Synthesising protease-stable isopeptides by proteases: An efficient biocatalytic approach on the basis of a new type of substrate mimetics

Article abstract of DOI:10.1039/b105752a

A biocatalytic route to the 'post-synthesis' formation of Asp/Glu-derived isopeptides is illustrated on the basis of the Staphylococcus aureus V8 protease used as the biocatalyst, a new type of substrate mimetics as the donor peptides, and several accepto

Full text of DOI:10.1039/b105752a

Synthesising protease-stable isopeptides by proteases: an efficient
biocatalytic approach on the basis of a new type of substrate
mimetics†
Nicole Wehofsky,a Mandy Alischb and Frank Bordusa*ab
a Max-Planck Society, Research Unit ‘Enzymology of Protein Folding’, Weinbergweg 22, D-06120
Halle/Saale, Germany. E-mail: bordusa@enzyme-halle.mpg.de; Fax: +49 345 551 1972;
Tel: +49 345 552 2806
b University of Leipzig, Department of Biochemistry, Talstr. 33, D-04103 Leipzig, Germany
Received (in Cambridge, UK) 2nd July 2001, Accepted 19th July 2001
First published as an Advance Article on the web 9th August 2001
A biocatalytic route to the ‘post-synthesis’ formation of Asp/
Glu-derived isopeptides is illustrated on the basis of the
Staphylococcus aureus V8 protease used as the biocatalyst, a
new type of substrate mimetics as the donor peptides, and
several acceptor peptides varying in length and sequence.
isoAsp- or isoGlu-Xaa bonds. Therefore, the development of
alternative catalytic and mild approaches to improve the current
synthesis methods of isopeptides is a challenging task.
This account reports on the use of a normal peptidase, i.e. the
Glu-specific endopeptidase from Staphylococcus aureus strain
V8 (V8 protease), that acts as an efficient biocatalyst for the
formation of isopeptides. The key feature of this approach is the
use of a novel type of substrate mimetics that directs the
synthesis activity of the protease to the side-chain carboxyl
moiety of Asp and Glu (Scheme 1). Similar to classical
substrate mimetics, the novel type donors bear a site-specific
ester leaving group that mediates the acceptance of non-specific
acyl moieties by the protease.9 To direct the activity of the
enzyme to the side-chain, the specific ester group, however, is
linked to the w-carboxyl moiety of Asp and Glu instead of being
at the C-terminus of the peptide donor. This different archi-
tecture was shown to lead to a shift in the synthetic activity of
the protease from the Ca-carboxyl group towards the side-chain
moiety of the two amino acids finally resulting in the synthesis
of isopeptides.
The general function of this biocatalytic approach was proved
by the use of several Asp- and Glu-containing peptides that
were esterified by carboxymethylthiol at their side-chain
carboxylates. The carboxymethyl thioester functionality (SCm,
cf. Scheme 1) itself was selected due to its function as a mimic
which was already shown to mediate the acceptance of non-
specific (non-Asp- and Glu-containing) linear donor peptides
by V8 protease and other highly Glu-specific peptidases as
well.10 The influence of the length of the donor peptide on the
synthesis reaction was investigated by using donor esters
derived from single amino acids up to pentapeptides. Fur-
thermore, the effect of the position of the Asp and Glu moieties
within the donor peptide on the acceptance by the enzyme was
studied by using peptides with either N-/C-terminal or endoge-
nous Glu/Asp-residues. The resulting 14 peptide esters have
served as donor components in V8 protease-catalysed isopep-
tide syntheses using amino acid amides, dipeptide amides, and
The trifunctional structure of Asp and Glu allows both amino
acids to be incorporated in the backbone of polypeptides by the
side-chain and the Ca-carboxyl moiety as well. While the latter
results in the formation of normal linear peptides, the involve-
ment of the side-chain carboxyl group in the peptide backbone
leads to the corresponding isopeptides. Isopeptides are ubiqui-
tously found in nature and are considered to be one of the most
common forms of non-enzymatic degradation of linear poly-
peptides under physiological conditions.1 Rearrangement of
normal peptides into isopeptides is catalysed by bases as well as
acids, and starts with the formation of cyclic succinimide
intermediates following dehydration. Re-opening of the ring
finally leads to a mixture of normal peptides and isopeptides in
a typical ratio that ranges from 40+60 to 15+85 along with small
amounts of D-isopeptides.2 Since isopeptide formation involves
the addition of extra carbon atoms to the polypeptide backbone,
its presence in a functional protein domain has profound effects
on biological activity and its susceptibility to proteolytic
degradation.3 Recent findings have demonstrated that wide-
spread accumulation of isopeptide linkages in cellular proteins
greatly increases the immunogenicity and disrupts a wide range
of important biochemical pathways by competitively displacing
normal proteins in protein–protein or protein–ligand inter-
actions.4 Because of this function, isopeptides were found to be
highly useful to study complex biological phenomena and for
the development of therapeutic agents resulting in a great
interest in their synthesis.
Although isopeptide bond formation is often an undesired
side reaction during the chemical synthesis of normal linear
peptides due to the need for strong bases and acids for
deprotection steps and peptide release,5 the directed synthesis of
isopeptides is difficult and strictly limited by competitive
cyclisation reactions. Like bases and acids, the use of coupling
reagents to yield side-chain activation of Glu- and Asp-
containing peptides further reinforces the formation of un-
wanted cyclic products and finally leads to a crude synthesis
product with substantial amounts of undesired material. A few
enzymes, i.e. protein carboxyl methyltransferases,6 isopepti-
dases,7 and transglutaminases,8 are known to be active on the
side-chain of Glu and Asp and, therefore, may be interesting
biocatalysts for the synthesis of isopeptide bonds. However, due
to the highly restricted substrate and reaction specificity, none
of these enzymes is capable of catalysing the synthesis of
Scheme 1 General structures of classical linear (1) and new type (2)
substrate mimetics. The site-specific carboxymethyl thioester group is
emphasized by bold letters. PG, protecting group; R1, R2, individual side
chains.
† Electronic supplementary information (ESI) available: complete details
for experimental procedures. See http://www.rsc.org/suppdata/cc/b1/
b105752a/
1602
Chem. Commun., 2001, 1602–1603
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b105752a

Table 1 Yields (%) of the V8 protease-catalysed synthesis of isopeptides. [donor]: 2 mM, [acceptor]: 15 mM, [V8 protease]: 3.61–10.33 mM
Acceptor peptide
Donor peptide
H-Met-NH2
H-Gly-Leu-NH2
H-Leu-Gly-NH2
Ile-Ala-Ala-Ala-Gly Leu-Ala-Ala-Ala-Gly
Z-Glu/Asp(SCm)-OH
41.2/41.5
41.3/59.5
44.9/53.9
39.0/51.7
44.7/58.7
42.5/60.9
40.3/41.6
39.8/56.9
45.1/56.6
39.6/54.2
45.4/55.8
43.9/62.8
41.5/61.9
47.7/48.5
47.0/60.3
50.4/61.9
46.5/59.7
52.6/61.7
50.2/65.7
51.9/64.0
49.8/49.3
47.2/67.9
49.4/62.3
48.6/66.8
54.6/67.5
50.1/67.4
52.4/66.4
50.5/52.2
50.3/65.9
51.0/61.8
49.3/64.3
54.9/64.8
51.6/68.0
52.8/65.0
Z-Ala-Glu/Asp(SCm)-NH2
Z-Ala-Ala-Glu/Asp(SCm)-NH2
Z-Glu/Asp(SCm)-Ala-NH2
Z-Glu/Asp(SCm)-Ala-Ala-NH2
Z-Ala-Glu/Asp(SCm)-Ala-NH2
Z-Ala-Ala-Glu/Asp(SCm)-Ala-Ala-NH2 40.8/59.1
pentapeptides as the amino components. The reactions them-
selves were performed under identical conditions at pH and
temperature optimal to the enzyme.11 As a control for
spontaneous hydrolysis and aminolysis of the peptide esters that
may interfere with the enzymatic syntheses, parallel reactions
without enzyme were analysed. On the basis of these control
experiments, non-enzymatic aminolysis could be ruled out and
the extent of spontaneous hydrolysis was found to be less than
5%. The results observed for the enzyme-catalysed syntheses
are summarised in Table 1. Generally, the data document that all
donor peptides show productive binding at the active site of the
biocatalyst resulting in isopeptide bond formation. In contrast,
model reactions using donor components lacking the specific
ester moiety, i.e. Z-Glu(SMe)-OH and Z-Asp(SMe)-OH (SMe,
methylthiol), gave no reaction. This finding indicates that the
negative charge of the SCm-leaving group is essential to
mediate the acceptance by V8 protease. Accordingly, the lack of
this charge causes a complete loss of specificity that inevitably
leads to a loss of synthesis activity of the enzyme. On analysis
of the efficiency of syntheses, yields of isopeptide products
within a range of about 40–70% were obtained that roughly
correspond with those of comparable reactions using normal
linear substrate mimetics.10 Interestingly, apart from the
formation of the respective hydrolysed donor peptides no
further side products could be detected. Addressing the
moderate variations in the yields, the Asp- and Glu-residue
itself appears to affect the efficiency of synthesis to the highest
extent. While the use of Z-Glu(SCm)-OH and Z-Asp(SCm)-OH
led to practically the same yields, the chain elongation of the
Asp ester either in the C- or N-terminal direction resulted in an
increase in the product yields of around approximately 10–20%.
This increase makes the Asp-containing peptides more efficient
donor components than those derived from glutamic acid.
Interestingly, the individual position of both the Asp and Glu
residue within the donor peptide does not affect the course of
isopeptide synthesis significantly. In the same way, there is only
a minor influence of the length of the donor peptide on the
reaction. Surprisingly, this also holds for the acceptor peptide as
it is reflected by the similar product yields which are less
affected by the sequence and the chain length of the respective
peptide. This atypical behaviour is in contrast to reactions using
classical linear substrate mimetics as the donor peptides that
generally displayed a more pronounced influence of the
acceptor peptide on the course of synthesis.10 Accordingly, in
particular, N-terminal Gly and Met residues usually led to a
decrease in product yield; an effect that can not be found for the
synthesis of isopeptides. The synthesis of glutathione (g-
glutamylcyteinylglycine) starting from Z-Glu(SCm)-OH and
Cys-Gly, which proceeds with a yield of about 62%, gave a
further hint to the small influence of the acceptor peptide on the
efficiency of catalysis. From a synthetic point of view, this
broad tolerance towards the sequence and chain length of both
the acceptor and donor peptide should make the approach
presented a rather general one for the synthesis of a wide variety
of isoaspartyl- and isoglutamyl-containing peptides. Further-
more, due to the mild reaction conditions, the weak carboxyl
activation, and abandoning the use of additional bases and
acids, the risk of undesired cyclisation reactions is reduced to a
minimum. The synthetic utility of the method is even not
narrowed by proteolytic side reactions. Model syntheses with
elongated reaction times up to several days do not gave any
hints to an undesired cleavage activity of V8 protease towards
the formed isoGlu- and isoAsp-bonds indicating an irreversible
course of the enzymatic ligation reaction (data not shown). Only
the presence of further Glu and Asp residues additional to those
that are to be modified may lead to undesired peptide cleavages.
In such cases, freezing of the reaction mixture, which was
shown to repress the proteolytic activity of enzymes closely
related to V8 protease while retaining their synthesis activity,12
may represent an efficient resource. From a biocatalytic point of
view, the approach presented broadens the scope of proteases in
organic synthesis and opens up a new field of synthetic
applications of these enzymes for the synthesis of isopeptides.
Finally, because of the universal applicability of substrate
mimetics it can be further expected that other proteases may
also be useful biocatalysts for isopeptide synthesis. Studies in
this direction are presently under investigation.
Generous financial support by the DFG (Innovationskolleg
‘Chemisches Signal und biologische Antwort’) and Fonds der
Chemischen Industrie (Liebig-scholarship, F. B.) is gratefully
acknowledged. The authors thank Professor K. Burger and co-
workers for helpful discussions and technical support and
Professors G. Fischer and A. Beck-Sickinger for hosting.
Notes and references
1 L. Gráf, S. Bajusz, A. Patthy, E. Barát and G. Cseh, Acta Biochim.
Biophys. Acad. Sci. Hung., 1971, 6, 415; B. A. Johnson, J. M.
Shirokawa, W. S. Hancock, M. W. Spellmann, L. J. Basa and D. W.
Aswad, J. Biol. Chem., 1989, 264, 14262.
2 T. Geiger and S. Clarke, J. Biol. Chem., 1987, 262, 785; B. A. Johnson,
E. D. Murray, Jr., S. Clarke, D. B. Glass and D. W. Aswad, J. Biol.
Chem., 1987, 262, 5622.
3 D. W. Aswad, M. V. Paranandi and B. T. Schurter, J. Pharm. Biomed.
Anal., 2000, 21, 1129.
4 M. J. Mamula, Immunol. Rev., 1998, 164, 231; M. J. Mamula, R. J. Gee,
J. I. Elliot, A. Sette, S. Southwood, P.-J. Jones and P. R. Blier, J. Biol.
Chem., 1999, 274, 22321.
5 M. Bodanszky and J. Z. Kwei, Int. J. Pept. Protein Res., 1978, 12, 69;
M. Bodanszky, J. C. Tolle, S. S. Deshmane and A. Bodanszky, Int. J.
Pept. Protein Res., 1978, 12, 57; G. Barany and R. B. Merrifield, in The
Peptides, ed. E. Gross and J. Meienhofer, Academic Press, New York,
1979, 193.
6 E. D. Murray, Jr. and S. Clarke, J. Biol. Chem., 1984, 259, 10722; P. N.
McFadden and S. Clarke, J. Biol. Chem., 1986, 261, 11503; I. M. Ota
and S. Clarke, J. Biol. Chem., 1989, 264, 54.
7 A. G. Loewy, J. K. Blodgett, F. R. Blase and M. H. May, Anal.
Biochem., 1997, 246, 111.
8 M. L. Fink and J. E. Folk, Methods Enzymol., 1983, 94, 347.
9 V. Schellenberger, H.-D. Jakubke, N. P. Zapevalova and Y. V. Mitin,
Biotechnol. Bioeng., 1991, 38, 104; H. Sekizaki, K. Itoh, E. Toyota and
K. Tanizawa, Chem. Pharm. Bull., 1996, 44, 1585; F. Bordusa, D.
Ullmann, C. Elsner and H.-D. Jakubke, Angew. Chem. Int. Ed. Engl.,
1997, 36, 2473; M. Thormann, S. Thust, H.-J. Hofmann and F. Bordusa,
Biochemistry, 1999, 38, 6056.
10 N. Wehofsky and F. Bordusa, FEBS Lett., 1999, 443, 220; N. Wehofsky,
J.-D. Wissmann, M. Alisch and F. Bordusa, Biochim. Biophys. Acta,
2000, 1479, 114.
11 G. R. Drapeau, Y. Boily and J. Houmard, J. Biol. Chem., 1972, 247,
6720.
12 N. Wehofsky, S. W. Kirbach, M. Haensler, J.-D. Wissmann and F.
Bordusa, Org. Lett., 2000, 2, 2027.
Chem. Commun., 2001, 1602–1603
1603