Solution Structures of Phenol-Soluble Modulins α1 ... - ACS Publications

Aug 15, 2016 - Phenomenex) was used with a 1 mL injection volume, a flow ... Figure 1. Alignment of amino acid sequences of (A) PSMα1, PSMα2,. PSMα...
0 downloads 9 Views 4MB Size
Download PDF
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article pubs.acs.org/biochemistry

Solution Structures of Phenol-Soluble Modulins α1, α3, and β2, Virulence Factors from Staphylococcus aureus Kaitlyn M. Towle, Christopher T. Lohans, Mark Miskolzie, Jeella Z. Acedo, Marco J. van Belkum, and John C. Vederas* Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada T6G 2G2 S Supporting Information *

ABSTRACT: Phenol-soluble modulins (PSMs) are peptide virulence factors produced by staphylococci. These peptides contribute to the overall pathogenicity of these bacteria, eliciting multiple immune responses from host cells. Many of the α-type PSMs exhibit cytolytic properties and are able to lyse particular eukaryotic cells, including erythrocytes, neutrophils, and leukocytes. In addition, they also appear to contribute to the protection of the bacterial cell from the host immune response through biofilm formation and detachment. In this study, three of these peptide toxins, PSMs α1, α3, and β2, normally produced by Staphylococcus aureus, have been synthesized using solid-supported peptide synthesis (SPPS) (PSMα1 and PSMα3) or made by heterologous expression in Escherichia coli (PSMβ2). Their three-dimensional structures were elucidated using nuclear magnetic resonance spectroscopy. PSMα1 and PSMα3 each consist of a single amphipathic helix with a slight bend near the N- and C-termini, respectively. PSMβ2 contains three amphipathic helices, which fold to produce a “v-like” shape between α-helix 2 and α-helix 3, with α-helix 1 folded over such that it is perpendicular to α-helix 3. The availability of three-dimensional structures permits spatial analysis of features and residues proposed to control the biological activity of these peptide toxins.

S

the role of individual amino acids in eliciting an immune response.9 Many of the PSMs have been shown to form and structure biofilms, surface-attached agglomerations that create a sticky coating around the bacterial cell.10−12 The delivery of nutrients to the biofilm leads to the growth and eventual detachment of the biofilms. Biofilm detachment results in clusters of cells breaking off, allowing the bacteria to spread to new infection sites.10,13 Several of the α-type PSMs have been shown to exhibit cytolytic activity against a variety of eukaryotic cells, including monocytes, leukocytes, erythrocytes, endothelial and epithelial cells, and osteoblasts.6,14,15 Some PSMs from S. aureus have been reported to have antimicrobial activity. In particular, PSMα1 and PSMα2 demonstrated a higher level of antimicrobial activity when the N-terminal formylmethionine and glycine residues were removed.16,17 It has been shown that the antimicrobial effects of the PSMs do not harm the producing staphylococci through interaction with the PSM transporter Pmt.16,18 Pmt is an ABC transporter that is responsible for the export of all PSMs.18 Deletion of the genes encoding Pmt resulted in a mutation in the agr locus, which encodes the regulatory quorum-sensing system and is involved in the expression of the PSM peptides.19 This indicates that Pmt is important for self-immunity against the

taphylococci are Gram-positive bacteria that commonly infect skin and soft tissues but are also the leading cause of osteomyelitis.1,2 Although there are many different species of staphylococci, Staphylococcus aureus is of particular interest because of its notoriety as a human pathogen.2,3 S. aureus can cause minor infections such as abscesses and boils but is also responsible for much more serious diseases such as cellulitis and endocarditis. When the bacterium breaches the mucosal barrier, S. aureus can also cause toxic shock and in severe cases sepsis.3 The pathogenicity of these bacterial infections is strongly related to the production of virulence factors.4 S. aureus produces a wide variety of virulence factors, but of particular interest are the virulence factors known as phenolsoluble modulins (PSMs).5 PSMs have garnered much attention because of their role in biofilm formation and detachment that protects the bacteria from human immune cells. In addition their cytolytic activity toward a wide variety of eukaryotic cells, they can cause tissue damage.6 PSMs produced by S. aureus are categorized by length; the α-type PSMs are shorter and contain ∼20−25 amino acids, whereas the β-type PSMs are longer and contain ∼44 amino acids.7 Both α-type and β-type PSMs have been shown to have a wide variety of functions on a human host cell, which are primarily regulated through their interaction with human formyl peptide receptor 2 (FPR2).8 Though the exact structural details for the interaction between the PSMs and FPR2 are not known, extensive studies of PSMα3 have led to an analysis of © XXXX American Chemical Society

Received: June 15, 2016 Revised: July 29, 2016

A

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry PSMs.18 Pmt was found to be present in all Staphylococcus species and absent from other bacteria, suggesting that Pmt is specific for staphylococci, as is PSM production.18 PSMs are produced and secreted without the aid of a leader sequence, nor are they cotranscribed with a dedicated immunity protein.20 Additionally, they are located in the core of the genome or in highly conserved genomic islands, which means they are produced across a wide variety of staphylococcal species.20 PSMγ, also known as δ-toxin, is a well-characterized virulence factor produced by S. aureus, and its structure has been elucidated.21 δ-Toxin, later recategorized to be included in the α-type PSMs, is frequently used as a model for the α-type PSMs.7,9 On the basis of helical wheel projections, CD spectroscopy, and the data available for δ-toxin, it was proposed that the α-type PSMs are comprised of one amphipathic α-helix and that the mode of action for these α-type PSMs is similar to that which has been elucidated for δ-toxin.9,22,23 The amphipathic nature of these peptides is thought to allow them to act as biological detergents, allowing them to disrupt membranes without the need for a receptor.5 Both α-type and β-type PSMs have been characterized via circular dichroism (CD) spectroscopy and helical wheel diagrams, giving a rough picture of the structure of these peptides.20,23 However, multiple studies have shown the importance of understanding the role of the various amino acid side chains in the overall biological activity of peptides.9,24 A shared structural motif, such as a single amphipathic α-helix found in the α-type PSMs, does not exclusively determine the biological activity of a peptide. Variations in the amino acid sequence may result in a difference in the overall hydrophobicity of a peptide or the electrostatic potential of its surface, which could lead to differing biological activities, regardless of a similar secondary structure. The differing levels of cytolytic activity exhibited among α-type PSMs are yet another example of peptides, which adopt similar secondary structures, that have vastly disparate levels of activity. The cytolytic properties of the α-type PSMs produced by S. aureus have been well characterized.23 PSMα3 has the greatest cytolytic activity, whereas PSMα1 displays one of the lowest levels of cytolysis within the α-type PSMs.9,23 Interestingly, PSMα1 and PSMα2 share more sequence homology with one another than with PSMα3 or PSMα4 (Figure 1A).23 Although PSMβ2 and PSMβ1 share a high degree of sequence homology (Figure 1B), PSMβ2 exhibits an overall neutral charge whereas PSMβ1 is slightly anionic.23 Determination of the threedimensional solution structures, in combination with the data already acquired for these peptide toxins, is essential for

detailed understanding of their biological function. It initially seemed that some PSMs (e.g., PSMβ) could potentially resemble antimicrobial peptides produced by bacteria,25 including leaderless bacteriocins whose three-dimensional structures have a saposin-like fold, such as enterocins 7A and 7B,26 lacticin Q, and aureocin A53.24 Here we report the NMR solution structures of two α-type PSMs, PSMα1 and PSMα3, and of one β-type PSM, PSMβ2.

■

MATERIALS AND METHODS Synthesis of PSMα1 and PSMα3. PSMα1 and PSMα3 were prepared through manual solid phase peptide synthesis (SPPS). The peptides were synthesized using a Wang resin on a 0.1 mmol scale utilizing a jacketed peptide synthesis reaction vessel. Coupling reactions were conducted using N-fluorenylmethyloxycarbonyl (Fmoc)-protected amino acids (5 equiv relative to resin loading) and activated with 1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate (HATU) (4.9 equiv relative to resin loading) and diisopropylethylamine (DIPEA) (5 equiv relative to resin loading). Subsequent deprotection of the coupled amino acid was achieved using a 20% piperidine solution in dimethylformamide (DMF). Because of ineffective deprotections of the last five and six residues (PSMα1 and PSMα3, respectively), deprotections were conducted using 20% piperidine in the “magic mixture” (1% Triton X-100 in a 1:1:1 DMF/NMP/DCM mixture at 60 °C).27 The coupling and deprotection reactions were monitored using the Kaiser test and matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS).28 The peptides were cleaved from the resin using a 9.5:0.25:0.25 trifluoroacetic acid:triisopropylsilane:water (TFA:TIPS:H2O) ratio. Upon cleavage of the peptides from the resin, the crude peptide was dissolved in the minimum amount of dimethyl sulfoxide (DMSO) and diluted with water to give a volume:volume percent of 5% DMSO in water. The solutions of crude peptide were partially purified through dialysis against water (20 h, 25 °C; molecular weight cutoff of 1000, Spectra/Por 7) followed by multiple iterations of reverse phase high-performance liquid chromatography (RP-HPLC) until the peptides were determined to be pure via MALDI-TOF MS. Briefly, for both PSMα1 and PSMα3, an analytical scale column (3.6 μm particle size, 200 Å, 250 mm × 4.6 mm, Aeris Widepore C4, Phenomenex) was used with a 1 mL injection volume, a flow rate of 1.5 mL/min, and a detection wavelength set at 220 nm. A gradient of solvent A (0.1% TFA in water) and solvent B (0.1% TFA in isopropyl alcohol) was used. For PSMα1, the level of solvent B was initially set to 40% and held for 5 min. Over the next 20 min, the level of solvent B was increased to 95% and held for an additional 5 min. For PSMα3, the level of solvent B was initially set to 5% and held for 5 min. Over the next 40 min, the level of solvent B was gradually increased to 60% and then sharply increased to 95% over 5 min. The level of solvent B was held at 95% for 15 min. According to these gradients, PSMα1 and PSMα3 co-eluted with truncated peptides at 30 and 56 min, respectively. The peaks containing PSMα1 and PSMα3 were collected and injected for a second time using the same gradient as described above. Fractions of PSMα1 determined to be pure through MALDI-TOF MS were concentrated, lyophilized, and stored at −20 °C (Figure S1). PSMα3 was collected and injected for a third time following the same gradient as described above, and the fractions determined to be most pure by MALDI-TOF MS were concentrated under

Figure 1. Alignment of amino acid sequences of (A) PSMα1, PSMα2, PSMα3, and PSMα4 and (B) PSMβ1 and PSMβ2. Sequence alignments were generated using Clustal Omega;36 conserved, conservative, and semiconservative substitutions are indicated using asterisks, colons, and periods, respectively. B

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry vacuum, lyophilized, and stored at −20 °C (Figure S2). Yields of 0.88 and 0.99 mg were obtained for PSMα1 and PSMα3, respectively. Construction of Expression Vectors. A gene sequence encoding PSMβ2 codon-optimized for Escherichia coli expression was purchased from BioBasic Inc. The gene was cloned into the pET SUMO (small ubiquitin-like modifier) expression vector according to the manufacturer’s instructions (Invitrogen). Sequencing of the DNA construct confirmed that the sequence was correct and in frame with the His-tagged SUMO fusion protein. The resulting pET SUMO-PSMβ2 plasmid was transformed into competent E. coli BL21(DE3) cells according to the manufacturer’s instructions. Expression of His-Tagged SUMO-PSMβ2. Protein expression in E. coli was performed in 1 L of Terrific Broth [12 g of tryptone, 24 g of yeast extract, 4 mL of glycerol, 2.31 g of potassium dihydrogen phosphate, and 12.64 g of potassium hydrogen phosphate (pH 7.4)]. Medium was inoculated [1% (v/v)] with an overnight starter culture of E. coli BL21(DE3) (pET SUMO-PSMβ2) and grown at 37 °C while being shaken at 225 rpm to an optical density (OD600) of 0.8−1.0. Kanamycin was used at a concentration of 50 μg/mL for selective pressure. Protein expression was induced by adding isopropyl β-D-1-thiogalactopyranoside to a final concentration of 1 mM, and the culture was incubated at 30 °C for 24 h. The cells were then harvested (5000g for 15 min at 4 °C), and the cell pellet was suspended in lysis buffer composed of 50 mM NaH2PO4 (pH 8.0), 500 mM NaCl, 10 mM imidazole, and 1% glycerol. The suspension was passed once through a Constant Systems Cell Disruptor, model TS (Constant Systems, Ltd.), operated at 20000 psi. The lysate was centrifuged (15000g for 30 min at 4 °C), and the supernatant, which contained the fusion protein, was collected for subsequent purification. Purification of Fusion Protein. The lysate supernatant was mixed with 2.0 mL of Ni-NTA agarose (Qiagen) resin and shaken for 1 h at 8 °C. It was then loaded onto a fritted column and allowed to flow by gravity. The resin was washed with 75 mL of buffer A [50 mM NaH2PO4, 500 mM NaCl, and 20 mM imidazole (pH 8.0)]. The fusion protein was eluted with 50 mL of buffer B [50 mM NaH2PO4, 500 mM NaCl, and 400 mM imidazole (pH 8.0)] and dialyzed for 3 h at 8 °C against a 20 mM Tris-HCl buffer (pH 8.0) with 150 mM NaCl. The purification was monitored by sodium dodecyl sulfate− polyacrylamide gel electrophoresis (SDS−PAGE). Cleavage of Fusion Protein. His-tagged SUMO protease (McLab, South San Francisco, CA) was used to cleave the SUMO tag from the fusion protein as suggested by the manufacturer. Briefly, a 200 μL cleavage cocktail contained 20 μg of fusion protein, 5 units of the His-tagged SUMO protease, 20 μL of 10× SUMO protease buffer [500 mM Tris-HCl (pH 8.0), 2% Igepal CA-360 (Sigma), and 10 mM dithiothreitol], and 150 mM NaCl. After cleavage (20 h, 25 °C), 1 mL of NiNTA agarose (Qiagen) resin was used to remove the Histagged SUMO and SUMO protease. PSMβ2 was further purified by RP-HPLC using a preparative scale C18 column (5 μm particle size, 100 Å pore size, 21.2 mm × 250 mm, Luna AXIA). The detector and flow rate were set at 220 nm and 8 mL/min, respectively. Ten milliliters of sample was injected per run. A gradient of solvent A (0.1% TFA in water) and solvent B (0.1% TFA in acetonitrile) was used. The level of solvent B was initially set at 30% for 10 min, gradually increased to 95% for 30 min, and maintained at 95% for 10 min. Fractions containing PSMβ2, which co-eluted with impurities at 41 min, were

combined and concentrated under vacuum, lyophilized, redissolved, and re-injected for a second round of RP-HPLC. The second round of RP-HPLC used a semipreparative C18 column (5 μm particle size, 100 Å, 250 mm × 10 mm, Luna model no. 517180-1, Phenomenex). The flow rate and detector were set to 5 mL/min and 220 nm, respectively. Ten milliliters of sample was injected per run using the same gradient method described above. The fractions containing PSMβ2, which eluted after 32 min, were collected, concentrated under vacuum, lyophilized, and stored at −20 °C (Figure S3). Approximately 2−2.5 mg of peptide was obtained per liter of culture. MALDI-TOF Mass Spectrometry. Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectra were acquired using a Perspective Biosystems Voyager Elite MALDI-TOF mass spectrometer in reflectron, delayedextraction, positive ion mode. Samples were prepared using sinapinic acid as the matrix. NMR Spectroscopy. PSMα1, PSMα3, and PSMβ2 were each dissolved in 500 μL of 50% d3-TFE and 50% water to give final concentrations of 0.77, 0.75, and 1.9 mM, respectively. All samples were run using a standard 5 mm NMR tube, and with 4,4-dimethyl-4-silapentane-1-sulfonic acid [0.01% (w/v)] as a reference. One-dimensional 1H NMR and two-dimensional homonuclear 1H−1H TOCSY and NOESY spectra were acquired for PSMα1, PSMα3, and PSMβ2. Experimental details for each NMR experiment can be found in Table S1. Briefly, the NMR spectra for PSMα1 were recorded on a 600 MHz Varian VNMRS spectrometer equipped with a z-axis pulsed field gradient triple-resonance HCN probe. NMR spectra for PSMα3 and PSMβ2 were recorded on a 700 MHz Varian VNMRS spectrometer using a z-axis pulsed field gradient tripleresonance HCN cold probe. The acquisition software was VNMRJ 4.2A on both spectrometers. As resonance overlap was observed in the amide region, temperature scans of PSMα3 and PSMβ2 revealed a greater dispersion of peaks at elevated temperatures. Therefore, the temperatures at which the NMR spectra were recorded were 25, 37, and 40 °C for PSMα1, PSMα3, and PSMβ2, respectively. Water suppression was achieved by presaturation during the relaxation delay for the spectra of PSMα1 and PSMα3. Water suppression in the spectra of PSMβ2 was achieved using excitation sculpting.29 The spectra were analyzed using NMRPipe and NMRView, followed by manual assignment of chemical shifts as described below. Full chemical shift assignments can be found in Tables S2−S4. Circular Dichroism Spectroscopy. PSMα1, PSMα3, and PSMβ2 were dissolved in 50% d3-TFE at concentrations of 0.88, 0.99, and 3.07 mg/mL respectively. Spectra were recorded on an OLIS DSM 17 (Olis) CD spectrophotometer at 25, 37, and 40 °C (for PSMα1, PSMα3, and PSMβ2, respectively) in a 0.2 mm quartz cell. Samples were scanned five times from 250 to 185 nm. The percent α-helicity was calculated as (3000 − θ222)/39000 × 100.30 Structure Calculations. The structures of PSMα1, PSMα3, and PSMβ2 were calculated using CYANA 2.1.31 NOE crosspeaks were almost entirely automatically assigned by CYANA, and minimal manually assigned cross-peaks were provided. For PSMα1, 636 cross-peak NOEs (0 long-range, 99 mediumrange, and 538 short-range) were used in the final structure calculation. Chemical shift assignments for PSMα1 were deposited in the Biological Magnetic Resonance Bank (BMRB) (accession number 30109), and coordinates for the structure were deposited in the Protein Data Bank (PDB) C

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry (accession number 5KHB). For PSMα3, 553 cross-peak NOEs (0 long-range, 101 medium-range, and 452 short-range) were used in the final structure calculation. Chemical shift assignments for PSMα3 were deposited in the BMRB (accession number 30106), and coordinates for the structure were deposited in the PDB (accession number 5KGY). PSMβ2 had 1320 cross-peak NOEs (34 long-range, 246 medium-range, and 1040 short-range) that were used in the final structure calculation. Chemical shift assignments for PSMβ2 were deposited in the BMRB (accession number 30107), and coordinates for the structure were deposited in the PDB (accession number 5KGZ). Backbone overlays of the 20 lowest-energy conformers of PSMα1, PSMα3, and PSMβ2 can be found in Figure S4.

■

RESULTS AND DISCUSSION Production and Purification of PSMα1, PSMα3, and PSMβ2. PSMα1, PSMα3, and PSMβ2 from S. aureus are virulence factors that are 21, 22, and 44 amino acids long, respectively.6 Fmoc-based SPPS was used to synthesize both PSMα1 and PSMα3. Upon cleavage of the peptides from the resin, the peptides were purified using a combination of dialysis and multiple rounds of RP-HPLC. Attempts to chemically synthesize PSMβ2 were unsuccessful; therefore, PSMβ2 was overexpressed in E. coli BL21(DE3) as a His-tagged SUMO fusion protein and partially purified using Ni-NTA chromatography. Subsequent purification of the cleaved PSMβ2 peptide after SUMO protease treatment was completed using a second Ni-NTA column and RP-HPLC. Yields of 0.88 mg, 0.99 mg, and 2−2.5 mg/L were obtained for PSMα1, PSMα3, and PSMβ2, respectively. Consistent with the natural peptides from S. aureus, the chemically synthesized PSMα1 and PSMα3 contained an N-terminal formylmethionine residue. PSMβ2, produced through heterologous expression in E. coli, did not contain the formylmethionine. While the formylmethionine may be important for the interaction of PSMs with some FPRs, it has also been shown that FPR2 also recognizes the unformylated PSMs.32 Solubility and Solvent Selection for NMR Studies Using Circular Dichroism Spectroscopy. The solubility of these peptides was limited, as all three were found to be insoluble in 100% water and were only sparingly soluble in 25% d3-TFE. They were, however, soluble in 50% d3-TFE; hence, their solution structures in this solvent system were analyzed using CD spectroscopy. PSMα1, PSMα3, and PSMβ2 were calculated to have moderate levels of α-helicity (25.8, 37.5, and 22.2%, respectively) in this structure-inducing solvent (Figure 2A,B). Temperature did not greatly affect the level of α-helicity observed in PSMα3 and PSMβ2 (39.5% at 25 °C vs 37.5% at 37 °C and 25.9% at 25 °C vs 22.2% at 40 °C, respectively). Though the calculated percent α-helicity was only moderate, a one-dimensional 1H NMR spectrum was obtained for all three peptides. While the amide region of PSMα1 appeared to have sufficient separation of proton signals at 25 °C, elevated temperatures (37 and 40 °C) were required to obtain sufficient separation of peaks for PSMα3 and PSMβ2, respectively. NMR Solution Structures of PSMα1, PSMα3, and PSMβ2. Hydrogen chemical shift assignments for all three peptides were obtained through manual identification of spin systems in the TOCSY data set. The spin systems were ordered using the standard interresidue NHNH(i, i + 1), αNH(i, i + 1), and βNH(i, i + 1) NOE cross-peaks from the NOESY data sets. Peak lists comprising all NOE cross-peaks (majority unas-

Figure 2. CD spectra of (A) PSMα1 and PSMα3 and (B) PSMβ2. Spectra were acquired in 50% d3-trifluoroethanol at 25, 37, and 40 °C, respectively.

signed) along with the chemical shift assignments were provided to CYANA 2.1 for structure calculations.31 The number of NOEs used by CYANA to calculate the solution structures of PSMα1, PSMα3, and PSMβ2 is summarized in Table 1. PSMα3 and PSMβ2 showed some variability in the termini of the peptide, which can be seen in the 20 lowestenergy conformers shown in panels B and C of Figure S3. The root-mean-square deviation (rmsd) values listed in Table 1 are reflective of residues 1−20 and residues 3−41 of PSMα3 and PSMβ2, respectively. The lack of long-range NOEs in the αtype PSMs is consistent with the primarily linear α-helix backbone. PSMβ2 showed relatively few long-range NOEs compared to other structured peptides of similar length. The long-range NOEs were primarily found between hydrophobic residues, creating a pocket of hydrophobicity on the inner part of the structure. For all three peptides, the Ramachandran plots show that there are no unfavorable ϕ or ψ backbone angles (Figure S5A−C). Structurally, PSMα1 and PSMα3 are similar to one another. Each contains a single α-helix with a slight bend in it (panels A and B of Figure 3, respectively). The lack of long-range NOEs supports the single α-helix with no significant kinks. The amphipathic helix in PSMα1 runs from residue 2 to 19. The bend in the helix occurs at residue 6, resulting in a slight curve to the backbone structure of the peptide. The hydrophobicity map of the surface reveals that PSMα1 has one hydrophobic and one hydrophilic side (Figure 4A). The helix in PSMα3 runs from residue 2 to 20. Like PSMα1, PSMα3 is amphipathic, and the hydrophobicity surface map reveals one hydrophobic side and one hydrophilic side (Figure 4B). Interestingly, the slight curve in the secondary structure of PSMα3 appears to have a greater impact on the surface structure and results in a slightly D

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry Table 1. Summary of NMR Solution Structure Calculations of PSMα1, PSMα3, and PSMβ2 total no. of NOE peak assignments short-range (|i − j| ≤ 1) medium-range (1 < |i − j| < 5) long-range (|i − j| ≥ 5) average target function value rmsd for full peptide backbone atoms (Å) heavy atoms (Å)

PSMα1

PSMα3

PSMβ2

636 538 99 0 2.10 × 10−3 ± 2.66 × 10−4

553 452 101 0 0.00 × 10−3 ± 0.00 × 10−4

1320 1040 246 34 1.99 × 10−2 ± 5.66 × 10−3

0.77 ± 0.17 1.23 ± 0.15

0.64 ± 0.19 1.28 ± 0.14

0.80 ± 0.21 1.23 ± 0.25

Figure 3. Ribbon diagrams for the NMR solution structures of (A) PSMα1, (B) PSMα3, and (C) PSMβ2 as calculated by CYANA and generated using PyMOL.31,37

Figure 4. NMR solution structures depicting the amphipathic helices and the hydrophobic surface maps of (A) PSMα1, (B) PSMα3, and (C) PSMβ2. Hydrophobic residues are colored white, while hydrophilic residues are colored green. The formylmethionines of PSMα1 and PSMα3 are depicted in traditional coloring: nitrogen, blue; carboxyl, red; and sulfur, yellow. Hydrophobicity surface map representations of the PSMs were generated using PyMOL.37 Electrostatic potential surface maps of (E) PSMα1, (F) PSMα3, and (G) PSMβ2. Electrostatic potential surface maps were generated using the APBS function of the PDB2PQR online pipeline.38 Briefly, negatively charged surfaces are colored red and positively charged surfaces blue. The intensity of the color represents the relative strength of the charge.

implications of this remain unclear. PSMβ2 contains three amphipathic helices (Figures 3C and 4C). The first begins at residue 5 and continues to residue 15. This helix has a slight

more curved tertiary structure. Curiously, whereas the bend in PSMα1 occurs near the N-terminus, the bend in PSMα3 occurs closer to the C-terminus at residue 16, though the biological E

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 5. PyMOL cartoon representation of (A) PSMα1, (B) PSMα3, and (C) PSMβ2. Side chains that may be participating in salt bridges are shown explicitly.37

PSMα3.23,33 Our NMR studies suggest that PSMα1, PSMα3, and PSMβ2 all form highly structured α-helices (Table 1). It is interesting to see that there appear to be two salt bridges present in PSMα3 between residues Glu2 and Lys6 and between residues Asp13 and Lys17 (Figure 5B). There also seems to be a single salt bridge present in both PSMα1 and PSMβ2. The locations of these salt bridges appear to be between residues Lys12 and Glu16 and between residues Asp19 and Lys22 (Figure 5A,C), respectively. While salt bridges may add to the propensity to form an α-helix, they do not seem to be a dominating factor in these PSM peptides. Previous studies have shown that the presence of aromatic, UVabsorbing side chains greatly contributes to the CD spectra of peptides in the amide region of 195−215 nm.34 Therefore, determination of secondary structure based exclusively on CD data can be misleading, and the three-dimensional NMR solution structures give a much more accurate representation of secondary and surface structure. The NMR solution structures of all three PSMs we studied show a high degree of α-helicity, even for the β-type PSM, which is essentially a noncytolytic PSM. It is more likely that the side chains of the amino acid residues in PSMs determine their ability to form pores within the membrane and therefore determine cytolytic capacity.35 This is also the case with some well-studied antimicrobial poreforming peptides. Leaderless and circular bacteriocins have recently been found to share the same structural motif but display differing antimicrobial potency thought to result from variation within the amino acid sequences.24,26 Indeed, Otto and co-workers point to the presence of phenylalanine, or the lack thereof, as a key determinant of cytolytic and antimicrobial activity in the α-type PSMs.9 In the case of PSMβ2, the surface is significantly less cationic than those of the α-type PSMs. This reduced positive charge may decrease the level of attraction between these peptides and negatively charged cell membranes. This would also result in a weakened ability of these peptides to permeate the membrane, resulting in lysis of the cell. Tertiary Structure and Its Influence on Biological Activity in PSMα3. In a previous study, a model structure for PSMα3 was proposed on the basis of the superimposition of the PSMα3 amino acid sequence on the NMR solution structure of δ-toxin.9 While the overall backbone structure of the model appears to be similar to our NMR solution structure, the positions of the amino acid side chains differ. The surface structure determined by NMR shows a slight bend in the tertiary structure, whereas the modeled structure does not. This shifts the positions of residues that have been determined to

bend and interacts with the third helix. The second helix is the shortest and runs from residue 17 to 23. Finally, the third and longest helix runs from residue 25 to 43. The interface between α-helicies 1 and 3 is such that hydrophobic residues interact, forming a hydrophobic core. The hydrophobic surface maps of PSMβ2 reveal a mostly hydrophilic surface with a few patches of hydrophobicity (Figure 4C). The fold that PSMβ2 assumes is primarily created by hydrophobic interactions between α-helix 1 and α-helix 3. Strong NOEs occur between the γ-protons of Ile8 and the entire side chain of Ile30. Additionally, a NOE occurs between the γ-protons of Ile8 and the amide of Val31. Further substantiating the interaction between α-helix 1 and α-helix 3 are the multiple NOEs present between the α- and β-protons of Ile27 and the γ-protons of Ile8. The second helix appears to interact with the third helix through the interaction of the amide and β-protons of Ala15 and the β-protons of Ser20. More NOEs occur between the α-proton of Ala15 and the amide proton of Ser20. Surprisingly, even though this peptide has relatively few long-range NOEs, the large number of shortand medium-range NOEs result in a highly structured peptide as evidenced by the low rmsd and the overlay of the 20 lowestenergy conformers (Table 1 and Figure S3). Comparison of Electrostatic Potential Surface Maps of PSMα1, PSMα3, and PSMβ2. Electrostatic potential surface maps of the peptides reveal a mostly cationic surface in PSMα1 and PSMα3. Interestingly, both PSMα1 and PSMα3 contain a small patch of strongly anionic charge on the Cterminus. PSMβ2 is significantly less charged, as can be visualized through the less intense blue and red color depicted in Figure 4E−G. It is interesting to note that PSMβ2 seems to have a strong cationic charge near the N- and C-termini of the peptide and a less intense cationic charge located on helix 2. The middle section of PSMβ2 has a minor negative charge. Implication of Secondary Structure for Activity. Despite the moderate percent α-helicity calculated from the CD spectra, all three peptides are primarily α-helical and highly structured in TFE, as seen from the low rmsd (Table 1). In 2014, Laabei and co-workers proposed that the PSMs with higher α-helical content, as determined by CD spectroscopy, appeared to have a higher cytolytic capacity.23 It was suggested that a negatively charged residue three or four residues from a positively charged residue may create a salt bridge that promotes and stabilizes α-helix formation. The authors proposed that these factors, α-helicity and salt bridges, may contribute to the increased lytic activity displayed by F

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

Figure 6. Location of amino acids in PSMα3 responsible for enacting a biological response, determined by Cheung and co-workers through an alanine scan.9 The top row reflects the hydrophilic side of PSMα3, and the bottom row reflects the hydrophobic side. Amino acids not enacting any biological response are colored green. Amino acids contributing to biofilm formation and detachment are colored orange, those contributing to antimicrobial activity light blue, those contributing to pro-inflammatory response red, and those contributing to cytolytic activity dark blue.

are comprised of amphipathic helices, which fold in such a way to give a hydrophilic outer surface and hydrophobic core. However, there are a few key differences between β-type PSMs and bacteriocins. Primarily, the electrostatic potential surface maps reveal that PSMβ2 has patches of cationic character among a slightly anionic surface, whereas bacteriocins have a strongly cationic surface. This may explain why PSMβ2 does not exhibit antibacterial activity despite other similarities with the bacteriocins. The cationic surface is thought to play an important role in the mode of action of the bacteriocins, as it may attract them to the negatively charged phospholipid cell membrane. A second noticeable difference between the bacteriocins and PSMβ2 is the particular orientation of the αhelicies. In the bacteriocins, the perpendicular helix is located at the C-terminus (e.g., helix 3 in the leaderless bacteriocin enterocin 7B), whereas in PSMβ2, the perpendicular helix (i.e., helix 1) is located at the N-terminus (Figure 7). The interaction

play an important role in cytolysis, biofilm formation, proinflammatory response, and antimicrobial activity (Figure 6).9 It was suggested that the hydrophilic side of the peptide is primarily responsible for the cytolytic properties and the hydrophobic side is primarily responsible for biofilm formation.9 Using our calculated structure of PSMα3 and the results produced from the alanine scan published by Cheung et al., we found that the amino acids responsible for the cytolytic activity are actually located on both the hydrophobic and hydrophilic sides.9 Moreover, the residues important for biofilm formation, previously thought to be due to the hydrophobic side of the peptide, appear to be more evenly spread out throughout the peptide and are on both the hydrophilic and hydrophobic sides.9 The amino acids responsible for antimicrobial activity are primarily located on the hydrophobic side as predicted.9 In the model structure of PSMα3, the amino acids responsible for pro-inflammatory response were located primarily on the hydrophilic side of the peptide, which we find is in agreement with our calculated structure. Comparison of PSMβ2 to Leaderless Bacteriocins. The genetic determinants for PSMs show certain similarities to the leaderless bacteriocins, a group of antibacterial peptides produced by bacteria. Both are ribosomally synthesized, do not require post-translational modification, require no leader sequence for secretion, and are not cotranscribed with a dedicated immunity protein.25 The genetic similarities led us to investigate the overall structure of these toxins. Our group has previously shown that leaderless bacteriocins with broadspectrum activity seem to adopt a similar overall fold independent of the sequence homology.24,26 Regardless of the low degree of sequence similarity between PSMβ2 and various leaderless bacteriocins, they share a somewhat similar structural motif. Specifically, they both exhibit two α-helices that form a “v-shape”, with a third α-helix that runs perpendicular to one of the two “v-shape” α-helices. Both PSMβ2 and the bacteriocins

Figure 7. Cartoon representation of amphipathic helices in leaderless bacteriocin enterocin 7B and peptide toxin PSMβ2. Both peptides contain three α-helices, two of which form a “v-shape” and a third perpendicular helix. In enterocin 7B, helix 1 and helix 2 form the “vshape” and helix 3 is perpendicular to the curved helix 2. In PSMβ2, helix 2 and helix 3 form the “v-shape” and helix 1 is perpendicular to the curved helix 3. G

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

chromatography; SDS−PAGE, sodium dodecyl sulfate−polyacrylamide gel electrophoresis; SPPS, solid phase peptide synthesis; SUMO, small ubiquitin-like modifier; TIPS, triisopropylsilane; TFA, trifluoroacetic acid; TFE, trifluoroethanol; TOCSY, total correlation spectroscopy.

between the perpendicular helix in PSMβ2 seems to be weaker in comparison to the interaction between the overlaying helix in the bacteriocins, as evidenced by the relatively few long-range NOEs observed.24,26

■

■

CONCLUDING REMARKS In summary, we have determined the three-dimensional NMR solution structures of three virulence factors produced by S. aureus, PSMα1, PSMα3, and PSMβ2. The backbones of PSMα1 and PSMα3 are primarily α-helical, forming a single amphipathic helix. The predominantly cationic surface of PSMα3 may attract this peptide to cell membranes, resulting in its ability to lyse certain eukaryotic cells. The slightly diminished cytolytic capabilities of PSMα1 may be due to its reduced cationic surface area in comparison to that of PSMα3. PSMβ2 is comprised of three amphipathic α-helices that fold to reveal a hydrophilic surface and create a hydrophobic core. The overall neutral surface may explain the lack of cytolytic activity for this peptide. With the structures of PSMα1, PSMα3, and PSMβ2 available, studies of the interactions between these peptide toxins and FPR2 would lead to an improved understanding of the active site of FPR2 to which the PSMs are binding and from which they are eliciting immune response.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.6b00615. NMR experimental parameters, chemical shift assignments for PSMα1, PSMα3, and PSMβ2, HPLC and MALDI-TOF spectra, 20 lowest-energy conformers, and Ramachandran plots of PSMα1, PSMα3, and PSMβ2 (PDF)

■

REFERENCES

(1) Arnold, S. R., Elias, D., Buckingham, S. C., Thomas, E. D., Novais, E., Arkader, A., and Howard, C. (2006) Changing patterns of acute hematogenous osteomyelitis and septic arthritis: emergence of community-associated methicillin-resistant Staphylococcus aureus. J. Pediatr. Orthop. 26, 703−708. (2) DeLeo, F. R., Otto, M., Kreiswirth, B. N., and Chambers, H. F. (2010) Community-associated meticillin-resistant Staphylococcus aureus. Lancet 375, 1557−1568. (3) Lowy, F. D. (1998) Staphylococcus aureus Infections. N. Engl. J. Med. 339, 520−532. (4) Novick, R. P., Schlievert, P., and Ruzin, A. (2001) Pathogenicity and resistance islands of staphylococci. Microbes Infect. 3, 585−594. (5) Otto, M. (2014) Staphylococcus aureus toxins. Curr. Opin. Microbiol. 17, 32−37. (6) Wang, R., Braughton, K. R., Kretschmer, D., Bach, T. H. L., Queck, S. Y., Li, M., Kennedy, A. D., Dorward, D. W., Klebanoff, S. J., Peschel, A., DeLeo, F. R., and Otto, M. (2007) Identification of novel cytolytic peptides as key virulence determinants for communityassociated MRSA. Nat. Med. 13, 1510−1514. (7) Cheung, G. Y. C., Joo, H., Chatterjee, S. S., and Otto, M. (2014) Phenol-soluble modulins - critical determinants of staphylococcal virulence. FEMS Microbiol. Rev. 38, 698−719. (8) Kretschmer, D., Gleske, A., Rautenberg, M., Wang, R., Köberle, M., Bohn, E., Schöneberg, T., Rabiet, M., Boulay, F., Klebanoff, S. J., van Kessel, K. A., van Strijp, J. A., Otto, M., and Peschel, A. (2010) Human Formyl Peptide Receptor 2 Senses Highly Pathogenic Staphylococcus aureus. Cell Host Microbe 7, 463−473. (9) Cheung, G. Y. C., Kretschmer, D., Queck, S. Y., Joo, H., Wang, R., Duong, A. C., Nguyen, T. H., Bach, T. L., Porter, A. R., DeLeo, F. R., Peschel, A., and Otto, M. (2014) Insight into structure-function relationship in phenol-soluble modulins using an alanine screen of the phenol-soluble modulin (PSM) alpha 3 peptide. FASEB J. 28, 153− 161. (10) Costerton, J. W. (1999) Bacterial Biofilms: A Common Cause of Persistent Infections. Science 284, 1318−1322. (11) Wang, R., Khan, B. A., Cheung, G. Y., Bach, T. H., Jameson-Lee, M., Kong, K. F., Queck, S. Y., and Otto, M. (2011) Staphylococcus epidermidis surfactant peptides promote biofilm maturation and dissemination of biofilm-associated infection in mice. J. Clin. Invest. 121, 238−248. (12) Tsompanidou, E., Denham, E. L., and van Dijl, J. M. (2013) Phenol-soluble modulins, hellhounds from the staphylococcal virulence-factor pandemonium. Trends Microbiol. 21, 313−315. (13) Periasamy, S., Joo, H. S., Duong, A. C., Bach, T. H., Tan, V. Y., Chatterjee, S. S., Cheung, G. Y., and Otto, M. (2012) How Staphylococcus aureus biofilms develop their characteristic structure. Proc. Natl. Acad. Sci. U. S. A. 109, 1281−1286. (14) Cheung, G. Y. C., Rigby, K., Wang, R., Queck, S. Y., Braughton, K. R., Whitney, A. R., Teintze, M., DeLeo, F. R., and Otto, M. (2010) Staphylococcus epidermidis Strategies to Avoid Killing by Human Neutrophils. PLoS Pathog. 6, e1001133. (15) Rasigade, J., Trouillet-Assant, S., Ferry, T., Diep, B. A., Sapin, A., Lhoste, Y., Ranfaing, J., Badiou, C., Benito, Y., Bes, M., Couzon, F., Tigaud, S., Lina, G., Etienne, J., Vandenesch, F., and Laurent, F. (2013) PSMs of Hypervirulent Staphylococcus aureus Act as Intracellular Toxins That Kill Infected Osteoblasts. PLoS One 8, e63176. (16) Joo, H. S., Cheung, G. Y., and Otto, M. (2011) Antimicrobial activity of community-associated methicillin-resistant Staphylococcus aureus is caused by phenol-soluble modulin derivatives. J. Biol. Chem. 286, 8933−8940.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: 780-492-5475. Fax: 780-492-8231. Funding

This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and Alberta Innovates Health Solutions (AIHS). Notes

The authors declare no competing financial interest.

■

ABBREVIATIONS APBS, Adaptive Poisson−Boltzmann Software; BMRB, Biological Magnetic Resonance Bank; CD, circular dichroism; DCM, dichloromethane; DMSO, dimethyl sulfoxide; DIPEA, diisopropylethylamine; DMF, dimethylformamide; HATU, 1[bis(dimethylamino)methylene]-1H-1,2,3-triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate; Fmoc, N-fluorenylmethyloxycarbonyl; FPR2, formyl peptide receptor 2; MALDITOF MS, matrix-assisted laser desorption ionization time-offlight mass spectrometry; Ni-NTA, nickel-nitrilotriacetic acid; NMP, N-methyl-2-pyrrolidone; NMR, nuclear magnetic resonance; NOESY, nuclear Overhauser effect spectroscopy; PSM, phenol-soluble modulin; PCR, polymerase chain reaction; PDB, Protein Data Bank; rmsd, root-mean-square deviation; RP-HPLC, reverse phase high-performance liquid H

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX

Article

Biochemistry

(36) Sievers, F., and Higgins, D. G. (2014) Clustal omega. Current Protocols in Bioinformatics 48, 3.13.1−3.13.16. (37) The PyMOL Molecular Graphics System, version 1.8 (2015) Schrödinger, LLC, Portland, OR. (38) Dolinsky, T. J., Nielsen, J. E., McCammon, J. A., and Baker, N. A. (2004) PDB2PQR: an automated pipeline for the setup of PoissonBoltzmann electrostatics calculations. Nucleic Acids Res. 32, W665−7.

(17) Gonzalez, D. J., Okumura, C. Y., Hollands, A., Kersten, R., Akong-Moore, K., Pence, M. A., Malone, C. L., Derieux, J., Moore, B. S., Horswill, A. R., Dixon, J. E., Dorrestein, P. C., and Nizet, V. (2012) Novel phenol-soluble modulin derivatives in community-associated methicillin-resistant Staphylococcus aureus identified through imaging mass spectrometry. J. Biol. Chem. 287, 13889−13898. (18) Chatterjee, S. S., Joo, H. S., Duong, A. C., Dieringer, T. D., Tan, V. Y., Song, Y., Fischer, E. R., Cheung, G. Y., Li, M., and Otto, M. (2013) Essential Staphylococcus aureus toxin export system. Nat. Med. 19, 364−367. (19) Cheung, G. Y. C., Wang, R., Khan, B. A., Sturdevant, D. E., and Otto, M. (2011) Role of the Accessory Gene Regulator agr in Community-Associated Methicillin-Resistant Staphylococcus aureus Pathogenesis. Infect. Immun. 79, 1927−1935. (20) Peschel, A., and Otto, M. (2013) Phenol-soluble modulins and staphylococcal infection. Nat. Rev. Microbiol. 11, 667−673. (21) Tappin, M. J., Pastore, A., Norton, R. S., Freer, J. H., and Campbell, I. D. (1988) High-resolution proton NMR study of the solution structure of.delta.-hemolysin. Biochemistry 27, 1643−1647. (22) Gladstone, G. P., and Yoshida, A. (1967) The cytopathic action of purified staphylococcal delta-hemolysin. Br. J. Exp. Pathol. 48, 11− 19. (23) Laabei, M., Jamieson, W. D., Yang, Y., van den Elsen, J., and Jenkins, A. T. A. (2014) Investigating the lytic activity and structural properties of Staphylococcus aureus phenol soluble modulin (PSM) peptide toxins. Biochim. Biophys. Acta, Biomembr. 1838, 3153−3161. (24) Acedo, J. Z., van Belkum, M. J., Lohans, C. T., Towle, K. M., Miskolzie, M., and Vederas, J. C. (2016) Nuclear Magnetic Resonance Solution Structures of Lacticin Q and Aureocin A53 Reveal a Structural Motif Convserved among Leaderless Bacteriocins with Broad-Spectrum Activity. Biochemistry 55, 733−742. (25) Cintas, L. M., Casaus, P., Holo, H., Hernandez, P. E., Nes, I. F., and Håvarstein, L. S. (1998) Enterocins L50A and L50B, two novel bacteriocins from Enterococcus faecium L50, are related to staphylococcal hemolysins. J. Bacteriol. 180, 1988−1994. (26) Lohans, C. T., Towle, K. M., Miskolzie, M., McKay, R. T., van Belkum, M. J., McMullen, L. M., and Vederas, J. C. (2013) Solution Structures of the Linear Leaderless Bacteriocins Enterocin 7A and 7B Resemble Carnocyclin A, a Circular Antimicrobial Peptide. Biochemistry 52, 3987−3994. (27) Zhang, L., Goldammer, C., Henkel, B., Zühl, F., Panhaus, G., Jung, G., and Bayer, E. (1994) in 3rd International Symposium on Innovation and Perspectives in SPPS, Oxford UK, Mayflower Scientific Ltd: Birmingham, Vol. 711. (28) Kaiser, E., Colescott, R. L., Bossinger, C. D., and Cook, P. I. (1970) Color test for detection of free terminal amino groups in solidphase synthesis of peptides. Anal. Biochem. 34, 595−598. (29) Hwang, T. L., and Shaka, A. J. (1995) Water Suppression That Works. Excitation Sculpting Using Arbitrary Wave-Forms and PulsedField Gradients. J. Magn. Reson., Ser. A 112, 275−279. (30) Morrow, J. A., Segall, M. L., Lund-Katz, S., Phillips, M. C., Knapp, M., Rupp, B., and Weisgraber, K. H. (2000) Differences in Stability among the Human Apolipoprotein E isoforms Determined by the Amino-Terminal Domain. Biochemistry 39, 11657−11666. (31) Güntert, P., Mumenthaler, C., and Wüthrich, K. (1997) Torsion angle dynamics for NMR structure calculation with the new program Dyana1. J. Mol. Biol. 273, 283−298. (32) Kretschmer, D., Nikola, N., Dürr, M., Otto, M., and Peschel, A. (2012) The Virulence Regulator Agr Controls the Staphylococcal Capacity to Activate Human Neutrophils via the Formyl Peptide Receptor 2. J. Innate Immun. 4, 201−212. (33) Kumar, S., and Nussinov, R. (2002) Close-Range Electrostatic Interactions in Proteins. ChemBioChem 3, 604−617. (34) Krittanai, C., and Johnson, W. C., Jr. (1997) Correcting the Circular Dichroism Spectra of Peptides for Contributions of Absorbing Side Chains. Anal. Biochem. 253, 57−64. (35) Verdon, J., Girardin, N., Lacombe, C., Berjeaud, J. M., and Hechard, Y. (2009) Delta-Hemolysin, an Update on a MembraneInteracting Peptide. Peptides 30, 817−823. I

DOI: 10.1021/acs.biochem.6b00615 Biochemistry XXXX, XXX, XXX−XXX