14-3-3 protein C-terminal stretch occupies ligand binding groove and is displaced by phosphopeptide binding.

14-3-3 proteins are important regulators of numerous cellular signaling circuits. They bind to phosphorylated protein ligands and regulate their functions by a number of different mechanisms. The C-terminal part of the 14-3-3 protein is known to be involved in the regulation of 14-3-3 binding properties. The structure of this region is unknown; however, a possible location of the C-terminal stretch within the ligand binding groove of the 14-3-3 protein has been suggested. To fully understand the role of the C-terminal stretch in the regulation of the 14-3-3 protein binding properties, we investigated the physical location of the C-terminal stretch and its changes upon the ligand binding. For this purpose, we have used Forster resonance energy transfer (FRET) measurements and molecular dynamics simulation. FRET measurements between Trp242 located at the end of the C-terminal stretch and a dansyl group attached at two different cysteine residues (Cys25 or Cys189) indicated that in the absence of the ligand, the C-terminal stretch occupies the ligand binding groove of 14-3-3 protein. Our data also showed that phosphopeptide binding displaces the C-terminal stretch from the ligand binding groove. Intramolecular distances calculated from FRET measurements fit well with distances obtained from molecular dynamics simulation of full-length 14-3-3zeta protein.

The 14-3-3 protein family represents one of the most important group among proteins recognizing phosphorylated targets (1)(2)(3)(4). Two canonical 14-3-3 binding motifs have been defined, RSXpSXP and RX(Y/F)XpSXP (5,6), where pS denotes phosphoserine. Many of the 14-3-3 protein binding partners identified so far contain one of these motifs. Through the functional modulation of a wide range of binding partners, 14-3-3 proteins are involved in many biologically important processes, including cell cycle regulation, metabolism control, apoptosis, and control of gene transcription (1)(2)(3)(4). All 14-3-3 proteins form very stable homo-and heterodimers with characteristic cup-like shape and a large, 40-Å-wide, deep central channel (5)(6)(7)(8)(9)(10). Each monomer consists of nine antiparallel ␣-helices and an amphipathic groove where the phosphorylated segment of the binding partner is bound. The primary structure of 14-3-3 proteins is highly conserved, with completely conserved residues forming either the dimer interface or the walls of the ligand binding groove. Maximal sequence diversity occurs within the flexible C-terminal stretch, which has been shown to be involved in the regulation of binding properties of 14-3-3 proteins (11)(12)(13)(14)(15)(16). In addition, the C-terminal stretch of two vertebrate 14-3-3 isoforms ( and ) contains a casein kinase I␣ phosphorylation site at position 232 (17). Phosphorylation of 14-3-3 proteins has been suggested to be an important regulatory mechanism of individual isoforms, and it has been shown that in vivo phosphorylation of the C-terminal phosphorylation site inhibits the interaction between 14-3-3 and Raf-1 kinase (17,18). The structure of the C terminus is unknown, because this region is disordered in all available 14-3-3 protein crystal structures (5-10). Liu et al. (7) suggest that, in the absence of the ligand, the 14-3-3 protein C-terminal stretch could occupy the ligand binding groove and thus has to be pushed away during the ligand binding. Moreover, Truong et al. (14) show that removal of the C terminus increases the binding affinity of 14-3-3 protein for several tested ligands and propose that the C terminus might function as an autoinhibitor by suppressing unspecific interactions between 14-3-3 protein and inappropriate ligands. In addition, we have recently shown that phosphopeptide binding changes the conformation and increases the flexibility of the 14-3-3 protein C-terminal stretch (16).
To fully understand the role of the C-terminal stretch in the regulation of 14-3-3 protein binding properties, we have attempted to provide evidence for the location of the missing C-terminal part of the 14-3-3 protein molecule. Förster resonance energy transfer (FRET) 1 measurements between Trp 242 (located within the C-terminal stretch) and a dansyl group (attached at two different cysteine residues) indicate that, in the absence of the ligand, the C-terminal stretch occupies the ligand binding groove of the 14-3-3 protein. Our data also showed that phosphopeptide binding displaces the C-terminal stretch from the ligand binding groove. Intramolecular distances calculated from FRET measurements fit well with distances obtained from molecular dynamics simulation of fulllength 14-3-3 protein.
* This work was supported by Grants 204/03/0714 and 309/02/1479 of the Grant Agency of the Czech Republic, by Grant B5011308 of the Grant Agency of the Czech Academy of Sciences, by Research Projects 1K03020, MSM 1131 00001, and 1132 00001 of the Ministry of Education, Youth, and Sports of the Czech Republic, and by Research Project AVOZ 5011922 from the Academy of Sciences of the Czech Republic. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Labeling of 14-3-3 Protein Mutants by 1,5-IAEDANS-Human 14-3-3 protein possesses three cysteine residues (Cys 25 , Cys 94 , and Cys 189 ). To prepare proteins suitable for FRET measurements, we constructed three mutants containing a single cysteine residue (either at position 25, 94, or 189, respectively) and a single Trp residue located at the end of the C-terminal stretch at position 242. Covalent modification of the 14-3-3 protein containing the single Trp 242 and single Cys residues either at position 25, 94, or 189, respectively, with 1,5-IAE-DANS was carried out as described previously (19). Briefly, the protein (50 -70 M) in 50 mM Tris (pH 7.5), 100 mM NaCl, and 1 mM EDTA and label were mixed at a molar ratio of 1:40 and incubated at 30°C for 2 h and then at 4°C overnight in the dark. The free unreacted label was removed by gel filtration in buffer containing 50 mM Tris (pH 7.5), 100 mM NaCl, and 1 mM EDTA. The incorporation stoichiometry was determined by the absorbance at 336 nm using an extinction coefficient of 5700 M Ϫ1 cm Ϫ1 (Molecular Probes, Eugene, OR).
Mass Spectrometry Analysis-MALDI-TOF mass spectrometry was used to check amino acid sequences of 14-3-3 protein constructs with mutated cysteines and verify the modification of cysteine residues by the AEDANS moiety. Samples were first separated by 12% SDS-PAGE, and the excised protein bands were digested directly in gel (20). The resulting peptide mixtures were extracted by 30% acetonitrile and 0.5% acetic acid and subjected to the MALDI-TOF mass spectrometer BI-FLEX (Bruker-Franzen, Bremen, Germany) equipped with a nitrogen laser (337 nm) and gridless delayed extraction ion source. Ion acceleration voltage was 19 kV, and the reflectron voltage was set to 20 kV. Spectra were calibrated externally using the monoisotopic [MϩH] ϩ ion of peptide standards angiotensin I (Sigma). A saturated solution of ␣-cyano-4-hydroxy-cinnamic acid in 30% MeCN 0.3% acetic acid was used as a MALDI matrix. A 1-l sample was loaded on the target, and the droplet was allowed to dry at ambient temperature, overlaid with 1 l of matrix solution, and then allowed to co-crystallize at ambient temperature. Negative and positive ion mass spectra of peptide maps were measured in the reflection mode to check amino acid sequences of tryptic peptides with mutated cysteines (Cys 25 or Cys 189 ) in used 14-3-3 protein mutants and to control the labeling of proteins with 1,5-IAEDANS. The detected peak in negative ion mass spectra having the mass of 896.4 (m/z) corresponds to peptide ACSLAK (590.3 m/z) with labeled Cys 189 by 1,5-IAEDANS, and similarly, the peak of 1351.6 (m/z) fits to peptide YDDMAACMK (1045.4 m/z) with labeled Cys 25 by 1,5-IAEDANS. The identity and structure of these labeled tryptic peptides were further corroborated by analysis of their post-source decay spectra (data not shown).
Fluorescence Peptide Binding Assay-Fluorescence energy transfer between AEDANS and fluorescein has been used to determine the dissociation constant of peptide binding to AEDANS-labeled 14-3-3 protein mutants (21). Fluorescence measurements were performed using a PerkinElmer LS50B spectrofluorometer. Peptides Raf-259 and pRaf-259 with a fluorescein group attached to the N terminus (PolyPeptide Laboratories, Prague, Czech Republic) were titrated into the cuvette containing 400 nM 14-3-3 protein mutants labeled by 1,5-IAE-DANS in 50 mM Tris (pH 7.5), 100 mM NaCl, and 1 mM EDTA buffer. After mixing, the samples were incubated for 1 min at room temperature, and then the dansyl fluorescence at 480 nm was recorded using an excitation wavelength of 336 nm. The relative change in the fluorescence signal, f B , was plotted as a function of peptide concentration and was fitted to Equation 1 (22) to determine the K D values, where K D is the equilibrium dissociation constant. Nonlinear data fitting was performed using the Origin 6.0 software package (Microcal Software Inc.).
Time-resolved Fluorescence Measurements-Fluorescence resonance energy transfer was observed between the single tryptophan residue Trp 242 and the AEDANS moiety covalently attached to Cys 25 or Cys 189 . Fluorescence intensity decays of Trp 242 were measured on an apparatus as described by Obsilova et al. (16). Fluorescence decays were acquired under "magic angle" conditions, where the measured intensity decay I(t) was independent of a rotational diffusion of the chromophore and provided unbiased information about lifetimes. The apparatus response function was done at the excitation wavelength measured with a diluted Ludox solution. The samples were placed in a thermostatic holder, and all experiments were performed at 22°C in buffer containing 50 mM Tris-HCl (pH 7.5), 100 mM NaCl, and 1 mM EDTA. 14-3-3 protein concentration was 20 M, and ligand (pRaf-259 or Raf-259 peptide) concentration was 50 M. Fluorescence data processing was performed as described previously using the singular value decomposition maximum entropy method (16,(23)(24)(25), and the mean lifetimes were calculated. The average efficiency of energy transfer E was calculated from the mean donor lifetime in the presence ( DA ) and absence of acceptor ( D ).
The average distance between the donor-acceptor pair R was calculated from Equation 3, where R 0 is the Förster critical distance (the distance at which the energy transfer occurs with 50% efficiency). R 0 is given by (26,27), where n is the refractive index of the medium, Q D is the emission quantum yield of the donor in the absence of the acceptor, and 2 is the orientation factor that accounts for relative orientation of the donor emission and acceptor absorption transition dipole. The spectral overlap integral J() of the donor fluorescence and acceptor absorption was calculated from Equation 5 (26,27), where F D () and ⑀ a () represent the fluorescence intensity of the donor and the molar extinction coefficient of the acceptor, respectively, at the wavelength expressed in centimeters. The orientation factor 2 was assumed to be equal to 2 ⁄3, which corresponds to randomly oriented dipole moments of donor-acceptor pairs. A value of 1.31 was used for the refractive index of the aqueous solution. Quantum yields of the donor, which is the Trp 242 in the single Trp mutant of the 14-3-3 protein, were calculated relative to the Q D of L-Trp in water (28). Molecular Modeling-Molecular dynamics (MD) simulations were performed according to the protocol published previously (16). The GROMACS version 3.1 (www.gromacs.org) molecular dynamics software package was used with the parameter set ffG43a1p (29). All MD trajectories were analyzed using GROMACS software. The cluster analysis was performed using the GROMOS software method (30). A model of the 14-3-3 protein with attached AEDANS moieties at residues Cys 25 and Cys 189 was built using a representative conformation of 14-3-3 protein (obtained using a cluster analysis of simulated trajectory with a root mean square deviation cutoff of 1 Å) and AEDANS coordinates obtained from the crystal structure of labeled RNase (Protein Data Bank accession code 1RAS) (31). The geometry of the resulting complex was optimized by energy minimization using the molecular modeling package Ghemical (32) and Desktop Molecular Modeler version 4.2 (Polyhedron Software).

Construction of 14-3-3 Protein Mutants for Förster Reso-
nance Energy Transfer-To provide evidence for the physical location of the C-terminal stretch using FRET, we have constructed three 14-3-3 protein mutants containing a single Trp residue within the C-terminal stretch at the position 242 (residues Trp 59 and Trp 228 were mutated to Phe) and a single Cys residue located at three different positions: Cys 25 (this mutant was named 14-3-3w242c25), Cys 94 (14-3-3w242c94), and Cys 189 (14-3-3w242c189). The Trp residue served as an energy transfer donor, and the Cys residues were selectively labeled with the extrinsic fluorophore 1,5-IAEDANS as energy acceptors. Labeling experiments revealed that only Cys 25 and Cys 189 can be completely modified by 1,5-IAEDANS. The stoichiometries of AEDANS/mol of these two 14-3-3 protein mutants were found to be ϳ1 (0.95 and 1.06, respectively). The third cysteine, Cys 94 , was modified partially by 1,5-IAEDANS (labeling stoichiometry was found to be only ϳ0.4) and therefore was not used for FRET measurements.
Modification of 14-3-3 protein mutants by 1,5-IAEDANS has been verified using MALDI-TOF mass spectrometry. A fluorescence binding assay was used to test the binding abilities of both 14-3-3 protein mutants (Fig. 1). These experiments revealed that both 14-3-3 protein mutants modified with 1,5-IAEDANS bind phosphorylated pRaf-259 peptide with comparable affinity as wild-type protein, which binds pRaf-259 peptide with K D value of 116 nM (5). No significant binding of unphosphorylated Raf-259 peptide within the used concentration range has been detected (data not shown).
Investigation of the C-terminal Stretch Location Using FRET Steady-state Fluorescence Measurements-FRET measurements between Trp 242 and AEDANS moiety attached to a cysteine residue of 14-3-3 protein mutants were used to investigate the physical location of the C-terminal stretch of the 14-3-3 protein. First, we tested the presence of the energy transfer between Trp 242 and AEDANS moieties using the steady-state fluorescence. Tryptophan fluorescence emission spectra of 14-3-3 protein mutants, unmodified and modified with 1,5-IAEDANS, are shown in Fig. 2. When excited at 295 nm, Trp 242 has a fluorescence emission maximum near 350 nm. In the presence of 1,5-IAEDANS, Trp 242 fluorescence intensities of both single cysteine-containing mutants were significantly reduced compared with unmodified proteins, indicating the presence of FRET. The sensitized fluorescence emission of AEDANS (induced by the nonradiative transfer of energy from Trp 242 ) was observed as a peak at 485 nm. In the case of the 14-3-3w242c25 mutant, the donor steady-state fluorescence was quenched by ϳ24% compared with the emission of unlabeled protein (Fig. 2A). Higher steady-state FRET efficiency was observed for the 14-3-3w242c189 mutant, with ϳ50% reduction of the donor fluorescence (Fig. 2C), indicating that AEDANS attached to Cys 189 is closer to Trp 242 than in the case of Cys 25 . Next, we tested the effect of the peptide binding on steady-state FRET efficiency between Trp 242 and AEDANS. The presence of the unphosphorylated peptide Raf-259 had no effect on steady-state FRET efficiency, irrespective of AEDANS position within the 14-3-3 protein molecule (data not shown).
On the other hand, the addition of the phosphorylated peptide pRaf-259, which contains optimal 14-3-3 protein binding motif and binds to 14-3-3 protein with high affinity, was found to affect steady-state FRET efficiency. The effect was dependent on the position of AEDANS. In the case of the 14-3-3w242c25 mutant, the pRaf-259 binding slightly reduced FRET efficiency (compare Fig. 2, A and B), whereas for the 14-3-3w242c189 mutant, the FRET remained practically unchanged (compare Fig. 2, C and D).
Investigation of the C-terminal Stretch Location Using FRET Time-resolved Fluorescence Measurements-To quantitatively investigate the physical location of the C-terminal stretch, we employed time-resolved intensity decays of Trp 242 to measure distances between residue Trp 242 and an AEDANS moiety attached at different positions within the 14-3-3 protein molecule. We chose the time-resolved fluorescence approach, because, unlike fluorescence intensities, lifetimes do not depend on the excitation intensity, excited sample volume, protein concentration, or photobleaching, etc. Consequently, time-resolved data are more reliable than and superior to fluorescence intensities for quantitative FRET measurements. To use FRET to measure distances between Trp 242 and AEDANS, the Förster critical distance R 0 of this donor-acceptor pair must be determined first. The R 0 is a distance between the donor and acceptor at which the energy transfer is 50% efficient. The R 0 was determined from Equations 4 and 5. The spectral overlap J() between Trp 242 and AEDANS (obtained by numerical integration of a product of an area-normalized emission spectrum of 14-3-3w242 protein containing a single tryptophan at position 242 and the absorption spectrum of AEDANS) was found to be 5.876 ϫ 10 Ϫ15 cm 3 M Ϫ1 . The quantum yield of Trp 242 fluorescence of 14-3-3 protein mutants at 22°C was determined to be 0.074 Ϯ 0.002 in the absence of the ligand, 0.071 Ϯ 0.002 in the presence of 25 M pRaf-259 peptide, and 0.073 Ϯ 0.003 in the presence of 25 M Raf-259 peptide, respectively. The quantum yields were determined relative to a standard solution of tryptophan in water (Q D ϭ 0.140) (28). This resulted in the calculated R 0 of 19.7 Å in both the absence of the ligand and the presence of unphosphorylated peptide Raf-259. In the presence of the phosphorylated peptide pRaf-259, the R 0 is 19.5 Å. These R 0 distances are somewhat lower compared with the value of 22 Å reported elsewhere (19), presumably because of the lower quantum yield of Trp 242 fluorescence.
Time-resolved fluorescence intensity decays were analyzed using a singular value decomposition maximum entropy method as described previously (16). The intensity decays of Trp 242 for all three 14-3-3 protein mutants can be adequately described by a lifetime distribution containing three lifetime components (data not shown). Mean excited-state lifetimes ( mean ) of Trp 242 in the absence and presence of acceptor and the efficiency of the energy transfer are presented in Table I. Upon labeling by 1,5-IAEDANS for both mutants, the mean of donor was reduced as a result of the energy transfer (Fig. 3). Values of the FRET efficiency were used to calculate the average distances between the donor and the acceptors. Distances between Trp 242 and AEDANS attached at Cys 25 and Cys 189 were calculated to be 30.4 Å and 23.8 Å, respectively. These distances strongly indicate that, in the absence of the ligand, the C-terminal stretch of 14-3-3 protein is physically located within the ligand binding groove (see "Molecular Modeling" for details).
Next, we tested the effect of phosphopeptide binding on the location of the C-terminal stretch. The addition of phosphorylated peptide pRaf-259 significantly decreased the FRET efficiency and thus increased the distance between Trp 242 and AEDANS attached to Cys 25 from 30.4 to 36 Å. On the other hand, the distance between Trp 242 and AEDANS-Cys 189 remained practically unchanged (Table I). These data were supported by observations obtained from the steady-state measurements and indicate that the binding of the phosphorylated peptide pRaf-259 induces structural rearrangement of the Cterminal stretch. This conformational change caused an increase in distance between Trp 242 and Cys 25 , whereas the distance between Trp 242 and Cys 189 remained unaltered. The presence of the unphosphorylated peptide Raf-259, which did not interact with the 14-3-3 protein had no effect on FRET efficiency as was documented in the case of the 14-3-3w242c25 mutant (Table I). These data were consistent with the steadystate fluorescence measurements where no effect of unphosphorylated Raf-259 peptide on FRET efficiency was found.
Molecular Modeling of Full-length 14-3-3 Protein-Molecular dynamics was used to investigate interactions between C-terminal stretch and the ligand binding groove of 14-3-3 protein and to confront intramolecular experimental distances obtained from FRET measurements. Recently, we have reported a 3-ns-long molecular dynamics simulation of fulllength 14-3-3 protein with the C-terminal stretch located within the ligand binding groove (16). To better sample the conformational space of the C-terminal stretch, the 3-ns-long MD simulation was extended to a trajectory of 20 ns. The time course of the root mean square deviations of backbone-heavy atoms during the production run shows that the 14-3-3 protein structure was stable during the simulation (Fig. 4A). Fig. 4B shows time evolution of distances between C␣ atoms of residue pairs 25O242 and 189O242. Average distances were calculated to be 23.9 Ϯ 1.1 Å for the C␣ 25 OC␣ 242 pair and 16.1 Ϯ 0.9 Å for the C␣ 189 OC␣ 242 pair. Representative conformation of the 14-3-3 protein has been calculated as an average conformation of the most populated cluster obtained using a cluster analysis of simulated trajectory with a root mean square deviation cutoff of 1 Å (30). For better comparison with proteins used in our FRET experiments, we built a model of full-length 14-3-3 protein with AEDANS moiety attached at Cys 25 and Cys 189 (Fig. 5A). This model revealed that distances between the dansyl group and the indole group of Trp 242 in the absence of the phosphopeptide are ϳ30.0 and 23.5 Å for AEDANS moiety attached at Cys 25 and Cys 189 , respectively (Fig. 5A). These values are in an excellent agreement with experimental distances provided by FRET measurements (Table I), suggesting that conformation with the C-terminal stretch located within the ligand binding groove is an appropriate model for the full-length 14-3-3 protein structure. Next we attempted to model the position of the C-terminal stretch with bound phosphopeptide. The crystal structure of the 14-3-3 protein with bound phosphopeptide containing the sequence around phosphoserine 259 from Raf-1 kinase (7) has been used to position the phosphopeptide into the ligand binding groove. To meet average distances obtained from FRET measurements (Table  I), the Trp 242 has to be located somewhere above the helix H9 as suggested in Fig. 5B. Only in this conformation, will the model fulfill the measured distances 23.5 and 36 Å. Taken together, our data indicate that the C-terminal stretch is physically located within the ligand binding groove of the 14-3-3 protein molecule and is displaced when phosphorylated peptide binds.
Possible contacts between the C-terminal stretch and the ligand binding groove of the 14-3-3 protein and their comparison with the crystal structures of the 14-3-3 protein⅐ phosphopeptide complex (6) are shown in Fig. 6. The main chain of the C-terminal stretch adopts an extended conformation. On the basis of our MD simulation the 11-residue sequence, DEAEAGEGGEN, from Asp 235 to Asn 245 , constitutes the region of the C-terminal stretch interacting with the ligand binding groove of the 14-3-3 protein.
The key interactions of the C-terminal stretch binding seem to be contacts between side chains of negatively charged residues Glu 241 and Glu 244 and several residues located within helices H3, H5, and H7 (Arg 127 , Tyr 128 , and Asn 173 ). Residues Arg 127 and Tyr 128 are known to participate in the coordination of the phosphate group of 14-3-3 protein binding partners, whereas the side chain of Asn 173 makes contact with the ligand main chain (Fig.  6B) (6). The second acidic residue within the C-terminal stretch, which is predicted to interact with the ligand binding groove of 14-3-3 protein is Glu 244 (Fig. 6A). Other possible contacts between the ligand binding groove of the 14-3-3 protein and the C-terminal stretch involve residues Asn 42 , Ser 214 , and Asn 224 .

DISCUSSION
Location of the C-terminal Stretch in the Absence of the Ligand-To fully understand the role of the C-terminal stretch in the regulation of 14-3-3 protein binding properties, we decided to investigate the localization of the C-terminal stretch and its changes upon the ligand binding. For this purpose, we used FRET measurements and MD simulations. FRET reports on distances in the range of ϳ10 -100 Å and its efficiency depends on the sixth power of the distance between the energy  donor and the energy acceptor (26,27). Therefore, FRET is a very sensitive "ruler" for measurements of distances on the molecular scale. To study the localization of the C-terminal stretch by FRET we created two 14-3-3 protein mutants containing a single energy donor Trp 242 and a single Cys residue labeled by energy acceptor 1,5-IAEDANS (Fig. 5). Average distances between Trp 242 and AEDANS bound to Cys 25 and Cys 189 obtained from FRET measurements (Table I) indicate that, in the absence of the ligand, the C-terminal stretch is located within the ligand binding groove of the 14-3-3 protein.
Molecular modeling showed that only in this conformation both measured distances are fulfilled and that any different conformation of the C-terminal stretch would increase either one of the measured distances or both. Molecular dynamics simulation of full-length 14-3-3 protein revealed that the C terminus stayed bound within the ligand binding groove during the whole simulation and adopted an extended conformation. If we consider the size of the indole ring of the Trp residue and the AEDANS moiety attached at the Cys residues, the calculated average distances between C␣ atoms of residues used in this study (Fig. 4B) fit well with experimental distances obtained from FRET measurements. MD simulation of full-length 14-3-3 protein also suggests the possible mode of the C-terminal stretch interaction with the ligand binding groove (Fig. 6A) and shows that it can share some similarities with the binding of the phosphorylated peptides (5,6). Crystal structures of several 14-3-3 protein complexes demonstrated that residues Lys 49 , Arg 56 , Arg 127 , and Tyr 128 are involved in the coordination of the phosphate group of 14-3-3 binding partners (Fig.  6B), whereas residues Asn 42 , Asn 173 , Glu 180 , and Asn 224 provide additional contacts between the 14-3-3 protein ligand binding groove and the binding partner (5-10). Our MD simulation suggests that negatively charged residues Glu 241 and Glu 244 mimic the negative charge of the phosphate group, and together with residues Ser 45 , Lys 49 , Arg 127 , Tyr 128 , and Asn 173 Representative conformation has been calculated as an average conformation of the most populated cluster obtained using a cluster analysis (30). This model reveals that distances between the dansyl group and the indole group of Trp 242 in the absence of the phosphopeptide are ϳ30.0 and 23.5 Å for AEDANS moiety attached at Cys 25 and Cys 189 , respectively. B, model of fulllength 14-3-3 protein with bound phosphopeptide (shown in blue). Crystal structure of 14-3-3 protein with bound phosphopeptide containing the sequence around phosphoserine 259 from the Raf-1 kinase (Protein Data Bank accession code 1A37) (7) has been used to place the phosphopeptide into the ligand binding groove. C, C terminus; N, N terminus.
FIG. 6. Binding of the C-terminal stretch and its comparison with the crystal structure of the 14-3-3 protein⅐phosphopeptide complex. A, contacts between C-terminal stretch and ligand binding groove of 14-3-3 protein. Representative conformation of the 14-3-3 protein has been calculated as an average conformation of the most populated cluster obtained using a cluster analysis of simulated trajectory with a root mean square deviation cutoff of 1 Å (30). B, the "mode 1" phosphopeptide (ARSHpSYPA) bound to 14-3-3 protein (6). located within helices H3, H5, and H7 could be responsible for the interaction between the C-terminal stretch and the ligand binding groove of the 14-3-3 molecule.
Location of the C-terminal Stretch in the Presence of the Ligand-Next, we tested whether C-terminal stretch changes its localization upon the binding of the phosphorylated peptide. In the presence of the phosphopeptide, one of the measured distances (Trp 242 -AEDANS-Cys 25 ) increased significantly, whereas the second distance (Trp 242 -AEDANS-Cys 189 ) remained practically unchanged. Residue Cys 25 is located in the middle of the helix H2 (Fig. 5A), which is tightly packed with helices H1 and H3. In addition, helix H1 interacts with helices H3 and H4 of the second monomer and thus forms a dimer interface (5-10). The 14-3-3 molecule is very rigid because of an extensive number of interactions between the ␣-helices (5-10). Comparison of 14-3-3 protein structures in the absence of bound ligand (7, 8) and 14-3-3 protein complexes with bound phosphopeptides (5)(6)(7)10) or serotonin N-acetyltransferase (9) revealed a negligible amount of movement of the ␣-helices. Therefore, any conformational change involving the helix H2, which can also (in theory) increase the average distance between Trp 242 and AEDANS attached at Cys 25 , is very unlikely. Because both labeled 14-3-3 protein mutants are able to bind phosphorylated peptide (Fig. 1), we interpret the increase in the average distance between Trp 242 and AEDANS attached at Cys 25 for ϳ6 Å as a displacement of the C-terminal stretch from the ligand binding groove induced by phosphopeptide binding. The second Cys residue used as a site of attachment of AE-DANS fluorophore, Cys 189 , is located at the beginning of the helix H8. Phosphopeptide binding had no effect on average distance between Trp 242 and AEDANS attached at Cys 189 (Table I). We suggest that the C-terminal stretch upon its displacement from the ligand binding groove by peptide binding adopts a new conformation just above the helix H9 (Fig. 5B). Only in this conformation, will our model fulfill the measured distances 23.5 and 36 Å obtained from FRET measurements.
In conclusion, our results indicate that, in the ligand-free form, the C-terminal stretch of 14-3-3 protein occupies the ligand binding groove. Binding of the phosphopeptide to the 14-3-3 protein displaces the C-terminal stretch from the ligand binding groove. In addition, molecular dynamics simulation suggests some similarities in the binding of the C-terminal stretch and phosphorylated peptides.