A. Lang, P. Kuhl / Tetrahedron Letters 51 (2010) 3779–3781
3781
Table 3
interesting properties, which can easily be prepared by protease-
catalysed synthesis.
Protease-catalysed synthesis of Z-Ala-L-ACL (3b) in an acidic aqueous–organic
mediuma
Acyl donor–
nucleophile-ratio
Reaction time (h)
Yield of Z-Ala-L-ACL (%)
Acknowledgement
Papain
Bromelain
Ficin
The authors thank Ms A. Kühnel for some experimental
contributions.
1:5
2
24
47
43
31
23
82
69
a
20 mg protease, 40 °C, 2 mL buffer (0.1 M sodium citrate buffer, pH 5.0); 0.5 mL
References and notes
methanol; 0.05 M Z-Ala-OMe (1b, acyl donor).
1. Jakubke, H.-D. Enzyme Catalysis in Organic Synthesis; VCH Verlagsgesellschaft
mbH: Weinheim, 1995.
2. Bordusa, F. Chem. Rev. 2002, 102, 4817–4867.
3. Tai, D.-F. Curr. Org. Chem. 2003, 7, 515–554.
4. van’t Hoff, J. H. Z. Anorg. Chem. 1898, 18, 1–13.
highest yields, with Z-Ala-OMe the proteases show rather less
conversion than with Z-Gly-OMe. For papain this is surprising be-
cause of its known far wider substrate acceptance than that of ficin.
5. Mitin, Y. V.; Braun, K.; Kuhl, P. Biotechnol. Bioeng. 1997, 54, 287–290.
6. Clapés, P.; Infante, M. R. Biocatal. Biotransform. 2002, 20, 215–233.
7. Lang, A.; Hatscher, C.; Kuhl, P. Tetrahedron Lett. 2007, 48, 3371–3374.
8. Lang, A.; Hatscher, C.; Wiegert, C.; Kuhl, P. Amino Acids 2009, 36, 333–340.
9. Schechter, I.; Berger, A. Biochem. Biophys. Res. Commun. 1967, 27, 157–167.
10. Barbas, C. F.; Wong, C. J. Chem. Soc., Chem. Commun. 1987, 533–534.
11. Kuhl, P.; Jakubke, H.-D. Pharmazie 1990, 45, 393–400.
12. Mitin, Y. V.; Zapevalova, N. P.; Gorbunova, E. Y. Int. J. Peptide Protein Res. 1984,
23, 528–534.
The formation of Z-Gly-L-ACL catalysed by bromelain seems to be
the slowest process. After 2 h there is the lowest yield compared
to the other proteases but it increases during the ongoing reaction
until 24 h, whereas for papain and ficin the gain ratios remain
nearly constant. The reaction time of 2 h is rather short and reason-
able and gives for Z-Ala-L-ACL better conversion than in reactions
over 24 h, without much doubt due to partial secondary hydrolysis
in the latter case.
In additional experiments an aqueous–organic medium has
been tested for the synthesis of Z-Ala-L-ACL consisting of 20%
methanol and 80% buffer with a pH of 5.0 near the pH optimum
of thiol proteases. The results are presented in Table 3.
Compared to the product yields obtained in basic buffer this
medium can just compete in ficin-catalysed reactions. Papain-
and bromelain-catalysed conversions differ from the results under
basic conditions. In that acidic medium also secondary hydrolysis
occurs. This is apparent from the yields after 24 h. All tested buf-
fer-methanol compositions displayed this characteristic.
Summing up all reactions, ficin seems to be the best thiol
protease for these types of substrates and this nucleophile. The
difference with the three thiol proteases is significant in the case
13. Tai, D.-F.; Fu, S.-L. J. Chin. Chem. Soc. 2003, 50, 179–183.
14. Several reaction batches were pooled to gain an amount of about 0.5 mmol
Z-Xaa-L-ACL. The mixture was evaporated to dryness under vacuum. The crude
product was dissolved in 20 mL CHCl3. The solution was washed thrice with
5 mL saturated sodium bicarbonate solution, and the two phases were
separated. The aqueous phase was discarded to waste, and from the organic
layer ACL was extracted twice with 5 mL 1 M HCl. Then, the organic phase was
washed twice with 2 mL distilled water and dried afterwards with sodium
sulfate, filtered and evaporated. The obtained product was desiccated
overnight at 40 °C. 1H and 13C NMR spectra were recorded with DRX 500,
Bruker. Polarimetry was performed with Model 341 LC, Perkin Elmer
Instruments and ESI/APCI-MS data were acquired by LC/MS equipment HP
1100—Bruker Esquire Ion Trap.
15. Z-Gly-
L
-ACL: white solid; ½a D20
ꢂ
8.3 (c 0.4, MeOH); 1H NMR [500 MHz, CDCl3] d
(in ppm) 1.37–1.47 (m, 1H, CH2 (ACL)), 1.47–1.52 (m, 1H, CH2 (ACL)), 1.79–1.82
(m, 1H, CH2 (ACL)), 1.82–1.86 (m, 1H, CH2 (ACL)), 1.98–2.01 (m, 1H, CH2 (ACL)),
2.05–2.08 (m, 1H, CH2 (ACL)), 3.23–3.28 (m, 2H, CH2 (ACL)), 3.86–3.97 (dd, 2H,
CH2 (Gly)), 4.50–4.53 (q, 1H, CH (ACL)), 5.12 (s, 2H, CH2 (Z)), 5.41 (t, 1H, NH
(Gly)), 6.01 (t, 1H, NH (ACL)), 7.22 (d, 1H, NH (ACL-Ala)), 7.29–7.33 (m, 3H,
C6H5 (Z)), 7.35 (d, 3J = 4.4 Hz, 2H, C6H5 (Z)); 13C NMR [125.75 MHz, CDCl3] d (in
ppm) 27.86 (CH2, ACL), 28.82 (CH2, ACL), 31.42 (CH2, ACL), 42.16 (CH2, ACL),
44.33 (CH2, Gly), 52.25 (CH, ACL), 67.14 (CH2, Z), 128.11, 128.17, 128.53, 136.18
(C6H5, Z), 156.39 (CO, Z), 167.77 (CO, Gly), 175.00 (CO, ACL); ESI-MS m/z 320.1
[M+H+], m/z 342.1 [M+Na+], C16H21N3O4 requires 319.4
of Z-Ala-L-ACL, but more balanced in the formation of Z-Gly-L-ACL.
In summary, this work emphasises the far wider synthetic
potential of proteases than that known so far. It gives proof of
introducing another 1,2-amino ketone besides AAP as the nucleo-
phile in the amidation of amino acid derivatives. For the first time
16. Z-Ala-
L
-ACL: white solid; ½a D20
ꢂ
ꢃ13.7 (c 0.4, MeOH); 1H NMR [500 MHz, DMSO-
d6] d (in ppm) 1.20 (d, 3J = 12.6 Hz, 3H, CH3 (Ala)), 1.13–1.24 (m, 2H, CH2 (ACL)),
1.71–1.81 (m, 2H, CH2 (ACL)), 1.85–1.88 (m, 2H, CH2 (ACL)), 3.03–3.07 (m, 2H,
CH2 (ACL)), 4.02–4.06 (quin, 1H, CH (Ala)), 4.31–4.35 (q, 1H, CH (ACL)), 5.02 (s,
2H, CH2 (Z)), 7.30–7.33 (m, 1H, C6H5 (Z)), 7.33–7.38 (m, 4H, C6H5 (Z)), 7.63 (d,
3J = 7.6 Hz, 1H, NH (ACL-Ala)), 7.78 (d, 3J = 6.5 Hz, 1H, NH (Ala)), 7.88 (t, 1H, NH
(ACL)); 13C NMR [125.75 MHz, DMSO-d6] d (in ppm) 18.06 (CH3, Ala), 27.68
(CH2, ACL), 28.87 (CH2, ACL), 31.11 (CH2, ACL), 40.65 (CH2, ACL), 50.40 (CH,
Ala), 51.38 (CH, ACL), 65.44 (CH2, Z), 127.71, 127.83, 128.43, 137.07 (C6H5, Z),
155.74 (CO, Z), 171.49 (CO, ACL), 174.14 (CO, Ala); ESI-MS m/z 334.2 [M+H+],
m/z 356.2 [M+Na+], C17H23N3O4 requires 333.4
Z-L-aminoacyl-L-caprolactam amides could be synthesised by using
thiol proteases as catalysts, because there is nothing reported so far
about these compounds neither in enzymatic nor in chemical syn-
thesis. The findings accentuate the stereospecificity of proteases to
prefer the
L-configured isomer of a DL-ACL racemate. These results
encourage the idea of finding new substances with possibly