2.1. Materials

E. coli (CR 261) was a gift from Dr. Ch. Roessner of Texas A & M University. Production of E. coli and purification of PBGS has been described earlier [20]. Compounds 16, 17 ,18 ,19 ,20 ,21 ,23 ,24 and 29 were bought from Fluka and compound 22 from Aldrich. All the other inhibitors rac-25 [47-49], rac-26 [47-49], 27 [50], 28 [51,52], 30 [52,53], 31 [54], 32 [55], 33 [56] and 35 [57] were synthesized in our laboratories following the corresponding references.

Synthesis of 5-(3-nitropropionylamino)pentanoic acid (34)

In a two necked round-bottom flask were introduced under N2, 1.07 g (9 mmol) 3-nitroproionic acid, 40 ml dichloromethane and 2 drops of dimethylformamide. The suspension was cooled with an ice-bath and 1.26 g (9.9 mmol) oxalyl chloride was added over 5 min. After 3 h at r.t. the solution was cooled with an ice-bath and 2.73 g (9 mmol) 4-methoxycarbonylbutylammonium toluene-4-sulfonate and 911 mg (9 mmol) triethylamine in 25 ml dichloromethane were added dropwise followed by 911 mg (9 mmol) triethylamine neat over 15 min. After 2 h, the ice-bath was removed and the reaction was kept at r.t. for 1 h then quenched with 60 ml 1 M HCl. The aqueous phase was extracted with 60 ml dichloromethane and the combined organic phases were washed with 60 ml KHCO3 10%, 60 ml brine, dried over MgSO4 anhydrous, filtered and evaporated. The resulting oil (1.55 g) was separated by flash column chromatography (dichloromethane with 1 to 4% methanol) to obtain 1.33 g (64%) white solid methyl 5-(3-nitropropionylamino)pentanoate.

C9H16N2O5 (232.24) C 46.72 (46.55), H 7.18 (6.94), N 12.35 (12.06); Rf (CH2Cl2-MeOH 9 : 1, KMnO4) = 0.62; m.p. = 50-51°C; IR (KBr) : 3302m, 3105w, 3031w, 2957w, 2924m, 2854w, 1733s, 1646s, 1552vs, 1446m, 1413m, 1370m, 1226m, 1202m, 1165m; 1H-NMR (400 MHz, CDCl3) : 1.49-1.56 (m, 2H, H2C(4)); 1.60-1.67 (m, 2H, H2C(3)); 2.32 (t, 3J2-3 = 7.2, 2H, H2C(2)); 2.80 (t, 3J7-8 = 6.2, 2H, H2C(7)); 3.04 (td, 3J5-4 ~ 6.6, 3J5-HN ~ 5.6, 2H, H2C(5)); 4.69 (t, 3J8-7 = 6.2, 2H, H2C(8); 6.22 (s(br), 1H, HN); 13C-NMR (100 MHz, CDCl3, HETCOR) : 21.8 C(3); 28.7 C(4); 32.5 C(7); 33.3 C(2); 39.2 C(5); 51.5 C(9); 70.2 C(8); 168.3 C(6); 173.9 C(1); MS (DCI) : 250 (29, [M+18]+), 233 (100, [M+1]+), 201 (8, [M-CH3O]+), 100 (15, [C5H10NO]+).

In a 100 ml Erlenmeyer flask were introduced 1.2 g (5.17 mmol) methyl 5-(3-nitropropionylamino)pentanoate, 50 ml water and 3 mg hog liver esterase in 0.3 ml 3.2 M ammonium sulfate. For 4 h, pH = 7.3 was kept by pH-state controlled adding of 0.5 M NaOH, then the solution was acidified with conc. HCl, saturated with NaCl and extracted 4 times with 50 ml ethyl acetate. The combined organic phases were dried over MgSO4, filtrated and evaporated. The resulting orange solid (1.01 g) was crystallized from tetrahydrofurane and diethylether to obtain 810 mg (72%) white solid 5-(nitropropionylamino)pentanoic acid.

C8H14N2O5 (218.21) C 44.24 (44.03), H 6.53 (6.47), N 12.67 (12.84); Rf (ethyl acetate-MeOH 2 : 1, KMnO4) = 0.52; m.p. = 82-83°C; IR (KBr) : 3298s, 3071m, 2948m, 2878m, 1692s, 1638s, 1551vs, 1477m, 1464m, 1416m, 1377m, 1334m, 1280m, 1218m, 1192m, 924m, 872m, 681m; 1H-NMR (400 MHz, d8-DMSO) : 1.35-1.42 (m, 2H, H2C(4)); 1.45-1.52 (m, 2H, H2C(3)); 2.20 (t, 3J2-3 = 7.3, 2H, H2C(2)); 2.73 (t, 3J7-8 = 6.0, 2H, H2C(7)); 3.04 (td, 3J5-4 = 7.0, 3J5-HN = 5.3, 2H, H2C(5)); 4.68 (t, 3J8-7 = 6.0, 2H, H2C(8); 8.04 (t, 3JHN-5 = 5.3, 1H, HN); 12.00 (s, 1H, HO); 13C-NMR (100 MHz, d8-DMSO, HETCOR) : 21.9 C(3); 28.5 C(4); 31.4 C(7); 33.2 C(2); 38.2 C(5); 70.7 C(8); 168.3 C(6); 174.4 C(1).

2.2. PBGS Assay and Determination of Kinetic Constants

The PBGS assay is a colorimetric assay based on the reaction between PBG and 4-dimethylaminobenzaldehyde [3].

The assay for E. coli PBGS contained 4 - 6.4 µg PBGS, the inhibitor in 1.5ml of 0.1M KP, at pH=8.1, 12.3 mM mercaptoethanol, 10 mM MgCl2.6H2O and 10 µM ZnCl2. The preincubation took place at 37° for 30-45 minutes. The substrate was added in varying concentrations and the solution was incubated for 14 minutes. After the 14 minutes of incubation the PBGS-catalysed reaction was stopped by 1 ml of the stop reagent (20% TCA, 10 mM HgCl2) at 0°C. After centrifugation (4 min, 1900 g) 1 ml of the supernatant was treated with 1 ml of Ehrlich's reagent (4-dimethylaminobenzaldehyde in perchloric acid 70% / acetic acid solution). This solution was centrifuged (4 min, 1900 g) for a second time. The quantity of product formed was determined by measuring the absorbance at 554 nm (e = 60'000 [mol-1 * cm1]).

The type of inhibition was determined using Eadie-Hofstee, Lineweaver Burk and Hanes plots. The Ki shown in the table 1 and table 2 are the average of at least three independent assays. Ki-values were calculated from the Kmapp deduced from the hyberbola plot, typical for Michaelis-Menten kinetics.

2.3. Estimation of the specific activity

The specific activity is estimated by the Bio-Rad Protein Assay using the color change of the Coomassie Brilliant Blue-G-250. This color change is followed by measuring the absorbance at 596 nm. The enzyme quantity is deduced by comparison with a standard curve.

2.4. Dialysis Assay

4 - 6.4 µg PBGS were dissolved in 10 ml phosphate buffer (0.1M KP, pH=8.0, 12.3 mM mercaptoethanol, 10 mM MgCl2.6H2O and 10 µM ZnCl2). 4 ml of this solution were placed in two different 5 ml tubes, one containing 5.8 mg of 4-oxo-sebacic acid (30) and the second (without inhibitor) was used as a reference. After 24 hours at room temperature, the specific activities were determined. The solution containing the inhibitor showed only 22.4% of the specific activity compared to the blank experiment.

1 ml of each solution is placed in a dialysis tube and were seperately dialysed against 2.5 l phosphate buffer solution (0.1M KP, pH=8.0, 12.3 mM mercaptoethanol, 10 mM MgCl2.6H2O and 10 µM ZnCl2). After 66 hours dialysis at 4°C, the specific activities were again determined (the solution containing the inhibitor showed 28.5% of the specific activity compared to the blank experiment).

The same measures were repeated with the compound (35).

3.1. Inhibition of E. coli PBGS with Diacids

A series of diacids were tested as potential analogues of the intermediates. We hoped thereby to obtain further hints on the mechanism of the PBG biosynthesis.

Almost all of the diacids tested showed competitive behaviour (Table 1). The inhibition constants of the competitive inhibitors are ranging from values close to 10'000 µM to more then 20'000 mM for the 2-nitro benzoic acid (20) and the compound 32. These competitive inhibitors all show only a moderate inhibition potency. It is interesting to compare pimelic acid (23) and 4-oxo pimelic acid (24). The introduction of a keto group in position 4 in pimelic acid is lowering the Ki-value from 12'400 µM to 8'600 µM. Much to our surprise there is no significant difference as expressed by the inhibition constant between maleic acid (18) and fumaric acid (17). Even the more sterically demanding phthalic acid (19) is in the same range of inhibition.

The most evident difference is observed between sebacic acid (29) and the 4-oxo sebacic acid (30). The introduction of the keto function did not only improve the Ki value, like it was the case for pimelic acid, but a change of the type of inhibition occured. Sebacic acid (29) showed a moderate competitive inhibition (Ki=7'975 mM) whereas the 4-oxo-sebacic acid (30) became an irreversible inhibitor just due to the introduction of the keto function. Others cases of irreversible inhibition were found for the compounds 35 and 31. Even after 70 hours dialysis the inhibitor could not be removed from the enzyme.

Three compounds of this series rac-25, rac-26 et 29 showed competitive behaviour. Using lower concentrations of inhibitor and higher concentrations of ALA these compounds activated the enzyme. The different plots of the inhibition test allowed to recognize this unusual behaviour clearly. Testing of the two racemic analogues of the intermediate postulated by Shemin showed that the difference between the inhibition constants of the two diastereoisomers was only a factor of 1.5. This difference of inhibition potencity has been more visible when these compounds where tested with PBGS from Rhodobacter spheroïdes. The introduction of a nitro group in 34 compared to 32 improves slightly the Ki value. It was difficult to estimate correctly the inhibition constants because the compound 34 showed an activator effect (activation in the same conditions as rac-25, rac-26 and 29). The replacement of the carboxylate group by a nitro group in general decreases the Ki value. However the 2-nitro benzoic acid (20) shows a considerably increased Ki value compared to phthalic acid (19).

Table 2: Results of the inhibition tests