Comparative analysis of the virulence characteristics of methicillin – resistant and – 1 susceptible Staphylococcus pseudintermedius isolates isolated from small animals : a 2 RNA - Seq - based transcriptome analysis 3 4

20 Staphylococcus pseudintermedius is often associated with pyoderma, which can turn into a life21 threatening disease. The dissemination of highly resistant isolates has occurred in the last 10 22 years and has challenged antimicrobial treatment of these infections considerably. We have 23 compared the carriage of virulence genes and biofilm formation between methicillin-resistant 24 and methicillin-susceptible S. pseudintermedius isolates (MRSP and MSSP, respectively) and 25 additionally the in vitro gene expression profiles of MRSP and MSSP by RNA-seq. 26 Isolates were relatively unevenly distributed among the four agr groups, and agr type III 27 predominated in MRSP. Five virulence genes were detected in all isolates. Only the spsO gene 28 was significantly associated with MSSP isolates (p=0.04). All isolates produced biofilm in 29 BHIB+4% NaCl. MSSP produced more biofilm on BHIB and BHIB+1% glucose media than 30 MRSP isolates (p=0.03 and p=0.02, respectively). Virulence genes encoding surface proteins 31 and toxins (spsA, spsB, spsD, spsK, spsL, spsN, nucC, coa, Iuk-I) and also prophage genes 32 (encoding phage capsid protein, phage infection protein, two phage portal proteins and a 33 phage-like protein) were highly expressed in the MRSP isolate (compared with the MSSP), 34 suggesting they may play a role in the rapid and widespread dissemination of MRSP. 35 This study indicates that MRSP may upregulate surface proteins, which may increase the 36 adherence of MRSP, especially ST71, isolates to corneocytes. MSSP may have an increased 37 ability to form biofilm in acidic circumstances, through upregulation of the entire arc operon. 38 Complete understanding of S. pseudintermedius pathogenesis and host-pathogen signal 39 interaction during infections is critical for the treatment and prevention of S. pseudintermedius 40 infections. 41 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom Introduction 42 Methicillin-resistant Staphylococcus pseudintermedius (MRSP) isolates have emerged as one of 43 the leading causes of infectious diseases (including pyoderma, otitis and urinary tract infections) 44 in companion animals, accounting for 20%-47% of all clinical S. pseudintermedius isolated from 45 dogs and cats (1). Moreover, some MRSP isolates are resistant to the antimicrobials regularly 46 used for treatment (β-lactams, fluoroquinolones, tetracyclines, lincosamides, potentiated 47 sulphonamides) in small animal practice (1,2). The mecA gene, encoding resistance to β48 lactams, has been acquired by several S. pseudintermedius clonal lineages on independent 49 occasions, however two clones, MRSP ST68-SCCmec V and MRSP ST71-SCCmec II-III, are 50 the dominant ones and have spread globally (1,3,4). This dissemination was rapid but the 51 reasons for the fast emergence and success of these lineages are not fully understood (2). 52 Genomic and proteomic studies, conducted in the last few years, are giving the first clues on the 53 pathways by which MRSP have become successful. A recent genomic report suggested that 54 multidrug-resistance evolved rapidly in MRSP due to the acquisition of a very limited number of 55 mobile genetic elements and mutations (1). Thus, the use of different antimicrobial classes co56 selected for the spread and emergence of the multidrug-resistant MRSP isolates (1). The 57 frequent carriage of prophages in MRSP ST71 and ST68 genomes suggested they have a role 58 in the fitness of MRSP and that the predominant transfer of genetic material in these isolates is 59 through bacteriophage transduction, rather than plasmid conjugation as happens in methicillin60 resistant S. aureus (MRSA) (1). MRSP are able to produce biofilm and MRSP ST71 isolates, in 61 particular, are better biofilm producers than other MRSP clones (5,6). The icaA gene can be 62 significantly upregulated in biofilm samples, suggesting a role in the biofilm production by S. 63 pseudintermedius (7). The ability to form biofilm may play an important role in the 64 pathophysiology of bacterial infections and can be related to survival and persistence of S. 65 pseudintermedius, namely MRSP, in the environment (5,6). The MRSP ST71 isolates also show 66 greater adherence to corneocytes than MRSP non-ST71 and MSSP, and thus it has been 67 suggested that the enhanced adherence of ST71 might be a factor contributing to the 68 epidemiological success of this MRSP lineage (2). Furthermore, an MRSP ST71 isolate of 69 human origin adhered evenly well to canine and human corneocytes, implying that MRSP ST71 70 may also be capable of adapting to the human skin (2). Two proteins, SpsD and SpsO, can 71 mediate adherence to canine corneocytes (8); however, the genetic factors responsible for the 72 enhanced in vitro adherence of MRSP ST71 are not yet known (2). 73 In order to understand the epidemiological success of MRSP isolates, our goal was to 74 understand if the phenotype (biofilm) and genotype (virulence genes) related to virulence factors 75 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom was different between MRSP and methicillin-susceptible S. pseudintermedius (MSSP) isolates. 76 Furthermore, we compared the in vitro transcriptional profiles by RNA-seq of one MRSP and 77 one MSSP isolate to test the hypothesis that MRSP could have altered expression of virulence 78 genes, by comparing to MSSP gene expression, which could have contributed to its rapid 79 spread. 80 81 Materials and methods 82 Genotypic characterization of the MRSP and MSSP isolates 83 Twenty-one consecutive methicillin-resistant S. pseudintermedius (MRSP) isolated over a 784 year period from 2007 to 2014 were included in the study. Twenty-one matched (in terms of 85 isolation year, isolation site and host) methicillin-susceptible (MSSP) were also included. These 86 isolates were from 18 asymptomatic carriers (9 MRSP and 9 MSSP), 12 patients with pyoderma 87 (6 MRSP and 6 MSSP), 6 patients with urinary tract infection (3 MRSP and 3 MSSP), 5 patients 88 with otitis (2 MRSP and 3 MSSP) and 1 with a surgical site infection (1 MRSP). Five isolates 89 were from cats and 37 were from dogs. Isolates were characterized by Multi-locus sequence 90 typing (MLST) (9). The eBURST algorithm identified groups of related sequence types (ST) (10). 91 Specific sequences for virulence genes involved in biofilm formation (bap, icaA, icaB, icaC, 92 icaD), enterotoxin production (se-int, seccanine, seh), host adherence (ebpS, spsD, spsL, spsO), 93 toxin production (lukS, lukF, siet, speta, expA, expB) were detected by PCR on a Mastercycle 94 thermocycler (Eppendorf, New York, USA) with the primers, product size and annealing 95 temperatures shown in Supplementary Table 1. The primers designed in this study were 96 generated using the Primer-BLAST tool from NCBI. All PCR products were analyzed by 97 electrophoresis through 1.2% agarose gels (NZYTech, Lisbon, Portugal). The primers agrD-F 98 (5’GGG GTA TTA TTA CAA TCA TTC -3’) and agrD-R (5’CTG ATG CGA AAA TAA AGG 99 ATT G -3’) (STABvida, Monte da Caparica, Portugal) were used as previously described to 100 amplify a 300-bp agr fragment encompassing the 3’ end of agrB, all of agrD, and the 5’ end of 101 agrC. Amplification was carried out on a Mastercycler thermocycler (Eppendorf) under 102 the following conditions: an initial 5-min denaturation step at 94°C; followed by 30 cycles of 1 103 min of denaturation at 94°C, 1 min of annealing at 45°C, and 1 min of extension at 72°C; 
and 104 a final extension step at 72°C for 10 min. The PCR products were purified by using the 105 NZYGelpure (NZYTech) and sequenced with the same primers used for the PCRs 
(STABvida). 106 The 42 isolates were assigned to one of the four agr groups by comparing the predicted product 107 of AgrD and the N-terminal half of AgrC with those of four control isolates (GenBank accession 108 numbers EU157336, EU157366, EU157334 and EU157330). 109 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom 110 Biofilm-producing ability on polystyrene 111 The capacity of the isolates to form biofilm was investigated by a method described by 112 Stepanović and colleagues (11) and Pettit and colleagues (12) with minor modifications, 113 and was determined by the ability of S. pseudintermedius isolates to adhere to 96-well 114 polystyrene microtitre plates (Greiner bio-one, Frickenhausen, Germany). In brief, the study was 115 carried out using brain-heart infusion broth (BHIB; Biokar), BHIB with 4% NaCl and BHIB with 116 1% glucose as the growth media. The plates were incubated at 37oC for 24h. Following 117 incubation, the Alamar Blue solution was added to each well. After 30 minutes at room 118 temperature, the optical densities at 570 nm (OD570) were measured. Staphylococcus 119 epidermidis RP62A strain (ATCC 35984) was used as a positive control. We defined the cut-off 120 OD (ODC) for the microtiter-plate test as three standard deviations above the mean OD of the 121 negative control as described previously (11). All isolates were classified into the following 122 categories: non-adherent (0) if the OD ≤ ODC, weakly adherent (+) if the ODC < OD ≤ 2 x ODC, 123 moderately adherent (++) if the 2xODC < OD ≤ 4 x ODC, or strongly adherent (+++) if the OD ≥ 4 124 x ODC, based upon the ODs of bacterial films (11). 125 126 RNA isolation, sequencing and gene expression analyses 127 To test the hypothesis that MRSP and MSSP isolates differ in their expression of virulence 128 genes, we compared the in vitro transcriptional profiles of a clinical MRSP isolate and a clinical 129 MSSP isolate using RNA-seq. We attempted to choose 2 representative of the S. 130 pseudintermedius collection: isolated from skin swabs of dogs with pyoderma (the most frequent 131 clinical specimen where S. pseudintermedius was isolated from); isolated in the same period of 132 time; agr type III (the most frequent agr type found in this study); at least one ST71 (the most 133 frequent ST found in this study); and similar virulence profiles (considering the virulence genes 134 tested by PCR). Bacterial cells were grown until the mid-log phase growth (OD600=0.5), since it 135 has been shown that the majority of surface proteins are produced during this phase (13). RNA 136 was isolated using the RNeasy kit from Qiagen (Hilden, Germany). Briefly, 2x10 cells were 137 removed from growing cultures and 2 volumes of RNAprotect® Bacteria Reagent (Qiagen) were 138 added and incubated for 5 minutes at room temperature. Cells were then centrifuged and 139 incubated with TE buffer (30 mM Tris.Cl, 1 mM EDTA pH 8.0, Sigma) containing 0.5-mg/ml 140 lysostaphin (Sigma) and 15-mg/ml lysozyme (Sigma) for 20 minutes at 37oC. Proteinase K (20 141 mg/ml, Sigma) was added and incubated for 10 minutes. After, the procedure was carried out 142 according to the manufacturer’s specifications. The purified RNA was quantified using a 143 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom spectrometer (NanoDrop, ThermoScientific). RNA quality was assessed by visualization on an 144 agarose gel. The rRNA was removed using the MICROBExpress kit (Ambion). RNA quality was 145 then evaluated on a ByoAnalyzer (Agilent). Bacterial mRNA was fragmented (yield fragments 146 were in the size range of 200–250 bp) and the double-stranded cDNA was generated using the 147 Ion Total RNA-seq kit v2 (Life Technologies, Thermo Fisher Scientific) according to the 148 manufacturer’s instructions. The samples were sequenced using the Ion PGM (Life 149 Technologies, Thermo Fisher Scientific) sequencer at STABvida. 150 Mapping to the reference genomes and normalization of gene expression were performed by 151 CLC Genomics Workbench v8.0.1. RNA-seq reads were aligned with the three available S. 152 pseudintermedius reference genomes ED99 (ST25, agr type III, lacks spsF, spsO, spsQ), 153 HKU10-03 (ST308, agr type III, lacks nanB) and E140 (ST71, agr type III, lacks nanB, lukF-S) 154 (RefSeq accession numbers CP002478, CP002439 and ANOI01000001, respectively). Gene 155 expression was normalized by calculating reads per kilobase per million mapped reads (RKPM), 156 given by dividing the total number of reads by the number of mapped reads (in millions) times 157 the length (kb) (14). 158 Differentially expressed genes were identified using the Baggerley’s test (binomial test), which 159 compares the proportions of counts in a group of samples against those of another group of 160 samples (15) with false discovery rate (FDR) correction applied (16). Genes with an adjusted 161 p≤0.05 were identified as being differentially expressed. This study focused particularly in the 162 expression of virulence genes, but expression of other relevant genes (e.g. antimicrobial 163 resistance genes) was also evaluated. 164 165 Statistical analysis 166 All data analysis was carried out using IBM SPSS Statistics Version 20.0 (IBM, New York, USA). 167 Differences between the two groups, MRSP and MSSP, were calculated by the Fisher’s exact 168 test for categorical comparisons and Student’s t-test for continuous outcome. A p-value of ≤ 169 0.05 was considered to be statistically significant. 170 171 Results 172 The results of MLST are shown in Table 1. The MSSP isolates were divided into 21 different 173 STs, while 15 MRSP isolates were assigned to the ST71, three to ST203; one to ST196; one to 174 ST213, and one to ST195. Yet ST203 and ST195 belonged to the clonal complex (CC) 71, as 175 detected by the eBURST analysis. Equally, ST196 and ST213 differed only by one allele and 176 belonged to CC196. 177 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom All isolates were classified as part of one of the four agr groups, and the distribution was highly 178 uneven, with 2 isolates belonging to agr group I, 7 belonging to group II, 30 belonging to group 179 III, and 3 belonging to group IV (Table 1). There was a significant difference in the agr groups’ 180 distribution between MRSP and MSSP (p=0.025), with allele III being significantly more 181 associated with MRSP than with MSSP (p=0.014). 182 The virulence genes detected in the MRSP and MSSP isolates are detailed in Table 2. Genes 183 se-int, speta, siet, spsL and ebpS were present in all 42 isolates. The genes lukF and lukS, 184 encoding for leukocidin Luk-I, were found in all isolates except for two MRSP isolates (ST196, 185 ST213). Gene expB was only detected in 3 isolates. Only two MSSP isolates carried the 186 enterotoxin seccanine. No isolates harbored genes seh and expA. Eight isolates carried the spsO 187 gene and by statistical analysis, this gene was significantly more associated with MSSP than 188 with MRSP (p=0.04). No differences were found between clinical isolates and isolates from 189 carriage. 190 Results of the biofilm-forming ability on polystyrene are shown in Table 3. All isolates produced 191 biofilm in the BHIB+4% NaCl medium. Two and nine isolates did not produce biofilm on BHIB 192 and BHIB+1% glucose, respectively. Biofilm production in the BHIB and BHIB+1% glucose 193 media was significantly higher in MSSP than in MRSP isolates (p=0.03 and p=0.02, 194 respectively), but there were no differences between clinical isolates and isolates from carriage. 195 The ica genes were detected in all 42 isolates. 196 The number of mapped reads assigned by using each reference genome (ED99, HKU10-30 197 and E140) is shown in Supplementary Table 2. Of these mapped reads, the amount of S. 198 pseudintermedius genes with altered expression also varied when using the three different 199 reference genomes as shown in Figure 1. The MSSP isolate had higher expression in 200 transcription of regulatory genes agrB and agrD. On the other hand, the MRSP isolate had 201 higher transcription of regulatory genes sigB, srrA, sarA, rot and saeRS system. The signal 202 transduction protein TRAP gene (traP) was also highly expressed. Considering genes encoding 203 surface proteins, only one, spsC encoding an autolysin, was highly expressed in the MSSP 204 isolate, while 6, spsA, spsB, spsD, spsK, spsL, spsN, were highly expressed in the MRSP 205 isolate. The gamma-hemolysin component B gene (hlgB), both subunits of luk-I gene (lukF-I 206 and lukS-I), the coagulase and thermonuclease genes (coa and nucC, respectively) were 207 upregulated in the MRSP isolate. The arc genes (arcA, arcB, arcC, and arcD) were upregulated 208 in the MSSP isolate. Several genes associated with antimicrobial resistance were highly 209 expressed in the MRSP isolate: the norA, gyrA and gyrB genes associated with resistance to 210 quinolones; the aadE and the bifunctional aacA-aphD genes associated with aminoglycoside211 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom resistance; the mecA, mecR1 and blaI genes associated with beta-lactam resistance; and tet(M) 212 gene associated with resistance to tetracycline. The MRSP isolate upregulated several phage213 associated genes (encoding phage capsid protein, phage infection protein, two phage portal 214 proteins and a phage-like protein), and an integrase gene located in the superantigen-encoding 215 pathogenicity islands SaPI (SPSINT_0063). 216 217 Discussion 218 In the last 10 years MRSP have become highly frequent in clinical samples from infected 219 animals and as colonizers of healthy ones (1). However, it is still not clear why MRSP, 220 especially certain lineages like ST71, have spread so quickly. To understand the rapid evolution 221 that led to the dissemination of MRSP isolates we assessed the virulence determinants present 222 in a collection of MSSP and MRSP isolates, and compared the ability to form biofilm in 3 223 different media. Finally, we performed and compared the in vitro gene expression analysis of 224 one MSSP and one MRSP isolate. 225 Analysis of the virulence genotype of the MRSP and MSSP isolates revealed a strong 226 conservation of genes: five genes (ebpS, se-int, siet, speta, spsL) were carried by all S. 227 pseudintermedius isolates, and five genes (expB, luk-I, seccanine, spsD, spsO) were only present 228 in some isolates. Two studies have reported the existence of some specific toxins (e.g. coa, lip, 229 geh, htrA, nuc, clpX, hlb, se-int, speta, spsA, spsB, spsC) present in several S. 230 pseudintermedius isolates that might be important for the canine host tropism, in particular the 231 skin (1,17). However, variation was found in others (e.g. spsF, spsO, spsP, spsQ, luk-I, nanB), 232 suggesting that a difference in virulence factors in the core genome was probably lineage233 associated (1). For example, in one of these studies the five ST71 isolates lacked the nanB and 234 the lukF-S genes (1). Still, in our study and in a previous study conducted in Spain (18), all the 235 ST71 isolates carried the lukF-S genes, suggesting that variation may also be related to the 236 region of isolation. It would be interesting to collect a large collection of ST71 isolates from 237 different countries to study these variations. In other lineages, however, this will be difficult to 238 ascertain, since only a few isolates in each lineage have been reported so far. 239 The capacity of bacteria to form biofilms is an important virulence factor not only in the 240 development of device-related infections but also in a range of chronic infections (17). This 241 capacity might further complicate the treatment of already challenging infections due to the 242 decrease in effectiveness of antimicrobials on biofilms (5). In one study that all S. 243 pseudintermedius isolates produced biofilms, suggesting that biofilm-production might be 244 essential for the pathogenicity of S. pseudintermedius (6). Yet, the study failed to find 245 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom differences in the biofilm formation between MRSP and MSSP isolates. The number of MSSP 246 isolates that was studied was low and the authors suggested that further experiments with a 247 larger number of isolates were warranted (6). By using a larger set of isolates, we observed that 248 biofilm production in the BHIB and BHIB+1% glucose media was significantly higher in MSSP 249 than in MRSP isolates. This is a phenomenon that has been observed in S. aureus, when 250 comparing methicillin-resistant and methicillin-susceptible isolates, and is due to different 251 triggering mechanisms leading to biofilm formation, including ica-dependent and -independent 252 mechanisms (19). In our study, all isolates produced biofilm and all were positive for the ica 253 genes, suggesting this operon has a crucial role in biofilm formation. However, the mechanisms 254 triggering the higher biofilm production in the BHIB and BHIB+1% glucose media by MSSP 255 strains remain unknown. One clue to this occurrence may be related to the upregulation of the 256 entire arc operon in the MSSP isolate studied here. A similar operon has been found in other 257 staphylococcal species and, in S. aureus, arcA (belongs to the arc operon) encodes an arginine 258 deaminase, which allows for enhanced survival in acidic environments (20). The upregulation of 259 this operon may improve survival and promote biofilm formation of MSSP in acidic 260 circumstances, such as in BHIB medium with glucose (which has a more acidic pH than BHIB 261 medium alone or with NaCl). 262 During the early emergence of community acquired MRSA, the USA300 (ST8) lineage 263 disseminated rapidly and was considered hypervirulent, compared with lineages like MRSA 264 USA400 (ST1) (20). However, USA300 does not contain much more virulence genes than 265 USA400, but it does have an alteration in the expression of regulatory genes and an increased 266 expression of certain virulence genes (20). By microarray analysis, USA300 displayed an 267 increased expression of genes encoding cell envelope proteins (including lipoproteins and 268 superantigen-like proteins), genes residing in the prophage φSa3usa, several genes encoded in 269 pathogenicity islands vSAα and vSAß, proteases and the gene encoding the IgG binding protein 270 Sbi (20). Interestingly our MRSP isolate also had increased expression of several genes 271 including spsK, which encodes the IgG binding protein Sbi, the toxins nucC and coa, prophages 272 and several virulence regulatory genes including saeRS. The higher expression of the prophage 273 genes might be one of the factors contributing to the rapid dissemination of MRSP, particularly 274 ST71 isolates. The higher expression of the genes spsD and spsL (encoding fibronectin-binding 275 proteins able to adhere to the extracellular matrix) found in this study, may explain the higher 276 adherence of MRSP ST71 isolates to corneocytes previously detected (2). We observed a very 277 different expression of virulence regulatory genes between the two isolates, with agr highly 278 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom expressed in MSSP and saeRS in MRSP. This may explain the differences observed in the 279 expression of the genes encoding surface proteins and toxins. 280 One of the most important bacterial defenses against uptake of foreign DNA is restriction281 modification (R-M) systems (21). These systems, comprising restriction endonucleases and 282 methyltransferases, recognize and modify specific DNA sequences, protecting “own” DNA from 283 restriction while eliminating potentially harmful foreign DNA (21). In S. pseudintermedius, type I 284 R-M systems have been recognized, including one that was carried on all SCCmec II-III 285 elements of MRSP ST71 (1). One study suggested that MRSP were not more efficient or 286 inefficient than MSSP in acquiring mobile genetic elements due to the wide distribution of Type I 287 and Type II R-M systems in S. pseudintermedius isolates (1). In our study, however, we found 288 that the Type I restriction-modification system restriction subunit R (hsdR) was highly expressed 289 in the MSSP isolate, suggesting it blocks DNA horizontal gene transfer into methicillin290 susceptible isolates. Lower expression of the subunit R in the MRSP isolate could also suggest 291 a more efficient way of acquiring mobile genetic elements. In fact it has been shown that MRSP 292 genomes carry more prophages than MSSP isolates. Our results showed that the MRSP isolate 293 also upregulates several phage-associated genes, which could be linked to the upregulation of 294 the integrase located in the superantigen-encoding pathogenicity islands SaPI. The upregulation 295 of prophage particles is also concordant with the suggestion that transfer in MRSP is 296 predominantly made by transduction (1). 297 In summary, this is the first study to document the global transcription differences between 298 MSSP and MRSP isolates during in vitro growth. This study indicates that MRSP may 299 upregulate surface proteins, which may increase the adherence of MRSP, especially ST71, 300 isolates to corneocytes. Although MRSP and MSSP have the capacity to form biofilm, MSSP 301 may have an increased ability to form biofilm in acidic circumstances, through upregulation of 302 the entire arc operon. Complete understanding of S. pseudintermedius pathogenesis and host303 pathogen signal interaction during infections is critical for the treatment and prevention of S. 304 pseudintermedius infections. 305 306 Funding 307 This work was funded by National funds through the FCT – Fundação para a Ciência e 308 Tecnologia, Project PTDC/CVT-EPI/4345/2012 and a PhD grant SFRH/BD/68864/2010 to 309 Natacha Couto from the same institution. Manuela Oliveira is a researcher from the program 310 “Ciência 2007” from FCT, Portugal. 311 312 on N ovem er 3, 2017 by gest httpaac.asm .rg/ D ow nladed fom Acknowledgements 313 The authors would like to thank Elena Gómez-Sanz, Myriam Zarazaga and Carmen Torres
for 314 providing control isolates for some of the virulence genes; Rui Seixas from FMV-UL and 315 Magdalena Lewicka from STABvida for excellent technical assistance.316 317References3181. McCarthy AJ, Harrison EM, Stanczak-Mrozek K, Legget B, Waller A, Holmes MA, Lloyd319DH, Lindsay JA, Loeffler A. 2015. Genomic insights into the rapid emergence and evolution of320MDR in Staphylococcus pseudintermedius. J. Antimicrob. Chemother. 70:997-1007.3212. Latronico F, Moodley A, Nielsen SS, Guardabassi L. 2014. Enhanced adherence of322methicillin-resistant Staphylococcus pseudintermedius sequence type 71 to canine and 
human323corneocytes. Vet. Res. 45:70.3243. Perreten V, Kadlec K, Schwarz S, Grönlund Andersson U, Finn M, Greko C, Moodley A,325Kania SA, Frank LA, Bemis DA, Franco A, Iurescia M, Battisti A, Duim B, Wagenaar JA,326van Duijkeren E, Weese JS, Fitzgerald JR, Rossano A, Guardabassi L. 2010. Clonal spread327of methicillin-resistant Staphylococcus pseudintermedius in Europe and North America: an328international multicentre study. J. Antimicrob. Chemother. 65:1145-54.3294. Kadlec K Schwarz S, Perreten V, Andersson UG, Finn M, Greko C, Moodley A, Kania330SA, Frank LA, Bemis DA, Franco A, Iurescia M, Battisti A, Duim B, Wagenaar JA, van331Duijkeren E, Weese JS, Fitzgerald JR, Rossano A, Guardabassi L. 2010. Molecular analysis332of methicillin-resistant Staphylococcus pseudintermedius of feline origin from different European333countries and North America. J. Antimicrob. Chemother. 65:1826-8.3345. Osland AM, Vestby LK, Fanuelsen H, Slettemeås JS, Sunde M. 2012. Clonal diversity and335biofilm-forming ability of methicillin-resistant Staphylococcus pseudintermedius. J. Antimicrob.336Chemother. 67:841-8.3376. Singh A, Walker M, Rosseau J, Weese SJ. 2013. Characterization of the biofilm forming338ability of Staphylococcus pseudintermedius from dog. BMC Vet. Res.9:93.3397. Crawford EC, Singh A, Metcalf D, Gibson TWG, Weese SJ. 2014. Identification of340appropriate reference genes for qPCR studies in Staphylococcus pseudintermedius and341preliminary assessment of icaA gene expression in biofilm-embedded bacteria. BMC Res.342Notes 7:451.3438. Bannoehr J, Brown JK, Shaw DJ, Fitzgerald RJ, van den Broek AHM, Thoday KL. 2012.344Staphylococccus pseudintermedius surface proteins SpsD and SpsO mediate adherence to ex345vivo canine corneocytes. Vet. Dermatol. 23:119-24.346onNovemer3,2017bygesthttpaac.asm.rg/Downladedfom 9. Solyman SM, Black CC, Duim B, Perreten V, van Duijkeren E, Wagenaar JA, Eberlein347LC, Sadeghi LN, Videla R, Bemis DA, Kania SA. 2013. Multilocus sequence typing for348characterization of Staphylococcus pseudintermedius. J. Clin. Microbiol. 51:306-354.34910. Spratt BG, Hanage WP, Li B, Aanensen DM, Feil EJ. 2004. Displaying the relatedness350among isolates of bacterial species–the eBURST approach. FEMS Microbiol. Lett. 241:129-34.35111. Stepanovic S, Vukovic D, Dakic I, Savic B, Svabic-Vlahovic M. 2000. A modified352microtiter-plate test for quantification of staphylococcal biofilm formation. J. Microbiol. Meth.35340:175-9.35412. Pettit RK, Weber CA, Kean MJ, Hoffmann H, Pettit GR, Tan R, Franks KS, Horton, ML.3552005. Microplate Alamar Blue assay for Staphylococcus epidermidis biofilm susceptibility testing.356Antimicrob. Agents Chemother. 49:2612-7.35713. Bannoehr J, Ben Zakour NL, Reglinski M, Inglis NF, Prabhakaran S, Fossum E, Smith358DG, Wilson GJ, Cartwright RA, Haas J, Hook M, van den Broek AH, Thoday KL, Fitzgerald359JR. 2011. Genomic and surface proteomic analysis of the canine pathogen Staphylococcus360pseudintermedius reveals proteins that mediate adherence to the extracellular matrix. Infect361Immun. 79:3074-86.36214. Mortazavi A, Williams BA, McCue K, Schaeffer L, Wold B. 2008. Mapping and363quantifying mammalian transcriptomes by RNA-seq. Nat. Methods 5:621-8.36415. Baggerly K, Deng L, Morris J, Aldaz C. 2003. Differential expression in SAGE: accounting365for normal between-library variation. Bioinformatics 19:1477-83.36616. Pawitan Y, Michiels S, Koscielny S, Gusnanto A, Ploner A. 2005. False discovery rate,367sensitivity and sample size for microarray studies. Bioinformatics 22:295-6.36817. Zakour NLB, Beatson SA, van den Broek AHM, Thoday KL, Fitzgerald JR. 2012.369Comparative genomics of the Staphylococcus intermedius group of animal pathogens. Front.370Cell Infect. Microbiol. 2:44.37118. Gómez-Sanz E, Torres C, Lozano C, Zarazaga M. 2013. High diversity of Staphylococcus372aureus and Staphylococcus pseudintermedius lineages and toxigenic traits in healthy pet-373owning household members. Underestimating normal household contact? Comp. Immunol.374Microbiol. Infect. Dis. 36:83-94.37519. O’Neill E, Humphreys H, O’Gara JP. 2009. Carriage of both the fnbA and fnbB genes and376 growth at 37oC promote FnBP-mediated biofilm development in methicillin-resistant377Staphylococcus aureus clinical isolates. J. Med. Microbiol. 58:399-402.37820. Jones MB, Montgomery CP, Boyle-Vavra S, Shatzkes K, Maybank R, Frank BC,379Peterson SN, Daum RS. 2014. Genomic and transcriptomic differences in community acquired380onNovemer3,2017bygesthttpaac.asm.rg/Downladedfom methicillin resistant Staphylococcus aureus USA300 and USA400 isolates. BMC Genomics38115:1145.38221. Murray NE. 2000. Type I restriction systems: sophisticated molecular machines (a legacy of383Bertani and Weigle). Microbiol. Mol. Biol. Rev. 64:412-34.38422. Becker K, Roth R, Peters G. 1998. Rapid and specific detection of toxigenic385Staphylococcus aureus: use of two multiplex PCR enzyme immunoassays for amplification and386hybridization of staphylococcal enterotoxin genes, exfoliative toxin genes, and toxic shock387syndrome toxin 1 gene. J. Clin. Microbiol. 36:2548-53.38823. Monday SR, Bohach GA. 1999. Use of multiplex PCR to detect classical and newly389described pyrogenic toxin genes in staphylococcal isolates. J. Clin. Microbiol. 37:3411-14.39024. Futagawa-Saito K, Sugiyama T, Karube S, Sakurai N, Ba-Thein W, Fukuyasu T. 2004.391Prevalence and characterization of leukotoxin-producing Staphylococcus intermedius in isolates392from dogs and pigeons. J. Clin. Microbiol. 42:5324-6.39325. Cucarella C, Solano C, Valle J, Amorena B, Lasa I, Penadés JR. 2001. Bap, a394Staphylococcus aureus surface protein involved in biofilm formation. J. Bacteriol. 183:2888-96.39526. Lautz S, Kanbar T, Alber J, Lämmler C, Weiss R, Prenger-Berninghoff E, Zschöck M.396 2006. Dissemination of the gene encoding exfoliative toxin of Staphylococcus intermedius397among isolates isolated from dogs during routine microbiological diagnostics. J. Vet. Med. B39853:434-8.39927. Futagawa-Saito K, Makino S, Sunaga F, Kato Y, Sakurai-Komada N, Ba-Thein W,400Fukuyasu T. 2009. Identification of first exfoliative toxin in Staphylococcus pseudintermedius.401FEMS Microbiol. Lett. 301:176-80.402onNovemer3,2017bygesthttpaac.asm.rg/Downladedfom Figure 1. Number of S. pseudintermedius genes with altered expression identified by the403Baggerley’s test with the FDR correction applied, using the three different reference genomes404available (ED99 ST25-agr III, HKU10-30 ST308-agr III, E140 ST71-agr III).405 406Table 1. Epidemiological characteristics of MRSP and MSSP isolates used in this study. The407number of isolates is shown inside parenthesis.408 409Table 2. Virulence traits of the methicillin-resistant and methicillin-susceptible S.410pseudintermedius isolates.411 412Table 3. Overall results of the microtiter-plate test according to the pattern of methicillin-413resistance in S. pseudintermedius.414 onNovemer3,2017bygesthttpaac.asm.rg/Downladedfom