NMR Solution Structure of the Tandem Src Homology 3 Domains of p47phox Complexed with a p22phox-derived Proline-rich Peptide*

The phagocyte NADPH oxidase plays a crucial role in host defense against microbial infections by generating reactive oxygen species. It is a multisubunit enzyme composed of membrane-bound flavocytochromeb558 as well as cytosolic components, including p47phox, which is essential for assembly of the complex. When phagocytes are activated, the cytosolic components of the NADPH oxidase translocate to flavocytochrome b558 due to binding of the tandem Src homology 3 (SH3) domains of p47phox to a proline-rich region in p22phox, a subunit of flavocytochrome b558. Using NMR titration, we first identified the proline-rich region of p22phox that is essential for binding to the tandem SH3 domains of p47phox. We subsequently determined the solution structure of the p47phox tandem SH3 domains complexed with the proline-rich peptide of p22phox using NMR spectroscopy. In contrast to the intertwined dimer reported for the crystal state, the solution structure is a monomer. The central region of the p22phox peptide forms a polyproline type II helix that is sandwiched by the N- and C-terminal SH3 domains, as was observed in the crystal structure, whereas the C-terminal region of the peptide takes on a short α-helical conformation that provides an additional binding site with the N-terminal SH3 domain. Thus, the C-terminal α-helical region of the p22phox peptide increases the binding affinity for the tandem SH3 domains of p47phox more than 10-fold.

The p47 phox protein contains in order from the N-to the C termini, a Phox homology (PX) 2 domain, tandem SH3 domains, a polybasic or autoinhibitory region (PBR/AIR), and a proline-rich region (PRR). The membrane-bound protein p22 phox also contains a PRR in its C-terminal cytosolic region. In the resting state, the tandem SH3 domains of p47 phox are masked by an intramolecular interaction with the PBR/AIR, which maintains the enzyme in an inactive form (10 -12). Upon cell stimulation, some serine residues in the PBR/AIR domain of p47 phox are phosphorylated by a specific kinase, which induces a conformational change. This leads to changes in intramolecular interactions (13)(14)(15)(16) and exposure of the tandem SH3 domains, permitting their binding with the PRR of p22 phox (7,11,12). In this way, the p47 phox -p67 phox -p40 phox complex can be tethered to flavocytochrome b 558 . The p47 phox -p22 phox interaction is thought to be essential for activation of NADPH oxidase because p22 phox mutants incapable of this interaction cause severe chronic granulomatous disease in humans (17)(18)(19)(20).
We and another group (21,22) recently reported a crystal structure of the tandem SH3 domains in the autoinhibited form of p47 phox . The tandem SH3 domains form an intertwined dimer in which the distal loop of the N-terminal SH3 domain serves as a hinge. Both crystal structures suggested that the autoinhibited form is a globular module that is split in half by the hinge region. However, we found that p47 phox exists in solution not as a dimer but rather as a monomer that is similar in structure to this globular module (23). Thus, the globular module rather than the intertwined dimer is thought to be the physiologically relevant structure. The crystal structure of the activated form of the tandem SH3 domains was also reported (21), but it again showed an intertwined dimer. This prompted us to use small angle x-ray scattering (SAXS) and nuclear magnetic resonance (NMR) spectroscopy to examine the solution structure of the tandem SH3 domains complexed with the PRR of p22 phox as a model of the activated form of p47 phox . The results presented here should help clarify the mechanism by which NADPH oxidase is activated.

EXPERIMENTAL PROCEDURES
Sample Preparation-A truncated form of human p47 phox containing residues 151-286 (p47 phox -(151-286)), which corresponds to the tan-* This work was supported in part by grants-in-aid for scientific research and by the National Project on Protein Structural and Functional Analyses from the Ministry of Education, Culture, Sports, Science, and Technology of Japan. 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. The atomic coordinates and structure factors ( Small Angle X-ray Scattering-SAXS data were collected for p47 phox -(151-286) in the presence and absence of p22 phox -(149 -168). The concentration of the proteins was varied from 2 to 16 mg/ml to assess the effects of the protein concentration on the structural parameters. All measurements were made using a SAXS diffractometer in the BL-10C Beamline at the Photon Factory in Tukuba, Japan (25,26). The wavelength of the x-ray was 1.488 Å. We used a 50-l sample cell with quartz windows and a 1-mm path length. The data acquisition time was 300 s for each measurement. Scattering by the protein was determined by subtracting scattering for the solvent (buffer alone) from the scattering for the protein solution.
The scattering data were first analyzed using the Guinier approximation I(Q) ϭ I(0) exp(ϪR g 2 Q 2 /3), where Q is the momentum transfer, R g is the radius of gyration, and I(0) is the intensity at a zero scattering angle (27). Q was calculated from the following relationship: Q ϭ 4 sin/, where 2 and are the scattering angle and wavelength of the X-rays, respectively. I(0) and R g were calculated from the linear regression intercept and slope (ϪR g 2 /3), respectively, from Guinier plots in the range of Q ϫ R g Ͻ 1.8 (26 -28). The relative molecular weight of the scattering species was estimated using I(0), obtained from the scattering data for bovine carbonic anhydrase (Sigma-Aldrich), a monomeric protein with a molecular mass of 29 kDa. The distance distribution function, P(r), defined as P(r) ϭ 1/2 2 ͐I(Q)Qr sin(Qr) dQ, corresponds to the distribution of the distance, r, between the volume elements. P(r) was calculated by indirect Fourier transformation using the GNOM program (29). The Q values used for the P(r) analysis were between 0.03 and 0.19 Å Ϫ1 . The R g value, calculated from R g 2 ϭ ͐r 2 P(r) dr/(2͐P(r)dr), was estimated from the distance distribution function, P(r). D max was the maximum dimension. Structural parameters were derived using both Guinier analysis and the distance distribution function, P(r).
Fluorescence Titration-The binding affinity of p22 phox peptides for p47 phox -(151-286) was monitored by fluorescence spectroscopy using a Shimadzu RF-5300PC fluorometer (excitation at 295 nm and emission at 330 nm). Aliquots of synthetic p22 phox peptides were incubated for 2 h with 3.0 M p47 phox -(151-286) in 20 mM sodium phosphate buffer, pH 7.0. All measurements were performed at 25°C. Dissociation constants (K d ) were determined by nonlinear regression analysis using Kaleidagraph (Synergy Software) assuming a simple bimolecular association.
Residual Dipolar Couplings-Residual dipolar couplings (RDC) were collected for a solution containing 0.5 mM 15 N-labeled p47 phox -(151-286) complexed with unlabeled p22 phox -(149 -168) in a liquid crystalline medium consisting of 5% polyethylene glycol/hexanol (r ϭ 0.96) (37), 25 mM sodium phosphate, and 150 mM NaCl, pH 6.5. One-bond 1 H-15 N coupling constants were measured on a Unity Inova 800 spectrometer at 25°C using two-dimensional 1 H-15 N IPAP-HSQC spectra (38,39). The RDC values were determined by comparing the spectra of the sample in the isotropic medium with those in the weakly oriented medium, which ranged from Ϫ17 Hz to ϩ16 Hz. Analysis of the powder pattern of dipolar couplings revealed an axial component of the 1 D NH alignment tensor of Ϫ8.5 Hz and a rhombicity of 0.4. The dipolar coupling restraints were not included in structural calculations for residues having hetNOE values less than 0.7, suggesting an appreciable amidebond vector reorientation on a subnanosecond time scale.
Structure Calculation-The cross-peak intensities on the NOESY spectra were integrated and partially assigned using NMRDraw (36). The NOE data, along with a table of chemical shifts, were used as an input to ARIA 1.2 (40) implemented in CNS 1.1 (41). A total of 48 and 48 torsion angle restraints were predicted by the TALOS program (42). The 24 distance restraints for hydrogen bonds were incorporated based on the characteristic NOE patterns. The RDC restraints were used for each iteration of ARIA. The calculations were performed starting from random templates, using the standard parameters of ARIA. The restraints used for the structural calculations are summarized in Table 1. The structural calculations in combination with iterative NOE peak assignments were performed in nine cycles, and a total of 100 structures were finally obtained. The mean structure (referred as the final structure) was obtained by averaging the coordinates of the 20 lowest energy structures. The Ramachandran plot of the final structure was analyzed using the Procheck-NMR program (43), and the molecular figures were generated using PyMol software (www.pymol.org).
In Vitro Pull-down Binding Assays-Mutations in the cDNA fragments of p47 phox -(151-286) and p22 phox -(132-195) leading to the indicated amino acid substitutions were introduced by PCR-mediated, sitedirected mutagenesis. All of the constructs were sequenced for confirmation of their identities. The PCR products were ligated to the following vectors: pGEX-2T (Amersham Biosciences) for the generation of GST fusion proteins; or pMALc2 (New England Biolabs) for the generation of maltose-binding protein (MBP) fusion proteins. GST-or MBP-tagged proteins were expressed in E. coli strain DH5 and purified using glutathione-Sepharose 4B (Amersham Biosciences) or amylose resin (New England Biolabs), respectively, according to the manufacturers' protocols. For in vitro pull-down binding assays, a GST and an MBP fusion protein (5 g each) were mixed in PBS (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 , and 1.4 mM KH 2 PO 4 , pH 7.4) containing 1% Triton X-100 and incubated for 30 min at 4°C. To pull down the GST fusion protein, a slurry of glutathione-Sepharose 4B beads was added, and the mixture was incubated for 30 min at 4°C. After washing three times with PBS, the GST fusion protein and associated proteins were eluted from the glutathione-Sepharose 4B beads with 20 mM Tris-HCl, pH 8.0, containing 10 mM glutathione. The eluates were subjected to SDS-PAGE, after which the proteins were stained with Coomassie Brilliant Blue.

RESULTS AND DISCUSSION
Identification of the Region of p22 phox Essential for p47 phox Binding-Based on an in vitro overlay binding assay, the region containing residues 151-160 of p22 phox (p22 phox -(151-160)) was initially proposed as a core region essential for the binding of p47 phox (7). In contrast, Dahan et al. (10) showed that its C-terminal flanking region enhanced the interaction with p47 phox . Groemping et al. (21) used a peptide corresponding to residues 149 -166 of p22 phox (p22 phox -(149 -166)), which has a higher binding affinity than p22 phox -(151-160), in a crystallographic study of the activated form of p47 phox . The crystal structure indicated that the tandem SH3 domains form an intertwined dimer with p22 phox -(149 -166) Furthermore, because an electron density was observed only for residues 151-160 of p22 phox -(149 -166), the contribution of non-core regions to binding has remained elusive.
Therefore, in the current studies, we used NMR spectroscopy to examine the solution structure of the tandem SH3 domains of p47 phox complexed with the p22 phox peptide. To identify the region of p22 phox essential for p47 phox binding in solution, we performed NMR titration experiments using two synthetic peptides, p22 phox -(149 -162) and p22 phox -(149 -168). Aliquots of each peptide solution were added to the sample solution of the 15 N-labeled tandem SH3 domains of p47 phox , and the two-dimensional 1 H-15 N HSQC spectra were analyzed (Fig. 1). We found that many of the signals decreased and finally disappeared as the concentration of p22 phox -(149 -162) increased (Fig. 1a). Notably, three of five tryptophan signals disappeared, leaving only two observed signals (indicated by asterisks in Fig. 1a). This shows that a structural change occurs in the tandem SH3 domains upon complex formation. This produces an intermediate slow-exchange limit in the NMR spectra so that some signals cannot be observed because of exchange broadening. In contrast, an increase in concentration of p22 phox -(149 -168) caused changes in the chemical shifts for a number of resonances (Fig. 1b). In particular, all five tryptophan signals (indicated by asterisks in Fig 1b), which consist of two Trp residues in the Nterminal SH3 domain and three Trp residues in the C-terminal SH3 domain, showed significant changes in their chemical shifts during titration with p22 phox -(149 -168), indicating that p22 phox -(149 -168) stably interacts with both SH3 domains.
The K d values for complex formation between p47 phox -(151-286) and the two p22 phox peptides were subsequently determined by fluorescence titration method. The fluorescence data gave K d values of 8.67 M for p22 phox -(149 -162) and 0.64 M for p22 phox -(149 -168), respectively (Fig. 1, c and d), indicating that residues 163-168 of p22 phox enhance the binding affinity of p22 phox for p47 phox by more than 10-fold.
Comparison of the Molecular Dimensions of p47 phox -(151-286) in the Presence and Absence of p22 phox -(149 -168) Using SAXS-We next used SAXS to determine the molecular masses and dimensions of free and complexed p47 phox -(151-286). Guinier analysis (27) showed that the molecular masses of the tandem SH3 domains in both forms were 15 and 18 kDa, respectively (data not shown). We therefore concluded that p47 phox -(151-286) exists as a monomer in solution, regardless of whether it is free or in a complex. Next, we calculated the distance distribution function, P(r), for p47 phox -(151-286) from a set of scattering data using indirect Fourier transformation (29). The peptide-free form showed a curve with a peak top located at ϳ25 Å and a skirt of distribution extending up to 80 Å (Fig. 2a), suggesting a relatively elongated structure. In contrast, the P(r) for the complex showed a peak at ϳ20 Å and a skirt extending up to 60 Å (Fig. 2b). From these findings, we concluded that p47 phox -(151-286) has a compact globular structure when complexed with p22 phox -(149 -168), whereas the peptide-free form has an extended structure. Consistent with the P(r) analysis, Guinier analysis determined that the radius of gyration is 25.3 Å for the free form and 19.5 Å for the complex form.
Comparison of Backbone Dynamics for p47 phox -(151-286) in the Presence and Absence of p22 phox -(149 -168)-To characterize the dynamics of p47 phox -(151-286), we measured the hetNOE spectra of the free and complexed forms. To assign backbone resonances for the tandem SH3 domains, we first performed a series of NMR experiments using 0.3 mM 13 C/ 15 N-labeled protein in 20 mM sodium phosphate buffer, pH 7, at 25°C in the presence or absence of p22 phox -(149 -168). More than 92% (excluding proline) of the resonances could be assigned, and the 1 H-15 N backbone resonances were utilized as probes of the dynamics of the free and complexed forms of p47 phox -(151-286). Fig. 3, a and b, shows the hetNOE intensities for each residue of p47 phox -(151-286) in the free and complex states, respectively. In general, the residues with large NOE values (0.7-1.0) are expected to have lower flexibilities and to be located in the rigid core, whereas those with smaller NOE values (Ͻ0.6) are expected to be located at loop or linker regions that have higher flexibility (30). The present results revealed that the region corresponding to N-SH3 (residues 159 -212) and C-SH3 (residues 229 -282) had large NOE intensities of more than 0.7, indicating that, in solution, each SH3 domain behaves as a structural core. However, the linker region (resi-  FEBRUARY 10, 2006 • VOLUME 281 • NUMBER 6 dues 213-228) had different NOE intensities in the free and complex forms. In the free state, the linker region showed negative NOE intensities (Fig. 3a), indicating that this region is highly flexible. Both SH3 domains are joined with the flexible linker and are mobile. In contrast, in the complex (Fig. 3b), the linker region showed small positive NOE intensities (ϳ0.3), suggesting that the motion of the linker region was restricted. On the basis of these findings, we propose that the two SH3 domains can move freely in the free state but that, upon binding to the peptide, both undergo structural rearrangements causing the complex to become rigid. This view is consistent with the SAXS data, which show that the tandem SH3 domains have an extended structure in the free state but a compact structure in the complex state.
Structure Description-In contrast to the intertwined dimer in the crystal, the solution structure of p47 phox -(151-286) complexed with p22 phox -(149 -168) was a monomer. In this monomeric structure, p22 phox -(149 -168) bundles the tandem SH3 domains and the linker to form a compact structure. This is consistent with the conclusions of the SAXS and NOE studies. As expected from the domain structure, the solution structure of the tandem SH3 domains is composed of N-SH3 (blue), the linker (orange), and C-SH3 (green) (Fig. 4a). Fig. 4 also shows p22 phox -(149 -168) in red. Whereas both SH3 domains form well defined structural cores, the linker region is moderately disordered due to a lack of long-range NOEs, consistent with the smaller hetNOE intensities (Fig. 3b). Both SH3 domains form canonical SH3 folds, composed of a ␤-barrel (two ␤-sheets containing five antiparallel ␤-strands) and a 3 10 ␣-helix (Fig. 4b). The conserved ligand-binding surfaces of N-SH3 and C-SH3 are arranged face-to-face and form a single binding groove. This requires spatial contact of the two SH3 domains at the n-Src loops. Indeed, interdomain NOE connectivities were observed between the following residues: Glu 190 (n-Src in N-SH3) and Leu 260 (n-Src in C-SH3); and Ser 191 (n-Src in N-SH3) and Asp 261 (n-Src in C-SH3).  The residues from Lys 149 to Arg 164 of p22 phox -(149 -168) form a relatively well defined structure, although the C-terminal four residues (165-168) are disordered. Residues Pro 155 -Pro 160 form a left-handed polyproline type II helix (PPII) that is sandwiched by both SH3 domains in manner similar to that observed in the crystal structure (21). However, the crystal structure only revealed the core structure of p22 phox -(149 -168) corresponding to the region from Pro 151 to Pro 160 . The solution structure reveals that residues Ala 161 -Arg 164 form a short ␣-helix (Fig. 4). The interaction between this short ␣-helix and p47 phox -(151-286) could explain the ability of the C-terminal extension to substantially (Ͼ10-fold) increase the affinity of p22 phox toward p47 phox -(151-286). N-SH3 for p22 phox -(149 -168)-Fig. 5a shows a close-up view of the binding surface between p47 phox -(151-286) and p22 phox -(149 -168). The residues from Pro 155 to Pro 160 form PPII. In N-SH3, the PPII ligand-binding surface, which is composed of Tyr 167 , Pro 206 , Trp 193 , Phe 209 , and Ile 164 , interacts with residues Pro 156 , Pro 157 , Pro 159 , and Pro 160 of p22 phox -(149 -168), which are located on the basal plane of the PPII prism in a plus orientation (class I ligand) (45). Pro 155 and Arg 158 are located at the apexes of the prism and interact with C-SH3. Asn 154 and Ser 153 take on an extended conformation, which enables hydrogen bond formation between the main-chain carbonyl group of Asn 154 and NH of the indole ring in Trp 193 . Pro 152 of p22 phox -   Recognition of p22 phox -(149 -168) by p47 phox -(151-286). N-SH3, linker, C-SH3 (in ribbon model), and p22 phox -(149 -168) (in wire model) are colored blue, orange, green, and red, respectively. a, recognition of p22 phox - (149 -168) by N-SH3. b, recognition of p22 phox -(149 -168) by C-SH3. The side chains of the N-and C-SH3 domains that recognize p22 phox -(149 -168) are also shown as a wire model.

Binding Surface of
(149 -168) is inserted into another hydrophobic pocket formed by Trp 193 and Trp 204 that fixes the ligand in a plus orientation. These interactions are similar to those observed in the crystal structure (21), but in the crystal structure, the regions outside of the PRR (Pro 151 -Pro 160 ) are missing.
The most significant feature of p22 phox -(149 -168) in solution is that the residues from Ala 161 to Arg 164 , located at the C-terminal region of PPII, form a short ␣-helix. Although the exposed Ala 161 has some weak NOE connectivities with Phe 209 , Glu 162 of p22 phox -(149 -168) has multiple and strong NOEs with the side chain of Ile 164 in N-SH3. Moreover, Ala 163 of p22 phox -(149 -168) strongly interacts with Phe 209 , Ile 164 , and Ala 165 of N-SH3. The main-chain atoms of Arg 164 have weak NOE connectivities with Ile 164 , but its side chain does not interact with either of the SH3 domains. The C-terminal residues after Lys 165 have no inter-molecular NOEs with the tandem SH3 domains. Therefore, the C-terminal region after Lys 165 of p22 phox does not form a stable conformation and does not play a role in the molecular recognition. Thus, the short ␣-helix composed of Ala 161 -Arg 164 of p22 phox is stabilized by hydrophobic interactions, especially between Ala 163 and N-SH3, and enhances the binding affinity of p22 phox toward p47 phox . Another feature of the solution structure is that Lys 149 of p22 phox -(149 -168) interacts with residues Trp 204 , Phe 195 , and Val 186 . This further increases the binding affinity between N-SH3 and p22 phox -(149 -168).
Comparison of the Solution and Crystal Structures-To compare the solution and crystal structures of the activated tandem SH3 domains, we measured the 1 H-15 N RDC of each backbone amide bond. RDC is a sensitive probe for the structure in solution and has been utilized as a direct and simple tool for evaluating the similarity between solution and crystal structures. Recently, analysis of the alignment tensor from RDC values has been applied successfully to the determination of the relative domain orientations in multidomain proteins (23,44,(47)(48)(49)(50). After excluding highly flexible residues based on NOE experiments (hetNOE values of Ͻ0.65), the RDC values for 91 residues were obtained and used for analysis of the principal axes and determination of the principal values of the alignment tensors. Fig. 6 shows the correlations between the experimental RDCs and back-calculated RDCs using the coordinates of the solution and crystal structures. We obtained a strong correlation between the experimental and the back-calculated RDCs for the solution structure ( Fig. 6a) but only a fairly good correlation for the crystal structure (Fig. 6b), because RDCs were incorporated in the NMR structure calculation. Subsequently, we superimposed the backbone atoms of the crystal and solution structures (Fig. 6c) that revealed a pairwise r.m.s.d. of only 0.92 Å for backbone carbon and nitrogen atoms of residues involved in the secondary structure. Systematic deviations were observed in the distal loop region of the N-SH3 domain and the linker region. Because the hetNOE values (Fig. 3b) indicate that the linker region undergoes appreciable structural fluctuations, it is not surprising that there is a structural difference in the linker region between the NMR and crystal structures. The difference in the distal loop region of N-SH3, on the other hand, is due to an artifact of the crystal structure, because the two chains in each unit cell are interchanged with each other at the distal loop of N-SH3. Except for these regions, the solution and crystal structures are quite similar.
Comparison with the Autoinhibited Form- Fig. 7 compares the molecular contacts between p47 phox -(151-286) and p22 phox -(149 -168) in the solution form ( Fig. 7a) with those in the crystal structure of the autoinhibited form of p47 phox -(151-340) (21, 22) (Fig. 7b). Despite the presence of an intertwined dimer in the crystal state, the recognition mode between the SH3 domains and Ala 298 , Pro 299 , Pro 300 , and Arg 301 in the autoinhibited form (Fig. 7b) is quite similar to that between the SH3 domains and Pro 155 , Pro 156 , Pro 157 , and Arg 158 in the activated form (Fig. 7a). These residues are sandwiched by both SH3 domains, each of which recognizes the different basal plane of the PPII helix. The Role of the Short ␣-Helix C-terminal to the PRR in Binding to p47 phox -(151-286)-The present NMR study revealed not only that the core PRR of p22 phox , which forms a PPII helix, plays a central role in the interaction with p47 phox but also that its C-terminally flanking region of FIGURE 8. Interaction of p47 phox -(151-286) with the ␣-helical region C-terminal to the PRR of p22 phox . MBP fusion protein of the wild-type (wt) or I164A or F209A mutants of p47 phox -(151-286) (5 g) was incubated with 5 g of GST, a GST fusion protein of the wild-type p22 phox C-terminal region (amino acids 132-195), or a GST fusion protein of the C-terminal region carrying the A163G/R164G substitutions. The proteins were precipitated with glutathione-Sepharose beads and eluted from the beads with glutathione. The eluates were subjected to SDS-PAGE and stained with Coomassie Brilliant Blue. Representative results are shown from three independent experiments. amino acids 161-164 (Ala-Glu-Ala-Arg) adopts an ␣-helical structure that appears to enhance the affinity of p22 phox for p47 phox . To investigate the role of the ␣-helix in binding to p47 phox , we replaced Ala 163 and Arg 164 with glycine, a residue known to destabilize ␣-helical structures. As shown in Fig. 8, the A163G/R164G substitution led to decreased binding to p47 phox -(151-286), implying a significant role for the p22 phox ␣-helical region.
Because the ␣-helix appears to interact directly with Ile 164 and Phe 209 in the p47 phox N-SH3 (Fig. 7a), we next investigated the roles of these residues. For this purpose, we prepared mutants of p47 phox -(151-286) carrying substitutions of alanine for Ile 164 or Phe 209 and tested their ability to bind to p22 phox . Compared with the wild-type p47 phox -(151-286), the mutant protein with the F209A substitution weakly bound to p22 phox (Fig. 8). The I164A substitution resulted in a more impaired interaction with p22 phox (Fig. 8), which is consistent with the present finding that Ile 164 makes stronger contacts with the p22 phox ␣-helix than does Phe 209 (Fig. 7a). These mutant p47 phox proteins also interacted more weakly with the ␣-helix-defective p22 phox (A163G/R164G) than with the wild-type p22 phox (Fig. 8). Thus, the ␣-helical conformation of the region C-terminal to the p22 phox PRR (amino acids 161-164) participates in the full interaction with p47 phox -(151-286).
Conclusion-We determined the solution structure of p47 phox -(151-286) complexed with p22 phox -(149 -168) using NMR spectroscopy. In contrast to the intertwined dimer in the crystal, the solution structure is a monomer. The p22 phox -(149 -168) forms PPII, in which the basal planes are recognized simultaneously by two SH3 domains. This mode of interaction provides specificity and high affinity of the peptide toward the tandem SH3 domains of p47 phox . Moreover, the solution structure reveals the presence of a C-terminal short ␣-helix in p22 phox -(149 -168). The interaction between the short ␣-helix and the tandem SH3 domains enhances the binding affinity by more than 10-fold. The mutational studies confirm that the residues on the short ␣-helix are crucial for high-affinity binding between p47 phox and p22 phox .