Discovery of carboxypeptidase Y as a catalyst for the incorporation of sterically demanding α-fluoroalkyl amino acids into peptides

Article abstract of DOI:10.1016/j.tetlet.2003.12.007

The first example of a direct enzymatic coupling of two different, sterically demanding C^α-fluoroalkyl amino acids to amino acid nucleophiles is reported. N-Protected Ala methyl ester derivatives bearing a methyl-, difluoromethyl-, or trifluoro

Full text of DOI:10.1016/j.tetlet.2003.12.007

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 1163–1165
Discovery of carboxypeptidase Y as a catalyst for the incorporation
of sterically demanding a-fluoroalkyl amino acids into peptides
Sven Thust and Beate Koksch*
Department of Organic Chemistry, Faculty of Chemistry and Mineralogy, University of Leipzig, Johannisallee 29,
D-04103 Leipzig, Germany
Received 7 November 2003; revised 25 November 2003; accepted 2 December 2003
Dedicated to Prof. Dr. H.-D. Jakubke on the occasion of his 70th birthday
Abstract—The first example of a direct enzymatic coupling of two different, sterically demanding Ca-fluoroalkyl amino acids to
amino acid nucleophiles is reported. N-Protected Ala methyl ester derivatives bearing a methyl-, difluoromethyl-, or trifluoromethyl
group, respectively, instead of the a-proton were accepted as substrates by carboxypeptidase Y and could, therefore, be coupled
directly to various nucleophiles.
ꢀ 2003 Published by Elsevier Ltd.
Effective site-specific incorporation of a wide variety of
non-natural amino acids into peptides and proteins
remains a topic of high interest. Highly functionalized
amino acid residues can serve as valuable tools to be
used as biophysical probes for investigation of struc-
ture–function relationships or for the construction of
tailor-made biomolecules. Incorporation of fluorine
usually influences the physical properties of amino acids
and proteins dramatically.1–3 Increased resistance to-
ward proteolysis as well as the stabilization of secondary
structures are mentioned here amongst others.4 More-
over, strategic placement of fluoroalkyl groups within
peptides or proteins, respectively, provides the oppor-
tunity for studying conformational properties, peptide/
protein-membrane interactions, or metabolic processes
by 19F NMR.5;6
Enzymatic peptide coupling represents an attractive
alternative to chemical peptide synthesis methods as the
latter often suffer from problems such as racemization
and expensive, time consuming side-chain protection/
deprotection strategies. We could already show the
applicability of several proteases for the synthesis of
peptides containing a-fluoroalkyl amino acids as long as
the a-fluoroalkyl amino acids do not appear in positions
P1 and P01 (notation according to Schechter and Ber-
ger7).8–11 Recently, we succeeded in incorporating
a-methyl as well as a-fluoroalkyl amino acids into the P1
position of peptides applying the substrate mimetic
concept.12 However, direct enzymatic coupling of
a-fluoroalkyl amino acids has failed so far.
Carboxypeptidase Y (CPY) is known to possess high
catalytic activity and enantioselectivity for kinetic reso-
lutions of many a-tertiary substituted carboxylic acid
esters.13 In contrast to the above mentioned proteases,
CPY has been shown to be provided with a broad S1
and S01 specificity while hydrophobic enzyme–substrate
interactions are of high importance.14–17
In our attempt to provide a broad variety of methods
for the fast and simple incorporation of fluorinated
amino acids into peptides, a commercially available
protease has been found that, for the first time, makes
possible the direct enzymatic coupling of the sterically
demanding a-fluoroalkyl alanine derivatives.
Protease catalyzed peptide synthesis can be considered
only on condition that substrate esters are hydrolyzed
by the protease within a reasonable period of time.
Therefore, several N-protected alanine methyl esters
bearing a methyl- (Me), difluoromethyl- (Dfm), or tri-
fluoromethyl- (Tfm) group, respectively, instead of the
a-proton were incubated with CPY. a-Amino isobutyric
Keywords: Carboxypeptidase Y; Dialkyl-substituted amino acid;
Enzyme; Enzymatic peptide ligation; Peptide.
* Corresponding author. Tel.: +49-341-9736532; fax: +49-341-9736-
599; e-mail: koksch@chemie.uni-leipzig.de
0040-4039/$ - see front matter ꢀ 2003 Published by Elsevier Ltd.
doi:10.1016/j.tetlet.2003.12.007

1164
S. Thust, B. Koksch / Tetrahedron Letters 45 (2004) 1163–1165
Scheme 1. General model of the CPY-catalyzed peptide synthesis using C a-alkylated alanine methyl ester and different nucleophilic amino com-
ponents. R1: CH3; R2: H, CH3, CH(CH3)2, CH2CH(CH3)2; x: 2, 3; y: 0, 1, 3.
for reaction of Z-R,S-(aDfm)Ala-OMe with H-Leu-
NH2 and reaction of Z-R,S-(aTfm)Ala-OMe with
H-Leu-NH2 (45%), respectively.
For the interpretation of the differences in product yields
between the Aib substrate and both of the a-fluoroalkyl
substituted amino acid esters, the use of racemic mix-
tures in the case of the latter ones has to be taken into
account. Therefore, the influence of the absolute con-
figuration of the a-fluoroalkyl amino acid on the effi-
ciency of peptide bond formation by CPY was studied
for the Tfm derivative in more detail. Separation of the
diastereomeric products of the enzymatic reaction
would have been the easiest way to do so. Unfortu-
nately, separation of the diastereomeric product mix-
tures by HPLC failed. Therefore, the dipeptide methyl
ester Z-Phe-R,S-(aTfm)Ala-OMe was chemically syn-
Figure 1. Carboxypeptidase Y catalyzed peptide synthesis of and dif-
ferent amino acid amides. Reagents and conditions: 0.1 M HEPES
thesized from Z-Phe-OH and racemic H-(aTfm)Ala-
OMe. Diastereomers of this dipeptide derivative could
be separated by flash chromatography (diastereomer I
buffer (4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid), 0.2 M
NaCl, 0.02 M CaCl2, 20% DMSO, pH 8.0, 37 ꢁC; [acyl donor]: 4 mM,
[acyl acceptor]: 40 mM, [CPY]: 1.6–3.2 · 10ꢀ6 M, all errors are less than
5%.
was identified as the product, which was eluted first
from the column, diastereomer II was eluted as the
second product). Both diastereomeric dipeptide esters
were reacted separately with all four nucleophilic amino
acid amides under CPY catalysis to give eight different
N-protected tripeptide amides. As an example, both
diastereomers of the tripeptide Z-Phe-(aTfm)Ala-Ala-
NH2 were synthesized in semi-preparative scale to
acid (Aib) was included in this study as a point of refe-
rence to be able to distinguish between electronic effects
of the fluorinated group and steric demands of a second
a-alkyl substituent. All derivatives were hydrolyzed by
CPY within short reaction times to give the free
N-protected amino acids.
1
enable H as well as 19F NMR characterization. The
absolute configuration of (aTfm)Ala within this tripep-
tide sequence could be assigned based on 19F NMR
analysis in combination with an earlier published X-ray
structure analyses.9 According to these data, diastereo-
mer I of the tripeptide Z-Phe-(aTfm)Ala-Ala-NH2 refers
to the all-S form while diastereomer II bears the
R-enantiomer of (aTfm)Ala.
Based on these results, the suitability of the Ca-fluoro-
alkyl alanine methyl ester to serve as acyl donors for
CPY-catalyzed peptide bond formation was studied.
Acyl acceptors were chosen following the known P10
specificity of CPY for hydrophobic amino acid amides.18
Therefore, H-Leu-NH2, H-Val-NH2, H-Ala-NH2, and
H-Gly-NH2, were used as nucleophilic components
(Scheme 1). The results are summarized in Figure 1.
CPY catalyzes the acyl transfer to all of the four amino
acid amides to give the desired N-protected dipeptide
amides. However, the extent of peptide bond formation
was considerably influenced by the second substituent at
the Ca-atom. In general, CPY was found to be consid-
erably more efficient in catalyzing peptide bond forma-
tion in the case of the Aib substrate. Highest peptide
yields (80% and 70%, respectively) were obtained for
reactions using nucleophiles with bulky residues (Leu
and Val, respectively). The efficiency of acyl transfer was
substantially diminished in the case of H-Ala-NH2
(40%) and even lower for H-Gly-NH2 (35%). This result
is in agreement with the known specificity of the pro-
tease for hydrophobic residues within both binding
sites.14–17 The same tendency was observed for both of
the corresponding Ca-fluoroalkyl amino acid esters.
Here, the highest peptide yields (50%) could be obtained
Comparison of product yields show that, in general,
CPY accepts both of the diastereomeric dipeptides while
certainly always preferring the S-configured (aTfm)Ala
in P1 position (Fig. 2). The strongest discrimination
between both diastereomeric substrates was observed
for the synthesis of Z-Phe-(aTfm)Ala-Ala-NH2. Reac-
tion of the diastereomeric peptide containing the
S-enantiomer of the fluorinated amino acid in P1 posi-
tion gave 55% yield while product formation dropped to
20% when S-(aTfm)Ala was replaced by the R-enan-
tiomer. Therefore, product formation increased up to
40% by applying the pure diastereomer of Z-Phe-S-
(aTfm)Ala-OMe and is now even slightly higher than in
the case of the corresponding Aib peptide. The highest
peptide yields out of the eight reactions could be
obtained for the synthesis of Z-Phe-S-(aTfm)Ala-Leu-
NH2 (75%). Interestingly, peptide yields for reactions of
Z-Phe-S-(aTfm)Ala-OMe with H-Gly-NH2 and H-Ala-

S. Thust, B. Koksch / Tetrahedron Letters 45 (2004) 1163–1165
1165
amides were used as nucleophiles. The application of
amino acids19 as nucleophiles will open up the possi-
bility of incorporating a variety of fluorinated residues
into biologically relevant peptides by enzymatic frag-
ment condensation.
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft
(Innovationskolleg ÔChemical Signal-Biological AnswerÕ
of the University of Leipzig) for financial support.
Figure 2. Influence of the absolute configuration of (aTfm)Ala on the
efficiency of CPY-catalyzed peptide synthesis. Reagents and condi-
tions: 0.1 M HEPES buffer, 0.2 M NaCl, 0.02 M CaCl2, 40% DMSO,
pH 8.0, 37 ꢁC; [acyl donor]: 4 mM, [acyl acceptor]: 40 mM, [CPY]:
3.0 · 10ꢀ7–1.6 · 10ꢀ6 M, all errors are less than 5%.
References and notes
1. Tang, Y.; Ghirlanda, G.; Vaidehi, N.; Kua, J.; Mainz, D.
T.; Goddard, W. A., III; DeGrado, W. F.; Tirell, D. A.
Biochemistry 2001, 40, 2790–2796.
2. Tang, Y.; Tirell, D. A. J. Am. Chem. Soc. 2001, 123,
11089–11090.
3. Bilgicer, B.; Xing, X.; Kumar, K. J. Am. Chem. Soc. 2001,
123, 11815–11816.
4. Koksch, B.; Sewald, N.; Jakubke, H.-D.; Burger, K. In
Biomedical Frontiers of Fluorine Chemistry; Ojima, I.,
McCarthy, J. R., Welch, J. T., Eds.; ACS Symposium
Series 639; American Chemical Society: Washington, DC,
1996; pp 42–58.
5. Entress, R. M. H.; Dancer, R. J.; OÕBrien, D. P.; Try, A.
C.; Cooper, M. A.; Williams, D. H. Chem. Biol. 1998, 5,
329–337.
6. Ulrich, A. S. In Encyclopedia of Spectroscopy and Spec-
trometry; Lindon, J., Tranter, G., Holmes, J., Eds.;
Academic: 2000; p 813.
7. Schechter, I.; Berger, A. C. Biochem. Biophys. Res.
Commun. 1967, 27, 157–162.
8. Bordusa, F.; Dahl, C.; Jakubke, H.-D.; Burger, K.;
Koksch, B. Tetrahedron: Asymmetry 1999, 10, 307–313.
9. Koksch, B.; Sewald, N.; Burger, K.; Jakubke, H.-D.
Amino Acids 1996, 11, 425–434.
10. Jakubke, H.-D. J. Chin. Chem. Soc. 1994, 41, 355–370.
11. Burger, K.; Muetze, K.; Hollweck, W.; Koksch, B.; Kuhl,
P.; Jakubke, H.-D.; Riede, J.; Schier, A. J. Prakt. Chem.
1993, 335, 321–331.
12. Thust, S.; Koksch, B. J. Org. Chem. 2003, 68, 2290–2296.
13. Kallwass, H. K. W.; Yee, C.; Blythe, T. A.; McNabb, T.
J.; Rogers, E. E.; Shames, S. L. Bioorg. Med. Chem. 1994,
2, 557–566.
14. Bai, Y.; Hayashi, R.; Hata, T. J. Biochem. (Tokyo) 1975,
78, 617–626.
15. Breddam, K. Carlsberg Res. Commun. 1984, 49, 535–554.
16. Breddam, K. Carlsberg Res. Commun. 1986, 51, 83–128.
17. Endrizzi, J. A.; Breddam, K.; Remington, S. J. Biochem.
1994, 33, 11106–11120.
18. Mortensen, U. H.; Stennicke, H. R.; Raaschou, M.;
Breddam, K. J. Am. Chem. Soc. 1994, 116, 34–41.
19. Eckstein, H.; Renner, H.-J. Peptides. In Proceedings of the
24th European Peptide Symposium, Edinburgh, 1996;
Mayflower Scientific: Kingswinford, UK, 1998; pp 355–
356.
NH2, respectively, are much higher now than observed
when reacting each of the nucleophiles with Z-Aib-OMe
(Fig. 1). These results imply that substitution of a
methyl group for a Tfm group at the Ca-atom of Aib
can result, depending on the absolute configuration, in
an improved binding of the substrate within the active
site of CPY. Remarkably, in the case of the a-Tfm
analogues, reaction of the diastereomer possessing the
fluorinated, bulkier side chain in the same position as it
would be for the side chain in the case of a natural
amino acid results in a lower peptide yield. Surprisingly,
CPY accepts the bulky Tfm group within the binding
site for the a-proton. The assumption can be made that
the a-Dfm substituted Ala derivative will show a similar
reaction behavior, however, this has to be further
investigated.
In conclusion, we present here the first example of a
direct enzymatic coupling of two different, sterically
demanding Ca-fluoroalkyl amino acids to amino acid
nucleophiles. N-Protected Ala methyl ester derivatives
bearing a methyl-, difluoromethyl-, or trifluoromethyl
group, respectively, instead of the a-proton were
accepted as substrates by CPY and could, therefore, be
coupled directly to various nucleophiles. Peptide yields
between 20% and 75% were obtained depending on the
nucleophilic amino acid derivative and the enantiomer
of the fluoroalkyl amino acid. To our knowledge, this is
the first time that sterically demanding Ca-dialkylated
amino acids have been coupled as substrates directly
with amino acid nucleophiles without any further acti-
vation of the electrophilic substrate and without any
medium or enzyme engineering. The synthetic strategy
introduced here, extends the scope of methods available
for site-specific peptide and protein modification by
fluorinated amino acids using a simple and environ-
mentally attractive route. In this study, amino acid