IAI Accepted Manuscript Posted Online 6 February 2017 Infect. Immun. doi:10.1128/IAI.01047-15 Copyright © 2017 American Society for Microbiology. All Rights Reserved. 1 2 3 4 5 6 Defining the Ail ligand-binding surface: hydrophobic residues in two 7 extracellular loops mediate cell and extracellular matrix binding to 8 facilitate Yop delivery 9 10 D o 11 Tiffany M. Tsang1, Jeffrey S. Wiese2, Jamal A. Alhabeil2, Lisa D. Usselman3, Joshua w 12 J. Thomson2, Rafla Matti3, Malte Kronshage3, Natalie Maricic1, Shanedah Williams3, nlo 13 Naama H. Sleiman2, Suleyman Felek3 and Eric S. Krukonis1,2,4* a d 14 e d 15 1Department of Microbiology and Immunology, University of Michigan Medical School, Ann Arbor, f r 16 Michigan 48109 o 17 2Department of Biomedical and Diagnostic Sciences, University of Detroit Mercy School of m h 18 Dentistry, Detroit, Michigan 48208 t 19 3Department of Biologic and Materials Sciences, University of Michigan School of Dentistry, Ann tp : 20 Arbor, Michigan 48109 // 21 4Department of Immunology and Microbiology, Wayne State University School of Medicine, Detroit iai. a 22 Michigan 48201 s 23 m . 24 o r 25 g / 26 Running Title: Y. pestis Ail adhesion mutants o n 27 A p r 28 * To whom correspondence should be addressed: Eric S. Krukonis, Department of Biomedical and il 29 Diagnostic Sciences, Room 444, University of Detroit Mercy School of Dentistry, Detroit, Michigan 1 , 30 48208 Tel: 313-494-6851, E-mail: [email protected] 2 0 31 1 9 32 Keywords: plague, Type III secretion, serum-resistance b 33 y g u e s t 1 34 Abstract 35 Yersinia pestis, the causative agent of plague, binds host cells to deliver cytotoxic Yop 36 proteins into the cytoplasm that prevent phagocytosis and generation of pro-inflammatory 37 cytokines. Ail is an eight-stranded β-barrel outer membrane protein with four extracellular loops 38 that mediates cell binding and resistance to human serum. Following deletion of each of the four 39 extracellular loops that potentially interact with host cells, the Ail-∆loop 2 and Ail-∆loop 3 mutants D 40 had no cell binding activity while Ail-∆loop 4 maintained cell binding (the Ail-∆loop 1 protein was o w n 41 unstable). Using the codon mutagenesis scheme SWIM (Selection Without Isolation of Mutants) lo a d 42 we identified individual residues in loops 1-3 that contribute to host cell binding. While several e d f 43 residues contributed to binding host cells, purified fibronectin, laminin and Yop delivery, three ro m 44 mutants, F80A (loop 2), S128A (loop 3) and F130A (loop 3), had particularly severe defects in cell h t t p : 45 binding. Combining these mutations led to an even further reduction in cell binding and severely // ia i. 46 impaired Yop delivery with only a slight defect in serum-resistance. These findings demonstrate a s m . 47 that Y. pestis Ail uses multiple extracellular loops to interact with substrates important for adhesion o r g / 48 via polyvalent hydrophobic interactions. o n A 49 p r il 1 , 2 0 1 9 b y g u e s t 2 50 Introduction 51 Ail is a multifunctional outer membrane protein of Yersinia species that is expressed under 52 various environmental conditions (1-7). Specifically, Y. pestis Ail mediates adhesion to host cells, 53 leads to autoaggregation, confers resistance to human serum and facilitates delivery of cytotoxic 54 Yop proteins to host cells (8-10). As a consequence of these various activities, an ail mutant is 55 highly attenuated, with a >3000-fold increase in LD when compared to parental Y. pestis KIM5 50 56 strain in a mouse model of septicemic plague (8) and shows a similar defect in a mouse model of D o w 57 bubonic plague using a fully virulent KIM5+ strain (11). In a rat model of pneumonic plague, where n lo a 58 serum-mediated killing of Y. pestis is more robust, an ail mutant is more than 105-fold attenuated d e d 59 for LD50 (10, 12). An ail mutant is also highly attenuated in rats in a bubonic plague model fro m 60 (>3,000-fold attenuated) although the LD50 has not been precisely determined (11). Thus, Ail is a h t t p 61 significant virulence factor for Y. pestis pathogenesis. :/ / ia 62 A number of ligands have been identified for Ail, including fibronectin (Fn) (13), laminin (Ln) i.a s m 63 and heparan sulfate proteoglycans (HSPGs)(14). The binding site for Ail on Fn has been . o r g 64 determined to be within the Fn Type III repeats 9 and 10 (15), while on laminin Ail binds the C- / o n 65 terminal 40kD fragment LG4-5 (14). Additionally, when the Ail/Fn and/or Ail/Ln interaction was A p r 66 blocked with antibodies to Fn or both Fn and Ln, Y. pestis was defective for the delivery of il 1 , 2 67 cytotoxic Yop proteins (13, 14). This indicates efficient delivery of Yops is dependent on Ail-ECM 0 1 9 68 (extracellular matrix) interactions, presumably by using ECM as a bridge to host cell receptors, b y g 69 similar to the well-studied YadA/Fn interaction that bridges enteropathogenic Yersinia binding to u e s t 70 host cell integrins (16). Thus, understanding Ail-ECM interactions will elucidate how Ail facilitates 71 the efficient delivery of critical Yop proteins to host cells. 72 Since Ail plays a critical role in pathogenesis, we sought to determine domains that 73 contribute to the various functions of Ail. Many bacteria express outer membrane proteins that are 74 predicted to be structurally similar to Ail (an eight-stranded flattened β-barrel), however, these 3 75 homologues have modest similarity at the amino acid level (17-21). Thus, defining amino acids 76 required for the various functions is not possible by simple homology alignments. 77 Ail from Y. enterocolitica has been studied extensively. One study identified regions of Ail 78 required for invasion (a process dependent on adhesion) and serum resistance (4). In this study, 79 key residues to examine were chosen by comparison with Ail homologues and natural Ail variants 80 as well as alanine substitution of charged residues. Two residues in the C-terminal end of 81 extracellular loop 2, D67 and V68 (numbered according to the proteolytically processed form D o w 82 following secretion; D90 and V91 of unprocessed form), were required for invasion into CHO cells n lo a 83 (4). Ail from Y. enterocolitica and Y. pestis have only 65%, 30%, 60% and 60% identity within the d e d 84 four extracellular loops, respectively, making analogies between the two proteins difficult. fr o m 85 However, we also included two possible analogous residues of Y. enterocolitica to include in our h t t p 86 study of Y. pestis Ail, D93 and F94 (analogous to Y. enterocolitica Ail D90 and V91). Two other :/ / ia 87 important findings from the studies with Y. enterocolitica Ail were that often multiple mutations i.a s m 88 were required to acquire a strong cell invasion phenotype (suggesting a polyvalent interaction) . o r g 89 and a peptide corresponding to the C-terminal 12 amino acids of loop 2 inhibited Ail-mediated / o n 90 invasion of CHO cells (4). A p r 91 To specifically define the molecular details of Y. pestis Ail interaction with the host, we il 1 , 2 92 determined Ail residues responsible for host cell attachment, Fn binding, Ln binding and Yop 0 1 9 93 delivery. We utilized the mutagenesis technique, SWIM (Selection Without Isolation of Mutants b y g 94 (22)), to generate mutant pools of ail that were subjected to a functional enrichment. Using this u e s t 95 technique and site-directed mutagenesis, we have identified residues in loops 1, 2, and 3 that play 96 a role in the various functions of Ail. 97 4 98 Materials and Methods 99 Strains and culture conditions. Y. pestis strains were cultivated in heart infusion broth (HIB) 100 overnight or on heart infusion agar (HIA) for 48 hours at 28°C. Escherichia coli strains were 101 cultured in Luria-Bertani (LB) broth or LB agar at 28°C or 37°C. Antibiotics were used at the 102 following concentrations: chloramphenicol (Cm) 25 µg/ml and ampicillin (Amp) 100 µg/ml. 103 Isopropyl-β-D-thiogalactopyranoside (IPTG) was used at a 100 µM concentration unless otherwise D o 104 noted. Strains and plasmids used in this study are listed in Table S1. w n 105 HEp-2 cells were cultured at 5% CO (37°C) in modified Eagle's medium (MEM, Life lo 2 a d 106 Technologies) supplemented with 10% fetal bovine serum (FBS) (Life Technologies), 1% sodium ed f r o 107 pyruvate (Life Technologies), and 1% non-essential amino acids (Life Technologies). m h 108 tt p : / / 109 SWIM mutagenesis. SWIM mutagenesis was performed similarly to a previously described study ia i. a 110 (22). Loop mutagenesis primers were ordered from Invitrogen and utilized a sequence in which s m . o 111 two nucleotides (WT and mutant) could be incorporated at various positions to encode wild-type r g / o 112 and mutant Ail proteins. The primer sequences for each loop mutant pool are found in Table S2 n A 113 with the forward primer listed (the reverse primer is exactly complementary to the forward primer). p r il 1 114 A pSK-Bluescript-derived pSK-ail construct was subjected to PCR amplification mutagenesis , 2 0 115 using the SWIM mutagenesis oligos. The resulting amplified plasmid was digested with DpnI to 1 9 b 116 degrade template plasmid and the digestion product was transformed into E. coli DH5α. DH5α y g u e 117 transformants were allowed to grow in liquid culture for two hours at 28°C before the addition of s t 118 ampicillin. The cultures were then grown overnight at 28°C. The next day, the liquid culture was 119 diluted 1:100 in LB + Amp and allowed to grow another day at 28°C. The two-day grown culture is 120 designated the input pool and 2ml were taken for isolation of the pSK-ail plasmid (Miniprep, 121 Qiagen) and sequenced. This culture was then diluted to an OD =0.6 and 100µl was used in 600 122 adhesion enrichment assays with HEp-2 cells. After 90 min of binding to host cells, the unbound 5 123 bacteria were moved to a fresh well of HEp-2 cells and incubated another 90min. After a total of 124 four enrichments, pools of bound and unbound bacteria were collected and cultured in fresh LB + 125 Amp overnight to re-grow both the bound and unbound pools. The next day, the plasmids were 126 isolated (Qiagen Miniprep kit) and sent for sequencing. These pools were designated as bound 127 and unbound mutant pools. 128 129 Generation of single point mutations. PCR-mutagenesis was performed using the enzyme Pfu D o w 130 (Stratagene) and primer pairs encoding the mutations. The primers used were complementary to n lo a 131 one another and the forward primers are listed in Table S3. Following PCR amplification using the d e d 132 pSK-ail plasmid as a template, PCR reactions were digested with DpnI to degrade the template fr o m 133 DNA and transformed into DH5α. Potential mutant clones were sequenced to confirm that only the h t t p 134 target site was mutated and a BamHI/PstI fragment containing the entire open reading frame and :/ / ia i. 135 ribosome-binding site, was liberated, purified and ligated into the IPTG-inducible plasmid a s m 136 pMMB207 (CmR) (23). . o r g 137 / o n 138 Introduction of ail alleles onto the Y. pestis chromosome. To facilitate introduction of point A p r 139 mutations into the ail locus of Y. pestis KIM5, we first used λ-RED recombination to mark the il 1 , 2 140 normal ail locus with a kanamycin-resistance/sucrose-sensitivity cassette (neo-sacB) from the 0 1 9 141 IVET plasmid pRES, (24). Using primers Ail-sacBf and Ail-sacBr (Table S3) the neo-sacB cassette b y g 142 was amplified by PCR and recombined at the ail locus using standard recombineering in Y. pestis u e s t 143 as previously described (8). The ail::neo-sacB strain was selected on HIA + 30µg/ml kanamycin 144 and replacement of the ail allele was confirmed by PCR. ail alleles of interest were then PCR 145 amplified from the plasmid pMMB207-ail using primers Ailf2 and Ailr2 and the PCR product was 146 used to replace the neo-sacB marked ail allele by selection on HIA + 5% sucrose. Proper 147 insertions were confirmed initially by PCR size and restoration of kanamycin sensitivity. The ail 6 148 allele from each candidate strain was then PCR amplified and sequenced to confirm the presence 149 of the desired mutation. 150 151 Fibronectin binding assay. This assay was described previously (13). Briefly, 96-well plates were 152 coated with 40µg/ml of plasma fibronectin (Sigma) in PBS overnight at 4°C. Y. pestis KIM5 153 ∆ail∆pla derivatives expressing various forms of Ail from the inducible plasmid pMMB207-ail were D 154 cultured overnight at 28°C in HIB with 20µM IPTG and Cm. The following day, wells were washed o w n 155 with PBS before blocking with PBS + 10mg/ml BSA (blocking buffer). Bacterial cells were pelleted lo a d 156 and resuspended at an OD of 1.0 and 50µl of the bacterial suspension was added to triplicate e 620 d f r 157 wells. The plate was then incubated for 2 hours at 37°C. The wells were washed 3 times with o m h 158 PBS before fixing with 100µl methanol and staining with 0.01% crystal violet for 20 min. After t t p : / 159 washing away excess crystal violet 3x with PBS, the bacterial-associated crystal violet stain was /ia i. a 160 solubilized with 100µl of 80% ethanol/20% acetone solution. The absorbance was measured at s m . o 161 A595 to quantify binding of bacteria to Fn. rg / 162 o n A 163 Laminin binding assay. Glass coverslips were added to 24-well tissue culture plates (Costar) and p r il 1 164 UV irradiated for 2 hours. Each coverslip was then coated with 40µg/ml human laminin (Sigma , 2 0 1 165 #L6274) in 300µl of PBS. Coverslips were coated for two hours at room temperature and washed 9 b y 166 once with 1 ml PBS. Coverslips were then blocked with 500µl PBS containing 10mg/ml BSA g u e 167 overnight at 4°C. Y. pestis strains expressing various forms of Ail from the plasmid pMMB207 s t 168 were then grown overnight at 28°C in HIB + 20µM IPTG and Cm. Bacteria were pelleted and 169 resuspended at an OD =5 in PBS + 0.4% BSA and 300µl bacterial cells were added to each well 620 170 for two hours at 28°C. Unbound bacteria were then removed by three washes with 1ml PBS and 171 bacteria were fixed with 300µl methanol for 20 min at room temperature. Methanol was then 172 removed and the plates were air dried for about 2 hours. Wells were then stained with 400µl 7 173 0.01% crystal violet for 20 min at room temperature followed by two washes with 500µl water. The 174 stained bacteria were then lysed with 500µl of a 20% acetone/80% ethanol mixture and 100µl of 175 the lysate was read for its absorbance at 595nm. Levels of binding were presented relative to the 176 binding by a strain expressing wild-type Ail. 177 178 Cell adhesion assays. This assay was described previously (8, 13). Briefly, HEp-2 cells were 179 cultured in MEM + 10% FBS 24-well tissue culture plates until reaching 80-90%. Y. pestis KIM5 or D o w n 180 E. coli AAEC185 (fim-) derivatives were cultured overnight at 28°C in HIB + Cm (Y. pestis) or lo a d 181 37°C in LB + Cm (E. coli) and then diluted 1:50 the next day in fresh media with Cm and 100µM e d f 182 IPTG to induce plasmid-based Ail expression. Strains were allowed to grow for an additional 3 ro m 183 hours at 28°C (Y. pestis) or 37°C (E. coli). Tissue culture cells were washed once with 1 ml PBS h t t p : 184 and then 400µl serum-free MEM was added followed by 100µl of bacteria resuspended in serum- // ia i. 185 free MEM at an OD620 (Y. pestis) or OD600 (E. coli) of 0.6 for an MOI ~100. Plates were incubated as m 186 with bacteria at 37°C in 5% CO for 90 min. Cells were then washed 3x with PBS, and cell- .o 2 r g / 187 associated bacteria were liberated by the addition of sterile H O containing 0.1% Triton X-100 for o 2 n A 188 10 min. Percent adhesion was calculated by dividing bound CFU by total bacteria in parallel wells p r il 189 and multiplying by 100. 1 , 2 0 190 1 9 b 191 Cytotoxicity assay. This assay was described previously (8, 13). Briefly, HEp-2 cells were y g u 192 cultivated in MEM + 10% FBS until they reached about 50-80% confluence in 24-well tissue e s t 193 culture plates. Y. pestis KIM5 derivative strains were cultured overnight in HIB (with Cm if 194 required). Overnight cultures were diluted 1:10 in fresh media and incubated for 3.5 hours at 28°C 195 (with Cm and 100µM IPTG if required for plasmid based expression). Tissue culture wells were 196 washed once with 1ml PBS and 490µl serum-free MEM was added followed by 10µl of bacteria 197 resuspended in serum-free MEM at an OD =0.6 to obtain an MOI of ~10. Plates were incubated 620 8 198 at 37°C in 5% CO for 3 hours. Cells were washed twice with 1ml PBS then fixed with methanol, 2 199 dried and stained with 0.076% Giemsa stain for 20 min and then washed four times with H 0. 2 200 Rounding was observed and photographs were taken with a phase-contrast microscope at 20X 201 magnification. Cytotoxicity was enumerated by counting total cells and the number of round dark 202 purple (shrunken cytoplasm) cells experiencing cytotoxicity in three microscopic fields (~150 203 cells/field). Percent cytotoxicity was calculated by dividing rounded cells by total cells and 204 multiplying by 100. D o w 205 n lo a 206 Autoaggregation assay. Y. pestis cultures were induced for Ail expression overnight in HIB + Cm d e d 207 and 100µM IPTG. After overnight induction, stationary phase cultures were allowed to settle for 60 fr o m 208 min. OD620 measurements were taken at time 0 (before settling) and at 60 min. The OD620 at 60 h t t p 209 min was subtracted from the starting OD620 and expressed as a percentage of the starting OD620 :// ia i. 210 to quantify the amount of autoaggregation. Assessment of Ail-mediated co-aggregation with a s m 211 fluorescently labeled strains was performed by transforming the KIM5 ∆ail strain SF777 with either .o r g / 212 the GFP-encoding plasmid pFVP25.1 (25) or the RFP-encoding plasmid pGEN-Pem7-DsRedT3- o n A 213 MUT (26). Each fluorescent strain was then transformed with a plasmid encoding one of several p r 214 alleles of ail on the vector pMMB207 (or with the empty vector). Strains were grown overnight at il 1 , 2 215 28°C to saturation. The following day 25µl of each of two strains (one red, one green) was 0 1 9 b 216 inoculated into 5mls of fresh HIB + Amp (selecting for the fluorescence-encoding plasmids) + Cm y g u 217 (selecting for pMMB207) and 100µM IPTG to induce ail expression and grown overnight again at e s t 218 28°C. The following day cultures were removed from the incubator, resuspended by gentle 219 vortexing and 1ml was gently pelleted for 5 min at 7,000 rpm and resuspended in PBS to prevent 220 autofluorescence from the medium and 100µl were moved to a black 96-well plate (USA 221 Scientific) to measure initial levels of both the green-fluorescent strain and the red-fluorescent 222 strain before clumping using a SpectraMax M3 spectrophotometer (Molecular Devices) and 9 223 excitation/emission 488nm/511nm for green, 557nm/592nm for red. The remaining 5ml cultures 224 were left to settle at room temperature for 60 min. At the end of 60 min 100µl of the supernatant 225 was moved to the black 96-well plate for fluorescent measurements to assess autoaggregation of 226 both red and green-labeled cells in the mixture. For fluorescence microscopy, red and green- 227 labeled cells were grown similarly, but after overnight growth at 28°C, bacteria were allowed to 228 settle for ~30 min, the supernatant was discarded and 50µl of the settled bacterial pellet was D 229 placed on a glass coverslip for visualization of Y. pestis aggregated cells. Imaging was performed o w n 230 using a 60X-magnifying objective on an Olympus BX63 microscope. lo a d 231 e d f 232 Protein expression analysis. Bacterial cultures were resuspended in Laemmli sample buffer ro m 233 normalizing for OD (E. coli) or OD (Y. pestis). Bacterial cell extracts were boiled and run on h 600 620 t t p : 234 a 15% SDS-polyacrylamide gel electrophoresis (PAGE) gel. Proteins were stained with / / ia i. 235 Coomassie blue dye where Ail can be visualized as a unique protein band at about 15 kDa (8). a s m 236 Other proteins bands in the cell lysate served as protein loading controls. For studies with the Ail .o r g / 237 loop deletions expressed in E. coli a previously described anti-Ail antibody (27) as well as o n A 238 Coomassie blue staining were employed to detect Ail. p r il 239 1 , 2 240 Serum resistance assay. Y. pestis strains expressing chromosomally-encoded mutant forms of Ail 0 1 9 241 were grown overnight at 28°C in HIB. Bacteria were diluted 1:50 the next day and grown in HIB for b y g 242 3-4 hours. After reading the read OD , 1ml of each strain were pelleted and resuspended in PBS u 620 e s t 243 at OD =0.5. The strains were then diluted 1:10 in PBS to reach a concentration of ~1.5 x 107/ml. 620 244 50µl of diluted bacteria (~7.5 x 105 bacteria) were mixed with 200µl normal human serum (NHS, 245 Sigma) or 200µl heat-inactivated serum (HIS), resulting in a concentration of 80% serum. Bacteria 246 and serum were incubated with rolling for 1 hour at 37°C. Bacteria were then plated at various 10
Description: