Introduction

Over the last few decades, the emergence and spread of antibiotic resistance is accelerated by the misuse and overuse of antibiotics (Laxminarayan and Chaudhury 2016). Most pathogenic bacteria, including methicillin-resistant Staphylococcus aureus (MRSA), multidrug-resistant (MDR) Salmonella Typhimurium, and carbapenem-resistant Klebsiella pneumoniae, possess several antibiotic resistance mechanisms such as enzymatic inactivation, efflux pump, and membrane permeability barrier, and metabolic pathway modification (Nikaido 2009). The infections caused by antibiotic-resistant bacteria have become a serious clinical and public health problem due to the frequent failures in chemotherapeutic treatments (Currie et al. 2014; Ferri et al. 2017). Specifically, the MDR bacterial infections are the leading causes of mortality and morbidity (van Duin and Paterson 2016). Therefore, the development of rapid, sensitive, and accurate detection methods is a primary concern for the MDR bacterial infections.

The molecular-based techniques such as PCR and microarray hybridization techniques have been used to detect antibiotic-resistant bacteria (Anjum et al. 2017). The PCR-based diagnostic methods including standard, real-time, and multiplex PCRs have been developed for detecting the presence of antibiotic resistance genes in bacteria. The high-throughput diagnostic microarray is a reliable qualitative and quantitative method for detecting various antibiotic resistance genes (Friedrich et al. 2010). These are well known as the most sensitive and reliable methods for detecting antibiotic-resistant bacteria. However, the PCR-based methods are only suitable for detecting biomarkers, which are identified based on prior knowledge (Anjum 2015; Moran et al. 2017). These detection methods tend to underestimate or overestimate due to the discrepancy between phenotype resistance and genotype resistance (Card et al. 2013; Strauss et al. 2015).

Recently, proteomic analysis has received a great attention as a promising tool for identifying antibiotic-resistant bacteria (Chen et al. 2017). Proteomic analysis helps to uncover the metabolic pathways in antibiotic-resistant bacteria (da Costa et al. 2015). The comparative proteomic studies can contribute to identify differential expressed proteins in specific antibiotic-resistant bacteria and also understand the antibiotic resistance mechanisms in regard with the protein expression patterns (Vranakis et al. 2014; Park et al. 2016). However, more systematic approach is needed to investigate protein profiles in association with antibiotic resistance in bacteria. Therefore, the objective of this study was to compare the proteomic profiles of laboratory-driven and clinically isolated antibiotic-resistant S. Typhimurium, K. pneumoniae, and S. aureus.

Materials and methods

Bacterial strains and culture conditions

Strains of S. Typhimurium ATCC 19585 (STWT), K. pneumoniae ATCC 23357 (KPWT), and S. aureus ATCC 15564 (SAWT) were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). The stepwise selection method was used to induce the antibiotic-resistant S. Typhimurium ATCC 19585, K. pneumoniae ATCC 23357, and S. aureus ATCC 15564, which were assigned to ciprofloxacin-induced S. Typhimurium (STCIP), ciprofloxacin-induced K. pneumoniae (KPCIP), ciprofloxacin-induced S. aureus (SACIP), and oxacillin-induced S. aureus (SAOXA), respectively (Birošová and Mikulášová 2009; Kim et al. 2016). The clinically acquired antibiotic-resistant S. Typhimurium CCARM 8009 (STCCARM), K. pneumoniae CCARM 10237 (KPCCARM), and S. aureus CCARM 3080 (SACCARM) were obtained from Culture Collection of Antibiotic-Resistant Microbes (CCARM, Seoul, Korea). All strains were cultured aerobically in TSB at 37 °C for 20 h and then collected by a centrifugation at 3000×g for 20 min at 4 °C. The harvested cells were washed twice with phosphate-buffered saline (PBS, pH 7.2).

Antibiotic susceptibility assay

The antibiotic susceptibilities of STWT, STCIP, STCCARM, KPWT, KPCIP, KPCCARM, SAWT, SACIP, SAOXA, and SACCARM were determined according to broth microdilution assay. Antibiotic stock solutions, including gentamicin, tobramycin, chloramphenicol, erythromycin, ciprofloxacin, levofloxacin, norfloxacin, tetracycline, ampicillin, cefotaxime, cefoxitin, ceftriaxone, and piperacillin, were prepared at the level of 1024 µg/ml dissolving in sterile distilled water. Each antibiotic solution (100 µl) was diluted to 1:2 in 96-well microtiter plates and inoculated with each test strain at approximately 105 cfu/ml in 100 µl. The 96-well microtiter plates were incubated for 18 h at 37 °C to determine minimum inhibitory concentration (MIC) at which there was no visible growth. The susceptible (S), intermediate (I), and resistant (R) strains were assigned based on the MIC breakpoints (CLSI 2015).

Protein extraction

The cultured bacterial cells were harvested at 3000×g at 4 °C for 20 min and washed three times with PBS (pH 7.2). The collected cells were suspended in lysis buffer-containing 7 M urea, 4% CHAPS, 100 mM DTT, 40 mM Tris (pH 8.8), and protease inhibitor cocktail (Complete™; Roche, Mannheim, Germany) for 2-DE analysis. The mixtures were disrupted by sonication at 4 °C for 30 s and centrifuged at 36,000×g at 4 °C for 50 min to remove cell debris. Protein concentration was measured by Bradford method (Bradford 1976).

2-DE analysis

The protein extracts (100 mg each) suspended in sample buffer-containing 7 M urea, 2 M thiourea, 4.5% CHAPS, 100 mM DTE, 40 mM Tris, and pH 8.8) were loaded to the immobilized pH 3-10 nonlinear gradient strips (Amersham Bio-science, Uppsala, Sweden). Isoelectric focusing (IEF) was performed for 80,000 Vh. A 9–16% linear gradient polyacrylamide gel (18 cm × 20 cm × 1.5 mm) was prepared at a constant current of 40 mA for the second dimension. Proteins were fixed in 40% methanol and 5% phosphoric acid for 1 h. The gels were stained with CBB G-250 for 12 h, decolored with water, and scanned using a Bio-Rad (Richmond, CA, USA) GS710 densitometer and converted into electric files. Image Master Platinum 5.0 image analysis program (Amersham Biosciences) was used to analyze the scanned gels (Cho et al. 2005; Lee et al. 2010).

In gel digestion

The highly up- and down-regulated paired and unpaired protein spots (> twofold) were transferred to a 1.5 mL tube, washed with 100 μL of distilled water (DW) for protein digestion, and then mixed with 100 μl of 50 mM NH4HCO3 (pH 7.8) and acetonitrile (6:4) for 10 min. After destaining the Coomassie brilliant blue G250 dye, the bands were dried for 10 min using a high-speed vacuum concentrator (LaBoGeneAps, Lynge, Denmark). The digestion was performed by shaking at 37 °C for 16 h using sequence-grade modified trypsin (Promega Co., Madison, WI, USA) (Cho et al. 2005; Lee et al. 2010).

LC–MS/MS

Nano LC–MS/MS analysis was carried out with an Easy n-LC (Thermo Fisher San Jose, CA, USA) interfaced to a nano-electrospray source and LTQ Orbitrap XL mass spectrometer (Cho et al. 2005; Lee et al. 2010). The capillary column (150 × 0.075 mm; Proxeon, Odense M, Denmark) and Magic C18 stationary phase were used for LC–MS/MS analysis. The mobile phase A and B were 0.1% formic acid, respectively, in deionized water and acetonitrile. The gradient profile was programmed to obtain a linear increase (%B, min: 5–40, 50; 40–60, 20; 60–80, and 5) at a flow rate of 300 nL/min. Mass spectra were collected using data-dependent acquisition with full mass scan (400–1800 m/z). The MS/MS scans were 1 microscan on the linear ion trap (LTQ). The temperature, spray, and collision energy were 200 °C, 1.5–2.0 kV, and 35%, respectively. The peptide sequence was identified using the MASCOT software.

Protein identification

The LC–MS/MS data were processed using Mascot Search (Matrix Science Ltd., London, UK). Database search criteria were taxonomy; Salmonella (622410 sequences), Klebsiella (7517848 sequences), Staphylococcus (18486459 sequences), fixed modification; carbamidomethylated at cysteine residues, variable modification; oxidized at methionine residues, and maximum allowed missed cleavage; 2, MS tolerance; 10 ppm, MS/MS tolerance; 0.8 Da. The peptides were searched with a statistically significant threshold values (p < 0.05).

Results and discussion

The emergence of antibiotic-resistant bacteria is of great public health concern due to the limited choice of antibiotic therapy (Muroi et al. 2012). The prevention and treatment of infectious diseases caused by antibiotic-resistant bacteria result in the loss of revenues. Therefore, a reliable discrimination of antibiotic resistance in bacteria is essential for the prevention and treatment of infectious diseases. The resistance-associated proteins could be useful biomarkers for detecting antibiotic-resistant pathogens. Thus, it is worth investigating for the proteomic discrimination between antibiotic-sensitive and antibiotic-resistant strains. The laboratory-driven antibiotic-resistant S. Typhimurium, K. pneumoniae, and S. aureus were induced to compare the differentially expressed proteins with antibiotic resistance.

Classification of differentially expressed proteins in S. Typhimurium, K. pneumoniae, and S. aureus

The antibiotic susceptibilities of wild-type (WT), ciprofloxacin (CIP) and/or oxacillin (OXA)-induced, and clinically isolated resistant (CCARM) S. Typhimurium (STWT, STCIP, and STCCARM), K. pneumoniae (KPWT, KPCIP, and KPCCARM), and S. aureus (SAWT, SACIP, SAOXA, and SACCARM) were determined after resistance induction by ciprofloxacin or oxacillin (Table 1). The wild-type strains were sensitive to all antibiotics used in this study with the exception of erythromycin for STWT, ampicillin and erythromycin for KPWT, and cefotaxime, and ceftriaxone for SAWT. KPCCARM and SACCARM were highly resistance to most antibiotics, showing MICs of more than 512 μg/ml (Table 1). The protein expression profiles in STWT, STCIP, STCCARM, KPWT, KPCIP, KPCCARM, SAWT, SACIP, SAOXA, and SACCARM were compared, as shown in Fig. 1. In total, 1034, 1013, and 877 protein spots were found in STWT, STCIP, and STCCARM, respectively (Fig. 1a). There were 960, 1126, and 696 protein spots observed in KPWT, KPCIP, and KPCCARM, respectively (Fig. 1b). As shown in Fig. 3b, a total of 444, 551, 581, and 488 protein spots were identified in SAWT, SACIP, SAOXA, and SACCARM, respectively. The Venn diagram analysis illustrated the relations between paired and unpaired protein expression. The unique proteins in STWT, STCIP, and STCCARM were 97, 107, and 57, respectively, accounting for 9.4%, 10.6%, and 6.5%. The unique proteins in KPWT, KPCIP, and KPCCARM were 147 (15.3%), 302 (26.8%), and 80 (11.5%), respectively. The unique proteins in SAWT, SACIP, SAOXA, and SACCARM were 49 (11.0%), 67 (12.2%), 137 (23.6%), and 68 (13.9%), respectively. After antibiotic induction, the specific protein spots were increased in STCIP, KPCIP, SACIP, and SAOXA when compared to the wild-type strains. The results suggest that antibiotic resistance can cause the modification in bacterial protein synthesis. The paired and unpaired protein spots obtained from S. Typhimurium (STWT, STCIP, and STCCARM), K. pneumoniae (KPWT, KPCIP, and KPCCARM), and S. aureus (SAWT, SACIP, SAOXA, and SACCARM) were exhibited, as shown in Figs. 2, 3, and 4, respectively. The protein spots were distributed in broad pH range from 3 to 10 and molecular weight from 10 to 200 kDa. More than 90% of total protein spots were shared between S. Typhimurium strains, while 97, 57, and 107 spots were specific to STWT, STCIP, and STCCARM, respectively (Fig. 2). In all protein spots found in K. pneumoniae strains, the specific spots were 147, 302, and 80 in KPWT, KPCIP, and KPCCARM, respectively (Fig. 3). In the pairwise comparison of the proteins found in S. aureus strains, the commonly expressed proteins of SAWT, SACIP, SAOXA, and SACCARM were 49, 67, 137, and 68, respectively (Fig. 4).

Table 1 Minimum inhibitory concentrations (MICs, μg/ml) of S. Typhimurium (STWT, STCIP, and STCCARM)a, K. pneumoniae (KPWT, KPCIP, and KPCCARM)a, and S. aureus (SAWT, SACIP, SAOXA, and SACCARM)a
Full size table
Fig. 1
figure 1

Venn diagram of protein distribution between strains of S. Typhimurium (A; STWT, STCIP, and STCCARM), K. pneumoniae (B; KPWT, KPCIP, and KPCCARM), and S. aureus (C; SAWT, SACIP, SAOXA, and SACCARM). STWT, KPWT, and SAWT indicate the wild-type strains of S. Typhimurium ATCC 19585, K. pneumoniae ATCC 23357, and S. aureus ATCC 15564, respectively. STCIP, KPCIP, SACIP, and SAOXA represent ciprofloxacin-induced S. Typhimurium, K. pneumoniae, S. aureus, and oxacillin-induced S. aureus. STCCARM, KPCCARM, and SACCARM denote clinically acquired antibiotic-resistant S. Typhimurium, K. pneumoniae, and S. aureus, respectively

Full size image
Fig. 2
figure 2

Two-dimensional gel electrophoresis patterns of wild-type S. Typhimurium (a; STWT), ciprofloxacin-induced S. Typhimurium (b; STCIP), and clinically acquired antibiotic-resistant S. Typhimurium (c; STCCARM). Each number indicates the protein spot identified in this study

Full size image
Fig. 3
figure 3

Two-dimensional gel electrophoresis patterns of wild-type K. pneumoniae (a; KPWT), ciprofloxacin-induced K. pneumoniae (b; KPWT), and clinically acquired antibiotic-resistant K. pneumoniae (c; KPCCARM). Each number indicates the protein spot identified in this study

Full size image
Fig. 4
figure 4

Two-dimensional gel electrophoresis patterns of wild-type S. aureus (a; SAWT), ciprofloxacin-induced S. aureus (b; SACIP), oxacillin-induced S. aureus (c; SAOXA), and clinically acquired antibiotic-resistant S. aureus (D; SACCARM). Each number indicates the protein spot identified in this study

Full size image

Characteristics of the paired proteins identified in S. Typhimurium, K. pneumoniae, and S. aureus

The paired protein spots up- or down-regulated by more than twofold were identified using LC–MS/MS (Tables 2, 3, 4, 5, 6, 7, and 8). The identified proteins were characterized by pI, pH, and molecular mass, as shown in Tables 2, 3, 4, 5, 6, 7, and 8. The pI values affect the protein solubility at particular pH values, inducing the fractions of negative-charged acidic (pH > pI) and positive-charged basic (pH < pI) proteins (Weiller et al. 2004).

Table 2 Identification of paired spots that were differentially expressed in ciprofloxacin-induced resistant S. Typhimurium ATCC 19585 (STCIP) compared to wild-type S. Typhimurium ATCC 19585 (STWT)
Full size table
Table 3 Identification of paired spots that were differentially expressed in clinically isolated S. Typhimurium CCARM 8009 (STCCARM) compared to wild-type S. Typhimurium ATCC 19585 (STWT)
Full size table
Table 4 Identification of paired spots that were differentially expressed in ciprofloxacin-induced resistant K. pneumoniae ATCC 23357 (KPCIP) compared to wild-type K. pneumoniae ATCC 23357 (KPWT)
Full size table
Table 5 Identification of paired spots that were differentially expressed in clinically isolated K. pneumoniae CCARM 10237 (KPCCARM) compared to wild-type K. pneumoniae ATCC 23357 (KPWT)
Full size table
Table 6 Identification of paired spots that were differentially expressed in ciprofloxacin-induced resistant S. aureus ATCC 15564 (SACIP) compared to wild-type S. aureus ATCC 15564 (SAWT)
Full size table
Table 7 Identification of paired spots that were differentially expressed in oxacillin-induced resistant S. aureus ATCC 15564 (SAOXA) compared to wild-type S. aureus ATCC 15564 (SAWT)
Full size table
Table 8 Identification of paired spots that were differentially expressed in clinically isolated S. aureus CCARM 3080 (SACCARM) compared to wild-type S. aureus ATCC 15564 (SAWT)
Full size table

Protein identification in S. Typhimurium

The protein folding-, stress response-, cell motility-, and efflux pump-related proteins were up-regulated in the STCIP compared to the STWT (Table 2). The outer membrane protein (OmpX, Spot 141) and flagellin (FliC, Spot 536) were up-regulated by more than fourfold when compared to the STWT. The expression levels of arginine deiminase (ArcA, Spot 617), phosphopyruvate hydrate (Eno, Spot 640), and ribose-phosphate pyrophosphokinase (PrsA, Spot 823) were increased by more than three, five, and tenfold, respectively. However, the outer membrane protein assembly factor (YaeT, Spot 235) and tail protein (Rbp, Spot 887) were down-regulated by more than four and sevenfold in STCIP. The expression of MalE (maltose ABC transporter periplasmic protein, Spot 739) was decreased by more than eightfold when compared to the STWT. The OmpX is a small outer membrane protein (18 kDa) contributing to the modulation of outer membrane permeability and adaptability. The overexpression of OmpX is associated with the down-regulation of several porin proteins (OmpC, OmpF, LamB, and Omp36), resulting in the increased resistance to various classes of antibiotics such as β-lactams and fluoroquinolones (Dupont et al. 2004, 2007). The OMP profile in antibiotic-resistant bacteria was similar to that in parent strain exposed to the same antibiotic (Morita et al. 2014). In addition, the alteration in OMP profile was directed by the plasmid-encoding antibiotic-resistant genes (Peng et al. 2017). The OMPs associated with antibiotic resistance involve various regulatory mechanisms (Peng et al. 2017), highlighting the importance of identification and characterization of OMPs. The arginine deiminase (arcA) is associated with the efflux pump activity, leading to the enhanced resistance to antibiotics, dyes, and other molecules (Webber et al. 2009). The outer membrane porins and efflux pump systems play a vital role in mediating antibiotic resistance (Kim et al. 2006; Peng et al. 2017). The significant change in the expression levels of PrsA (10-fold) was observed in STCIP (Table 2), which might involve the increased resistance of STCIP to ciprofloxacin (128-fold), levofloxacin (32-fold), norfloxacin (64-fold), and piperacillin (> 16-fold) (Table 1). The bacterial proteins substantially contribute to the alteration in the resistance mechanisms to different classes of antibiotics (Lee et al. 2015). In contrast, the metabolism-, cell envelope-, and phage receptor-related proteins were down-regulated in STCIP. Maltoporin (LamB, Spot 609) and phage receptor-binding protein (Rbp, Spot 887) act as receptors for phages (Berkane et al. 2006), which were relatively predominant in STWT (Table 2). The outer membrane-related protein was not observed in the STCCARM compared to the STWT, whereas DNA starvation/stationary phase protection proteins (Dps, Spot 1153, Spot 1183, Spot 1290), alcohol dehydrogenase (AdhE, Spot 142), and DNA-directed RNA polymerase subunit beta (RpoB, Spot 87) were up-regulated by more than five, five, and fourfold, respectively (Table 3). The oxidation–reduction process-related proteins, SodB (Spot 1073) and YhhX (Spot 769), were commonly up-regulated in STCIP and STCCARM, respectively. However, the oligopeptide ABC transporter protein (MppA, Spot 462), maltose ABC transporter periplasmic protein (MalE, Spot 743), and ribose ABC transporter protein (RbsB, Spot 968) were down-regulated in STCCARM. Compared to STWT, STCCARM was highly resistant to ampicillin (128-fold), and piperacillin (> 16-fold), which might be associated with the high expression levels of stress response-related proteins (Dps, five-to-eightfold) (Table 3). The Dps regulated by the stationary phase sigma factor plays an important role in the resistance to oxidative stress, survival in macrophages, and induction of virulence factors (Halsey et al. 2004). The ABC transporter proteins were down-regulated in STCCARM (Table 3). The decreased expression of ABC transporter proteins (MppA, MalE, and RbsB) has been reported to lead the multiple antibiotic resistance in bacteria (Moussatova et al. 2008; Jones et al. 2014).

Protein identification in K. pneumoniae

The relative expression of paired protein spots in KPCIP compared to the KPWT was evaluated, as shown in Table 4. The paired protein spots including fimbrial-like protein (YadC, Spot 185), fimbrial subunit type 3 (FimA, Spot 778), heat shock protein (DnaK, Spot 447), and translation-related protein (RplB, Spot 1369) were up-regulated in KPCIP. However, the trigger factor (Tig, Spot 573), acetoin reductase (BudC, Spot 1390), and isocitrate dehydrogenase (AceK, Spot 675) were down-regulated in KPCIP (Table 4). The up-regulated proteins (YadC, FimA, and RplB) in KPCIP (Table 4) are associated with the cell adhesion and antibiotic resistance mechanisms (Lima et al. 2013). These proteins can function as primary receptors in phage infection (Rakhuba et al. 2010). These proteins might be associated with the resistance of KPCIP to chloramphenicol (> 32-fold), ciprofloxacin (32-fold), levofloxacin (6-fold), norfloxacin (128-fold), and sparfloxacin (64-fold) when compared to KPWT. The overexpressed proteins might be involved in fluoroquinolone resistance. In Table 5, the fimbrial protein (YadC, Spot 1375), energy production-related proteins (DhaD, Spots 821, 822, 874, and 1499, AdhE, Spot 874), and small heat shock protein 20 (Spot 1228) were up-regulated in the KPCCARM by more than 10-, 6-, and 18-folds, respectively. Similar to STCCARM, the expression levels of oligopeptide transport protein (MppA, spot 487) and periplasmic-binding protein (Spot 1046) were decreased in the KPCCARM (Table 5). The up-regulation of energy-related proteins (DhaD and AdhE) indicates the energy requirement of cells exposed to antibiotic stress (Table 5). Bacteria need high energy to repair damaged proteins and DNA and/or to pump antibiotics out of the cells. The stress-related proteins play an important roles in preventing protein aggregation by properly folding proteins and repairing misfolded proteins (Susin et al. 2006). The heat shock proteins contribute to the maintenance of antibiotic-induced damage to proteins. The highly up-regulated proteins (AdhE and DnaK) in KPCCARM could be used as biomarkers for detecting multidrug-resistant K. pneumoniae.

Protein identification in S. aureus

The differentially expressions of paired protein spots were evaluated in SACIP, SAOXA, and SACCARM compared to the SAWT (Tables 6, 7, and 8). The stress-related proteins (DnaK, GroEL), efflux pump-related protein (ArcB), translation-related proteins (Tuf, RpoB, Tsf), energy production-related proteins (IdhA, enolase, AdhE, GapA, and AckA), amino acid transport-related proteins (LeuS, ThrS, and Dat) were up-regulated by more than twofold in SACIP, SAOXA, and SACCARM. The L-lactate dehydrogenase (IdhA, Spot 1911) in SAOXA (Table 7) and ornithine transcarbamoylase (ArcB, Spot 1868) in SACCARM were up-regulated by more than three and tenfold, respectively (Table 8). The ArcB is the efflux pump-related protein belonging to the RND family involves in multiple antibiotic resistance in bacteria (Kumar et al. 2013). The energy is required for the synthesis of cell components and activity of efflux pump systems (Pieper et al. 2006). The up-regulated unique proteins were PflB in SACIP resistant to ciprofloxacin (64-fold), levofloxacin (128-fold), norfloxacin (32-fold), sparfloxacin (64-fold), RpoB in SAOXA resistant to ampicillin (8-fold), ceftriaxone (8-fold), cephalothin (8-fold), and oxacillin (32-fold), and AcrB and RpoB in SACCARM resistant to multiple antibiotics. The up-regulation and down-regulation of proteins varied with the degree of antibiotic resistance (Muroi et al. 2012).

Characteristics of the unpaired proteins identified in S. Typhimurium, K. pneumoniae, and S. aureus

The unpaired protein spots were identified in S. Typhimurium, K. pneumoniae, and S. aureus, as shown in Tables S1 to S3, respectively. The spermidine/putrescine ABC transporter periplasmic-binding protein (PotD, Spot 826) was up-regulated in STCIP, while PflE, YeaG, PudC, AhpF, TalA, and Dps were up-regulated in STCCARM (Table S1). The unique PotD in STCIP negatively regulates the polyamine uptake system (Igarashi and Kashiwagi 1999). Spermidine and putrescine play an important role in ion homeostasis. The stress-related proteins were distinctively expressed in STCIP (GroEL, AlhF, and AhpC) and STCCARM (Dps), suggesting that the activation of stress-related proteins might be attributed to the development of antibiotic resistance (Fehri et al. 2005). The metabolism-related proteins (SucA, AdhE, SdhB, Pck, AceF, PepB, Tkt, Gap, and Map) were up-regulated in KPCIP, whereas the cysteine transport protein (Spot 1006), hyperosmotically inducible periplasmic protein (Spot 1100), and outer membrane protein A (Spot 1127) were distinctively up-regulated in KPCCARM (Table S2). The transporter protein (TolC) up-regulated in KPCCARM (Table S2) is involved in ArcAB-TolC multidrug efflux pump responsible for antibiotic resistance mechanisms (Webber et al. 2009). The OMP can act as a selective barrier that can successfully protect from antibiotic treatment (Fernández et al. 2009). The expressions of cell division protein (FtsZ, Spot 1776), elongation factor G (FusA, Spot 1494), and conserved hypothetical protein (Spot 1913) were exceptionally increased in SACIP (Table S3). The membrane transporter protein (PrsA, Spot 2020), phosphoglucosamine-mutase GlmM (FemD, Spot 1712), elongation factor G (FusA, Spot 1531), and other translation-related proteins were overexpressed in SAOXA. The β-lactamase regulatory protein (MecR1, Spot 1973) was highly expressed in SACCARM. The multiple antibiotic-resistant bacteria are attributed to the overexpression of efflux pumps (Rumbo et al. 2013). Furthermore, the β-lactam or aminoglycoside resistant bacteria are induced by the antibiotic-modifying enzymes. The fluoroquinolone resistant bacteria mutated in DNA gyrases or type IV topoisomerases are mainly due to the alteration in antibiotic target sites (Webber et al. 2013). The MecR1 regulates penicillin-binding protein 2a (PBP2a) and β-lactamase production in S. aureus (Lowy 2003). The foldase precursor (PrsA, Spot 1937) is a membrane-anchored peptidyl-prolyl cis–trans isomerase protein that can stabilize PPB2a (Hyyryläinen et al. 2010). The PBP2a encoded by mecA gene on mobile SCCmec cassette chromosome. The acquisition and stabilization of PBP2a result in the low affinity of β-lactam antibiotics in methicillin-resistant S. aureus (MRSA) (Pinho et al. 2001).The PrsA is also overexpressed in oxacillin- and glycopeptide-resistant S. aureus (Hao et al. 2012; Jousselin et al. 2012). The phosphoglucosamine-mutase GlmM helps to catalyze the conversion of glucosamine-6-phosphate to glucosamine-1-phosphate, which is the initial cytoplasmic step in peptidoglycan biosynthesis. The overexpression of GlmM can lead to high-level methicillin resistance (Glanzmann et al. 1999).The FemD is also involved in biosynthesis of peptidoglycan precursor (Chambers 1997). The overexpression of elongation factor G is involved in methicillin, linezolid, and daptomycin resistance in S. aureus (Lee et al. 2015). The expressions of FemD, AhpF, elongation factor G, and conserved hypothetical protein are responsible for methicillin-resistant S. aureus (Enany et al. 2014).

In conclusion, this study describes the discrepancy in protein profiles between antibiotic-sensitive and antibiotic-resistant S. Typhimurium, K. pneumoniae, and S. aureus. The most significant finding was that not only antibiotic resistance-related proteins, but also other bacterial membrane proteins were associated with the development of antibiotic resistance in bacteria. The enhanced expression level of PrsA in STCIP led to the increased resistance to ciprofloxacin (128-fold), levofloxacin (32-fold), norfloxacin (64-fold), and piperacillin (> 16-fold). YadC, FimA, and RplB were in KPCIP resistant to chloramphenicol (> 32-fold), ciprofloxacin (32-fold), levofloxacin (6-fold), norfloxacin (128-fold), and sparfloxacin (64-fold). AcrB and RpoB were in SACCARM resistant to multiple antibiotics. The overexpressed proteins in the antibiotic-resistant strains identified could be used as biomarkers for detection of antibiotic-resistant bacteria and optimization of antibiotic regimen in chemotherapy. Therefore, the protein profiles obtained in this study can provide useful information to differentiate between bacterial strains with various levels of antibiotic resistance. The proteomic approach will help to discriminate multidrug-resistant strains in association with differentially expressed protein profiles, leading to discovering novel biomarkers for early detection of antibiotic resistance.