вход по аккаунту


The Architecture of ProteinЦLigand Binding Sites Revealed through Template-Assisted Intramolecular PeptideЦPeptide Interactions.

код для вставкиСкачать
Protein–Ligand Interactions
The Architecture of Protein–Ligand Binding Sites
Revealed through Template-Assisted
Intramolecular Peptide–Peptide Interactions**
Chao Yu, Miroslav Malesevic, Gnther Jahreis,
Mike Schutkowski, Gunter Fischer,* and
Cordelia Schiene-Fischer
Many fundamental processes of life are based on protein–
peptide or protein–protein interactions. An array of discontinuous polypeptide segments within a given protein may
account for the specificity and free energy change of complex
formation.[1] Characterization of the polypeptide segments
that compose the ligand-binding surface requires expensive
structural investigations such as multidimensional NMR
spectroscopy, X-ray crystallography, and phage-displayed
shotgun scanning.[2]
Herein we report a simple method for obtaining a lowresolution three-dimensional picture of protein–peptide- or
protein–protein-interaction sites. The method involves the
generation of a library of short oligopeptides derived from the
primary sequences of interacting proteins. The peptides are
spotted onto cellulose membranes as template-bound pairs.
Spot synthesis[3] of peptide pairs is followed by an assay for
intramolecular interactions. We have termed this combined
[*] Dr. C. Yu,[+] Dr. M. Malesevic, Dr. G. Jahreis, Dr. M. Schutkowski,[++]
Prof. G. Fischer, Dr. C. Schiene-Fischer
Max Planck Research Unit for Enzymology of Protein Folding
Weinbergweg 22, 06120 Halle/Saale (Germany)
Fax: (+ 49) 345-551-1972
[+] Present address: Nuffield Department of Clinical Medicine
John Radcliff Hospital
Headington, Oxford, OX3 9DU (UK)
[++] Present address: Jerini AG
Invalidenstrasse 130, 10115 Berlin (Germany)
[**] This work was supported by the Deutsche Forschungsgemeinschaft
(SFB 610) and the Fonds der Chemischen Industrie.
Supporting information for this article is available on the WWW
under or from the author.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
technique the IANUS (induced organization of structure by
matrix-assisted togetherness) spot array. Each IANUS spot is
synthesized to comprise two peptide blocks. One block bears
a constant peptide sequence in all spots, in which the
sequence represents either the ligand peptide of a protein–
peptide complex or an interacting segment of a protein–
protein complex. The second block consists of variable
sequences of constant length that represent overlapping
regions of the entire partner protein sequence. After the
synthesis is complete, the spot array includes the entire
collection of linear motifs, including those of the binding cleft
of the protein–ligand complex.
In aqueous solution, short peptides exist in multiple
rapidly interchanging conformational states, and are expected
to exhibit a full range of structural elements specific for any
type of peptide–peptide, peptide–protein, or peptide–interface interaction.[4] Peptide chain arrangements that contain
multiple conformational constraints can induce a population
of molecules to favor restricted conformations through
mutual conformational induction. Typical constraints within
IANUS spot arrays include template-assisted intramolecular
interactions within the IANUS peptide pairs, high effective
concentration brought about by intramolecularity, and a
microenvironment of decreased polarity. In effect, matrixassisted togetherness of peptide chains could produce a large
population of binding-favorable conformations when the
peptide segments originate from the interacting sites of a
protein–protein complex.
Conceptually, IANUS spots that arise from protein segments distant from the active site must somehow be
distinguished from those obtained from binding-site-derived
peptide pairs. As the recognition site of a peptide becomes
buried upon protein–peptide-ligand complexation, a similar
type of masking is expected to be significant for spots
containing natively interacting peptide pairs. Noninteracting
IANUS peptide pairs, in contrast, leave the peptide ligands
unmasked. The resulting different binding properties for the
applied soluble protein allows a readout of the IANUS spot
arrays and the detection of conformational induction. The
result would be a collection of all protein segments that
correspond to the interaction sites of the protein–peptide or
protein–protein complexes.
In this study, spots were assayed by the application of
soluble protein, which competes with the protein-derived
peptides for binding to the matrix-bound ligand. After this
treatment, IANUS peptide pairs should produce a spot
pattern that reveals either: a) strong protein binding to
spots of noninteracting peptide pairs, or b) diminished protein
binding to spots containing peptide pairs that interact with
each other through conformational induction (Figure 1).
To test the method, an IANUS spot array experiment was
conducted with the complex that is formed between streptavidin and Strep-tag II (Stt II). The streptavidin homotetramer
from Streptomyces avidinii forms a complex with the 10residue (SNWSHPQFEK) Stt II peptide that is bound to the
biotin binding pocket.[5] Our IANUS spot array consists of
variable streptavidin-derived 12-residue peptides paired with
the constant Stt II peptide. Both were synthesized on the
same bifunctional lysyl template at a molar ratio of 1:1 for
DOI: 10.1002/anie.200460991
Angew. Chem. Int. Ed. 2005, 44, 1408 –1412
Figure 1. Treatment of spots with an interacting protein that competes
with protein-derived peptides for binding to a peptide ligand. Right:
noninteracting peptide pairs; strong protein binding. Left: ligand
peptides are masked through conformational induction, and protein
binding is weak. IP = interacting protein; IgG = protein-specific
each spot. The set of overlapping 12-mer peptides that
collectively span the streptavidin sequence with a shift of
three amino acid residues totals 50 individual spots, and was
synthesized on the Fmoc site of the orthogonally protected
template 1 (Scheme 1). The Stt II peptide was synthesized on
the Dde site of the template at every spot, and both peptides
were N-terminally acetylated.
Scheme 1. Templates 1 and 2 are attached to the cellulose membrane
by the (b-Ala)2 spacer. Boc = tert-butoxycarbonyl; Fmoc = 9-fluorenylmethoxycarbonyl; Dde = 2-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl.
To analyze the quality of the syntheses, peptide pairs of
four representative spots, both with and without intramolecular interactions, were synthesized in the dark on predefined
spots with a photolabile nitrobenzyl-based linker placed
between each of the peptide chains and template 2.[6] After
photolytic cleavage of the peptide pairs from the solid
support, MALDI-TOF mass spectrometric analysis was
performed. For every spot, each of the two peptides analyzed
showed a molecular mass in agreement with calculated
predictions (Supporting Information). To investigate the
influence of the matrix anchoring site on the IANUS peptide
pairs on template 1, streptavidin–Stt II peptides were also
Angew. Chem. Int. Ed. 2005, 44, 1408 –1412
synthesized on template 2 (Scheme 1). The streptavidin
peptide block of 12-mers was synthesized on the Fmoc site
and spanned the streptavidin sequence with a shift of two
amino acid residues, which required 75 individual spots.
Conformational induction was detected by treatment of the
IANUS spot array with soluble streptavidin; spots in which
the binding face of Stt II was freely accessible gave rise to
Stt II-mediated protein absorption, in contrast to spots in
which the Stt II binding face was blocked through conformational induction. Streptavidin bound to IANUS spot arrays
was detected with Western blot analysis. In fact, characteristically dark, strong Western blot signals occurred for the
majority of spots prepared with template 1. However, two
regions showed weaker signals for two or more adjacent spots,
which indicates a blocked Stt II binding face (Figure 2 a, blue
and pink underlining). These were termed IANUS-positive
spots, and were quantitatively identified by the individual
deviation from the reciprocal of mean spot intensity (Figure 2 b). The competing streptavidin–Stt II-like interaction of
the IANUS peptide pair, which precludes soluble streptavidin
protein binding, can be hypothesized to account for the
appearance of IANUS-positive spots. Owing to sequence
overlap of the streptavidin peptides, only two or more
Figure 2. Binding of streptavidin to the streptavidin–Stt II peptide array
attached to template 1. The streptavidin sequence is collectively represented by 12-mer peptides that overlap with a shift of three amino acid
residues. a) The cellulose membrane was probed with streptavidin
(100 nm) and bound streptavidin was detected by Western blot analysis. b) Densitometry analysis of Figure 2 a. The intensity of each spot
was analyzed with a GS-700 imaging densitometer (Bio-Rad). The ordinate represents the reciprocal of the intensity of each analyzed spot
minus the reciprocal of the average intensity of all spots tested. The
large positive arbitrary unit values correspond to a weak blot signal,
and indicates a potential interaction of the peptides in a pair. The
dashed line represents the scatter-derived deviation 3n manifested by
the mean deviation from zero of the negative values n. c) The cellulose
was probed with streptavidin (100 nm) in the presence of biotin
(600 nm) and the bound streptavidin was detected by Western blot
analysis. Ispot = intensity of each analyzed spot; Imean = average intensity
of all spots tested.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
adjacent IANUS-positive spots were considered indicative of
a nativelike interaction in the respective peptide pair. Switching Stt II attachment to the Fmoc site and streptavidin
sequences to the Dde site on template 1 did not markedly
alter the spot array staining patterns (Supporting Information).
When peptides were synthesized on template 2 the
pattern of IANUS-positive spots was quite similar to those
obtained with template 1 (Figure 3, blue and pink under-
Figure 3. Binding of streptavidin to the streptavidin–Stt II peptide array
attached to template 2. The streptavidin sequence is collectively represented by 12-mer peptides that overlap with a shift of two amino acid
residues. a) The cellulose was probed with streptavidin (100 nm) and
the bound streptavidin was detected by Western blot analysis. b) Densitometry analysis; the intensity of each spot was analyzed as described in Figure 2 b.
lining, Table 1). However, the identification of three additional interacting protein segments (Figure 3, yellow underlining) is consistent with a model in which the conformational
flexibility of the template facilitates more nativelike interactions of peptide pairs. Apparently template 2 is more
appropriate for the complete identification of the architecture
of the streptavidin–Stt II binding site. The additional three
covalent bonds in the linker region of template 2 over that of
template 1 enables a majority of streptavidin-derived peptides attached to template 2 to be placed in register with the
resident Stt II peptide.
Two additional IANUS-positive spots (spots 41 and 44 in
Figure 2 a, green underlining) indicate conformational induction but are isolated, and therefore do not fit the neighboring
spot rule mentioned above. We hypothesize that a continuous
stretch of four IANUS-positive spots may exist (spots 41–44).
The reason behind the intense staining of spots 42 and 43
must therefore be explained. IANUS spot array assays
conducted in the presence of the streptavidin ligand biotin
revealed that the staining of spots 42 and 43 was a result of an
interaction between soluble streptavidin and the streptavidinderived peptide. This indicates that these spots bear a
sequence region directly involved in the formation of
streptavidin tetramers. The oligomerization interface is
centered around H127,[7] and the peptides of spot 42 and 43
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
correspond to this region. Consequently, streptavidin retains
binding affinity for both spots despite blockage of the active
site with biotin. With the exception of spots 42/43, and 45/46,
all other spots were unstained in the presence of biotin
(Figure 2 c). Consequently, the readout by Western blot
analysis of the IANUS spot array bears some limitations for
homooligomeric proteins, in that the intense signal for spots
containing peptides of the oligomerization interface prevents
the detection of some IANUS-positive spots.
Application of the potential of mean force (PMF) can
generate a complete description of the binding interface of
protein–ligand complexes in which direct hydrogen bonding,
water bridging, and hydrophobic interactions are thought to
make up the binding energy.[8] In this model, a distance of
5.0 between the atoms of interacting molecules is assumed
to comprise all attractive protein–ligand interactions, thus
defining the binding interface. On the basis of streptavidin
residues that lie within a 5.0- radius around the Stt II atoms
(Table 1, residues in boldface), the three-dimensional binding
interface of the streptavidin–Stt II complex was identified
from X-ray crystallographic structure information (PDB
entry: 1RSU).[5a] Notably, the two N-terminal residues of
Stt II were not localized in the crystal structure, and as a
result, the unequivocal identification of their contact sites was
not possible.
With the exception of the Y43–E44 segment, the streptavidin-derived peptides that correspond to the IANUS-positive spots of both templates, encompass the complete binding
interface derived from the X-ray crystallographic structure.
The Y43–E44 stretch, which is close to the N terminus of the
IANUS-positive spot 23 (Table 1), was identified to reside
within the 5.0- radius of Q7 of Stt II, but did not produce an
IANUS-positive spot. Most importantly, the IANUS array
did not generate false-positive regions associated with the
detection of noninteracting sequences.
Critical to the appearance of some IANUS-positive spots
was the sole presence of either D128, W79, or R84 in the peptide
pair (Table 1, spots 41–44, 25–27, 35–42, 56–59). It appears
that for these amino acids, only a limited structural context
was required for the intramolecular interaction with Stt II. In
the case of D128 and R84, exceptionally strong hydrogenbonding contacts to Stt II have already been noted.[5a] The
appearance of other IANUS-positive spots was consistent
with multiple interaction sites of the peptide pairs, in which
the sequence context dominates (Table 1, spots 17–18, 25–27,
23–25, 10–12, 51–52). In these cases, inspection of the threedimensional structure of the streptavidin–Stt II complex
revealed a modular arrangement of interacting residues in
which both the position of the amino acid and peptide
composition are necessary for an IANUS-positive spot.
Single amino acid substitutions on IANUS-positive interacting pairs provided a clear indication of the specificity of the
IANUS method (Supporting Information). Positions 53 (R)
and 54 (Y) were individually substituted with every other
gene-coded amino acid in the streptavidin peptide segment
NAESRYVLTGRY60. These peptide preparations were subjected to the IANUS spot assay, and the paired peptide Stt II
was held constant. Indeed, the IANUS spot array proved
sensitive to single-point substitutions. Substitutions at posi-
Angew. Chem. Int. Ed. 2005, 44, 1408 –1412
Table 1: Sequences of peptides from IANUS-positive spots and streptavidin–Stt II contact regions derived from X-ray crystallographic data.[a]
Spot number
template 1
template 2
[a] The amino acid residues identified in X-ray crystallographic data to
reside in a sphere of 5.0- radius about the Stt II atoms in the
streptavidin–Stt II complex are printed in bold face (PDB entry: 1RSU).
They were considered to represent residues involved in attractive
protein–ligand interactions. [b] Assuming a continuous stretch of four
IANUS-positive spots (see text for details). [c] Limited information is
available on the contact sites for these segments of streptavidin. This
results from poorly defined atomic locations for either the streptavidin
sequence S136–Q159, or the S1 and N2 residues of Stt II in the X-ray
crystallographic data.
tion 53 led, in most cases, to IANUS-negative spots. Including
the wild-type streptavidin peptide sequence (R53), IANUSpositive spots were obtained with only K, Q, Y, and V.
Substitutions at position 54, on the other hand, gave IANUSpositive results with almost all amino acid residues except E,
V, and N.
The substitutional analysis of the Stt II peptide in
positions 5 (H) and 9 (E) paired with a constant streptavidin
peptide sequence 75VAWKNNYRNAHS88 was more difficult to
perform, as Stt II is not only the interacting partner in the
IANUS experiment, but also provides the probe for detecting
IANUS-positive spots by streptavidin binding. Additional
experiments showed that most Stt II variants are not able to
interact with native streptavidin. In fact, an IANUS-positive
spot was observed in the case of a conservatively substituted
Stt II E9 variant that allowed streptavidin binding.
Table 1 lists the discontinuous segments of streptavidin
that form the architecture of the streptavidin–Stt II binding
surface, as determined by the IANUS spot array. The need for
an extended sequence context in the IANUS-positive peptide
pairs is variable. Modified IANUS peptide pairs were
evaluated to determine the influence of reverse reading of
the streptavidin sequence, starting at the C-terminal end.
IANUS-positive spots were obtained with reverse streptavidin peptides containing the D128, W79, and R84 residues. Given
the very flexible linker regions, conformations with normal
orientation may be frequently populated in the reversedpeptide arrangement (Supporting Information).
The role of template-assisted chain orientation was also
investigated in two related approaches: the template assembled synthetic proteins (TASP) approach,[9] and the modular
assembly of amphiphilic helical peptides on a cyclic-peptide
template for the construction of functional four-helix-bundle
heme proteins.[10] In a similar method, the identification of the
Elk-1 docking domain in assisting Elk-1 phosphorylation at
Ser383 was based on random distribution of two peptide
Angew. Chem. Int. Ed. 2005, 44, 1408 –1412
chains synthesized on monofunctional linkers.[11] In all cases,
nativelike assembly indicates the loss of conformational
multiplicity of the component peptide chains in favor of
freezing out interactive conformations.
The readout described herein requires a sample of
biologically active streptavidin. This practical restriction on
IANUS peptide arrays likely mandates the need for an
alternative detection system that allows the direct observation
of peptide–peptide interactions on a solid support. Examination of the streptavidin–Stt II IANUS peptide pairs involving
N-terminally dansyl-labeled streptavidin peptides and Nterminally fluorescein-labeled Stt II peptide showed that
IANUS-positive spots have a high fluorescence intensity at
510–530 nm. It is likely that efficient quenching of fluorescein
emission is correlated to conformational flexibility of noninteracting peptide chains, and therefore occurs in IANUSnegative spots. In the presence of biotin, the fluorescence of
the IANUS-positive spots is quenched to that of IANUSnegative spots. This observation indicates competition
between biotin and Stt II for the interactive conformation
of the streptavidin peptide of the IANUS pair (Supporting
Information). The evidence presented herein suggests that
the biotin ligand interferes directly with mutual conformational induction by abolishing chain togetherness with
consequent alterations in the fluorescence signal. This result
confirms the idea that use of a protein-free detection method
for IANUS-positive spots, which would then create a powerful screen for active-site directed small-molecule ligands and
inhibitors, is a promising new option for the IANUS peptide
In conclusion, a novel strategy that focuses on the
induction of complementary molecular surfaces of peptide
pairs has identified a complete low-resolution picture of the
streptavidin–Stt II contact interface. This picture closely
resembles that obtained by X-ray crystallography. IANUS
peptide arrays that circumvent the need for a protein-based
readout yield a low-resolution picture of protein–protein
interactions on the sole basis of sequence information, thus
obviating the need for native proteins in the detection step.
Experimental Section
Peptide Synthesis: Template 1 was prepared by the stepwise coupling
of Fmoc-Lys(Dde)-OH and Boc-Lys(Fmoc)-OH to the (b-Ala)2
spacer on the membrane. The amino acids were each activated with
PyBOP coupling reagent (1 equiv; PyBOP = 1-benzotriazolyloxytris(pyrrolidino)phosphonium hexafluorophosphate) and DIEA
(2 equiv; DIEA = diisopropylethylamine) as base in DMF (DMF =
dimethylformamide). Template 2 was prepared by solid-phase synthesis on the Rink amide MBHA resin (MBHA = 4-methylbenzhydrylamine), followed by cleavage of Fmoc-Glu-Lys(Dde)-CONH2
from the resin by using TFA (trifluoroacetic acid). The Fmoc-GluLys(Dde)-CONH2 sequence was anchored to the (b-Ala)2 spacer by
the glutamic acid g-carboxyl group with the coupling procedure used
for template 1. After completing the synthesis of template 2 by
coupling of Boc-Lys(Fmoc)-OH, the free amino and hydroxy
positions were acetylated with Ac2O (5 %) and DIEA (2 %) in
DMF for 1 h. Peptide chains were synthesized under the standard
spot synthesis protocols.[3a] The Dde protecting group was removed
with hydrazine solution (2 %) in DMF, in preparation for the next
peptide chain to be synthesized.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
MALDI-TOF MS: Pieces of the peptide spots (4 mm2 ; spots 22,
29, 39, and 40; Figure 3) were irradiated with UV light for 2 min. A
solution of a-cyano-4-hydroxycinnamic acid in CH3CN/H2O (0.1 %
TFA) was then added directly onto the pieces, which were dried at
room temperature. MALDI-TOF MS was carried out with the dried
pieces on a Bruker Reflex II mass spectrometer (Bruker Daltonik
GmbH, Bremen, Germany).
Western blots: Dry cellulose membranes were rinsed in methanol
for 10 min and for 3 20 min in TBS buffer (Tris-HCl pH 7.6 (30 mm),
NaCl (170 mm), and KCl (6.4 mm)). Streptavidin (100 nm) in MP
buffer (TBS buffer supplemented with Tween (0.05 %), and sucrose
(20.5 %)) was allowed to react with wet membranes overnight at 4 8C
under gentle shaking. Unbound streptavidin was removed by washing
with TBS buffer (4 8C), and spot-bound protein was electrotransferred onto nitrocellulose membranes (0.45 mm, PALL Gelman, Germany) with a semi-dry blotter (Biometra, Germany). The nitrocellulose membranes were sandwiched between blotting paper
soaked with transfer buffer (Tris-HCl pH 8.3 (25 mm), glycine
(150 mm), and methanol (10 %)) kept at 4 8C. Electrotransfer was
performed at a constant power of 0.8 mA cm 2 with suitable time
courses (first electrotransfer step for 30–45 min, second electrotransfer step for 60–90 min). Transferred streptavidin was detected
with rabbit anti-streptavidin antibodies (Sigma–Aldrich) and peroxidase-conjugated anti-rabbit IgG. The final visualization was performed by using an enhanced chemiluminescence ECL system.
Densitometric analysis was performed with a GS-700 imaging
densitometer (Bio-Rad).[12]
[9] a) M. Mutter, P. Dumy, P. Garrouste, C. Lehmann, M. Mathieu,
C. Peggion, S. Peluso, A. Razaname, G. Tuchscherer, Angew.
Chem. 1996, 108, 1588 – 1591; Angew. Chem. Int. Ed. Engl. 1996,
35, 1482 – 1485; b) S. Peluso, P. Dumy, C. Nkubana, Y. Yokokawa, M. Mutter, J. Org. Chem. 1999, 64, 7114 – 7120.
[10] H. K. Rau, N. DeJonge, W. Haehnel, Proc. Natl. Acad. Sci. USA
1998, 95, 11 526 – 11 531.
[11] X. Espanel, S. Walchli, T. Ruckle, A. Harrenga, M. HugueninReggiani, R. H. van Huijsduijnen, J. Biol. Chem. 2003, 278,
15 162 – 15 167.
[12] S. Rdiger, B. Bukau, Biospektrum 1998, 4, 35 – 37.
Received: June 16, 2004
Revised: October 4, 2004
Published online: January 21, 2005
Keywords: combinatorial chemistry · conformational induction ·
peptides · spot synthesis
[1] a) P. F. Cook, J. S. Blanchard, W. W. Cleland, Biochemistry 1988,
27, 4853 – 4858; b) Y. Chen, D. Xu, Curr. Protein Pept. Sci. 2003,
4, 159 – 181.
[2] a) S. W. Homans, Angew. Chem. 2004, 116, 292 – 303; Angew.
Chem. Int. Ed. 2004, 43, 290 – 300; b) Z. Dauter, V. S. Lamzin,
K. S. Wilson, Curr. Opin. Struct. Biol. 1997, 7, 681 – 688; c) S. K.
Avrantinis, R. L. Stafford, X. Tian, G. A. Weiss, ChemBioChem
2002, 3, 1229 – 1234.
[3] a) R. Frank, Tetrahedron 1992, 48, 9217 – 9232; b) M. Lebl,
Biopolymers 1998, 47, 397 – 404; c) A. Kramer, U. Reineke, L.
Dong, B. Hoffman, U. Hoffmller, D. Winkler, R. VolkmerEngert, J. Schneider-Mergener, J. Pept. Res. 1999, 51, 319 – 327;
d) H. Wenschuh, R. Volkmer-Engert, M. Schmidt, M. Schulz, J.
Schneider-Mergener, U. Reineke, Biopolymers 2000, 55, 188 –
206; d) R. Frank, J. Immunol. Methods 2002, 267, 13 – 26.
[4] E. T. Kaiser, F. J. Kezdy, Science 1984, 223, 249 – 255.
[5] a) T. G. M. Schmidt, J. Koepke, R. Frank, A. Skerra, J. Mol. Biol.
1996, 255, 753 – 766; b) L. A. Klumb, V. Chu, P. S. Stayton,
Biochemistry 1998, 37, 7657 – 7663; c) I. P. Korndrfer, A.
Skerra, Protein Sci. 2002, 11, 883 – 893; d) S. Freitag, I. Le Trong,
L. Klumb, P. S. Stayton, R. E. Stenkamp, Protein Sci. 1997, 6,
1157 – 1166; e) A. Chilkoti, P. H. Tan, P. S. Stayton, Proc. Natl.
Acad. Sci. USA 1995, 92, 1754 – 1758; f) S. Freitag, I. Le Trong,
A. Chilkoti, L. A. Klumb, P. S. Stayton, R. E. Stenkamp, J. Mol.
Biol. 1998, 279, 211 – 221.
[6] F. Guillier, D. Orain, M. Bradley, Chem. Rev. 2000, 100, 2091 –
[7] G. O. Reznik, S. Vajda, C. L. Smith, C. R. Cantor, T. Sano, Nat.
Biotechnol. 1996, 14, 1007 – 1011.
[8] L. Jiang, Y. Gao, F. Mao, Z. Liu, L. Lai, Proteins Struct. Funct.
Genet. 2002, 46, 190 – 196.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 1408 –1412
Без категории
Размер файла
174 Кб
architecture, site, assisted, interactions, intramolecular, proteinцligand, revealed, template, binding, peptideцpeptide
Пожаловаться на содержимое документа