Báo cáo Y học: Conformational analysis by CD and NMR spectroscopy of a peptide encompassing the amphipathic domain of YopD from Yersinia potx

Conformational analysis by CD and NMR spectroscopy of a peptide encompassing the amphipathic domain of YopD from Yersinia Tobias Tengel 1 , Ingmar Sethson 1 and Matthew S. Francis 2 1 Departments of Organic Chemistry and 2 Molecular Biology, Umea ˚ University, Umea ˚ , Sweden To establish an infection, Yersinia pseudotuberculosis utilizes a plasmid-encoded type III secretion machine that permits the translocation of several anti-host factors into the cytosol of target eukaryotic cells. Secreted YopD is essential for this process. Pre-secretory stabilization of YopD is mediated by an interaction with its cognate chaperone, LcrH. YopD possesses LcrH binding domains located in the N-terminus and in a predicted amphipathic domain located near the C-terminus. This latter domain is also critical for Yersinia virulence. In this study, we designed synthetic peptides encompassing theC-terminal amphipathicdomain of YopD. A solution structure of YopD 278)300 , a peptide that strongly interacted with LcrH, was obtained by NMR methods. The structure is composed of a well-defined amphipathic a helix ranging from Phe280 to Tyr291, followed by a type I b turn between residues Val292 and His295. The C-terminal trun- cated peptides, YopD 278)292 and YopD 271)292 , lacked helical structure, implicating the b turn in helix stability. An inter- action between YopD 278)300 and its cognate chaperone, LcrH, was observed by NMR through line-broadening effects and chemical shift differences between the free peptide and the peptide–LcrH complex. These effects were not observed for the unstructured peptide, YopD 278)292 ,which confirms that the a helical structure of the YopD amphi- pathic domain is a critical binding region of LcrH. Keywords: YopD; amphipathic helix; LcrH; NMR solution structure; 2,2,2-trifluoroethanol. Injection of anti-host factors into eukaryotic cells by numerous economically important animal- and plant-inter- acting Gram-negative bacteria is achieved by functionally homologous Ôtype III secretion systemsÕ (TTSS) [1,2]. This TTSS-dependent process is essential to establish bacterial infections. The enteropathogen Yersinia pseudotuberculosis is a model system used to study the basic molecular mechanisms of type III secretion. All pathogenic Yersinia spp. harbor a  70-kb virulence plasmid that encodes numerous Yop (Yersinia outer protein) and Lcr (low calcium response) virulence determinants that are secreted by the Ysc (Yersinia secretion) type III apparatus [3,4]. Two protein classes are secreted by the Ysc apparatus; antihost Yop-effector proteins and those required for their efficient injection into target cells. Collectively, these determinants co-operate to allow Yersinia to resist uptake by both professional and nonprofessional cells [5–7] and subvert host cell signalling that would normally lead to effective bacterial clearance [8]. YopD is a crucial TTSS component during a Yersinia infection being essential for the injection of antihost Yop-effectors into target cells, possibly through stabilization of a YopB–LcrV pore complex in the plasma membrane through which Yop-effectors are injected into host cells [4,9]. However, involvement of YopD in pore formation is only transitory, because a portion of YopD is also localized to the host cell cytosol [10]. In addition, we and others observed that a yopD null mutant is constitutively induced for synthesis of Yops in vitro, while Yop synthesis in wild type bacteria remained tightly regulated in response to temper- ature and Ca 2+ [10,11]. This highlights important dual roles for YopD in both negative regulation of Yop synthesis and injection of Yop-effectors into target cells. While the mechanism of YopD function is unknown, it is dependent on an interaction with the nonsecreted TTSS chaperone LcrH [12,13]. This interaction is responsible for the presecretory stabilization and efficient secretion of YopD [12,13], and is important for control of yop regulation [14,15]. It follows that protein interactions involving several TTSS chaperones and their cognate secreted partner are now recognized as having pivotal roles in temporal and spatial control of virulence [16,17]. Therefore, to better understand this relationship, we have chosen to analyze the YopD–LcrH complex because functional homologues exist in other systems [18,19] and their interactive domains have already been mapped in vitro [13]. In fact, we have previously identified several hydrophobic residues within a putative C-terminal amphipathic domain of YopD that are necessary for binding LcrH [13]. This finding was significant as it coincides with the additional requirements for this Correspondence to M. S. Francis, Department of Molecular Biology Umea ˚ University, SE-901 87 Umea ˚ ,Sweden. Fax: + 46 90 77 14 20, Tel.: + 46 90 785 25 36, E-mail: matthew.francis@molbiol.umu.se or I. Sethson, Department of Organic Chemistry, Umea ˚ University, SE-901 87 Umea ˚ ,Sweden. Fax: + 46 90 13 88 85, Tel.: + 46 90 786 99 76, E-mail: ingmar.sethson@chem.umu.se Abbreviations: CSI, chemical shift index; SA, simulated annealing; TTSS, type III secretion system(s). Note: Web pages are available at http://www.cmb.umu.se and http://www.chem.umu.se/Department/orgchem Note: Individual amino acids are indicated by the three-letter abbreviation followed by a number indicating sequence position relevant to the full length YopD protein. Complete peptide sequences are presented in one-letter amino acid code. (Received 18 March 2002, revised 6 June 2002, accepted 17 June 2002) Eur. J. Biochem. 269, 3659–3668 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03051.x domain in Yersinia pathogenesis, being essential for both regulation of Yop production and injection of antihost effectors into host cells [10]. Thus, in this initial structural study of YopD, we focused on the putative amphipathic domain. This strategy was advantageous because full length YopD is susceptible to aggregation [20] and the amphipathic domain is clearly biologically relevant [10,13]. The utilization of small peptides to evaluate smaller domains to build up the tertiary structure of large polypeptides has made a substantial contribution to the understanding of protein structures and initial protein folding events [21,22]. In this study, we therefore designed synthetic peptides that encompassed the C-terminal amphipathic domain of YopD. The peptide structures were examined using CD spectroscopy and 2D homonuclear/heteronuclear NMR spectroscopy. Using these peptides, the interaction between the amphipathic domain of YopD and its cognate chaperone LcrH was investigated by NMR. EXPERIMENTAL PROCEDURES Materials Peptides spanning the C-terminal amphipathic domain of YopD were purchased from Chemical R & D Laboratory (Copenhagen, Denmark). Peptide purity was confirmed by HPLC. The peptide sequences were as follows, YopD 278)292 (DNFMKDVLRLIEQYV); YopD 271)292 (EEAMNYND NFMKDVLRLIEQYV); YopD 278)300 (DNFMKDVLRL IEQYVSSHTHAMK) and YopD 271)300 (EEAMNYN DNFMKDVLRLIEQYVSSHTHAMK). 2,2,2-trifluoro- ethanol-d 3 (99%) was purchased from Larodan Chemicals (Malmo ¨ , Sweden) and nondeuterated trifluoroethanol used in CD experiments was obtained from Sigma-Aldrich. All other chemicals were analytical grade and obtained from various manufacturers. Cloning, expression and purification of LcrH The lcrH gene was amplified by PCR on a 540-bp NdeI/BglII DNA fragment using the primer combination of plcrH10: 5¢-CGAGGTACATATGCAACAAGAGACG-3¢ and plcrH11: 5¢-ACGTACAGATCTCCTTGTCGTCGTCGT CTGGGTTATCAACGCACTC-3¢.Thisfragmentwas then cloned into the expression vector pET30a (Novagen, Wisconsin, USA) giving rise to pMF322, encoding LcrH containing a C-terminal enterokinase cleavage site upstream of a His 6 -tag. To express this recombinant protein, an overnight culture of Escherichia coli BL21(DE3)/pMF322 grown at 26 °C in Luria–Bertani broth (1% (w/v) NaCl, 0.5% (w/v) yeast extract, 1% (w/v) tryptone) was subcul- tured (0.1 volume) into 500 mL fresh medium. After 1.5 h incubation at 26 °C, protein expression from pMF322 was induced by the addition of isopropyl thio-b- D -galactoside to 1m M for a further 3.5 h. Cells were pelleted by centrifuga- tion at 9820 g and stored overnight at )80 °C, from which 10 mL of cleared lysate was prepared under native condi- tions using the QIAexpressionist protocol (Qiagen, CA, USA). To the cleared lysate, 1.5 mL of nickel-nitrilotriacetic acid slurry (Qiagen) was added, followed by a 1-h incubation at 4 °C on a rotary shaker. The sample was then loaded on a Poly Prep chromatography column (Bio-Rad, CA, USA) and each subsequent flow-through collected. The column was washed twice with wash buffer (50 m M sodium phos- phate,pH8,300m M NaCl, 0.8 m M imidazole) containing complete protease inhibitor cocktail (Roche Molecular Biochemicals, Basel, Switzerland) and once with wash buffer without inhibitors. The LcrH::His protein was eluted from the column in 50 m M sodium phosphate, pH 8, 300 m M NaCl, 8 m M Imidazole. SDS/PAGE analysis and Coomassie Brilliant Blue staining was used to assess the purity of LcrH::His contained in each column flow-through fraction. Pure fractions were combined and dialyzed for several days in large volumes of 50 m M sodium phosphate buffer, pH 8. The concentration of LcrH::His was determined with the Bradford Reagent (Sigma) using known concentrations of bovine serum albumin (New England Biolabs, Massachu- setts, USA) as the standard. Size-exclusion chromatography Size-exclusion chromatography was performed on Super- dex 75 HR 10/30 columns using an FPLC-system (Amer- sham Pharmacia Biotech, New Jersey, USA). The mobile phase for the SEC experiments was 50 m M sodium phosphate buffer, pH 8, 150 m M NaCl with a flow rate of 0.75 mLÆmin )1 . Circular dichroism Samples for CD were either 60 l M peptide in 5 m M sodium phosphate buffer at pH 4.5 and 6 or 60 l M LcrH in 10 m M buffer at pH 8. CD experiments were conducted on YopD 278)292 ,YopD 271)292 and YopD 278)300 using different concentrations of 2,2,2-trifluoroethanol, 0–40%. In addition, a temperature study between 25 and 60 °C was performed on YopD 278)300 in 40% 2,2,2-trifluoroeth- anol. No CD data was collected for YopD 271)300 because this peptide was difficult to solubilize in phosphate buffer. CD spectra were recorded on a CD6 spectrodichrograph (Jobin-Yvon Instruments SA, Longjumeau, France). Spectra were collected between 185 and 260 nm at 25 °C using a 0.5-mm quartz cell. Data were collected at 0.5-nm intervals with an integration time of 2 s. Three spectra per sample were acquired and averaged, followed by subtraction of the CD signal of the solvent. Ellipticity is expressed in terms of mean residue molar ellipticity [h] (degÆcm 2 Ædmol )1 ). Nuclear magnetic resonance Peptide samples for NMR were 2–4 m M in 20 m M sodium phosphate buffer and 1 m M NaN 3 , pH 4.5. However, the YopD 278)300 and YopD 271)300 peptides were also examined in 2,2,2-trifluoroethanol/water mixtures. YopD 278)300 was studied in 40% 2,2,2-trifluoroethanol-d 3 /H 2 O/D 2 O solution (4 : 5 : 1, v/v/v) at pH 4.5 and 6.3, whereas experiments involving YopD 271)300 were carried out in a 40% 2,2,2- trifluoroethanol/water mixture at pH 3.8. When analyzing the peptide–LcrH interaction, 0.25 m M samples of YopD 278)300 and YopD 278)292 were prepared in 10% 2,2,2-trifluoroethanol at pH 6.3 and purified LcrH was added in sequential steps to a final peptide/protein molar ratio of 2 : 1. The appropriate pH was corrected by the addition of small aliquots of HCl and NaOH. NMR 3660 T. Tengel et al. (Eur. J. Biochem. 269) Ó FEBS 2002 experiments were also conducted between 10 and 50 °Cto elucidate the appropriate temperature for further NMR analysis. A temperature of 40 °C was chosen in order to minimize peptide aggregation and obtain a better resolved spectrum. All NMR spectra were recorded on a Bruker DRX and a Bruker AMX2 spectrometer operating at a proton frequency of 600.13 MHz and 500.13 MHz, respectively. Both were equipped with a triple resonance gradient probe. The spectra used for resonance assignments and structure elucidation included phase sensitive DQF-COSY [23], TOCSY [24], NOESY [25] and gradient enhanced HSQC [26]. In TOCSY and NOESY experiments the solvent signal was suppressed just before the FID acquisition using the WATERGATE pulse sequence [27]. The DIPSI pulse sequence with a spin lock time of 85 ms was used in the TOCSY experiments and the NOESY spectra was recorded with a mixing time of 150 ms. Data were processed on a Silicon Graphic workstation using the XWINNMR software (Bruker). Prior to Fourier transformation, the data were multiplied by appropriate window functions. Zero-filling was applied in both dimensions and linear prediction in the indirect dimension. The chemical shift of the water signal was used as a reference and calibrated to 4.60 p.p.m. at 40 °C. The HSQC spectra were calibrated using the ratio 13 C/ 1 H ¼ 0.25144953 for carbon and 15 N/ 1 H ¼ 0.101329118 for nitrogen [28]. Derivation of distance and dihedral restraints Distance restraints for YopD 278)300 were obtained from the NOESY spectrum recorded at 40 °C, pH 4.5 and 40% 2,2,2-trifluoroethanol, using a mixing time of 150 ms. Assigned NOE cross peaks were volume integrated and converted to distance restraints using MARDIGRAS [29]. An extended structure of YopD 278)300 was subjected to unre- strained molecular dynamics calculations at 1000 K to generate 10 different structures. These 10 divergent struc- tures served as a representation of the conformational space, and each of them was used in the MARDIGRAS calculations. The extreme values were used as upper and lower bonds in the structure calculation. As no specific assignment could be made for the methyl and methylene protons, appropriate pseudoatom correction was applied [30]. The rotational correlation time, s c ,usedinthe MARDIGRAS calculations, was calculated from experimental spin-lattice (T 1 )andspin- spin (T 2 ) relaxation time measurements of well resolved peaks in YopD 278)300 . T 1 and T 2 values were obtained for residues 5–8, 11–16, 20 and 22. The rotational correlation time was calculated for each residue using the equation s c ¼ 2x )1 (3T 2 /T 1 ) )1/2 [31] resulting in s c values between 6 and 10 ns. The average value, 8 ns, was used in the following MARDIGRAS calculations. Backbone / dihedral angle restraints were obtained using the program TALOS [32]. Structure calculations Structure calculations were carried out using X - PLOR 3.851 [33]. This involved simulated annealing (SA) [34] and SA refinement. The starting structures for the SA calculations were varied to ensure that the resulting structure represented a global energy minimum in the conformational space. From three structures with a pair-wise rmsd of 2 A ˚ or more for the backbone heavy atoms, 150 structures were calcu- lated using the SA and SA refinement protocols. To describe the quality of the solution structure of YopD 278)300 , rmsd values between all the accepted struc- tures and the average structure were studied. The structures were analyzed using INSIGHT II (Accelrys Inc., California, USA), MOLMOL [35] and VMD [36]. In order to verify that no residues were in disallowed regions, Ramachandran plot analysis was conducted using the program PROCHECK - NMR [37]. RESULTS AND DISCUSSION CD and NMR studies Computer analysis of YopD primary sequence predicts a central hydrophobic membrane spanning domain and a C-terminal amphipathic domain (Fig. 1A) [38]. This latter region can be presented on a helical wheel projection to reveal an amino-acid sidedness (Fig. 1B) [13]. While the Fig. 1. Overview and helical wheel projection of biologically significant domains in YopD (306 amino acids). (A) Computer prediction [38] was used to define the central hydrophobic and the C-terminal amphipathic domains of YopD. (B) A helical wheel projection of the amphipathic domain of YopD incorporates residues 278–292 [13]. Amino acids are presented in one-letter amino-acid code with hydrophobic residues boxed. Ó FEBS 2002 Tertiary structure of the YopD amphipathic domain (Eur. J. Biochem. 269) 3661 spatial distribution of these amino acids appeared crucial for binding the LcrH chaperone [13], we wished to extend these findings using a chemical approach. In particular, this initial study aimed at obtaining the secondary structure of the predicted C-terminal amphipathic domain of YopD. To overcome the risk of YopD aggregation [20] we designed small YopD-specific peptides that encompassed the pre- dicted C-terminal amphipathic domain. As an efficient means to confirm the presence of a helical structure of these peptides, CD experiments were conducted on YopD 278)292 , YopD 271)292 and YopD 278)300 . The CD spectrum of YopD 278)300 , in aqueous buffer, showed two minima at 208 and 222 nm and an isodichroic point at 200 nm, which are characteristics of a a helical conformation (Fig. 2). We were unable to detect any secondary structure for the peptides, YopD 278)292 and YopD 271)292 , even in the pres- ence of 2,2,2-trifluoroethanol (data not shown). The fact that neither peptide displayed any helical structure indicates that the amino acids downstream of residue 292 may be essential for helical stability. Because no secondary structure was detected for the YopD 278)292 and YopD 271)292 peptides, NMR structural characterization was conducted on YopD 278)300 . However, in the first attempts to determine the structure in aqueous buffer at pH 4.5, the peptide severely aggregated. Accord- ingly, under these conditions the spectrum of YopD 278)300 showed extensive line broadening (Fig. 3A). Several studies have reported that the addition of organic solvents can reduce the incidence of peptide aggregation [39,40]. In view of this, 2,2,2-trifluoroethanol was added to this sample to give different final concentrations in the range of 0–40%. When NMR experiments were recorded to monitor the effects of adding 2,2,2-trifluoroethanol, a resolved NMR spectrum indicative of the disruption of large aggregates was observed even at low concentrations of 2,2,2-trifluoroethanol (Fig. 3B). However, as 2,2,2-trifluoroethanol is known to stabilize helical structure [41], additional CD experiments were conducted to investigate whether this solvent induced structural changes in the YopD peptide. Furthermore, the helical structure may be pH-dependent due to variations in charge distribution of the histidine side chains. This fact was taken into account by conducting CD experiments at both pH 4.5 (data not shown) and pH 6 (Fig. 2) as well as NMR experiments at pH 4.5 and 6.3. Collectively, no significant change in peptide helical content was observed, indicating that, at the pH conditions used in this study, the addition of 2,2,2-trifluoroethanol did not significantly alter the secondary structure of YopD 278)300 . As NMR spectra of YopD 278)300 were recorded at 40 °C, we used CD spectroscopy to verify that only minimal variations in helical content of the peptide occurred when the temperature was varied between 25 and 60 °C(datanot shown). Thus, in this range, temperature had no significant impact on the secondary structure. 1 H resonance assignment and secondary structure All NMR spectra were assigned according to classical procedures including spin system identification and sequen- tial assignment [42]. Initial spin system assignments of YopD 278)300 were obtained using COSY and TOCSY spectra and a NOESY spectrum was used to identify sequential backbone connectivities. A comparison of the H a and C a chemical shift deviation from random coil values Fig. 2. Plot of the residual molecular ellipticity from 185 to 260 nm of YopD 278)300 peptide samples at different 2,2,2-trifluoroethanol concen- trations. From below at 222 nm, the spectra are of peptide in 0, 30, 20, 40 and 10% 2,2,2-trifluoroethanol, respectively. All spectra were obtained with 60 l M of the peptide in 5 m M sodium phosphate buffer, pH 6 and conducted at 25 °C. Fig. 3. 1D 1 H NMR spectrum of YopD 278)300 . (A) Spectrum of a 3 m M peptide sample prepared in 50 m M sodium phosphate buffer pH 4.5, obtained with a probe temperature of 40 °C (amide region is shown). (B) Spectra of a 3 m M peptide sample containing 2,2,2-trifluoroethanol at a concentration of 0, 10, 20, 25, 30, 35 or 40% (percent 2,2,2- trifluoroethanol shown for each individual spectrum). All experiments were conducted at pH 4.5 in 20 m M sodium phosphate buffer at 40 °C (amide region is shown). 3662 T. Tengel et al. (Eur. J. Biochem. 269) Ó FEBS 2002 according to the chemical shift index (CSI) [43], highlighted a region of the peptide incorporating residues 280–295 where an ahelical structure was predicted (Fig. 4). These observations support the presence of an a helix as suggested from the CD analysis and define the location of the helical region. Structural restraints Several medium range NOEs, d aN (i,i +3); d aN (i,i +4) and d ab (i,i + 3), and strong sequential NOEs between amide protons also support a helical structured peptide (Fig. 4). NOEs assigned from the NOESY spectrum were converted to distance restraints using MARDIGRAS [29] and used as input for the structure calculations. The final number of restraints, after removal of those that according to the relaxation matrix originated from spin diffusion, was 242, which consisted of 134 intraresidue, 53 sequential and 55 medium range restraints. The proton, carbon and nitrogen chemical shifts of each residue were used to extract the / dihedral restraints using the program TALOS [32]. The / dihedral restraints obtained by TALOS were then used for residues in which TALOS indicated a good prediction relative to a known structure. These values were collected for residues 279 through to 295 and used as dihedral restraints in the structure calculations. NOEs were found that were not compatible with a monomeric structure. Accordingly, these NOEs were ascribed to intermolecular interactions and have been excluded in the structure calculations conducted in this study. However, it is still possible that the intermolecular interactions do not only appear as resolved peaks in the NOESY spectrum. They may also have the same frequen- cies as NOEs reflecting intramolecular interactions and thereby affecting the NOE intensities. Such influences will obviously affect the calculated monomeric structure. For- tunately, however, these influences appear minor for two reasons. Firstly, structure calculations with the used NOEs proceed without violations. Secondly, the dihedral restraints, which only represent intramolecular interactions, are completely compatible with the NOEs. Taken together, this implies that the structure of YopD 278)300 presented in this work does represent the monomeric structure, even though intermolecular interactions are present. Description and quality of the calculated structures The peptide YopD 278)300 adopts a well-defined helical structure with a more flexible C-terminal region (Fig. 5), with the hydrophobic and hydrophilic residues mainly located at opposite sides of the helix (Fig. 6). The a helix incorporates residues Phe280 to Tyr291 with the following four residues, Val292 to His295, forming a type I b turn. The exclusion of the TALOS dihedral restraints from the structure calculations generates an almost identical structure containing an a helix with a C-terminal turn. The presence of the b turn is also supported by the lowfield shift of 9.2 p.p.m. for the Val292 amide proton, as such shifts are rarely found in helical regions of Fig. 5. Superposition of the backbone atoms for the 25 lowest energy structures of YopD 278)300 . The structures were aligned for the best overlap of the backbone of residues 280–295 and superimposed on the lowest energy structure. This image was constructed with the VMD software [36]. Fig. 6. NMR structure of the amphipathic domain of YopD illustrating the hydrophobic and hydrophilic sidedness of the peptide. Thesidechains are displayed for residues 279–295 with the hydrophobic residues colored in red, the hydrophilic in blue and the tyrosine in grey. This image was constructed using the MOLMOL software [35]. Fig. 4. Overview of NOE connectivities and chemical shift data of YopD 278)300 in 40% 2,2,2-trifluoroethanol at pH 4.5. The relative intensities of each filled bar indicate the strength of the NOE restraints. Distance restraints were derived from a NOESY spectrum with a mixing time of 150 ms. The chemical shift indices are indicated by an index with values of )1, 0 and +1, which corresponds to upfield, random coil and downfield, respectively. This image was constructed using the VINCE software (The Rowland Institute for Science, MA, USA). Ó FEBS 2002 Tertiary structure of the YopD amphipathic domain (Eur. J. Biochem. 269) 3663 polypeptides. Significantly, the amino acids involved in this turn appear to be essential for stabilizing the a helical structure because the two synthesized peptides lacking this motif (YopD 278)292 and YopD 271)292 ) did not contain any helical structure (data not shown). Therefore, the formation of the b turn may act as a stabilizer, capping the C-terminal end of the a helix. This phenomenon has been previously described by Forood and colleagues [44]. However, we have not been able to identify any specific hydrogen bonds or other favourable interactions within the calculated struc- tures that would support this conclusion. Rather, the stabilizing effect may well occur via intermolecular interac- tions within the observed aggregates of YopD 278)300 .The presence of these putative intermolecular interactions would be consistent with the fact that the most significant chemical shift changes upon aggregation state variation occurred for the amide protons in the b turn (Fig. 3B). However, further detailed structural descriptions of the aggregate are needed to better understand the stabilizing function of the b turn. To examine whether the helix extended upstream of the N-terminus, the properties of the longer YopD 271)300 peptide in a 2,2,2-trifluoroethanol/water preparation were analyzed. The chemical shifts and NOE patterns of this peptide, compared to those of YopD 278)300 , confirmed that the a helix begins at residue Phe280 (data not shown). Of the 150 calculated peptide structures, 145 were accepted. The criteria for acceptance were as follows: rmsd for bonds < 0.01 A ˚ ; rmsd for angles < 2°;noNOE violation > 0.3 A ˚ and no constraint dihedral violation >5°. To verify the quality of the YopD 278)300 solution structure, rmsd values between all 145 accepted structures and the average structure as well as pair-wise rmsd were studied. When superposition was performed using residues 280–295, this region displayed a well-defined structure with a rmsd value of 0.18 A ˚ for the backbone atoms. An illustration of the rmsd on a per residue basis compared to the number of NOE restraints per residue is presented in Fig. 7. Ramachandran plot analysis using the program PROCHECK - NMR [37] was used to verify that no residues were located in disallowed regions. From the minimized average structure, 81% of the residues were in most favoured regions with the remaining 19% in additional allowed regions. Structural restraints, rmsd values and the results from Ramachandran plot analysis are summarized in Table 1. Fig. 7. Distribution of distance restraints and rmsd values for YopD 278)300 . (A) Distribution of NOE restraints in YopD 278)300 on a per residue basis. Three types of restraints are specified: black, intra- residue; light grey, sequential; dark grey, medium range. All interres- idue NOEs are plotted twice. NOE data was obtained from a 150 ms NOESY spectrum conducted in 40% 2,2,2-trifluoroethanol at pH 4.5 with a sample temperature of 40 °C. (B) Distribution of rmsd values on a per amino acid basis. The structures were superpositioned according to the best fit of the backbone of residues 280–295 and the rmsd value was calculated for all of the accepted 145 structures (see Table 1). Table 1. Summary of the structural statistics and rmsd differences. Unless stated, all 145 accepted structures have been used to calculate structural statistics. NOE statistics a Intraresidual 134 Sequential 53 Medium range 55 Dihedral angle restraints a / 17 Ramachandran plot analysis b Residues in most favorable regions 81% Residues in additional allowed regions 19% rmsd from average structure (A ˚ ) c All residues 1.41/2.15 Residues 280–295 0.18/1.08 rmsd pair-wise (A ˚ ) c All residues 1.92/2.95 Residues 280–295 0.24/1.43 a No structure exhibited distance violations greater than 0.3 A ˚ or dihedral angle violations greater than 3°. b The minimized average structure was used to perform Ramachandran plot analysis. c Backbone/heavy atoms. 3664 T. Tengel et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Aggregation of YopD The NMR experiments conducted on YopD 278)300 revealed that the peptide severely aggregated in aqueous buffer, pH 4.5. The 1D NMR spectrum of YopD 278)300 was poorly resolved with extensive line broadening (Fig. 3A). This is indicative of an aggregate having a size well above the limit allowing high resolution NMR. By adding 2,2,2-trifluoro- ethanol it was possible to disrupt this large aggregate without any significant changes in peptide secondary structure (see above). However, even though well resolved spectra were recorded, several observations indicate that the aggregate was not completely disrupted but forms smaller aggregates. We observed long range NOEs from the aromatic protons of Phe280 to the side chains of Ile288 and Val292. This indicates that the peptide forms an aggregate with the a helices oriented in an antiparallel direction with their hydrophobic sides facing each other. In addition, the formation of small aggregates is also supported by the rotational correlation time (s c ). In our case, s c was determined to be 8 ns, which is too long to be explained by the viscosity of the 2,2,2- trifluoroethanol/water mixture [45]. Hence, this value sug- gests that the peptide is not monomeric but rather forms a smaller aggregate that reflects this s c value. It is also noteworthy that even in 40% 2,2,2-trifluoro- ethanol, the aggregation state can be affected by changing the temperature. When the temperature was lowered, extensive line broadening and an upfield shift of the a-protons occurred for all the residues. We interpret these findings to further indicate the formation of a larger aggregate that stabilizes the helical structure of the peptide. In addition, analysis of the 2,2,2-trifluoroethanol titration of YopD 278)300 identified two different amide proton behavioural patterns, residues that experienced a downfield shift upon the addition of 2,2,2-trifluoroethanol and those which displayed an upfield shift. Interestingly, most residues on the hydrophobic side of the helix experienced a large downfield shift whereas those on the hydrophilic side of the helix experienced a minor downfield shift or in some cases an upfield shift. This supports the hypothesis that hydro- phobic residues form the aggregate, as these residues are likely to be most affected when the aggregate is destabilized. It follows that the aggregation of full length secreted YopD observed by Michiels and colleagues [20], may involve the same interaction between the amphipathic C-terminal helices. Because the nonsecreted LcrH chaper- one prevents premature aggregation of presecretory YopD [12,13] and binds to a region incorporating the amphipathic a helix [13], it would be very interesting to determine if LcrH-binding modulates the extent of peptide aggregation. In fact, such a biophysical study would provide valuable information towards understanding the biological relevance of YopD multimerization and even the general role of TTSS chaperone function, because functional homologues of YopD and LcrH exist in other bacterial pathogens [18,19]. Information on the latter would be an important develop- ment because dual roles for these specialized molecules have been recently proposed [16,17]. Structure and aggregation of LcrH Several observations support the fact that LcrH forms higher order structures in aqueous solution. These multimers may explain why high quality NMR spectra, even at LcrH concentrations above 1 m M , were difficult to obtain. Nev- ertheless, the fact that the chemical shifts of the amide protons and a-protons were found in a relatively restrained area does indicate that LcrH is a a helical protein (Fig. 8). This is consistent with the CD spectra of 60 l M LcrH, which also displayed characteristics of a helical conformation (Fig. 9). Interestingly, size-exclusion chromatography of Fig. 8. NOESY spectrum of LcrH. A0.5m M sample of LcrH in 50 m M phosphate buffer, pH 8, was used for the experiment. The spectrum was recorded with a mixing time of 150 ms at 25 °C(amide region is shown). Fig. 9. Plot of the residual molecular ellipticity from 195 to 240 nm of LcrH. The spectrum was obtained with a protein concentration of 60 l M in 10 m M sodium phosphate buffer, pH 8 and conducted at 25 °C. Ó FEBS 2002 Tertiary structure of the YopD amphipathic domain (Eur. J. Biochem. 269) 3665 LcrH suggested that the protein forms aggregates in aqueous solution because two main fractions were detected corres- ponding to 2.5 and 3.7 times the expected monomeric mass (data not shown). This implied the presence of multimers ranging from dimers (most abundant) to tetramers. Homod- imer formation by LcrH is consistent to that observed for other TTSS chaperones [46–51]. Significantly, this feature is apparently required to necessitate substrate secretion in diverse TTSS associated with both pathogenesis and flagellar biogenesis [49,50]. Interaction of the YopD peptide with LcrH In the yeast two hybrid assay, the C-terminal amphi- pathic domain of YopD was required for LcrH binding [13]. Because the YopD–LcrH complex is important for regulatory control of virulence gene expression in Yersinia infections [14,15], a detailed structural analysis of this complex is required. As part of this initial structural study we wanted to confirm the involvement of the C-terminal YopD domain in LcrH binding. For interaction studies with the YopD 278)300 peptide, LcrH was purified as a His-tag recombinant fusion by Nickel exchange chromatography. The peptide was prepared in a concentration of 0.25 m M in buffer supplemented with 10% 2,2,2-trifluoroethanol. These conditions minimized peptide aggregation and provided a better resolved spectrum (Fig. 3B). A 10% 2,2,2-trifluoroethanol con- centration was specifically chosen because LcrH precip- itated at higher concentrations. In addition, interaction studies were performed at pH 6.3 to avoid precipitation of LcrH under acidic conditions. Importantly, although the conditions chosen to examine the peptide–LcrH interaction are different from those used to describe the YopD 278)300 peptide solution structure, we clearly con- firmed that they did not influence the peptide structure (see above). LcrH was added in a stepwise manner to the peptide sample to give a final peptide:protein molar ratio of 2 : 1. An interaction between YopD 278)300 and LcrH was observed from line-broadening and chemical shift differ- ences within the amide region from a 1D NMR spectrum of the free peptide and the peptide-protein solution (Fig. 10). The amide-proton resonances of Tyr291 and Val292 are considerably broadened in the presence of LcrH, consistent with their induced chemical shift differences upon addition of LcrH. In addition, we observed a decreased relaxation time of the peptide in the presence of LcrH, which supports peptide/LcrH binding (data not shown). Moreover, when we examined the unstructured YopD 278)292 peptide for the ability to bind LcrH, no such interaction was observed (data not shown). We interpret this finding to indicate the absolute requirement of the YopD helical structure for the YopD–LcrH interaction. It follows that line broadening and chemical shift differences between the bound and the unbound YopD 278)300 peptide indicate that at least Tyr291 and Val292 are directly involved in the peptide–LcrH complex. This is consistent with the view that hydrophobic residues within the amphipathic domain of YopD are required for LcrH binding [13]. The fact that amino acid replacements of Tyr291 and Val292 did reduce LcrH binding to YopD in the yeast two-hybrid assay also corroborates with this study [13]. Moreover, the chemical shift behaviour of YopD 278)300 in the presence of LcrH is similar to the behaviour when peptide aggregates increase in size. In particular, assigned amide protons in the YopD 278)300 peptide experienced an upfield frequency shift upon LcrH addition and this same trend is also observed when the 2,2,2-trifluoroethanol concentration is lowered. These findings imply that the peptide aggregate mimics the interaction between the peptide and LcrH. Therefore, analysis of LcrH binding on the dynamics of peptide aggregation warrants further investigation. CONCLUSIONS In this study, we have initialized a means to understand the role of the YopD–LcrH complex in Yersinia pathogenesis by determining the a helical structure of the biologically relevant C-terminal amphipathic domain of YopD. Impor- tantly, this domain precedes a type I b turn that is essential for stability of the helical structure. An interesting feature of the peptide encompassing this domain was its tendency to form small aggregates that were likely composed of a helices layered in an antiparallel manner. In addition, we confirmed that this domain interacts with LcrH through hydrophobic interactions that include at least two residues, Tyr291 and Fig. 10. 1D 1 H NMR experiment at 20 °C of a 0.25 m M YopD 278)300 sample in 10% 2,2,2-trifluoroethanol and 50 m M phosphate buffer at pH 6.3. (A) In the absence of purified LcrH, and (B) In the presence of purified LcrH to give a peptide/protein molar ratio of 2 : 1. The peptide residues Tyr291 and Val292 identified to bind LcrH are indi- cated. 3666 T. Tengel et al. (Eur. J. Biochem. 269) Ó FEBS 2002 Val292. Although our laboratory and others have recently proposed new roles for TTSS chaperones, it is clear that chaperone-substrate complexes are fundamental to the process of functional type III secretion and ultimately for successful infection by the bacterium. Based on the recent crystal structure determination of a TTSS chaperone/ effector protein complex from Salmonella spp., it is likely that at least one function of chaperones is to maintain their cognate partner in an elongated unfolded state, presumably as a prerequisite for efficient secretion [50]. We have begun to reveal the secrets of a biologically relevant YopD–LcrH complex in Yersinia infections. However, a detailed struc- tural study of this intriguing TTSS complex is ongoing. ACKNOWLEDGEMENTS This work was supported by grants from the Swedish Medical Research Council, Swedish Natural Science Research Council and Swedish Foundation for Strategic Research. We are indebted to Hans Wolf- Watz for insightful discussions, financial assistance and critical reading of this manuscript. We also thank Peter Stenlund and Gull-Britt Trogen for excellent technical assistance. REFERENCES 1. Hueck, C.J. (1998) Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Biol. Rev. 62, 379–433. 2. Schesser, K., Francis, M.S., Forsberg, A ˚ . & Wolf-Watz, H. (2000) Type III secretion systems in animal- and plant-interaction bac- teria. In Cellular Microbiology (Cossart, P., Normark, S. & Rap- pouli, R., eds), pp. 239–236. ASM Press, Washington, DC. 3. Cornelis, G.R., Boland, A., Boyd, A.P., Geuijen, C., Iriarte, M., Neyt, C., Sory, M.P. & Stainier, I. (1998) The virulence plasmid of Yersinia, an antihost genome. Microbiol. Mol. Biol. Rev. 62, 1315– 1352. 4. Cornelis, G.R. (2000) Molecular and cell biology aspects of pla- gue. Proc. Natl Acad. Sci. USA 97, 8778–8783. 5. Rosqvist, R., Bolin, I. & Wolf-Watz, H. (1988) Inhibition of phagocytosis in Yersinia pseudotuberculosis: a virulence plasmid- encoded ability involving the Yop2b protein. Infect. Immun. 56, 2139–2143. 6. Hanski, C., Kutschka, U., Schmoranzer, H.P., Naumann, M., Stallmach, A., Hahn, H., Menge, H. & Riecken, E.O. (1989) Immunohistochemical and electron microscopic study of interac- tion of Yersinia enterocolitica serotype O8 with intestinal mucosa during experimental enteritis. Infect. Immun. 57, 673–678. 7. Simonet, M., Richard, S. & Berche, P. (1990) Electron microscopic evidence for in vivo extracellular localization of Yersinia pseudo- tuberculosis harboring the pYV plasmid. Infect. Immun. 58, 841– 845. 8. Fa ¨ llman, M., Persson, C., Schesser, K. & Wolf-Watz, H. (1998) Bidirectional signaling between Yersinia and its target cell. Folia. Microbiol. 43, 263–273. 9. Holmstro ¨ m,A.,Olsson,J.,Cherepanov,P.,Maier,E.,Nordfelth, R., Pettersson, J., Benz, R., Wolf-Watz, H. & Forsberg, A ˚ . (2001) LcrV is a channel size-determining component of the Yop effector translocon of Yersinia. Mol. Microbiol. 39, 620–632. 10. Francis, M.S. & Wolf-Watz, H. (1998) YopD of Yersinia pseud- otuberculosis is translocated into the cytosol of HeLa epithelial cells: evidence of a structural domain necessary for translocation. Mol. Microbiol. 29, 799–813. 11. Williams, A.W. & Straley, S.C. (1998) YopD of Yersinia pestis plays a role in negative regulation of the low-calcium response in addition to its role in translocation of Yops. J. Bacteriol. 180, 350– 358. 12. Wattiau, P., Bernier, B., Deslee, P., Michiels, T. & Cornelis, G.R. (1994) Individual chaperones required for Yop secretion by Yersinia. Proc. Natl Acad. Sci. USA 91, 10493–10497. 13. Francis, M.S., Aili, M., Wiklund, M.L. & Wolf-Watz, H. (2000) A study of the YopD–LcrH interaction from Yersinia pseudotuber- culosis reveals a role for hydrophobic residues within the amphi- pathic domain of YopD. Mol. Microbiol. 38, 85–102. 14. Francis, M.S., Lloyd, S.A. & Wolf-Watz, H. (2001) The type III secretion chaperone LcrH co-operates with YopD to establish a negative, regulatory loop for control of Yop synthesis in Yersinia pseudotuberculosis. Mol. Microbiol. 42, 1075–1094. 15. Anderson, D.M., Ramamurthi, K.S., Tam, C. & Schneewind, O. (2002) YopD and LcrH regulate expression of Yersinia enterocolitica YopQ by a posttranscriptional mechanism and bind to yopQ RNA. J. Bacteriol. 184, 1287–1295. 16. Lloyd, S.A., Forsberg, A ˚ ., Wolf-Watz, H. & Francis, M.S. (2001) Targeting exported substrates to the Yersinia TTSS: different functions for different signals? Trends Microbiol. 8, 367–371. 17. Francis, M.S., Wolf-Watz, H. & Forsberg, A ˚ . (2002) Regulation of type III secretion systems. Curr. Opin. Microbiol. 5, 166–172. 18. Rosqvist, R., Ha ˚ kansson, S., Forsberg, A ˚ . & Wolf-Watz, H. (1995) Functional conservation of the secretion and translocation machinery for virulence proteins of Yersiniae, Salmonellae and Shigellae. EMBO J. 14, 4187–4195. 19. Frithz-Lindsten, E., Holmstro ¨ m,A.,Jacobsson,L.,Soltani,M., Olsson, J., Rosqvist, R. & Forsberg, A ˚ . (1998) Functional con- servation of the effector protein translocators PopB/YopB and PopD/YopD of Pseudomonas aeruginosa and Yersinia pseudo- tuberculosis. Mol. Microbiol. 29, 1155–1165. 20. Michiels, T., Wattiau, P., Brasseur, R., Ruysschaert, J.M. & Cornelis, G. (1990) Secretion of Yop proteins by Yersiniae. Infect. Immun. 58, 2840–2849. 21. Pristovsek, P., Lucke, C., Reincke, B., Ludwig, B. & Ruterjans, H. (2000) Solution structure of the functional domain of Paracoccus denitrificans cytochrome c552 in the reduced state. Eur. J. Biochem. 267, 4205–4212. 22.Yeagle,P.L.,Salloum,A.,Chopra,A.,Bhawsar,N.,Ali,L., Kuzmanovski, G., Alderfer, J.L. & Albert, A.D. (2000) Structures of the intradiskal loops and amino terminus of the G-protein receptor, rhodopsin. J. Pept. Res. 55, 455–465. 23. Rance, M., Sorensen, O.W., Bodenhausen, G., Wagner, G., Ernst, R.R. & Wu ¨ thrich, K. (1983) Improved spectral resolution in COSY 1H NMR spectra of proteins via double quantum filtering. Biochem. Biophys. Res. Commun. 117, 479–485. 24. Braunschweiler, L. & Ernst, R.R. (1983) Coherence transfer by isotropic mixing: application to proton correlation spectroscopy. J. Magn. Reson. 53, 521–528. 25. Jeener, J., Meier, B.H., Bachmann, P. & Ernst, R.R. (1979) Investigation of exchange processes by two-dimensional NMR spectroscopy. J. Chem. Phys. 71, 4546–4553. 26. Kay, L.E., Keifer, P. & Saarinen, T. (1992) Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114, 10663–10665. 27. Piotto, M., Saudek, V. & Sklena ´ r, V. (1992) Gradient-tailored excitation for single-quantum NMR spectroscopy of aqueous solutions. J. Biomol. NMR 2, 661–665. 28. Wishart, D.S., Bigam, C.G., Yao, J., Abildgaard, F., Dyson, H.J., Oldfield, E., Markely, J.L. & Sykes, B.D. (1995) 1 H, 13 Cand 15 N chemical shift referencing in biomolecular NMR. J. Biomol. NMR 6, 135–140. 29. Borgias, B.A. & James, T.L. (1990) MARDIGRAS – a procedure for matrix analysis of relaxation for discerning geometry of an aqueous structure. J. Magn. Reson. 87, 475–487. 30. Wu ¨ thrich, K., Billerter, M. & Braun, W. (1983) Pseudo-structures for the 20 common amino acids for use in studies of protein conformations by measurements of intramolecular proton–proton Ó FEBS 2002 Tertiary structure of the YopD amphipathic domain (Eur. J. Biochem. 269) 3667 distance constraints with nuclear magnetic resonance. J. Mol. Biol. 169, 949–961. 31. Woessner, D.E. (1962) Spin relaxation processes in a two-proton system undergoing anisotropic reorientation. J. Chem. Phys. 36, 1–4. 32. Cornilescu, G., Delaglio, F. & Bax, A. (1999) Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J. Biomol. NMR 13, 289–302. 33. Bru ¨ nger, A.T. (1992) X-PLOR, Version 3.1. A System for X-Ray Crystallography and NMR. Yale Universtity Press, New Haven, CT. 34. Nilges, M., Clore, G.M. & Gronenborn, A.M. (1988) Determi- nation of three-dimensional structures of proteins from inter- proton distance data by dynamical simulated annealing. FEBS Lett. 239, 129–136. 35. Koradi,R.,Billeter,M.&Wu ¨ thrich,K.(1996)MOLMOL:a program for display and analysis of macromolecular structures. J. Mol. Graph. 14, 51–55. 36. Humphrey, W., Dalke, A. & Schulten, K. (1996) VMD: visual molecular dynamics. J. Mol. Graph. 14, 33–38. 37. Laskowski, R.A., Rullmannn, J.A., MacArthur, M.W., Kaptein, R. & Thornton, J.M. (1996) AQUA and PROCHECK-NMR: programs for checking the quality of protein structures solved by NMR. J. Biomol. NMR 8, 477–486. 38. Ha ˚ kansson, S., Bergman, T., Vanooteghem, J.C., Cornelis, G. & Wolf-Watz, H. (1993) YopB and YopD constitute a novel class of Yersinia Yop proteins. Infect. Immun. 61, 71–80. 39. Slupsky, C.M., Kay, C.M., Reinach, F.C., Smillie, L.B. & Sykes, B.D. (1995) Calcium-induced dimerization of troponin C: mode of interaction and use of trifluoroethanol as a denaturant of qua- ternary structure. Biochemistry 34, 7365–7375. 40. Slupsky, C.M., Reinach, F.C., Smillie, L.B. & Sykes, B.D. (1995) Solution secondary structure of calcium-saturated troponin C monomer determined by multidimensional heteronuclear NMR spectroscopy. Protein. Sci. 4, 1279–1290. 41. Buck, M. (1998) Trifluoroethanol and colleagues: cosolvents come of age. Recent studies with peptides and proteins. Q. Rev. Biophys. 31, 297–355. 42. Wu ¨ thrich, K. (1986) NMR of Proteins and Nucleic Acids. John Wiley & Sons, New York. 43. Wishart, D.S., Sykes, B.D. & Richards, F.M. (1991) Relationship between nuclear-magnetic-resonance chemical-shift and protein secondary structure. J. Mol. Biol. 222, 311–333. 44. Forood, B., Feliciano, E.J. & Nambiar, K.P. (1993) Stabilization of alpha-helical structures in short peptides via end capping. Proc. Natl Acad. Sci. USA 90, 838–842. 45. Kaiser, B., Laesecke, A. & Schmeck, M. (1989) Experimental study and correlation of the viscosity of 2,2,2-trifluoroethanol (TFE)-water mixtures. Int. J. Thermophys. 10, 713–726. 46. Darwin, K.H. & Miller, V.L. (2001) Type III secretion chaperone- dependent regulation: activation of virulence genes by SicA and InvF in Salmonella typhimurium. EMBO J. 20, 1850–1862. 47. Wattiau, P. & Cornelis, G.R. (1993) SycE, a chaperone-like pro- tein of Yersinia enterocolitica involved in the secretion of YopE. Mol. Microbiol. 8, 123–131. 48. Cheng, L.W. & Schneewind, O. (1999) Yersinia enterocolitica type III secretion. On the role of SycE in targeting YopE into HeLa cells. J. Biol. Chem. 274, 22102–22108. 49. Bennett, J., Thomas, J., Fraser, G. & Hughes, C. (2001) Substrate complexes and domain organization of the Salmonella flagellar export chaperones FlgN and FliT. Mol. Microbiol. 39, 781–791. 50. Stebbins, C.E. & Galan, J.E. (2001) Maintenance of an unfolded polypeptide by a cognate chaperone in bacterial type III secretion. Nature 414, 77–81. 51. Birtalan, S. & Ghosh, P. (2001) Structure of the Yersinia type III secretory system chaperone SycE. Nat. Struct. Biol. 8, 974–978. SUPPLEMENTARY MATERIAL The following material is available from http://www.black- well-science.com/products/journals/suppmat/EJB/EJB3051/ EJB3051sm.htm Table S2. Phi dihedral angles for YopD (278)300) . Table S1. Chemical shifts (p.p.m.) of YopD (278)300) . 3668 T. Tengel et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . Conformational analysis by CD and NMR spectroscopy of a peptide encompassing the amphipathic domain of YopD from Yersinia Tobias Tengel 1 , Ingmar Sethson 1 and Matthew S. Francis 2 1 Departments. putative amphipathic domain. This strategy was advantageous because full length YopD is susceptible to aggregation [20] and the amphipathic domain is clearly biologically relevant [10,13]. The. studies Computer analysis of YopD primary sequence predicts a central hydrophobic membrane spanning domain and a C-terminal amphipathic domain (Fig. 1A) [38]. This latter region can be presented on a helical