Membrane Protein Profiling of Acidovorax Avenae Subsp. Avenae Under Various Growth Conditions

Arch Microbiol (2015) 197:673–682 DOI 10.1007/s00203-015-1100-9

ORIGINAL PAPER

Membrane protein profiling of Acidovorax avenae subsp. avenae under various growth conditions

Bin Li1 · Li Wang1 · Muhammad Ibrahim1,3 · Mengyu Ge1 · Yanli Wang2 · Shazia Mannan3 · Muhammad Asif3 · Guochang Sun2

Received: 24 July 2014 / Revised: 1 February 2015 / Accepted: 2 March 2015 / Published online: 13 March 2015 © Springer-Verlag Berlin Heidelberg 2015

Abstract Membrane proteins (MPs) of plant pathogenic transport of small molecules, protein synthesis and secretion bacteria have been reported to be able to regulate many essen- as well as virulence such as NADH, OmpA, secretion pro- tial cellular processes associated with plant disease. The aim teins. Therefore, the result of this study not only suggests that of the current study was to examine and compare the expres- it may be an alternate method to analyze the in vivo expression sion of MPs of the rice bacterial pathogen Acidovorax avenae of proteins by using LE medium to mimic plant conditions, subsp. avenae strain RS-1 under Luria-Bertani (LB) medium, but also reveals that the two sets of differentially expressed M9 medium, in vivo rice plant conditions and leaf extract (LE) MPs, in particular the common MPs between them, might be medium mimicking in vivo plant condition. Proteomic analy- important in energy metabolism, stress response and virulence sis identified 95, 72, 75, and 87 MPs under LB, in vivo, M9 of A. avenae subsp. avenae strain RS-1. and LE conditions, respectively. Among them, six proteins were shared under all tested growth conditions designated as Keywords Acidovorax · Membrane proteins · In vivo · abundant class of proteins. Twenty-six and 21 proteins were Leaf extract · Rice expressed uniquely under in vivo versus LB medium and LE versus M9 medium, respectively, with 17 proteins common among these uniquely induced proteins. Moreover, most of Introduction the shared proteins are mainly related to energy metabolism, Membrane proteins (MPs) are important components of the bacterial membrane, which play an important role in Communicated by Shuang-Jiang Liu. transport of nutrients, cellular processes, energy transduc- Electronic supplementary material The online version of this tion, scaffolding of cell structure and cell-to-cell commu- article (doi:10.1007/s00203-015-1100-9) contains supplementary nications or interactions (Marreddy et al. 2011). Further- material, which is available to authorized users. more, interaction of bacterial plant pathogen with the host is directly associated with the expression of MPs by regu- * Bin Li [email protected] lating the movement of effector proteins from the bacterial cells directly into the eukaryotic plant cells (Ibrahim et al. * Guochang Sun [email protected] 2012; Knief et al. 2011). Indeed, several MPs from Gram- negative bacteria had been found to have the ability to rec- 1 State Key Laboratory of Rice Biology, Institute ognize essential virulence factors and host immune rec- of Biotechnology, Zhejiang University, Hangzhou 310058, ognition target (Schell et al. 2011). Therefore, identifying China abundant and/or novel MPs, and characterizing their func- 2 State Key Laboratory Breeding Base for Zhejiang tion in disease, pathogen physiology and resistance against Sustainable Plant Pest and Disease Control, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China the host are important steps for understanding the genetics and disease intensity of bacterial plant pathogens. 3 Department of Biosciences, COMSATS Institute of Information Technology, Sahiwal Campus, Sahiwal, Membrane proteome has been characterized and ana- Pakistan lyzed in bacteria under various synthetic medium such as

1 3 674 Arch Microbiol (2015) 197:673–682

Luria-Bertani (LB) and M9 medium. However, the expression suspension cultured in LB broth in our previous study of proteins associated with derivative enzymes and transport (Xie et al. 2011) and preserved in 25–35 % sterile glycerol systems and proteins involved in invading the host cells, e.g., (Sangon Biotech, Shanghai, China) at 80 °C. LE broth − type III secretion system (T3SS), are usually expressed during medium was prepared as described by Mehta and Rosato the bacterial infection process in plant (Barnard et al. 2007; (2001). Briefly, about 10.0 g of rice leaves were grinded Mole et al. 2007). Indeed, recently, comprehensive studies of into powder and macerated in 100 mL M9 medium. After proteomics of Methylobacterium extorquens have identified 10 min centrifugation at 5000 RPM, supernatant was more abundant proteins when the commensals colonized plant taken and filtered sequentially using 0.45- and 0.22-µm surface compared to the cells cultured in synthetic medium membranes (Millipore, Bedford, USA) and mixed to M9 (Knief et al. 2011; Gourion et al. 2006). In addition, proteomic broth to final concentration of 1.0 mg leaves/mL of M9 analysis of cellular outer membrane in bacterial pathogen of broth. rice has shown significant differential expression of proteins under in vivo and in vitro conditions (Ibrahim et al. 2012). Bacterial growth under different conditions Therefore, characterizations of bacterial membrane proteome under the in vivo condition could provide important role in A. avenae subsp. avenae strain RS-1 was cultured on LB understanding the pathogen physiology (Cash 2011). medium, following the incubation at 30 °C for 24 h. For In a previous study, a method for analyzing outer mem- synthetic growth culture medium, single bacterial colony brane proteome of in vivo cultivated bacteria has been was picked and inoculated into 10 mL fresh LB broth, LE developed (Ibrahim et al. 2012; Li et al. 2014). In general, and M9 broth and then incubated at 30 °C. Bacteria were however, due to the difficulty in collecting bacteria from harvested, and the final concentration was adjusted to infected tissues and the high risk of contamination, cellular 1.8–2.0 of OD600nm. The centrifugation was done at 8000 proteome of plant pathogenic bacteria in plants is less char- RPM and 4 °C for 10 min, and bacterial cells were pro- acterized. To solve this problem, some researchers use plant cessed to the steps for extraction of MPs. The MPs of the extracts to mimic the plant environment (Büttner and Bonas in vivo cells were extracted as described in our previous 2010). Proteomic response and changes to plant materials study (Ibrahim et al. 2012). Briefly, 6 days after bacterial have been revealed in previous studies (Knief et al. 2011; inoculation, leaves with visible and prevailing symptoms Andrade et al. 2008; Astua-Monge et al. 2005; Brown et al. were collected and washed with alcohol. The leaves were 2001). However, little is known about the mimic effect of incised to produce very small pieces by using sterile razor plant extracts on the membrane proteome of plant patho- blade. These small pieces of leaves were kept for 30 min genic bacteria Acidovorax avenae subsp. avenae (Xie et al. in contamination-free glass plates containing 20 mL double 2011; Liu et al. 2012; Yang et al. 2014). distilled water to allow leaching bacteria from leaf tissues. In this study, we present the membrane proteome sum- The bacterial suspension and leaf tissues were separated by mary of A. avenae subsp. avenae strain RS-1 under LB, in centrifugation following the wash of bacterial cell pellets vivo rice plant, M9 and leaf extract (LE) culture conditions with PBS and with water. The bacterial cell pellets were using LC–MS/MS and in silico functional characterizations of processed for protein extraction. membrane proteome. Furthermore, comparison of proteomic profiling revealed that the majority of differentially expressed Preparation of MPs MPs under in vivo versus LB medium are common to LE medium versus M9 medium. This result not only indicates that MPs were extracted as described by Jagannadham and it may be an alternate method to analyze the in vivo expression Chowdhury (2012) by using sucrose gradient techniques of proteins by using LE medium to mimic plant conditions, with minor modification. Cells were washed with Tris– but also suggests that the two sets of differentially expressed HCl (30 mM) with pH 8.0 following re-suspension in MPs, in particular the common MPs between them, may be sucrose solution (20 %) made in Tris buffer having 1 mg/ involved in pathogenicity, cell survival, energy metabolism mL lysozyme, DNase (300 μg/mL) and RNase (200 μg/ and stress response of A. avenae subsp. avenae strain RS-1. mL). Cells were lysed using sonication following centrifugation for 10 min at 10,000 RPM. The supernatant was taken and loaded onto a double-step sucrose gradi- Materials and methods ent (60 and 70 % sucrose layers) and ultra-centrifuged at 55,000 RPM (Beckman Coulter, USA) for 18 h. After Bacterial strain and LE medium centrifugation, MPs were separated as double bands: The upper band corresponding to the fraction of inner mem- A. avenae subsp. avenae strain RS-1 was isolated from brane, while the lower band to outer membrane. The two diseased rice plants following inoculation with bacterial bands having membrane fractions were collected precisely,

1 3 Arch Microbiol (2015) 197:673–682 675 combined, and subsequently diluted and centrifuged for Genome‑wide in silico functional characterization another round at 55,000 RPM for 1 h. The solubilization of of LC–MS/MS peptides the pellet containing MPs was performed in Triton X-100 (2 %) following the overnight precipitation with 10 % LC–MS/MS-identified proteins were further analyzed for TCA at 20 °C. The resultant pellet was centrifuged for protein subcellular localization using PSORTb (version − 10 min at 14,000 RPM and washed with acetone twice. 3.0.2) and Phobius (Kall et al. 2004; Yu et al. 2010). The Finally, the pellet solubilization for isoelectric focusing parameters selected in PSORTb and Phobius online tools (IEF) was performed using rehydration buffer having 2 % were of normal format with Gram-negative strains, and ASB-14. The protein was quantified using estimation kit score >7.5 was set as a significant score out of 10. Fur- (Bio-Rad protein estimation kit) with bovine serum albu- ther, BLAST analysis was performed for domain and fam- min as standard. ily search using Pfam and InterPro online tools. Moreover, each protein that was assigned a subcellular position was One‑dimensional SDS–PAGE also analyzed for N-terminal secretary signal peptides prediction using SignalP 3.0 server (Petersen et al. 2011). The precipitated mixture of MPs extracted from A. avenae subsp. avenae strain RS-1 was run on a 12 % SDS–PAGE Modeling and structural characterization of in vivo as described by Laemmli et al. (1970) with few changes. and LE‑identified proteins The protein samples were run in a gel electrophoresis apparatus (VE-180 vertical gel caster, Tanon, China). The Selected virulence proteins were further analyzed in sil- protein band appeared on SDS gel were observed with ico to corroborate the role of identified virulence protein Coomassie Brilliant Blue and silver staining. under in vivo and LE medium. The analysis was focused on sequence characterizations such as functional domain In gel digestion and LC–MS/MS analysis prediction, multiple sequence alignment and domain conservation. The names of these selected proteins are as The main protein bands of SDS–PAGE were sliced into follows: NADH dehydrogenase (locus tag Aave_1264), small pieces and performed in gel digestion. Briefly, the NADH-ubiquinone oxidoreductase subunit 4L (locus tag sliced gel bands were transferred into 0.5-mL eppendorf Aave_1273), NADH-quinone oxidoreductase subunit M tubes, destained and digested by using MultipPROBE (locus tag Aave_1275) and OmpA/MotB domain-contain- II Plus Ex robotic liquid handling system (PerkinElmer, ing protein (locus tag Aave_2800). The domains of all pro- USA) following extraction of digested peptides with ace- teins were determined by using InterPro, which provided tonitrile and formic acid solution with the concentration of functional analysis of proteins by classifying them into 50 and 2 %, respectively. The resultant peptides were dried families and predicting domains and important sites (Jones and 10 µL of them was separated on the UltiMate® 3000 et al. 2014) and Pfam database. Following that, protein and Nano LC systems. The peptides were examined using ama- domain sequences were BLAST and selected the Gram- Zon ETD ion trap MS having nanosource setting. The mass negative species reported with potential virulence charac- spectrometry scan range was adjusted at 300–1400 m/z teristics. After that the sequence of domain of correspond- and for LC–MS/MS was adjusted at 50–2200 m/z. The ing bacteria was compared with the domain of Aave_1264 sequence similarities of the obtained peptides with the by using their multiple sequence alignment file in GENE- Acidovorax species was searched by matching the pep- DOC (Nicholas et al. 1997) and determined the gene con- tides from NCBI database by using computer program, servation and multiple sequence alignment. the MASCOT ion search algorithmic tool. The entire MS spectra were explored for the selection of following parameters the enzyme trypsin, the precursor mass tolerance of Results and discussion 0.35 Da and fragment ion tolerance of 0.35 Da. Moreo- ver, searches were conducted with variable modification One‑dimensional profile of MPs under four growth and fixed modification, the oxidation of methionine and conditions carboxyamidomethylation of cysteine, respectively. The cross-correlation scores (X corr) (Olsen et al. 2005) were The result of one-dimensional SDS–PAGE revealed that fixed for protein identification if singly, doubly and triply the MPs profile under LB, M9, LE medium and in vivo charged peptides were greater than 1.8, 2.5 and 3.5, respec- conditions of A. avenae subsp. avenae strain RS-1 were tively. The list of putative OM proteins was made out either different (Figs. 1, 2). Four major bands at about 14, 20, 31 by satisfactory peptides score or by minimum one peptide and 98 kD (designated as 1, 2, 3 and 4, respectively) were with acceptable scores. observed under LB medium conditions, but only three main

1 3 676 Arch Microbiol (2015) 197:673–682

Fig. 3 Culture-dependent and independent distribution of membrane proteins in diverse growth conditions

about 14 and 45 kD (designated as 1 and 3, respectively) were visible under M9 medium conditions (Fig. 2).

Fig. 1 Membrane proteins profile of Acidovorax avenae subsp. ave- Global membrane protein profiling nae strain RS-1 under in vivo and Luria-Bertani (LB) medium conditions obtained by SDS–PAGE In total, 95, 72, 75 and 87 MPs were identified under LB, in vivo, M9 and LE growth conditions, respectively. These proteins were detected in two replicates under each condition with 96 % confidence level. A non-redundant list of MPs were assembled and designated as the A. avenae subsp. avenae strain RS-1 membrane proteome (Table S1–4; Fig. 3). The LC–MS/MS-identified proteins were further annotated based on the Cluster of Orthologus Groups (COG). Among these proteins, the most abundant, shared and unique MPs under LB, LE, M9 medium and in vivo conditions are as follows: The six most abundant proteins such as porin, efflux, ABC transporter and phasing family proteins were identified under all four growth conditions (Table 1). Furthermore, in vivo versus LB medium uniquely expressed MPs, and LE medium versus M9 medium uniquely expressed MPs are presented in Tables 2 and 3, respectively, while the 17 common MPs between the two sets of differentially expressed MPs (in vivo and LE) were marked with asterisk. Interestingly, there were more than 80 proteins whose GRAVY was hydrophobic. The hydrophobic effect can be used to separate mixtures of proteins based on their hydro- phobicity. To achieve better separation, a salt may be added Fig. 2 Membrane proteins profile of Acidovorax avenae subsp. ave- (higher concentrations of salt increase the hydrophobic nae strain RS-1 under leaf extract (LE) medium and M9 medium con- effect), and its concentration decreased as the separation ditions obtained by SDS–PAGE goes on. Moreover, at the molecular level, the hydrophobic effect is important in driving protein folding, formation of bands at about 20, 31 and 98 kD (designated as 1, 2 and lipid bilayers and micelles, insertion of membrane proteins 3, respectively) were observed under in vivo conditions into the nonpolar lipid environment and protein–small mol- (Fig. 1). In addition, three major bands at about 14, 20 and ecule interactions (Chandler 2005). 45 kD (designated as 1, 2 and 3, respectively) were visible Furthermore, cytoplasmic proteins, including riboso- under LE medium condition, but only two major bands at mal proteins, DNA topoisomerase, protein/dioxygenase,

1 3 Arch Microbiol (2015) 197:673–682 677

Table 1 List of the most abundant proteins detected under each Table 2 List of LC–MS/MS-identified differentially expressed mem- growth condition of Acidovorax avenae subsp. avenae strain RS-1 brane proteins under in vivo versus LB medium conditions of Acido- proteome vorax avenae subsp. avenae strain RS-1 proteome Name of proteins Pfam COG Locus tag Protein name COG

Porin PF00267 3203 Aave_0301* Extracellular solute-binding protein 0834 Phasin family PF09361 9913 Aave_0662 Respiratory nitrate reductase subunit beta 1140 OmpW PF03922 3047 Aave_0780* Unnamed protein product NF ABC transporter PF12849 0226 Aave_0925* General secretion pathway protein G 2165 Efflux proteins PF00529 1538 Aave_1010 Extracellular solute-binding protein 0834 Membrane protein insertase, YidC PF02096 0706 Aave_1264* NADH dehydrogenase 0377 Aave_1273* NADH-ubiquinone oxidoreductase subunit 4L 0713 Aave_1275* NADH-quinone oxidoreductase subunit M 1008 Aave_1519* Unnamed protein product 0589 glyoxalase resistance, tetraacyl disaccharide, elongation factor, chaperonin GroEL, Dead/Deah box helicase pro- Aave_1797* Peptidase M14, carboxypeptidase A 2866 teins, ribonuclease E, rng/rng family, gluconolactonase, Aave_1879 OmpA domain-containing protein NF pyruvate kinase and 3-isopropylmalate dehydratase, were Aave_2204* Succinate dehydrogenase subunit C 2009 also identified in this study, which is in agreement with the Aave_2223* Ubiquinol oxidase subunit II 1622 result of Jagannadham (2008). Aave_2800* OmpA/MotB domain-containing protein 2885 Aave_3154* UspA domain-containing protein 0589 Characterizations of abundant MPs Aave_3405* OmpA domain-containing protein 2885 Aave_4228 Carbon starvation protein CstA 1966 Porins (COG3203), which comprise the majority of pre- Aave_4565* Lipase 3240 dicted and pragmatic MPs by PSORTb and Phobius, were Acav_0258* Aromatic hydrocarbon degradation MPs 2067 detected abundantly under each growth condition. Func- Acav_1536* Glycine/betaine/l-proline ABC transporter 1125 tionally, porins cross a cellular membrane and act as a ATPase pore through which molecules can diffuse (Klebba 2005). Acav_2917 Acid phosphatase 3511 Another abundant class of proteins identified under each Acav_3188 Hypothetical protein NF growth condition was efflux proteins (COG1538). In agree- Acav_4039* Transport-associated protein 1652 ment with this result, previous study has revealed a house- Acav_4274 CheW protein 0835 keeping role for the efflux genes in all Gram-negative bac- Acav_4577 Preprotein translocase subunit YajC 1862 teria (Morita et al. 2006). Bacteria often come across many Aave_2113 Methyl-accepting chemotaxis proteins 0840 such kinds of substances that are lethal to them, e.g., lipo- * Common to both in vivo versus LE medium philic, and have the ability to infiltrate the cell quite eas- ily. Hence, bacteria must need to get rid of these toxic substances. The identification of efflux pump proteins showed polyhydroxyalkanoates (PHAs). These PHAs are storage that A. avenae subsp. avenae strain RS-1 may have very polyesters synthesized by various Gram-negative bacteria efficient way to eject toxic substances leading to the sur- as energy reserve material and intracellular carbon and vival of this pathogen. preserved as water-insoluble inclusions within the cells Members of OmpW (COG3047) families were also iden- (Matsumoto et al. 2002). YidC was also identified under tified under each growth condition. Previously, it has been each growth condition. YidC is necessary for membrane reported that OmpW was highly conserved in a variety of incorporation process of lately produced proteins that do bacterial species (Nandi et al. 2005). It has been shown that not need the classical Sec machinery (Nandi et al. 2005). the modulation of OmpW expression by environmental fac- In addition, proteins involved in energy production, con- tors may be linked to the adaptive response of the A. avenae version and assimilation were also detected under each subsp. avenae strain RS-1 under stress conditions. However, growth condition, which might be important for growth little information is available on their regulation of expres- and survival of Gram-negative A. avenae subsp. avenae sion that may be linked to their putative function (Mole strain RS-1. These proteins included methyl-accepting et al. 2007; Nandi et al. 2005; Nandi et al. 2000). chemotaxis (MAC) protein, outer membrane protein A, Phasin family protein with unknown location was fre- LPP_SERMA, major outer membrane lipoprotein, outer quently identified under four growth conditions. This fam- membrane porin protein 32, peptidase S45 penicillin ami- ily had been found to have PhaF and PhaI proteins, which dase, polyhydroxyalkanoate depolymerase and peptidase play role of transcriptional regulator for biosynthesis of M14 carboxy.

1 3 678 Arch Microbiol (2015) 197:673–682

Table 3 List of LC–MS/MS-identified differentially expressed mem- two domains: peptidoglycan domain of C terminus and brane proteins under leaf extract medium versus M9 medium condi- 8-stranded β-barrel domain of N-OmpA—an important tions of Acidovorax avenae subsp. avenae strain RS-1 proteome virulence factor in bacteria (Chen et al. 2010; Ibrahim et al. Locus tag Protein name COG 2012; Teng et al. 2006), which arbitrates bacterial conju- gation and functions for a range of bacteriophages as a Aave_0301* Extracellular solute-binding protein 834 receptor (Ozkanca and Flint 2002). In addition, chemotaxis Aave_0780* Unnamed protein product NF proteins, such as methyl-accepting chemotaxis proteins Aave_0925* General secretion pathway protein G 2165 (MCPs), of COG0840 that share a conserved C-terminal Aave_1264* NADH dehydrogenase subunit B 377 methyl-accepting domain involved in sensory adaptation Aave_1273* NADH-ubiquinone oxidoreductase subunit 4L 713 (Le Moual and Koshland 1996) were also detected under Aave_1275* NADH-quinone oxidoreductase subunit M 1008 in vivo conditions, although different MCPs have been Aave_1519* Unnamed protein product 589 detected under each growth condition, which share no Aave_1797* Peptidase M14, carboxypeptidase A 2866 sequence homology with COG0840. The members of Aave_2204* Succinate dehydrogenase subunit C 2009 COG0840 belong to a family of bacterial chemotaxis Aave_2223* Ubiquinol oxidase subunit II 1622 receptors that transmit information from the environment to Aave_2800* OmpA/MotB domain-containing protein 2885 the inside of the cells, which then triggers a two-compo- Aave_3154* UspA domain-containing protein 589 nent phosphorylation cascade (Varughese 2002). Aave_3405* OmpA domain-containing protein 2885 Therefore, the existence of MCPs only under in vivo Aave_4344 Porin 3203 conditions indicated that chemotaxis may be related to the Aave_4565* Lipase 3240 virulence of A. avenae subsp. avenae strain RS-1. Mutation Aave_4679 Tfp pilus assembly protein major pilin PilA 2165 of MCP-encoding genes in both Dickeya dadantii and Ral- Aave_4763 PpiC-type peptidyl-prolyl cis–trans isomerase NF stonia solanacearum are economically important pathogens Acav_0258* Aromatic hydrocarbon degradation membrane 2067 of plants, which results in significantly reduced virulence protein on host plants (Büttner and Bonas 2010). These reports Acav_1536* Glycine/betaine/l-proline ABC transporter 1125 ATPase suggest that motility and chemotaxis may be required for full virulence. Acav_3412 l-lactate transport 1862 Chew protein of COG0835 was also detected under in Acav_4039* Transport-associated protein 1652 vivo conditions. It plays a key role in pairing receptors to * Common to both in vivo versus LB medium and leaf extract the bacterial chemotaxis intracellular signaling system and medium versus M9 medium then to CheY, thereby affecting flagellar rotation (Jagan- nadham 2008). Furthermore, extracellular solute-binding Identification of proteins under different growth condi- proteins (COG0834) were also identified, which work as tions not only revealed the proteome profiling ofA. ave- chemoreceptor, in metabolism, in detection of transport nae subsp. avenae strain RS-1 (Table S1–4), but also pro- system’s constituent, and as signal transduction pathway vided a basis for comparative analysis of MPs in A. avenae inhibitors in gram-negative bacteria (Tam and Saler 1993). subsp. avenae strain RS-1 (Tables 2, 3). In addition, path- The presence of these extracellular solute-binding proteins ogenicity-related proteins could be identified by evaluat- under in vivo conditions rather than LB medium conditions ing the differentially expressed MPs under in vivo versus indicates that these proteins may play a role in host adapta- LB medium conditions (Table 2), whereas proteins associ- tion and virulence of A. avenae subsp. avenae strain RS-1. ated with the adaptation of bacteria to rice hosts could be Acquirement of iron, zinc or manganese by ABC trans- determined by examining the differentially expressed MPs porters for the growth of bacterial pathogens during inva- under LE medium versus M9 medium conditions (Table 3). sion to the host is a prerequisite (Brown et al. 2001; Wes- ton and Ryan 2012; Detmers et al. 2001; Shah and Swiatle MPs expression under in vivo versus LB medium 2008). In agreement with previous studies, peptides, amine and amino acid ABC transporters were identified in our A marked difference between in vivo and LB medium MPs study in in vivo conditions, which may also be important expression was noted by the one-dimensional SDS–PAGE for virulence of A. avenae subsp. avenae strain RS-1. The protein pattern of A. avenae subsp. avenae strain RS-1 identification of these membrane transport proteins gener- (Fig. 1). Further LC–MS/MS analysis revealed that 26 ally involved in the movement of small molecules, ions or MPs such as OmpA, NADH dehydrogenase, MCPs were macromolecules. The expression of these proteins under uniquely expressed under in vivo, but not LB medium con- in vivo proteome indicates that these proteins may sup- ditions (Table 2). OmpA (COG2885), which is found in port in the movement of substances that help A. avenae many Gram-positive and Gram-negative bacteria, encodes subsp. avenae strain RS-1 for survival and adaptation to

1 3 Arch Microbiol (2015) 197:673–682 679 the rice host. The type VI secretion system (T6SS) is bac- medium conditions (Table 3). These proteins are involved terial protein injection machinery with roles in virulence, in energy production, assimilation and conversion, includ- symbiosis, interbacterial interaction, antipathogenesis and ing NADH dehydrogenase subunit B, peptidoglycan gly- environmental stress responses. Interestingly, T6SS core cosyltransferase, glycine/betaine/l-proline ABC transporter component OmpA/MotB was also identified under in vivo ATPase, peptidase M14, carboxypeptidase A, peptidogly- conditions, suggesting that T6SS proteins could play sig- can glycosyltransferase and PpiC-type peptidyl-prolyl cis– nificant role in niche adaptation and virulence (Mougous trans isomerase. Some virulence-related proteins were also et al. 2006; Ibrahim et al. 2012). Another important family detected under LE medium conditions, including OmpA of protein identified under in vivo conditions by LC–MS/ domain-containing protein, general secretion pathway pro- MS includes NADH-quinone oxidoreductase (COG1008), tein G or OmpA/MotB domain-containing protein, and type NADH-ubiquinone oxidoreductase (COG0713) and NADH four pili (TFP). Most of these identified proteins are similar dehydrogenase (COG0377), which catalyzes the oxidation to the list of proteins identified in in vivo rice plant condition. of NADH, the reduction of ubiquinone and the transfer of It substantiates that proteomics studies in combination with 4H+/NADH across the coupling membrane, respectively artificial plant host mimic medium could be a potential alter- (Friedrich et al. 1995). Furthermore, Seong et al. (2005) native to study the host–pathogen interaction. revealed that NADH: ubiquinone oxidoreductase has been LC–MS/MS analysis in this study indicated that there was found to be involved in pathogenicity through restriction- a high similarity between LE medium versus M9 medium enzyme-mediated insertion (REMI) and directed mutagen- uniquely expressed proteins (Table 2) with in vivo versus esis in plants. Our study indicates that the presence of LB medium uniquely expressed proteins (Table 3). However, NADH not only could be involved in bacterial pathogenic- several important proteins were identified uniquely between ity, but also be related to energy metabolism of A. avenae the two conditions, indicating the complexity of interaction subsp. avenae strain RS-1. The several members of univer- between A. avenae subsp. avenae strain RS-1 and rice hosts. sal stress proteins (UspA) family of COG0589 are found in The proteins, which are identified in the presence of plant the genomes of bacteria, fungi, archaea plant and protozoa, materials, but not in vivo conditions, include lytic transgly- but the biological or biochemical functions of these pro- cosylase catalytic subunit, type 4 pilus assembly protein, teins are still unclear. However, the identification of univer- peptidyl-prolyl cis–trans isomerase, peptidoglycan glyco- sal stress protein UspA domain-containing protein in an in syltransferase and L-lactate transport proteins (Table 3). vivo condition raises new avenues of investigation. Interestingly, lytic transglycosylase catalytic subunit is a bac- In addition, several other proteins that are associated terial exo-muramidase that cleaves the cell wall peptidogly- with metabolism and regulation, like acid phosphatase can. This result may support that A. avenae subsp. avenae (COG3511), lipase (COG3240), respiratory nitrate reduc- strain RS-1 have catalytic activity for cell wall peptidoglycan tase subunit beta (COG1140), succinate dehydrogenase cleaves under LE medium conditions. subunit C (COG2009), and preprotein translocase subunit It is well known that the physiology of in vivo cultured YajC (COG1862), were also identified. bacteria was affected by several biotic and abiotic factors, while these factors have not yet been well analyzed. The MPs expression under LE medium versus M9 medium differences in few proteins between LE medium and in vivo conditions may correspond to these factors. Further- Plant environmental adaptation by bacteria, which occurs more, the LC–MS/MS method in this study mainly focus at different cellular levels, is not only very complicated but on the detection of the presence or absence of proteins, and also sometimes significant and sometimes devastating. A more attention should be paid to this differential expres- variety of advanced approaches are essential to acquire a sion in common proteins. However, result from this study broad understanding at a systems level (Schell et al. 2011). indicated that the proteome of bacteria whose genome has It has been well known that proteome has been regarded as been sequenced may be able to be analyzed in plant mimic the final products of the gene expression. In this study, we, medium instead of in vivo host conditions. here first time, have revealed that basic proteome profiling of plant pathogenic bacterium in the plant mimic condi- Protein structural and molecular characterization tions and in vivo plant conditions compared to LB and M9 media used as the control, respectively (Schell et al. 2011; The results of our studies showed that almost all proteins Mehta and Rosato 2001). such as NADH dehydrogenase (locus tag Aave_1264), Comparative analysis of the membrane proteome indi- NADH-ubiquinone oxidoreductase subunit 4L (locus tag cated that 66 proteins were shared by under both LE medium Aave_1273), NADH-quinone oxidoreductase subunit M and M9 medium (Table S3–4). Nonetheless, 21 proteins were (locus tag Aave_1275) and OmpA/MotB domain-contain- detected under LE medium conditions, but not under M9 ing protein (locus tag Aave_2800) have domain sequence

1 3 680 Arch Microbiol (2015) 197:673–682

Fig. 4 Traditional exploratory visual analysis of multiple sequence alignment. A depiction of the multiple sequence alignment of four selected proteins shows sequence and domain conservation leading to the role in virulence and various metabolic processes conservation with the proteins of other notorious disease- avenae strain RS-1 although many proteins designated as causing bacteria like Acidovorax citruli, Burkholderia ceno- abundant class of proteins were identified under all growth cepacia, Candidatus Burkholderia kirkii, Ralstonia solan- conditions. Indeed, 26 MPs were differentially expressed acearum, Burkholderia cepacia JBK9, Pseudomallei group under in vivo versus LB medium conditions, while 21 MPs as shown in Fig. 4. Most of the proteins showed significant were differentially expressed under LE medium versus M9 functional conservation with these devastating bacteria. The medium conditions. Furthermore, the 17 MPs were com- computational analysis not only revealed the functional monly expressed under both in vivo versus LB medium identification of virulence or cell metabolism proteins iden- conditions and LE medium versus M9 medium conditions, tified under in vivo or LE condition, but also supported this indicating that plant materials could provide an alternate observation that LE may be an alternate method to analyze for the identification of important protein candidates in the in vivo expression of proteins to mimic plant conditions. plant–microbe interactions. These identified proteins were not only involved in virulence, but also may be related with metabolic and regulatory processes such as energy and Conclusion stress response based on the functional prediction.

Our results clearly revealed that the expression of MPs was Acknowledgments This work was supported by Zhejiang Provin- dependent on the growth conditions of A. avenae subsp. cial Natural Science Foundation of China (R13C140001; Y3090150),

1 3 Arch Microbiol (2015) 197:673–682 681

National Natural Science Foundation of China (31371904), Zheji- Laemmli UK, Molbert E, Showe M, Kellenbe E (1970) Form-deter- ang Provincial Project (2014C32010), the Fundamental Research mining function of genes required for assembly of head of bacte- Funds for the Central Universities, the Agricultural Ministry of China riophage-T4. J Mol Biol 49:99–113 (nyhyzx 201303015; 201003029; 201003066) and Key Subject Con- Le Moual H, Koshland DE (1996) Molecular evolution of the C-ter- struction Program of Zhejiang for Modern Agricultural Biotechnol- minal cytoplasmic domain of a superfamily of bacterial receptors ogy and Crop Disease Control (2010DS700124-KF1101; KF1203; involved in taxis. J Mol Biol 261:568–585 KF1309; KF1410). Li B, Ibrahim M, Ge MY, Cui ZQ, Sun GC, Xu F, Kube M (2014) Transcriptome analysis of Acidovorax avenae subsp. avenae cultivated in vivo and co-culture with Burkholderia seminalis. Sci Rep 4:5698 References Liu H, Tian WX, Ibrahim M, Li B, Zhang GQ, Zhu B, Xie GL (2012) Characterization of pilP, a gene required for twitching motil- Andrade AE, Silva LP, Pereira JL (2008) In vivo proteome analysis ity, pathogenicity, and biofilm formation ofAcidovorax avenae of Xanthomonas campestris pv. campestris in the interaction subsp. avenae RS-1. Eur J Plant Pathol 134:551–560 with the host plant Brassica oleracea. FEMS Microbiol Lett Marreddy RKR, Pinto JPC, Wolters JC, Geertsma ER, Fusetti F, Per- 281:167–174 mentier HP, Kuipers OP, Bert P (2011) The response of Lactococ- Astua-Monge G, Freitas-Astua J, Bacocina G, Roncoletta J, Carvalho cus lactis to membrane protein production. PLoS One 6:e24060 SA, Machado MA (2005) Expression profiling of virulence and Matsumoto K, Matsusaki H, Taguchi K, Seki M, Doi Y (2002) Iso- pathogenicity genes of Xanthomonas axonopodis pv. citri. J Bac- lation and characterization of polyhydroxyalkanoates inclusions teriol 187:1201–1205 and their associated proteins in Pseudomonas sp. 61-3. Biomac- Barnard AML, Bowden SD, Burr T, Coulthurst SJ, Monson RE, Sal- romolecules 3:787–792 mond GPC (2007) Quorum sensing, virulence and secondary Mehta A, Rosato YB (2001) Differentially expressed proteins in the metabolite production in plant soft-rotting bacteria. Philos Trans interaction of Xanthomonas axonopodis pv. citri with leaf extract R Soc Lond B Biol Sci 362:1165–1183 of the host plant. Proteomics 1:1111–1118 Brown JS, Ogunniyi AD, Woodrow MC, Holden DW, Paton JC Mole BM, Baltrus DA, Dangl JL, Grant SR (2007) Global virulence (2001) Immunization with components of two iron uptake ABC regulation networks in phytopathogenic bacteria. Trends Micro- transporters protects mice against systemic Streptococcus pneu- biol 15:363–371 monia infection. Infect Immun 69:6702–6706 Morita Y, Sobel ML, Poole K (2006) Antibiotic inducibility of the Büttner D, Bonas U (2010) Regulation and secretion of Xanthomonas MexXY multidrug efflux system of Pseudomonas aeruginosa; virulence factors. FEMS Microbiol Rev 34:107–133 involvement of the antibiotic-inducible PA5471 gene product. Cash P (2011) Investigating pathogen biology at the level of the pro- Antimicrob Agents Chemother 188:1847–1855 teome. Proteomics 11:3190–3202 Mougous JD, Cuff ME, Raunser S, Shen A, Zhou M, Gifford CA, Chandler D (2005) Interfaces and the driving force of hydrophobic Goodman AL, Joachimiak G, Ordoñez CL, Lory S, Walz T, assembly. Nature 437:640–647 Joachimiak A, Mekalanos JJ (2006) A virulence locus of Pseu- Chen YY, Wu CH, Lin JW, Weng SF, Tseng YH (2010) Mutation domonas aeruginosa encodes a protein secretion apparatus. Sci- of the gene encoding a major outer-membrane protein in Xan- ence 312:1526–1530 thomonas campestris pv. campestris causes pleiotropic effects, Nandi B, Nandy RK, Sarkar A, Ghose GC (2005) Structural features including loss of pathogenicity. Microbiol 156:2842–2854 properties and regulation of the outer-membrane protein W Detmers FJM, Lanfermeijer FC, Poolman B (2001) Peptides and ATP (OmpW) of Vibrio cholera. Microbiol 151:2975–2986 binding cassette peptide transporters. Res Microbiol 152:245–258 Olsen JV, de Godoy LM, Li G, Macek B, Mortensen P, Pesch R, Friedrich T, Steinmüller K, Weiss H (1995) The proton-pumping res- Makarov A, Lange O, Horning S, Mann M (2005) Parts per mil- piratory complex I of bacteria and mitochondria and is homo- lion mass accuracy on an orbitrap mass spectrometer via lock logue of chloroplasts. FEBS Lett 367:107–111 mass injection into a C-trap. Mol Cell Proteomics 4:2010–2021 Gourion B, Rossignol M, Vorholt JA (2006) A proteomic study Ozkanca R, Flint KP (2002) The effect of starvation stress on the of Methylobacterium extorquens reveals a response regula- porin protein expression of Escherichia coli in lake water. Lett tor essential for epiphytic growth. Proc Natl Acad Sci USA App Microbiol 35:533–537 103:13186–13191 Petersen TN, Bruna SK, von Heijne G, Nielsen H (2011) SignalP 4.0: Ibrahim M, Shi Y, Qiu H, Li B, Jabeen A, Li LP, Liu H, Kube M, Xie discriminating signal peptides from transmembrane regions. Nat GL, Wang YL, Sun GC (2012) Differential expression of in vivo Methods 8:785–786 and in vitro protein profile of outer membrane of Acidovorax Schell MA, Zhao P, Wells L (2011) Outer membrane proteome of avenae subsp. avenae. PLoS One 7:e49657 Burkholderia pseudomallei and Burkholderia mallei from Jagannadham MV (2008) Identification of proteins from mem- diverse growth conditions. J Proteome Res 10:2417–2424 brane preparations by a combination of MALDI TOF-TOF and Seong K, Hou HM, Tracy M, Kistler HC, Xu JR (2005) Random LC-coupled linear ion trap MS analysis of an Antarctic bacte- insertional mutagenesis identifies genes associated with viru- rium Pseudomonas syringae Lz4W, a strain with unsequenced lence in the wheat scab fungus Fusarium graminearum. Phyto- genome. Electrophoresis 29:4341–4350 pathology 95:744–750 Jagannadham MV, Chowdhury C (2012) Differential expression of Shah P, Swiatle E (2008) A multifaceted role for polyamines in bacte- membrane proteins helps Antarctic Pseudomonas syringae to accli- rial pathogens. Mol Microbial 68:4–16 matize upon temperature variations. Proteomics 75:2488–2499 Tam R, Saler MH (1993) Structural, functional, and evolutionary rela- Kall L, Krogh A, Sonnhammer ELL (2004) A combined transmem- tionships among extracellular solute-binding receptors of bacte- brane topology and signal peptide prediction method. J Mol Biol ria. Microbiol Rev 57:320–346 338:1027–1036 Teng CH, Xie Y, Shin S, Di Cello F, Paul-Satyaseela M, Cai B, Kim Klebba PE (2005) The porinologist. J Bact 187:8232–8236 KS (2006) Effects of ompA deletion on expression of type 1 fim- Knief C, Delmotte N, Vorholt JA (2011) Bacterial adaptation to life in briae in Escherichia coli K1 strain RS218 and on the associa- association with plants—a proteomic perspective from culture to tion of E. coli with human brain microvascular endothelial cells. in situ conditions. Proteomics 11:3086–3105 Infect Immun 74:5609–5616

1 3 682 Arch Microbiol (2015) 197:673–682

Varughese KI (2002) Molecular recognition of bacterial phosphorelay Yang CL, Li B, Ge MY, Zhou KL, Wang YL, Luo J, Ibrahim M, Xie proteins. Curr Opin Microbiol 5:142–148 GL, Sun GC (2014) Inhibitory effect and mode of action of chi- Weston LA, Ryan PR (2012) Mechanisms of cellular transport and tosan solution against rice bacterial brown stripe pathogen Aci- release of allele chemicals from plant roots into the rhizosphere. dovorax avenae subsp. avenae RS-1. Carbohyd Res 391:48–54 J Exp Bot 63:3445–3454 Yu NY, Wagner JR, Laird MR, Melli G, Rey S, Lo R, Dao P, Sahi- Xie GL, Zhang GQ, Liu H, Lou MM, Tian WX, Li B, Zhou XP, nalp SC, Ester M, Foster LI, Brinkman FS (2010) PSORTb 3.0: Zhu B, Jin GL (2011) Genome sequence of the rice-pathogenic improved protein subcellular localization prediction with refined bacterium Acidovorax avenae subsp. avenae RS-1. J Bacteriol localization subcategories and predictive capabilities for all 193:5013–5014 prokaryotes. Bioinformatics 26:1608–1615

1 3