% yield) as a single diastereomer. Each of the N-substituted stereoisomers from the reaction above was subjected to a basic hydrolysis, which was followed by hydrogenolysis over Pd-C, where necessary, to afford the free amino acids (70-98% yields). Via the similar synthetic procedures as described above, the rest of the stereoisomers (2R,4R)- 57a-b and (2R,4S)-57a-b, (2R,4R)-104b-c and (2R,4S)-104b-c, (2R,4S)-63 and (2R,4S)-111 were obtained from (2R,4R)-39a-b, (2R,4R)-88b-c, (2R)-60 and (2R,4R)-73. The relative stereochemistry of the products obtained from the addition of the organometallic reagents to the 4-oxopyrrolidine derivatives was determined by chemical correlation and NOE measurements. NOE experiments performed with (2S,4R)-114 revealed that the phenyl group and H-2 are cis to each other. As important signals for the assignment overlapped in the 1H NMR spectrum of (2S,4R)-115, the NOE experiments were performed with (2S,4R)-111, which is the sodium salt of (2S,4R)-115. The NOE measurement revealed that the phenyl group is located cis to H- 2, thus, (2S,4R)-111 and (2S,4R)-115 being of the stereochemistry shown. The N-alkylation of (2S,4R)-114 with the bromide of 24a, and of (2S,4R)-115 with the bromide of 24c led to (2S,4R)-53a and (2S,4R)-102c, respectively. With these compounds as references, also the stereochemistry of all related compounds differing only in side chain on the amino nitrogen could be deduced. The target compounds obtained in this study were evaluated for their biological activities. Membrane fractions from frontal cortex of bovine brain (or porcine brain) were utilized to study the inhibitory potency of the test compounds regarding the GAT-1-mediated GABAuptake. For the determination of the potency as GAT-3 inhibitors, membrane fractions from brain stem of bovine brain (or porcine brain) were used. As compared to the corresponding 4-unsubstituted compounds with (2S) configuration (IC50 2.56 µM at GAT-1) and with (2R) configuration (IC50 18.5 µM), the 40a-b series containing a 4-hydroxy group showed a significant drop in the inhibitory potency at both GAT-1 and GAT-3, only one compound [(2R,4R)-40a] showed a reasonable potency at GAT-1 (IC50 9.4 µM) and no one of them for GAT-3 (IC50 > 100 µM). (2S,4S)-89a (IC50 3.29 µM at GAT-1) and (2S,4S)-89c (5.14 µM), (2S,4R)-89a (4.92 µM) and (2S,4R)-89c (3.15 µM) exhibiting a 4-hydroxypyrrolidine-2-acetic acid skeleton showed an inhibitory potency at GAT-1, which was only one order of magnitude lower than the potency of corresponding compounds with (2S) configuration without the 4-hydroxy group (with Nsubstituent 24a: IC50 0.40 µM and with N-substituent 24c: 0.34 µM). (2R,4S)-89b was the most potent inhibitor at GAT-3 (IC50 19.9 µM) of all the stereoisomers of series 89a-d and showed a much higher potency than its isomer (2R,4R)-89b (126 µM). According to these data a 4-hydroxy group is detrimental to the potency at both GAT-1 and GAT-3, and (2S)-configuration of the pyrrolidine-2-acetic acid unit is crucial for a reasonable potency at GAT-1. As compared to the 40b series, some stereoisomers of the 57b series, the latter exhibiting a (4-methoxy)phenyl group at C-4 of the pyrrolidine ring, showed an increased potency as inhibitors at GAT-1 and GAT-3 [e.g. (2S,4R)-57b: IC50 29.7 µM at GAT-3; (2R,4S)-57b: IC50 of 38.0 µM at GAT-3]. In contrast, the introduction of a (4-methoxy)phenyl group into C-4 of the 89b-c series, resulting in the compounds of 104b-c, gave rise to diverse biological results. As compared with (2R,4S)-89b, (2R,4R)-104b displayed a loss of inhibitory potency at GAT- 3 but some enhancement at GAT-1.