вход по аккаунту


Ligation of a Synthetic Peptide to the N Terminus of a Recombinant Protein Using Semisynthetic Protein trans-Splicing.

код для вставкиСкачать
Protein Modifications
DOI: 10.1002/anie.200600570
Ligation of a Synthetic Peptide to the N Terminus
of a Recombinant Protein Using Semisynthetic
Protein trans-Splicing**
Christina Ludwig, Martina Pfeiff, Uwe Linne, and
Henning D. Mootz*
Our ability to prepare chemically modified proteins with
defined covalent structure contributes significantly to the
understanding of protein function on the molecular level.
Unique biochemical, biophysical, and even cell-based investigations are enabled by the precise introduction of, for
example, fluorophores, biophysical probes, unnatural amino
acids (aa) with altered side chain or backbone composition,
and posttranslational modifications.[1] Several techniques
have been used to site-specifically incorporate such synthetic
building blocks into proteins.[2] Probably the currently most
widely used approach is based on the native chemical ligation
(NCL) methodology.[2b] It relies on the chemoselective
ligation of two polypeptides, either synthetic or recombinant,
one of which bears a C-terminal a-thioester and the other an
[*] C. Ludwig, M. Pfeiff, Dr. U. Linne, Dr. H. D. Mootz
Philipps-Universit*t Marburg
Fachbereich Chemie/Biochemie
Hans-Meerwein-Strasse, 35032 Marburg (Germany)
Fax: (+ 49) 6421-282-2191
[**] We thank Prof. Mohamed Marahiel for continuous support. This
research was funded by the DFG (Emmy Noether grant) and the
Fonds der Chemischen Industrie. C.L. is a recipient of a Ph.D.
fellowship from the Fonds der Chemischen Industrie.
Supporting information for this article is available on the WWW
under or from the author.
N-terminal cysteine residue, to give a fully synthetic or an Nor C-terminally modified semisynthetic protein (for reviews
see Ref. [3]). Nonetheless, the reaction conditions and the
nature of the required functional groups can limit the scope
and complicate the practical handling of NCL. Because it is a
bimolecular reaction, high reactant concentrations are
required. Problems can also be encountered during the
preparation of the reactants. In particular, the preparation
of peptides or proteins with an a-thioester can be technically
demanding. Protein a-thioesters are generated by thiolysis of
a fusion construct of the protein of interest with a modified
intein.[2e,f] Chemoenzymatic approaches to obtain semisynthetic proteins include reverse proteolysis,[2a,h,i, 4] which also
requires a peptide with an active ester, and the recently
reported use of sortase.[2k] Herein, we show that protein transsplicing can be exploited for the preparation of N-terminally
modified semisynthetic proteins. Our approach requires
neither an a-thioester at the synthetic peptide nor high
reactant concentrations.
Protein splicing describes the autocatalytic removal of an
internal protein domain, the intein, from a precursor protein,
while the flanking sequences, the N- and C-exteins, are
concomitantly linked with a native peptide bond.[5] In protein
trans-splicing, the intein domain is split into two protein
fragments, IntN and IntC, which form a complex and fold into
the active intein.[6] To use this process for the preparation of
semisynthetic proteins, we envisaged expression of the
recombinant protein of interest (POI) as C-extein in fusion
with an IntC fragment and chemical synthesis of a polypeptide
comprising the N-extein with the desired modification and the
IntN fragment (see Figure 1 A). However, the typical sizes of
split intein fragments with about 100–180 amino acids (aa) for
the IntN fragment and about 36–50 aa for the IntC fragment
have restricted their preparation by solid-phase peptide
synthesis (SPPS) to the IntC fragment.[7] Interestingly, a
recent genetic investigation on alternative split sites in the
Ssp DnaB intein showed that an IntN fragment of only 11 aa
and its corresponding 143-aa IntC counterpart were capable of
protein trans-splicing in vivo when co-expressed in E. coli.[8]
The key question at the beginning of this study was
whether the active intein could also be reconstituted from
these two fragments in vitro. Because the removed 11 aa of
the IntN fragment correspond to two short b strands that are
part of the catalytic pocket,[9] the 143-aa IntC fragment is most
likely prohibited to fold into a compact nativelike structure.
While the presence of endogenous chaperones and coexpression of the complementary fragment is likely to have
contributed to proper folding in vivo, the isolated fragments
would be expected to be more prone to misfolding and
aggregation, as has been observed for many split proteins and
In initial tests, we therefore prepared model intein fusion
proteins Trx-IntN (1) and IntC-His6 (2), which contain
thioredoxin (Trx) as the N-extein and a short sequence
including a hexahistidine tag (His6) as the C-extein (see
Figure 1 B). Following individual expression of 1 and 2 in
E. coli, both proteins could be purified from the soluble
fraction, suggesting at least partial folding of the IntC fragment. After 1 and 2 had been mixed in an equimolar ratio
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5218 –5221
Figure 1. Principle of the approach: A) Protein trans-splicing is mediated by N- and C-terminal intein fragments (IntN and IntC) and results in
ligation of the fused extein sequences with a native peptide bond. B) The group R, which represents the N-extein and may contain chemical
modifications, is ligated to the protein of interest (POI). The IntN fusion constructs can be either a synthetic peptide or a recombinant protein. Cterminal cleavage is a premature release of the POI from the IntC fragment and occurs as a side reaction. Calculated molecular masses (in Da): 1:
14 798.9, 2: 18 078.5, 3: 15 035.0, 4: 1320.5, 5: 16 539.0, 6: 2186.1, 7: 30 212.6, 8: 14 556.1, 9: 13 690.0, 10: 47 252.9, 11: 31 596.7, 12: 30 731.0.
bLac = b-lactamase, His6 = hexahistidine tag, Trx = thioredoxin.
kDa protein Trx represented the C-extein (Figure 1 B), was
(12 mm in each) and incubated at pH 7.0 and 25 8C, the
expressed in E. coli and purified from the soluble fraction.
formation of the expected splice products could be monitored
Incubation of 7 with synthetic peptide 6 yielded the desired
by SDS-PAGE and ESI-TOF MS (see the Supporting
semisynthetic protein Fl-Trx-His6 (8) in a time-dependent
Information). A kinetic analysis of the reaction revealed
that a splicing yield of about 25 % was reached after 40 h.
fashion. Also the intein products of the reaction, IntN (4) and
These data indicate that the IntN and IntC fragments are
IntC (5), could be identified (see Figure 2 A–C).
surprisingly robust and show that intein activity can be
The general assumption for protein trans-splicing is a fast
reconstituted in vitro from recombinant protein fragments
intein fragment association step followed by the rate-deterwithout further renaturation steps.
mining pseudo-monomolecular reaction of intein folding and
Encouraged by these findings, we replaced the recombiprotein splicing.[11] In agreement with this model, the kinetic
nant construct 1 with the synthetic peptide 5(6)-carboxyfluordata could be fitted by first-order kinetics. The calculated rate
escein- SEFSGCISGDSLISLASR (compound 6, see Figconstant of k = (4.1 0.2) B 10 5 s 1 is comparable to that
ure 1 B, the sequence of Int is underlined). 5(6)-Carboxydetermined for the naturally occurring split Ssp DnaE
fluorescein-SEFSG (abbreviated Fl)
represents the N-extein; the two
amino acids depicted in italics are
native residues flanking the intein.
The two C-terminal residues were
added to the peptide in analogy to
the sequence of 1. Upon incubation
of 6 and IntC-His6 (2), spontaneous
formation of the active intein from
the synthetic and recombinant fragments was again observed. The semisynthetic
FlSIEGSRSHHHHHH was detected
by UV light on a 20 % acrylamide
SDS-PAGE gel and by ESI-TOF MS
analysis (not shown; Mcalcd =
2423.3 Da,
Mexpt = 2423.8 Da).
Three native residues of the Cextein were included in all IntC constructs and are indicated in italics.
Figure 2. Semisynthetic protein trans-splicing. Analysis of the reaction of synthetic peptide 6 (80 mm) with
We moved on to explore the
recombinant protein IntC-Trx-His6 (7) (40 mm) at 25 8C to give semisynthetic Fl-Trx-His6 (8). A, B) Reaction
potential of our approach to prepare
progress monitored on an SDS-PAGE gel using Coomassie staining (A) or UV light (B) (* = impurity protein
semisynthetic proteins. The construct
band). C) ESI-TOF MS analysis of the reaction mixture after 48 h of incubation (mass signals of impurities
IntC-Trx-His6 (7), in which the 12are marked #). D) Time courses of the protein trans-splicing and the C-terminal cleavage reactions.
Angew. Chem. Int. Ed. 2006, 45, 5218 –5221
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
intein[11] but about 17-fold lower than that reported for the
Ssp DnaB intein split at the previously known permissive
site.[12] The likely explanation for this finding is that the more
radical split site employed in this study has a more significant
impact on association or folding of the two fragments. Despite
these kinetic differences the total achievable yield of protein
trans-splicing of 40–45 % is identical in both cases[12] and lies
in the typical range reported for various split inteins.[7a, 10c, 11–14]
Interestingly, another 40–45 % of IntC-Trx-His6 (7) was
transformed into the additional product 9, which was
identified by ESI-TOF MS analysis as the C-extein Trx-His6
(Figure 2 A–C). The formation of 9 by C-terminal cleavage
proceeded with a first-order rate constant of k = (2.1 0.9) B
10 5 s 1 (Figure 2 D). Although undesired, this common side
reaction in protein splicing is not problematic as long as the
splicing and cleavage products can be separated in a
subsequent purification step.[15]
Furthermore, we found that a molar excess of synthetic
peptide 6 over protein 7 did not improve the product yields
(not shown). The reaction is best performed at a molar ratio
of 1:1, which is favorable in economic terms, since only small
amounts of the synthetic peptide are consumed. In control
reactions using intein fragments with mutations in the key
catalytic residues we could confirm that the reaction still
proceeds according to the established pathway of protein
trans-splicing (see the Supporting Information).
To demonstrate the preparative utility of semisynthetic
protein trans-splicing, we undertook the synthesis, purification, and biochemical characterization of two fluoresceinlabeled proteins on a larger scale. For the semisynthesis of FlTrx-His6 (8), the above-mentioned 6 and 7 were mixed at
equimolar concentrations of 62 mm each in a total volume of
1.5 mL (corresponding to 0.2 mg of peptide and 2.8 mg of
protein). The reaction was allowed to proceed for 22 h at
25 8C, at which point about 40 % of 7 was converted to the
desired product. Pure Fl-Trx-His6 (8) was obtained after anion
exchange and a final step of Ni-NTA chromatography
(Figure 3 A–C) in an overall yield of about 30 % ( 0.4 mg).
Figure 3 D shows that the purified semisynthetic thioredoxin 8
exhibits slightly higher enzymatic activity in an insulinprecipitation assay[16] than the recombinant thioredoxin
positive control 1.
The 30.7-kDa enzyme b-lactamase (bLac) served as the
second example. The fusion protein IntC-bLac-His6 (10, see
Figure 1 B) was found in the inclusion bodies after expression
in E. coli. After solubilization in 8 m urea, it was purified using
Ni-NTA chromatography under denaturing conditions. Soluble protein was obtained by refolding in a one-step dialysis
against assay buffer. Incubation of peptide 6 with 10 at 25 8C
(12 mm in each, 1.7 mg of protein 10) for 42 h led to the
semisynthetic splice product Fl-bLac-His6 (11) in about 35 %
yield and similar amounts of the C-terminal cleavage product
12 (see Figure 4 A, B). After two-step purification involving
anion exchange and Ni-NTA chromatography, pure 11 was
obtained in an overall yield of about 30 % ( 0.3 mg,
Figure 4 C).
The enzymatic activity of the semisynthetic protein was
assayed photometrically by nitrocefin hydrolysis (Figure 4 D)
and was indistinguishable from that of an independently
Figure 3. Preparation and characterization of semisynthetic Fl-Trx-His6
(8). A, B) Reaction and purification monitored on a SDS-PAGE gel
using Coomassie staining (A) or UV light (B). Lane 1: peptide 6;
lane 2: protein 7; lane 3: 6 and 7 reacted at 25 8C and each 62 mm at
t = 0 h; lane 4: protein 7 alone under the same conditions for 22 h;
lane 5: same as for lane 3 but at t = 22 h; lane 6: purified semisynthetic product 8. C) ESI-TOF MS of pure product 8. D) Insulinprecipitation assay of enzymatic activity of semisynthetic 8 in comparison to that of the Trx control construct 1. Two negative controls, in
which either the Trx protein or the substrate DTT were omitted, are
Figure 4. Preparation and characterization of semisynthetic Fl-bLacHis6 (11). Analysis of the reaction and purification on an SDS-PAGE
gel using Coomassie staining (A) and UV light (B). Lane 1: peptide 6;
lane 2: protein 10; lane 3: 6 and 10 each 12 mm and t = 0 h; lane 4: 6
and 10 reacted at 25 8C and t = 42 h; lane 5: purified semisynthetic FlbLac-His6 (11). C) ESI-TOF MS of pure 11 (Mcalcd = 31 596.7 Da).
D) Enzymatic assay of 11 based on nitrocefin hydrolysis. The construct
b-lactamase-His6 was used as a positive control, construct 7 served as
a negative control.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 5218 –5221
prepared control construct bLac-His6 (KM(Fl-bLac) = (119.7 4.2) mm ;
KM(bLac) = (112.7 3.3) mm ;
kcat(Fl-bLac) = (212 30) s 1; kcat.(bLac) = (204 42) s 1). Thus, both semisynthetic
proteins were produced in pure form and good yields, and
their catalytic activities remained unaffected by the expression as an intein fusion and the subsequent protein-splicing
In conclusion, the approach described here is an attractive
and easy-to-perform alternative for the preparation of Nterminally modified semisynthetic proteins by protein transsplicing. As a main advantage, the synthesis of an a-thioester
peptide, as required for NCL, or other active esters, is
avoided. Also the generation of an N-terminal cysteine at the
recombinant protein, for example, by treatment with a
protease which entails an additional purification step, is
circumvented. Efficient reaction progress is obtained at
reactant concentrations as low as 10–60 mm due to the
inherent affinity of the self-cleaving intein fragments. The
amino acid left behind at the ligation junction is a serine,
which should generally be better tolerated in the protein of
interest than a cysteine in the case of NCL. Our approach has
the potential to be generally applicable, because inteins are
promiscuous with respect to their fused extein sequences.
Received: February 12, 2006
Published online: July 5, 2006
Keywords: intein · peptides · protein modifications ·
protein splicing
[1] a) M. E. Hahn, T. W. Muir, Trends Biochem. Sci. 2005, 30, 26;
b) L. Wang, P. G. Schultz, Angew. Chem. 2005, 117, 34; Angew.
Chem. Int. Ed. 2005, 44, 34.
[2] a) D. Y. Jackson, J. Burnier, C. Quan, M. Stanley, J. Tom, J. A.
Wells, Science 1994, 266, 243; b) P. E. Dawson, T. W. Muir, I.
Clark-Lewis, S. B. Kent, Science 1994, 266, 776; c) H. F. Gaertner, R. E. Offord, R. Cotton, D. Timms, R. Camble, K. Rose, J.
Biol. Chem. 1994, 269, 7224; d) D. A. Erlanson, M. Chytil, G. L.
Verdine, Chem. Biol. 1996, 3, 981; e) T. W. Muir, D. Sondhi, P. A.
Cole, Proc. Natl. Acad. Sci. USA 1998, 95, 6705; f) T. C.
Evans, Jr., J. Benner, M. Q. Xu, Protein Sci. 1998, 7, 2256;
g) L. Wang, A. Brock, B. Herberich, P. G. Schultz, Science 2001,
292, 498; h) N. Wehofsky, N. Koglin, S. Thust, F. Bordusa, J. Am.
Chem. Soc. 2003, 125, 6126; i) Z. Machova, R. von EggelkrautGottanka, N. Wehofsky, F. Bordusa, A. G. Beck-Sickinger,
Angew. Chem. 2003, 115, 5065; Angew. Chem. Int. Ed. 2003,
42, 4916; j) B. L. Nilsson, R. J. Hondal, M. B. Soellner, R. T.
Angew. Chem. Int. Ed. 2006, 45, 5218 –5221
Raines, J. Am. Chem. Soc. 2003, 125, 5268; k) H. Mao, S. A.
Hart, A. Schink, B. A. Pollok, J. Am. Chem. Soc. 2004, 126, 2670.
a) P. E. Dawson, S. B. Kent, Annu. Rev. Biochem. 2000, 69, 923;
b) T. W. Muir, Annu. Rev. Biochem. 2003, 72, 249; c) B. L.
Nilsson, M. B. Soellner, R. T. Raines, Annu. Rev. Biophys.
Biomol. Struct. 2005, 34, 91.
T. K. Chang, D. Y. Jackson, J. P. Burnier, J. A. Wells, Proc. Natl.
Acad. Sci. USA 1994, 91, 12 544.
C. J. Noren, J. Wang, F. B. Perler, Angew. Chem. 2000, 112, 458;
Angew. Chem. Int. Ed. 2000, 39, 450.
a) H. Wu, Z. Hu, X. Q. Liu, Proc. Natl. Acad. Sci. USA 1998, 95,
9226; b) H. Paulus, Annu. Rev. Biochem. 2000, 69, 447.
a) B. M. Lew, K. V. Mills, H. Paulus, Biopolymers 1999, 51, 355;
b) T. C. Evans, Jr., D. Martin, R. Kolly, D. Panne, L. Sun, I.
Ghosh, L. Chen, J. Benner, X. Q. Liu, M. Q. Xu, J. Biol. Chem.
2000, 275, 9091; c) I. Giriat, T. W. Muir, J. Am. Chem. Soc. 2003,
125, 7180.
W. Sun, J. Yang, X. Q. Liu, J. Biol. Chem. 2004, 279, 35 281.
Y. Ding, M. Q. Xu, I. Ghosh, X. Chen, S. Ferrandon, G. Lesage,
Z. Rao, J. Biol. Chem. 2003, 278, 39 133.
a) K. V. Mills, B. M. Lew, S. Jiang, H. Paulus, Proc. Natl. Acad.
Sci. USA 1998, 95, 3543; b) K. Shingledecker, S. Q. Jiang, H.
Paulus, Gene 1998, 207, 187; c) M. W. Southworth, E. Adam, D.
Panne, R. Byer, R. Kautz, F. B. Perler, EMBO J. 1998, 17, 918;
d) T. Yamazaki, T. Otomo, N. Oda, Y. Kyogoku, K. Uegaki, N.
Ito, Y. Ishino, H. Nakamura, J. Am. Chem. Soc. 1998, 120, 5591;
e) H. Yagi, T. Tsujimoto, T. Yamazaki, M. Yoshida, H. Akutsu, J.
Am. Chem. Soc. 2004, 126, 16 632.
D. D. Martin, M. Q. Xu, T. C. Evans, Jr., Biochemistry 2001, 40,
S. Brenzel, T. Kurpiers, H. D. Mootz, Biochemistry 2006, 45,
H. D. Mootz, T. W. Muir, J. Am. Chem. Soc. 2002, 124, 9044.
The residual fraction usually remains unreacted for unknown
reasons. Misfolding and aggregation are likely explanations.
We believe that the observed C-terminal cleavage reflects
incomplete folding of the intein under in vitro conditions,
because during co-expression of 1 and 7 in E. coli exclusively the
protein trans-splicing product was formed, whereas also Cterminal cleavage was observed when purified 1 and 7 were
incubated in in vitro reactions (see Figure S2 and data not
shown). This observation excludes potential sequence constraints as the reason for C-terminal cleavage. Several attempts
to eliminate the cleavage reaction by optimizing the reaction
conditions, that is, by varying the parameters temperature, salt
concentration, nature of reducing agent, and pH (see the
Supporting Information), were unsuccessful so far. However,
our results show that the IntN–IntC complex had formed almost
quantitatively, which should be a good starting point for future
optimization efforts to drive the reaction exclusively to the
splicing product.
A. Holmgren, J. Biol. Chem. 1979, 254, 9627.
2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Без категории
Размер файла
350 Кб
synthetic, using, ligation, recombinant, terminus, splicing, protein, semisynthetic, transp, peptide
Пожаловаться на содержимое документа