ISSN 0026-8933, Molecular Biology, 2009, Vol. 43, No. 3, pp. 444–452. © Pleiades Publishing, Inc., 2009. Original Russian Text © L.V. Gushchina, A.G. Gabdulkhakov, V.V. Filimonov, 2009, published in Molekulyarnaya Biologiya, 2009, Vol. 43, No. 3, pp. 483–491.

STRUCTURAL-FUNCTIONAL ANALYSIS OF BIOPOLYMERS AND THEIR COMPLEXES UDC 577.322

Design and Structural Thermodynamic Studies of the Chimeric Protein Derived from Spectrin SH3 Domain L. V. Gushchina, A. G. Gabdulkhakov, and V. V. Filimonov Institute of Protein Research, Russian Academy of Sciences, Pushchino, Moscow region, 142290 Russia; e-mail: [email protected] Received September 23, 2008 Accepted for publication November 7, 2008

Abstract—A number of the chimeric constructs with spectrin SH3 domain were designed for structural and thermodynamic studies of protein self-assembly and protein–ligand interactions. SH3 domains, components of many regulatory proteins, operate through weak interactions with proline-rich regions of polypeptide chains. The recombinant construct (WT-CIIA) studied in this work was constructed by linking the peptide ligand PPPVPPYSAG to the domain C-terminus via a long 12-residue linker to increase the affinity of this ligand for the spectrin domain, thereby ensuring a stable positioning of the polyproline helix to the conserved ligand-binding site in orientation II, which is regarded as untypical of the interaction between this domain and oligopeptides. A comparison of fluorescence spectra of the initial domain and the recombinant protein WT-CIIA suggests that the ligand sticks to the conservative binding site. However, analysis of the equilibrium urea-induced unfolding has demonstrated that this is an unstable contact, which leads to a two-stage unfolding of the chimeric protein. The protein WT-CIIA was crystallized; a set of X-ray diffraction data with a resolution of 1.75 Å was recorded from individual crystals. A preliminary analysis of these diffraction data has demonstrated that the crystals belong to space group P32 with the following unit cell parameters: a = b = 36.39, c = 112.17 Å, a = β = 90.0, and γ = 120.0. DOI: 10.1134/S0026893309030121 Key words: Src homology 3 (SH3) domain of protein kinase, α-spectrin SH3 domain (Spc-SH3), proline-rich peptide, protein–ligand interactions, X-ray analysis, structure stability

The up-to-date molecular genetic methods are widely applied to both the studies of protein selfassembly and the introduction of point amino acid substitutions into proteins, as well as the designing of more intricate artificial constructs, in particular, chimeric proteins. As the mechanisms providing the protein folding and protein–protein interactions are rather intricate, small readily soluble globular proteins, such as SH3 (Src homology 3) domain, are the best models for experimental studies. In an isolated from, these are β-structure proteins with orthogonally packed short βsheets, composed of 60–85 amino acid residues [1–3]. From a functional standpoint, SH3 domains are important components of regulatory proteins; it has been demonstrated that they are involved in the development of several pathologies, such as cancer, AIDS, leukemia, and osteoporosis [4]. This fact made SH3 domains the targets for designing various drugs [5, 6]. It is considered that SH3 domains function as regulators via a weak interaction (Kd ~ 5–100 µM) with proline-rich regions of their partner proteins; these regions are of 8–10 residues long and can be, in particular, interdomain linkers in multidomain proteins [4, 7–10].

In a complex with the domain, these peptide ligands acquire a polyproline II helix conformation [11], which can bind the conservative domain binding site in two opposite orientations, I and II. Each domain preferably binds the ligand in only one orientation. The hydrophobic ligand binding site of the domain comprises three shallow pockets confined between one tryptophan, two tyrosine, and one proline residues [10] (Fig. 1). Proline residues, separated by hydrophobic residues in the consensus PxxP motif, usually occupy the first two pockets. The third pocket, known as a specificity pocket, is formed by the side chains of RT and n-src loops of the domains [12, 13]. As a rule, aromatic residues, influencing the specificity of domain–ligand interaction, are located at its bottom. Analysis of the known spatial structures, such as Abl-SH3 in complex with the ligand in orientation I (PDB: 1ABO) and the SH3 domains of C-src and Sem-5 kinases in complex with the ligands in orientation II (PDB: 1QWE and 1SEM), has suggested that the preferable orientation of the ligand depends on the immediate neighborhood of the conserved tryptophan residue in the domain’s specificity pocket [14], and that the spectrin domain can bind the ligands in both

444

DESIGN AND STRUCTURAL THERMODYNAMIC STUDIES

445

Fig. 1. Ribbon diagram of the complex between PWT-SH3 (gray) with the ligand GPPPVPPYSA (black), the model was constructed with the help of the Swiss-Pdb Viewer by superposing the structures of PWT-SH3 and Grb2-SH3 in complex with the ligand (see text). The ligand is positioned in the binding site; its conserved residues are shown in black. The C-terminus of the ligand is located to the right. The linker is not shown.

orientations (although it has been considered so far that orientation I is preferable). The peptide libraries for spectrin domain have not been comprehensively screened; however, the decapeptide p41 (ASYPPVPPP), optimized for binding to the SH3 domain of Abl kinase [15] in orientation I, displays the highest affinity for it. The spectrin domain–oligopeptide dissociation constant is approximately 160 µM, which is almost by two orders of magnitude higher than the Kd for Abl–SH3 [13]. To test the possibility of binding between the palindrome of ligand p41 and the spectrin domain in orientation II, we constructed a chimeric protein, named WT-CIIA, carrying the decapeptide PPPVPPYSAG attached via a 12-residue linker to the C-terminus of the α-spectrin SH3 domain. According to preliminary assessments, the incorporation of the ligand into the protein amino acid sequence should elevate its affinity for the binding site, increase solubility of the partners (polyproline oligopeptides have a decreased watersolubility, hindering the studies of bimolecular reaction with a low binding constant), and simplify its structural NMR and X-ray analyses. MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

In this paper, we describe the results of designing, cloning, and crystallizing the protein WT-CIIA and the preliminary X-ray structure data for its crystals. In addition, we have obtained and analyzed the data on WT-CIIA fluorescence spectroscopy and urea denaturation. EXPERIMENTAL Reagents. We used Taq and Vent DNA polymerases; NcoI, HindIII, and BamHI restriction endonucleases and a mixture of deoxyribonucleotide triphosphates produced by SibEnzim (Russia); T4 DNA ligase (Fermentas, Lithuania); the Wizard Plus SV Minipreps DNA Purification System (Promega, United States); a QIAquick PCR Purification Kit and a QIAquick Gel Extraction Kit (Qiagen, United States); Crystal Screen and Crystal Screen II crystallization kits, siliconized glass slides, and Limbro plates (Hampton Research, United States) in the work. Protein design was performed with the help of the Swiss-Pdb Viewer (SPV) program and the known structures of the complexes with SH3 domains from the Protein Data Bank (PDB).

446

GUSHCHINA et al.

Vectors and strains of bacterial cultures. A multicopy vector for the expression of recombinant proteins, pBAT4 carries the lacIq gene, the T7 lac promoter and terminator, and a ribosome-binding site (g10L) [16]. E. coli strain XL1-Blue (Stratagene, United States) was used for the production of plasmid DNA and transformation for cloning [17]. E. coli strain BL21(DE3) (Novagen, United States), carrying a derivative of the λ phage genome with a T7 RNA polymerase gene under the control of a inducible lacUV5 promoter [18], was used for expressing the DNA encoding WT-CIIA protein cloned into the vector under the control of T7 phage promoter. Cloning of the gene encoding WT-CIIA protein. The amino acid sequence of chimeric protein is shown in the section Results and Discussion. The WT-CIIA gene was assembled in two stages. First stage. The fragment of the WT-CIIA protein homologous to the DNA sequence of the pseudowildtype protein (PWT-SH3) was produced by the standard polymerase chain reaction (PCR) [19]. The DNA molecule encoding PWT-SH3 protein was used as a template; the used primers—5'-CTAGCCATGGGTACTGGAAAAGAGCTTGTGCTAG-3' (forward) and 5'-CCCAAGCTTTTTCACATAGGCAGCTGGTAC-3' (reverse)—contained the sites for NcoI and HindIII restriction endonucleases. Second stage. Six overlapping synthetic oligonucleotides (Sintol, Russia), encoding the DNA sequences for linker and ligand, were annealed (see the underlined amino acid residues in the protein sequence). The oligonucleotides were constructed so that the correctly annealed products contained the 5'-ends protruding by four nucleotides, similar to the products of HindIII and BamHI hydrolysis products: (1) 5'-AGCTTGATCCGGCTCAATCTGCCTCCCGTGA-3', (2) 5'-pAAATCTTGGTGGCCCGCCGCCGGTCCCGCC-3', (3) 5'-pGTATTCCGCTGGTTAAG-3', (4) 5'-GATCCTTAACCAGCGGAATACGGCGGGACCG-3', (5) 5'-pGCGGCGGGCCACCAAGATTTTCACGGGAGG-3', and (6) 5'-pCAGATTGAGCCGGATCA-3'. Some primers (nos. 2, 3, 5, and 6) were also additionally phosphorylated from the 5'-ends to obtain full-sized, double-stranded DNA molecules after ligation of the annealed oligonucleotides. The six overlapping oligonucleotides were mixed in equimolar amounts, supplemented with MgCl2 to a final concentration of 2 mM, and denatured at 95°ë for 3 min; then the mixture was kept in an ice bath for 3 h. Ligation was performed in a volume of 10 µl. The reaction mix-

ture was supplemented with 1 µl of 10× buffer for T4 DNA ligase and 1 AU of T4 DNA ligase and incubated at 22°ë for 1 h; ligase was inactivated by heating at 65°ë for 10 min. After purification of all fragments (the PCR product and annealed ligand and linker DNA sequences) and ligation, the gene of WT-CIIA protein was cloned into the vector pBAT4 (EMBL, Germany) [16] at NcoI and BamHI sites. The resulting plasmid DNA of WT-CIIA protein was produced in E. coli strain XL1Blue. The gene primary structure was determined by sequencing (Institute of Protein Research) using the following primers: 5'-TAATACGACTCACTATAGGG-3' (forward) and 5'-GCTAGTTATTGCTCAGCGGT-3' (reverse). Expression of WT-CIIA protein. The gene cloned into the vector was expressed in the system of Studier et al. [18] in the strain BL21(DE3) grown in LB medium. E. coli BL21(DE3) competent cells were transformed according to a standard protocol [17]. The level of protein overproduction was estimated by 16.5% SDS-PAGE [20]. Gels were stained with Coomassie G-250. The purified protein preparations (20– 40 mg) were obtained as earlier described [21]. The protein was dialyzed against a weak HCl solution (pH 3.0), freeze-dried, and stored at –20°ë. Analytical gel filtration was performed in a FPLC system using a column (length, 30 cm; diameter, 1 cm; and elution rate, 0.4 ml/min) filled with a Superdex Peptide HR 10/30 (Amersham Pharmacia, United States) carrier. Tryptophan fluorescence. The fluorescence spectra of tryptophan residues were recorded in a Cary Eclips (United States) fluorescence spectrometer in standard quartz cuvettes with an optical path length of 1 cm, slit width of 2.5 nm, and protein concentration of 3 µM. The excitation wavelength was 293 nm. Emission spectra were recorded in the range of 300– 500 nm with a step of 0.2 nm. The center of mass of the spectra (λCM) was calculated as n

∑I λ j

λ CM =

j

j=1 -----------------, n

∑λ

(1)

j

j=1

where Ij is the fluorescence intensity at the wavelength λj (nm) [22]. Urea molar concentration was determined according to the refraction coefficient at a wavelength of 589 nm in an IFR-454B (Russia) refractometer [23]. The linear extrapolation method (LEM) was used to calculate the changes in free energy transitions depending on urea concentration [24]. The analysis of denaturation curves is detailed in Results and Discussion. MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

DESIGN AND STRUCTURAL THERMODYNAMIC STUDIES

Crystallization of WT-CIIA protein. The protein was crystallized by hanging drop vapor diffusion at 22°ë using the Crystal Screen and Crystal Screen II commercial kits. The freeze-dried protein was dissolved in the buffer containing 20 mM sodium acetate and 5 mM β-mercaptoethanol (pH 4.6) and dialyzed against the same buffer for 4 h. The WT-CIIA concentration in drops was 7.5–10 mg/ml; the volume of drops varied in the range of 3–5 µl; and the volumes of counter solutions varied from 500 to 700 µ]l. The best quality crystals with a size of 100 × 300 × 500 µm were produced using 2 M ammonium sulfate as a precipitant. Before freezing in liquid nitrogen, the crystals were placed into the cryosolution containing 20 mM sodium acetate buffer pH 4.6, 30% sodium malonate, and 2 M ammonium sulfate. Collection of diffraction data. The set of diffraction data was obtained from one crystal in the X12 (DESY, Hamburg, EMBL, Germany) synchrotron beamline at a wavelength of 1.072 Å in the resolution region of 27– 1.75 Å with a completeness of 97.5%. Statistical data of the collected set are listed in the table. RESULTS AND DISCUSSION Designing Chimeric Protein WT-CIIA The goal of the work was to obtain a single polypeptide chain able to fold in a compact tertiary structure imitating the interactions between SH3 domains and proline-rich peptides. Such linking of the ligand to the domain via a flexible loop has important advantages for experimental study: (1) the possibility to change the ligand specificity and affinity by the mutagenesis based on the structural design and (2) in an ideal situation, the binding constant must increase due to a gain in the entropy during the transition of bimolecular to monomolecular reaction. At the first stage, it was necessary to construct a model of the WT-CIIA protein. Using the SPV program, we analyzed the structures of the initial PWTSH3 domain (PDB: 1PWT) and the protein Grb2-SH3 in complex with the ligand (VPPPVPPRRR-NH2) in an orientation II (PDB: 1GBQ). Hereinafter, PWTSH3 (pseudowild type) means the recombinant variant of α-spectrin SH3 domain (Spc-SH3) where the N-terminal sequence MDETG is replaced with MGTG [27]. The program SPV makes it possible to align two homologous structures; consequently, we obtained the model where a foreign ligand in the necessary orientation was bound to the domain PWT-SH3. Note that such a graft did not lead to any pronounced steric collisions between the protein and the ligand. When analyzing the structures of other complexes with ligands in orientation II, for example, C-Src-SH3 (PDB: 1QWE), we found out that these ligands, as a rule, had the sequences PPP(V/L/I)PP; in addition, the aliphatic MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

447

Main diffraction characteristics of WT-CIIA protein Symmetry group Cell parameters Å, °C Radiation source Wavelength, Å1 Resolution, Å1 Total number of reflexes Number of unique reflexes Completeness of data set1, % I/σi(I)1 Rsym(I), %1 1

P32 a = b = 36.390, c = 112.17 α = β = 90.00, γ = 120.00 X12, DESY, Hamburg 1.072 27.47–1.75 (1.80–1.75)1 59244 16339 97.6 (93.7)1 20.05 (3.28)1 3.8 (38.5)1

The values for high resolution layer are parenthesized Rsym =

∑ ∑ I i ( hkl ) – 〈 I ( hkl )〉 ∑ ∑ I i ( hkl ) .

hkl i

hkl i

hydrophobic residue docked into one of the conserved pockets of the domain with a proline residue at its bottom (Pro53 in PWT-SH3). As valine is the smallest residue of the three possible, we decided to preserve this particular residue in the sequence of our ligand. Then the SPV subprogram providing for introducing point substitutions into the structure was used to replace the sequence of Grb2-SH3 ligand (VPPPVPPRRR) with the sequence GPPPVPPYSA, a quasipalindrome of ligand p41. Then some torsion angles of side residues of the ligand were optimized. The resulting constructed complex is shown in Fig. 1. As is noted above, the interaction between SH3 domains and ligands is not confined to only a conserved polyproline fragment and, as a rule, is supplemented with the binding of the other residues responsible for interaction specificity. In peptide p41, which binds to Abl-SH3 in an orientation I, such interaction involves the N-terminal sequence APSYS, and tyrosine residue plays a key role in the increase in ligand’s affinity for the domain [25]. Hoping that the inverted sequence SYSPA would also efficiently interact with the Spc-SH3 domain in orientation II, we decided to attach this sequence to the C-end of our ligand with a minor modification. Then, we selected the site for attaching the ligand as well as the length and composition of the linker. Theoretically, the linker had to be sufficiently hydrophilic and charged to enhance a good solubility of the chimeric protein and its length should be sufficient to fit the ligand to the binding site without a steric stress. When analyzing the α-spectrin primary structure, we found that the fragment flanking SH3 domain from the right, presumably, served as a natural linker between SH3 and the neighboring domains and, consequently, had to be sufficiently flexible yet resistant to proteases. Consequently, we decided to add two glycine residues to the natural fragment composed of 10 resi-

448

GUSHCHINA et al. (a)

2

(b)

n

(c)

Fig. 2. Models of possible conformational transitions of the chimeric protein WT-CIIA. The initial domain is shown as oval with a relief ligand binding site. The N-terminus of the domain is shown as a circle, the linker is shown as heavy broken line, and the ligand is colored gray. (a) Scheme of an equilibrium formation of an intertwined dimer, (b) scheme of a possible head-to-tail polymerization, and (c) scheme of a two-stage unfolding of chimeric protein induced by urea (see Eq. (2)).

dues to provide the optimal chain length and flexibility; thus, the linker sequence was PAQSASRENLGG and the complete amino acid sequence of chimeric protein, MGTGKELVLA10LYDYQEKSPR20EVTMKKGDI L30TLLNSTNKDW40WKVEVNDRQG50FVPAAYVK KL60DPAQSASREN70LGG-PPPVPPY80SAG The linker is italicized and ligand sequence, underlined. The Control for the Absence of Protein Oligomerization Note that a long linker can create a problem of oligomerization of the chimeric construct, illustrated in Fig. 2. Schemes 2a and 2b can occur at large enough protein concentrations and the presence of additional interactions that stabilize oligomeric structures. Examples of intertwined dimers are known for several proteins, including SH3 domains, even in the absence of sticky ligands [26]. The formation of long polymers is also quite likely, as we have observed such polymerization for another chimeric protein (NII-CP) con-

structed by attaching the same ligand to the N-terminus of a circular permutant (S19P20s-SH3) of the spectrin domain [27] via a short linker (unpublished data). Therefore, the control for the absence of oligomerization was a necessary condition for all subsequent experiments and, especially, for crystallization experiments. Consequently, in addition to SDSPAGE, the protein purity and degree of oligomerization was also controlled by analytical gel filtration under native conditions in an FPLC system. The gel filtration data (Fig. 3) suggest that the WT-CIIA protein in solution was a monomer at least to protein concentrations of approximately 0.1–0.2 mM, Urea Denaturation of WT-CIIA Protein To determine whether the ligand attached to the binding site within a monomer, we recorded the protein fluorescence spectra and conducted the experiments on protein unfolding induced by urea. We selected the tryptophan fluorescence rather that the scanning microcalorimetry for several reasons. First, one tryptophan residue is located directly on the ligand binding surface, and its spectrum must change MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

DESIGN AND STRUCTURAL THERMODYNAMIC STUDIES 0.16

OD 280 nm

0.14 0.12

OD 280 nm

0.10

449

0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0

0.08

0

5

10

15

20

25

V, ml

0.06 0.04 VT

V0 0.02 0 –0.02

0

5

10

15

20

25

V, ml

Fig. 3. Gel filtration profiles (FPLC system; carrier, Superdex Peptide HR 10/30) of the isolated proteins WT-CIIA (9 kDa) and PWT-SH3 (7 kDa); V0 is free volume and VT, total volume of Superdex Peptide HR 10/30. Inset: profile of the mixture of WT-CIIA and PWT-SH3.

when it contacts the ligand. Second, fluorescence assay is highly sensitive, allowing oligomerization to be avoided at low protein concentrations. Third, it appeared that the WT-CIIA thermal denaturation at neutral pH values was accompanied by the aggregation of unfolded chains, whereas the polypeptide chains were hydrolyzed at the Asp–Pro bond directly in a calorimetric cell in the acid pH range. The results of the urea-induced equilibrium unfolding of the initial PWT-SH3 and chimeric WTCIIA proteins in neutral and acid pH ranges are shown in Fig. 4. Note that the overall spectrum of native WTCIIA is shifted to the shortwave region as compared with the spectrum of the initial domain, thereby suggesting a decrease in the accessibility of tryptophan residues to solvent, presumably, due to their shielding by the ligand. The center of mass of the spectrum (λCM) for the initial protein is at 352 nm versus the region of 349 nm for WT-CIIA; compare also with the λCM at approximately 348 nm for the chimeric protein NII-CP in the absence of urea. As the structure of NIICP has been already determined by NMR (Gushchina et al., in press) and its analysis demonstrates that the ligand is stably bound to the binding site and actually tightly contacts one of the tryptophan residues, we can postulate an analogous contact in the chimeric protein WT-CIIA. MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

It is evident from the curves of urea-induced protein unfolding that the initial region of the curve for WT-CIIA at pH 7.0 displays an abnormally high inclination. It looks like this wide curve reflects two successive transitions rather than a single-stage transition; the first one corresponds to the dissociation of the ligand from the binding surface, as is shown in Fig. 2c. Based on this assumption, we attempted to apply a two-stage equilibrium scheme to the analysis of the melting curves for the chimeric protein: N

kNI

I

kIU

U,

(2)

where kNI and kIU are the equilibrium constants of the corresponding transitions, where the dependences of free energy on urea are considered linear, as in the classic LEM model, usually applied for single-stage transitions: ∆ X G = – RT 0 ln k XY = ∆ X G W – m XY C ur , Y

0

Y

0

(3)

where X and Y are states, w corresponds to the energy change in water, mXY is the linearity coefficient, index 0 corresponds to a standard temperature of 25°C, and ‡ ëur is the molar concentration of urea. Then the statistical sum of the system can be put down as Q = 1 + kNI + kNI kIU,

(4)

450

GUSHCHINA et al.

Then the position of the center of mass of protein spectrum at each urea concentration is

(a)

1.0

cm = fNcmN + fIcmI + fUcmU.

Populations

0.8

Changing the fitting parameters aX, bX, mXY, and ∆ X G W (in all, ten), it is possible using the SigmaPlot program to achieve the best description of the experimental curve by function (9). Unfortunately, when fitting such a smooth curve, the involvement of ten independent parameters provides too much freedom and leads to ambiguous results. Therefore, it was necessary to introduce some reasonable constraints, which appeared sufficient for obtaining a reliable solution for the problem. In particular, based on the known parameters of transition in isolated initial SH3 domain, we fixed the position of base lines for states I and U and the base inclination for state N (five fixed parameters). In addition, we introduced stringent limU 0 itations for the values mIU and ∆ I G W for the second transition, describing the unfolding of the domain itself, whose structural stability had to be weakly dependent on the presence of a “free tail”. Indeed, the fitting of the melting curve for PWT-SH3 at pH 7.0 according to the LEM model (single-stage transition) U 0 gives mNU = 2.7 kJ mol–1 M–1 and ∆ N G W = 16.8 kJ mol–1, whereas the analysis of the chimeric protein melting curve according to a two-stage scheme gives the following values for the second transition parameters: U 0 mIU = 2.6 kJ mol–1 M–1 and ∆ I G W = 16.97 kJ mol–1. For the first transition, the following values were U 0 obtained: mNI = 1.1 kJ mol–1 M–1 and ∆ N G W = 1.5 kJ mol–1 (the errors in all cases amounted to 10%). An analysis of the results described above and in Fig. 4 demonstrates that (a) the affinity of ligand for the domain within WT-CIIA protein is low, and the completely folded (native) state (N) even in the absence of urea is populated only for 70%, or slightly more, taking into account the analysis error; (b) a decrease in pH destabilizes the WT-CIIA structure mainly by destabilizing the structure of the domain itself (state I), approximately similar to how it takes place with the initial protein; and (c) a strong decrease in the stability of state I of the chimeric protein with a decrease in pH to 3.5 prevents the populating of this state, i.e., the denaturation transition approaches the model of two states. Y

0.6 0.4 0.2 0 (b)

365

λCM, nm

(9)

360

355

350 0

2

4 6 Curea, M

8

10

Fig. 4. Equilibrium unfolding curves for the chimeric and initial proteins in the presence of urea at 25°ë. (a) Dependences of the population of three states in model (2) during unfolding of the chimeric protein at pH 7.0 (solid line is the initial state N, dashed line is state I, and the dotted and dashed line is the unfolded state U). (b) Unfolding curves for PWT-SH3 (triangles) and WT-CIIA (circles); black symbols is pH 7.0, light circles is pH 3.5, and light triangles is pH 4.0; solid lines are best fit curves at pH 7.0 according to the singe-stage LEM model (triangles are the initial protein) and two-stage model (circles are the chimeric protein). The upper dashed curve shows the common base line for protein unfolded states; the lower dashed line corresponds to the changes in the center of mass of the spectrum for the first transition in the chimeric protein unfolding at pH 7.0; the dotted and dashed line corresponds to the changes in the center of mass of the spectrum for the second transition.

and the populations of individual states, as f N = 1/Q; f I = k NI /Q;

(5–7)

f I = k NI k IU /Q; The analysis of the single-stage melting curves for the initial domain demonstrates that the basic lines, i.e., the λCM dependences of each of the three states on urea concentration, are the linear functions cmX = aX + bXCur,

(8)

0

Note that according to our unpublished data, the chimeric protein NII-CP unfolds in a cooperative manner at all pH values under the action of both heat and urea; in this process, the binding of a ligand has a strong stabilizing effect on its structure. Presumably, these differences between the behaviors of two chimeric proteins are the result of different linker lengths: the longer the linker, the more unfavorable is the entropy contribution to the total free energy of the MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

DESIGN AND STRUCTURAL THERMODYNAMIC STUDIES

451

linker in order to increase the ligand affinity for the conserved binding site. We assumed that the ligand in such a construct would be stably bound to the domain in orientation II, untypical of the spectrin domain. The recombinant protein WT-CIIA has been isolated in a homogenous state and demonstrated to be a monomer in solution. The fluorescence spectra and the data on the urea denaturation of this protein suggest that the ligand is likely to bind to the conserved site; however, this contact is comparatively unstable. Crystals of this chimeric protein have been produced, and the set of diffraction data with a resolution of 1.75 Å has been collected. Fig. 5. Crystals of WT-CIIA protein with a size of 100 × 300 × 500 µm.

ligand binding and the smaller is its effect on the system stability. Crystallization of the Protein and Preliminary X-Ray Structure Data An unstable ligand binding to the binding site leads to a structural heterogeneity, thereby creating additional difficulties for both physicochemical studies and protein crystallization. In particular, despite the fact that the ligand is considerably more stably bound in the above mentioned protein NII-CP and its analog, chimeric protein Spc-p41 [28], neither we nor our Spanish colleagues succeeded in producing high quality crystals of these compact and stable proteins. Therefore, the structures of both chimeric proteins with short linkers were determined by NMR. The production of a pure homogeneous protein in a preparative amount allowed us to commence its crystallizing. For this purpose, we used Crystal Screen and Crystal Screen II commercial kits. The crystals appropriate for X-ray analysis were obtained only when using 2 M ammonium sulfate as a precipitant. Figure 5 shows the WT-CIIA crystals reflecting X-rays with a resolution of 1.75 Å. Using the data on symmetry group and cell parameters, we calculated the Matthews coefficient [29], which demonstrated that the asymmetric cell, with the highest likelihood for protein crystals, contained two to four protein molecules with the Matthews coefficients of 4.29–2.14 and solvent content of 71–42%, respectively. CONCLUSIONS Several chimeric proteins involving the SH3 domain of α-spectrin have been constructed to study the thermodynamics of protein–peptide interactions. In this work, we have studied the protein variant where the peptide ligand with a length of 10 residues is attached to the C-end of the domain via a 12-residue MOLECULAR BIOLOGY

Vol. 43

No. 3

2009

ACKNOWLEDGMENTS We are grateful to Prof. Serrano (Spain), who kindly provided the plasmid carrying the pseudowildtype gene of spectrin SH3 domain (PWT-SH3). The work was supported by the Presidium of the Russian Academy of Sciences under the program Molecular and Cell Biology; the Russian Foundation for Basic Research, project no. 03-04-48331; and INTAS, project no. 03-55-5569. REFERENCES 1. Mayer B.J., Hamaguchi M., Hanafusa H. 1988. A novel viral oncogene with structural similarity to phospholipase C. Nature. 332, 272–275. 2. Musacchio A., Gibson T., Lehto V.P., Saraste M. 1992. SH3: An abundant protein domain in search of a function. FEBS Lett. 307, 55–61. 3. Musacchio A., Noble M., Pauptit R., Wierenga R., Saraste M. 1992. Crystal structure of a Src-homology 3 (SH3) domain. Nature. 359, 851–855. 4. Brown M.T., Cooper J.A. 1996. Regulation, subsrates and function of src. Biochem. Biophys. Acta. 1287, 121– 149. 5. Vidal M., Gigoux V., Garbay C. 2001. SH2 and SH3 domains as targets for anti-proliferative agents. Crit. Rev. Oncol. Hematol. 40, 175–186. 6. Dalgarno D.C., Botfield M.C., Rickles R.J. 1997. SH3 domains and drug design: Ligands, structure, and biological function. Biopolymers. 43, 383–400. 7. Musacchio A., Wilmanns M., Saraste M. 1994. Structure and function of the SH3 domain. Prog. Biophys. Mol. Biol. 61, 283–297. 8. Musacchio A., Saraste M., Wilmanns M. 1994. Highresolution crystal structures of tyrosine kinase SH3 domains complexed with proline-rich peptides. Nature Struct. Biol. 8, 546–551. 9. Kuriyan J., Cowburn D. 1997. Modular peptide recognition domains in eukariotic signaling. Annu. Rev. Biophys. Biomol. Struct. 26, 259–288. 10. Kay B.K., Williamson M.P., Sudol M. 2000. The importance of being proline: The interaction of proline-rich motifs in signaling proteins with their cognate domains. FASEB J. 14, 231–241.

452

GUSHCHINA et al.

11. Macias M.J., Hyvonen M., Baraldi E., Schultz J., Sudol M., Saraste M., Oschkinat H. 2002. Structure of the WW domain of a kinase-associated protein complexed with a proline-rich peptide. Lett. Nature. 382, 646–649. 12. Mayer B.J. 2001. SH3 domains: Complexity in moderation. J. Cell. Sci.114, 1253–1263. 13. Martin-Sierra F.M., Candel A.M., Casares S., Filimonov V.V., Martinez J.C., Conejero-Lara F. 2003. A binding event converted into a folding event. FEBS Lett. 553, 328–332. 14. Fernandes-Ballester G., Blanes-Mira C., Serrano L. 2004. The tryptophan switch: Changing ligand-binding specificity from Type I to Type II in SH3 domains. J. Mol. Biol. 336, 619–629. 15. Pisabarro M.T., Serrano L., Wilmanns M. 1998. Crystal structure of the Abl-SH3 domain complexed with a designed high-affinity peptide ligand: Implications for SH3–ligand interactions. J. Mol. Biol. 281, 513–521. 16. Peranen J., Rikkonen M., Hyvonen M., Kaarliainen L. 1996. T7 vectors with a modified T7 lac promoter for expression of protein in Escherichia coli. Anal. Biochem. 236, 371–373. 17. Sambrook J., Fritsch E.F., Maniatis T. 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Lab. Press. 18. Studier F.W., Rosenberg A.H., Dunn J.J., Dubendorff J.W. 1990. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185, 60–89. 19. Higuchi R., Krummel B., Saiki R.K. 1988. A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res. 16, 7351–7367. 20. Schagger H., von Jagow G. 1987. Tricine–sodium dodecyl sulfate–polyacrylamide gel electrophoresis for the separation of proteins in the range from to 100 kDa. Anal. Biochem. 166, 368–379.

21. Viguera A.R., Martinez J.C., Filimonov V.V., Mateo P.L., Serrano L. 1994. Thermodynamic and kinetic analysis of the SH3 domain of spectrin shows a two-state folding transition. Biochemistry. 33, 2142–2150. 22. Wildes D., Meadowanderson L., Sabogal A., Marqusee S. 2006. Native state energetics of the Src SH2 domain: Evidence for a partially structured state in the denatured ensemble. Protein Sci. 15, 1769–1779. 23. Warren J.R., Gordon J.A. 1966. On refractive indices of aqueous solutions of urea. J. Phys. Chem. 70, 297. 24. Filimonov V.V., Azuaga A.I., Viguera A.R., Serrano L., Mateo P.L. 1999. A thermodynamic analysis of a family of small globular proteins: SH3 domains. Biophys. Chem. 77, 195–208. 25. Pisabarro M.T., Ortiz A.R., Viguera A.R., Gago F., Serrano L. 1994. Molecular modeling of the interaction of polyproline-based peptides with the Abl-SH3 domain: Rational modification of the interaction. Protein Eng. 7, 1455–1462. 26. Mongiovi A.M., Romano P.R., Panni S., Mendoza M., Wong W.T., Musacchio A., Cesareni G., Di Fiore P.P. 1999. A novel peptide–SH3 interaction. EMBO J. 19, 5300–5309. 27. Martinez J.C., Viguera A.R., Berisio R., Wilmanns M., Mateo P.L., Filimonov V.V., Serrano L. 1999. Thermodynamic analysis of alpha-spectrin SH3 and two of its circular permutants with different loop lengths: discerning the reasons for rapid folding in proteins. Biochemistry. 38, 549–559. 28. Candel A.M., Conejero-Lara F., Martinez J.C., Nico A.J., van Nuland N.A., Bruix M. 2007. The high-resolution NMR structure of a single-chain chimeric protein mimicking a SH3–peptide complex. FEBS Lett. 581, 687– 692. 29. Matthews B.W. 1968. Solvent content of protein crystals. J. Mol. Biol. 33, 491–497.

MOLECULAR BIOLOGY

Vol. 43

No. 3

2009