Temporal proteomic profiling reveals changes that support Burkholderia biofilms

ABSTRACT

Melioidosis associated with opportunistic pathogen Burkholderia pseudomallei imparts a huge medical burden in Southeast Asia and Australia. At present there is no available human vaccine that protects against B. pseudomallei infection and antibiotic treatments are limited particularly for drug-resistant strains and bacteria in biofilm forms. Biofilm forming bacteria exhibit phenotypic features drastically different to their planktonic states, often exhibiting a diminished response to antimicrobial therapies. Our earlier work on global profiling of bacterial biofilms using transcriptomics and proteomics revealed transcript-decoupled protein abundance in bacterial biofilms. Here we employed reverse phase liquid chromatography tandem mass spectrometry (LC-MS/MS) to deduce temporal proteomic differences in planktonic and biofilm forms of Burkholderia thailandensis, which is weakly surrogate model of pathogenic B. pseudomallei as sharing a key element in genomic similarity. The proteomic analysis of B. thailandensis in biofilm versus planktonic states revealed that proteome changes support biofilm survival through decreased abundance of metabolic proteins while increased abundance of stress-related proteins. Interestingly, the protein abundance including for the transcription protein TEX, outer periplasmic TolB protein, and the exopolyphosphatase reveal adaption in bacterial biofilms that facilitate antibiotic tolerance through a non-specific mechanism. The present proteomics study of B. thailandensis biofilms provides a global snapshot of protein abundance differences and antimicrobial sensitivities in planktonic and sessile bacteria.

Burkholderia rewires its proteome to support biofilm formation by decreasing abundance of metabolic proteins while increasing abundance of stress-related proteins.

INTRODUCTION

Melioidosis is caused by the highly pathogenic Gram-negative bacteria, Burkholderia pseudomallei (Chaowagul et al. 1989; Wiersinga et al. 2006) that is endemic to Southeast Asia, particularly Thailand, and Northern Australia. Melioidosis-associated mortality rates are as high as 70% (Tiangpitayakorn et al. 1997; Lipsitz et al. 2012) but possibly higher because cases of melioidosis are severely underreported (Limmathurotsakul et al. 2016). B. pseudomallei dwells in soil and non-potable water, which spread the bacteria to humans and animals through direct skin contact, inhalation and ingestion of contaminated water. B. pseudomallei is resistant to many antibiotics and in Thailand alone, melioidosis accounts for 40% of septic shock-associated deaths (White 2003; Wiersinga et al. 2006). Unfortunately, this deadly pathogen can survive for many years in nutrient-deficient, distilled water (Wuthiekanun, Smith and White 1995) and show clinical manifestations as late as six decades after exposure (Ngauy et al. 2005). In the host, B. pseudomallei achieves cell-to-cell spread and survives, among other strategies for intracellular persistence, through phagosome invasion (Allwood et al. 2011).

Antibiotic tolerance and resistance are global public health problems that add to treatment failure of pathogenic diseases and disease-related complications (Prestinaci, Pezzotti and Pantosti 2015). B.pseudomallei biofilm formation prevents efficacious use of antimicrobials in part because concentrations needed to kill must be roughly 1000 × higher when bacteria are in biofilms than the more susceptible planktonic forms (Sawasdidoln et al. 2010). Cases of clinical relapse due to tolerance of several antibiotics have been reported, including tolerance to ceftazidime, which is the mainstay of acute-phase melioidosis treatment (Sawasdidoln et al. 2010; Limmathurotsakul et al. 2014). Some early studies employed gene-based analysis of bacteria to deduce differences in planktonic versus biofilm forms that might point the way toward more effective therapies aimed at susceptibilities in the proteome. A recent comparative report showed that differences in gene profiles between high and low producer biofilms of B. pseudomallei were as high as ∼9.5% in the biofilm state (Chin et al. 2015). We and others previously showed that transcriptome to proteome correlation in biofilms was weak suggesting transcription-decoupled protein abundance (Whiteley et al. 2001; Freiberg et al. 2016). Notable, in the fight to eliminate bacteria in biofilms is that once bacteria reshape their proteome to support biofilms, antibiotic resistance could be acquired through phenotypic adaptations of the proteome and/or acquisition of plasmid-mediated resistance (Hu et al. 2016; Whiteley et al. 2001; Sawasdidoln et al. 2010). Thus, understanding protein abundance in the biofilm state is essential to complement to the transcriptomics data because the proteome more closely represents the functional phenotype of an organism in a given environment or state like biofilms. Numerous mass spectrometry (MS)-based proteomic and structural studies have facilitated our understanding of the bacterial biology, such as phenotypic changes that occur in bacteria (Foss et al. 2007; Perez-Llarena and Bou 2016; Khan et al. 2018a,b; Oyler et al. 2018).

Burkholderia thailandensis is a biosafety level 2 (BSL-2) weakly pathogenic species that is closely related phylogenetically (99%) as well as share conserved physiology to the virulent B. pseudomallei species (Brett, DeShazer and Woods 1998; Haraga et al. 2008). As such, B. thailandensis makes a much safer model strain to study the molecular functions of B. pseudomallei (Haraga et al.2008). The relevance of B. thailandensis as a surrogate model of B. pseudomallei is already established in an earlier work on the virulence-associated type III secretion system of B. pseudomallei (Haraga et al.2008). In addition to sharing conserved physiology to the virulent B. pseudomallei, B. thailandensis also has nearly identical quorum sensing systems to B. pseudomallei. In this study, we compared changes in the proteomic profiles of B. thailandensis at multiple time points in both the biofilm and the planktonic growth modes. These results represent a comprehensive proteomic data set of planktonic and biofilm states for B. thailandensis that may advance our understanding of the antibiotic tolerance phenomenon in Burkholderia during biofilm existence. Using a dynamic system for biofilm formation that uses a continuous bio-flow apparatus, proteome samples were obtained and analyzed by shotgun proteomics. Label-free quantitative (LFQ) measurements revealed differences in protein abundance temporally and between states. A better understanding of the molecular basis of antibiotic tolerance in biofilms may help fill multiple therapeutics gaps, facilitate advances in antimicrobial therapy and deduce specifics of the molecular basis of pathogenesis of biofilms.

MATERIALS AND METHODS

Bacterial isolate and growth conditions

A single colony of B. thailandensis E264 (ATCC 700 388) was isolated on Tryptic soy agar (TSA; Fluka Analytical, MO, USA). Overnight culture of bacteria was inoculated into 100 mL of Tryptic soy broth (TSB; Fluka Analytical, MO, USA) and left to grow with agitation at 200 rpm held at 37°C. Bacterial cell suspensions were measured at OD600 nm and then collected by centrifugation (8000 × gfor 5 min) at time points during Early-exponential phase (E-log), Mid-exponential phase (M-log), Early-stationary phase (E-sta) and Mid-stationary phase (M-sta) at 4, 8, 22 and 28 hr after starting inoculum, respectively (Fig. S1, Supporting Information). All growth conditions were prepared and collected as biological triplicates. Planktonic pellets were collected by centrifugation at 5000 × g at 4°C for 20 min and resuspended in an ice-cold buffer (50 mM Tris-Cl pH 8.0, 1 mM EDTA pH 8.0, 0.1% sodium azide, 2.8 mM phenylmethylsulfonyl fluoride) and stored at −80°C prior to downstream processing.

Antimicrobial susceptibility assay

Antimicrobial susceptibilities were determined according to the criteria of the broth microdilution assay of National Committee for Clinical Laboratory Standards (NCCLS). The planktonic, planktonic-shedding and biofilm forms susceptibility testing were done in 96-well microtiter plates represented as minimal inhibitory concentration (MIC), planktonic shedding (P-MIC) and minimum biofilm elimination concentration (MBEC) as previously described (Sawasdidoln et al. 2010) with some modification such as use of Nunc-Immuno TSP 96 pins in lid (ThermoFisher scientific, MA, USA). Antimicrobial agents ceftazidime (CAZ), ciprofloxacin (CIP), erythromycin (ERY) and trimethoprim (TMP) (LKT Laboratories, MN, USA; Sigma-Aldrich, MO, USA; Fluka Analytical, MO, USA) were serially diluted in Mueller Hinton broth (MHB; Fluka Analytical, MO, USA) within a concentration range of 0.5–1024 μg mL⁻¹ in the final volumes of 50 and 150 μL of media for planktonic and biofilm samples, respectively. Viability values of bacterial cells and associated MIC, P-MIC and MBEC were determined after 24 hr of incubation with antibiotics at 37°C subsequently reading the sample turbidity at 600 nm on a microtiter plate reader (Multimode detector DTX 880; Beckman coulter, CA, USA). A low optical density of medium in the wells (OD600 nm < 0.1) served as the end point for successful inhibition of bacterial growth.

Bacterial biofilm growth in a continuous flow reactor system

Bacterial biofilms were acquired using a flow reactor system as previously described (Freiberg et al. 2016). To retrieve adequate amounts of biofilms for proteomic studies, we used a continuous flow system wherein a pump-based continuous flow is achieved and the biofilm is formed in silicone tubing (Fig. S2, Supporting Information). In brief, a single isolate was inoculated in TSB medium. The overnight culture was diluted 1:100 into TSB and with agitation at 37°C, 200 rpm until exponential phase was reached. A 10-mL inoculum from the exponential phase was injected into a liquid flow reactor system containing 3% of TSB. The inoculum was allowed to acclimatize and attach without flow for 30 min after which point media flow through the tubing was restored at a flow rate 0.8 mL min⁻¹. Bacterial biofilms were subsequently allowed to grow inside the tubing at 37°C for 1, 3 and 6 days. All samples were prepared and collected in biological triplicates. Bacterial biofilm samples from each time point were harvested by squeezing the silicone tubing to force the cells out.

RNA isolation, cDNA libraries preparation and RNA Sequencing (RNA-Seq)

For the transcriptomics analysis, total RNA was extracted from B. pseudomallei K96243 that were grown under the transient biofilm tolerance, planktonic shedding and free-living cells as described (Chattagul et al. 2019). An overnight culture of the planktonic form (growth in shaking condition) was initially achieved by growth in enriched Luria Bertani broth (LB) 2% inoculum of overnight bacterial culture was grown in 20 mL LB at 37°C with shaking until the exponential phase was reached. Planktonic cell pellets were recovered by centrifugation at 8000 × g for 5 min and resuspended in 1 mL TRIzol reagent as recommended by the manufacturer (Ambion, USA). Bacterial biofilm cells were cultured in biofilm-inducing medium, which is modified Vogel and Bonner's medium (MVBM) in 24-wells plates (Thermo scientific, USA). An initial inoculum was diluted in MVBM to obtain 1.0 × 10⁷ CFU mL⁻¹ and grown at 37°C with shaking at 100 rpm for 24 hr. Bacterial biofilms were rinsed with phosphate buffer saline pH 7.4 (PBS [pH 7.4]) three times and bacterial cells were collected by scraping into 1 mL of TRIzol reagent. Planktonic shedding cells were obtained from the biofilm-inducing medium using the modification of Nunc-Immuno TSP 96 pins at the lid (Thermo scientific, USA) instead of the Calgary biofilm device (CBD). A 150 µL aliquot of bacterial inoculum in 1.0 × 10⁷ CFU mL⁻¹ was placed in each well of the 96-wells plate and incubated at 37°C for 24 hr. Bacterial biofilms that formed were rinsed with PBS for three times and transferred to 96-well plates that contained enriched medium after which they were incubated at 37°C for another 24 hr. The cell pellets in the medium were processed the same as planktonic cells. Bacterial pellets were stored at −80°C until the total bacterial RNA isolation could be performed.

Three biological replicates were obtained from each bacterial growth condition for a total of nine samples. The cDNA libraries from each RNA preparation were synthesized using the Illumina platform (Illumina, San Diego, CA, USA), and the libraries were subsequently sequenced and analyzed (Chattagul et al.2019). Finally, a total of output of normalized-genes was shown as fragments per kilobase per million mapped reads (FPKM).

Sample preparation, digestion and peptide extraction

Protein extraction and digestion were accomplished as previously described (Freiberg et al. 2016). All planktonic and bacterial biofilm pellets in PBS were thawed on ice, vortexed and genomic DNA removed with 30 μg mL⁻¹ DNase I treatment (New England Biolabs, MA). Cell lysis was performed using an Ultrasonic processor XL sonicator (Qsonica, CT, USA) at 4°C (6 rounds 20 sec with 10 sec rests). Unbroken cells and cell debris were eliminated by centrifugation at 5000 × g at 4°C for 20 min. Protein lysates were precipitated by ultracentrifugation at 39 000 × gat 4°C for 45 min after which fractionated proteins were dissolved in 50 mM ammonium bicarbonate (ambic) buffer. Protein concentration was determined by NanoDrop spectrophotometer (Thermo Fisher Scientific, MA). A 300 μg sample of protein was then denatured in 6 M Urea, 1.5 M Tris pH 8.8 and 5 mM TCEP followed by disulfide reduction in 40 mM dithiothreitol in a 25 mM ambic buffer. Post-reduction, alkylation was achieved by addition of 40 mM iodoacetamide in a 25 mM ambic buffer for 1 hr at room temperature. Finally, the urea buffer was diluted with 800 μL of 25 mM ambic and 200 μL methanol to facilitate proteolysis. Reduced, alkylated protein samples were digested overnight with trypsin (200:1, w/w; Trypsin Gold, MS grade; Promega, WI, USA) at 37°C. The next morning, 200 μL of water was added to the trypsin digested samples after which the liquid was removed by evaporation in a speed-vacuum concentrator (Eppendorf vacufuge; Eppendorf, NY, USA) and this process repeated three times. Finally, 90 μg of the digested protein samples were reconstituted in 5% ACN/0.1% TFA in water and desalted using microspin columns (The Nest Group, MA, USA). Eluted peptide samples were reconsitituted in 0.5% ACN/0.1% Formic acid at a final concentation 0.1 μg μL⁻¹ and stored at −20°C, until the MS data acquisition.

LC-MS/MS analysis, data acquisition, data processing and bioinformatics methods

For each sample, quantitative proteomics data was generated for biological triplicates and each biological replicate was analyzed as technical triplicates using a Waters NanoAcquity high-pressure liquid chromatography system (Waters Corporation, Milford, MA) coupled to a Thermo Orbitrap Fusion Tribrid Mass Spectrometer (Thermo Fisher, San Jose, CA). Digested peptides were desalted on a fused-silica precolumn (100 μm inner diameter, 365 μm outer diameter, 5-μm-diameter, 2 cm packing with 200-Å Magic C₁₈ reverse-phase particles (Michrom Bioresources, Inc., Auburn, CA)) and separated on an in-house packed 75-μm inner diameter, 180 mm long analytical column packed with 5-μm-diameter, 100-Å Magic C₁₈ particles. Analytical columns for peptide separation were constructed in-house using a Sutter P-2000 CO2 laser puller (Sutter Instrument Company, Novato, CA). Peptides were separated using a gradient mixture of mobile phase A (0.1% formic acid, water) and mobile phase B (0.1% formic acid, acetonitrile) in a 95-min period wherein buffer B was increased as following: 5%–35% in 60 min at 250 nL min⁻¹. Subsequently a 5-min wash in high organic phase (80% B) was achieved and column was re-equilibrated for 25-min in 5% buffer B. MS data was collected using Xcalibur (version 2.8; Thermo Scientific) in data dependent acquisition mode with a full MS scan in orbitrap followed by collision-induced dissociation (CID) fragmentation of precursors.

Tandem mass spectra were matched to peptide sequences by search against a B. thailandensis ATCC 700 388 database (UniProt) using software package MaxQuant (Cox and Mann 2008) supplied with the Andromeda search engine (Cox et al. 2011). Label-free quantitative data was generated from AUC measurements of peptides extracted from the MS1 peak intensity data (i.e. extracted ion current) and protein-assembled profiles from MaxQuant results were analyzed using Perseus software (version 1.5.1.6) to calculate differential protein abundance from the resulting LFQ intensity value data (Tyanova et al. 2016). LFQ values were calculated and transformed to log2 ratios as previously described (Cox et al.2014). Moreover, the differential protein abundance values with false-discovery-rate (FDR)-adjusted P-values less than 0.05 were accepted as significant after applying a global FDR (Cox et al.2014).

Proteins identified then were classified via a COGs (http://www.ncbi.nlm.nih.gov/COG) (Tatusov et al.2000). The RNAseq data was analyzed using the FastQC v0.11.4 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc), Trimmomatic v0.35 (Bolger, Lohse and Usadel 2014), and Cufflinks package v2.1.1 (Trapnell et al. 2010) software tools. The orthologous proteins of B. Thailandensis ATCC 700 388 were analyzed against B. pseudomallei K96243 using a reciprocal best-blast (RBB) search between proteins that were obtained from two genomes (http://burkholderia.com). Additionally, the RBB hits were further evaluated using Ortholuge (http://www.pathogenomics.ca/ortholuge/), which is a computational tool for generating precise ortholog predictions.

A description for each Ortholuge classification is available at the Pseudomonas Ortholog Database (http://pseudoluge.pseudomonas.com/). We manually curated virulence factors from the published literature as well as the online databases, including Virulence Factors Database (VFDB; www.mgc.ac.cn/VFs/), Victors virulence factor database (PHIDIAS; www.phidias.us/victors/) and the Pathosystems Resource Integration Center (PATRIC; http://www.patricbrc.org).

RESULTS

Antimicrobial susceptibility of B. thailandensis planktonic versus biofilm forms

Minimum inhibitory concentration (MIC) tests the antibiotic susceptibility of planktonic bacteria but doesn't accurately reflect bacterial susceptibility in biofilms. Instead, a MIC value for planktonic cells shed from a biofilm (i.e. a P-MIC) and minimum biofilm eliminating concentration (i.e. MBEC) can be determined, which provides a more precise measure of antibiotic susceptibility for planktonic bacteria that have been released from a biofilm and the bacteria residing within the biofilm, respectively. The MICs and P-MICs determined for planktonic free-living and biofilm-shedding cells from B. thailandensis indicated that in either form the planktonic bacteria were susceptible to multiple classes of antimicrobials including fluoroquinolone (CIP), beta-lactam (CAZ), trimethoprim (TMP) and macrolide antibiotic erythromycin (ERY) at 0.5, 1.0, 4.0 and 16 μg mL⁻¹, respectively (Fig. 1). Interestingly, the P-MIC value of TMP was 128-fold higher than the MIC value for planktonic bacteria (from 4 μg mL⁻¹ to 512 μg mL⁻¹). The MBEC values for the B. thailandenesis within a biofilm, however, showed high levels of tolerance to CAZ, ERY and TMP as they were over 1000-fold (1024 μg mL⁻¹) higher than the MICs. These results revealed that an observed decrease in susceptibility to antibiotics induced by biofilm growth is non-specific and not restricted to a single mechanism of action or class of antimicrobials. It is also likely that decreased susceptibility to antibiotics in biofilms forms could be due to reduced diffusion of antimicrobials in biofilms. However, as earlier shown, limited antibiotic diffusion is not the primary protective mechanism for decreased susceptibility to antibiotics in biofilms (Walters et al. 2003). Our results, shown in Fig. 1, are in agreement with earlier reports that show decreased antimicrobial susceptibility in bacteria in biofilms (Sawasdidoln et al. 2010).

Figure 1.

Antibiotic susceptibility evaluation of B. thailandensis to antimicrobial agents in planktonic and biofilm forms. The response of B. thailandensis to ceftazidime, ciprofloxacin, erythromycin and trimethoprim were shown (MIC; minimum inhibitory concentration, P-MIC; MIC of planktonic shedding cells and MBEC; minimum biofilm eliminating concentration).

Proteomic analysis of B. thailandensis biofilms and planktonic bacteria

As in biofilms the correlation between transcriptome and proteome is weak, it is suggested and expected that the transcription is decoupled from protein abundance (Whiteley et al. 2001; Freiberg et al.2016). Given the fact that antibiotic susceptibilities in B. thailandensis are extremely dependent on bacterial growth phases and Burkholderia has been shown to differentially express genes during biofilm growth that also facilitated antimicrobial tolerance (Sawasdidoln et al. 2010; Chattagul et al.2019), we performed temporal proteomic profiling to detect, quantify and characterize proteins important to the biofilm phenotype and to determine how these proteomes differed from free-floating cells. To do so, proteomic data from B. thailandensis biofilms grown in a continuous flow reactor were compared to planktonic data, harvested at four and three time points during growth, respectively. Principal component analysis (PCA) of these data at various time points clustered separately in biofilm and planktonic groups, exhibiting a clear separation in PCA between bacterial planktonic and biofilm states (Fig. 2). A total of 875 identified proteins, accounting for about 15.5% of protein coding genes, showed relative changes in their abundance between biofilm (1, 3 and 6 days) and planktonic samples (E-sta, M-sta, E-log and M-log) (Fig. 2). Analysis of variance (ANOVA) statistical analysis revealed 479 proteins (∼8.5%) (q < 0.05) to be differentially expressed and their changes in abundance to be statistically significant between biofilm and planktonic samples (Fig. S3 and Table S1, Supporting Information). These 479 proteins were further evaluated for any overrepresentation of Cluster of Orthologous Groups (COG) (Tatusov et al.2000) (Fig. 3). The COGs database contains predicted and known proteins from sequenced microbial genomes and when orthologous groups have members across ≥ 3 lineages the analysis indicates they likely correspond to an ancient conserved domain. COG analysis showed significant enrichment of proteins involved in amino acid transport and metabolism (E), energy production and conversion (C); translation, ribosomal structure and biogenesis (J) (Fig. 3).

Figure 2.

Principal component analysis (PCA) of biofilm and planktonic proteomics data (Log2 fold changes). The PCA plot represents 875 proteins with biological replicates that indicated clear proteomics profile differences between biofilm and planktonic states as in colors and matching symbols. PCA score plot was generated from planktonic (Early-exponential phase (E-log), Mid-exponential phase (M-log), Early-stationary phase (E-sta) and Mid-stationary phase (M-sta)) and biofilm sample (the early (B1D), maturing (B3D) and late (B6D) biofilm stages)

Figure 3.

Cluster of orthologous groups (COG) classification of 479 ANOVA significant and differentially abundant proteins with their classification (x-axis) and number of differentially regulated proteins (y-axis; both increased and decreased abundant). The total sum of COG classified proteins is greater than 479 proteins as some proteins fit in more than one COG classifiers: [R/S] Poorly characterized (R, S and uncharacterized); [X] Mobilome: prophages, transposons; [Q] Secondary metabolites biosynthesis, transport and catabolism; [P] Inorganic ion transport and metabolism; [I] Lipid transport and metabolism; [H] Coenzyme transport and metabolism; [G] Carbohydrate transport and metabolism; [F] Nucleotide transport and metabolism; [E] Amino acid transport and metabolism; [C] Energy production and conversion; [T] Signal transduction mechanisms; [U] Intracellular trafficking, secretion, and vesicular transport; [O] Posttranslational modification, protein turnover, chaperones; [N] Cell motility; [M] Cell wall/membrane/envelope biogenesis; [D] Cell cycle control, cell division, chromosome partitioning; [L] Replication, recombination and repair; [K] Transcription; [J] Translation, ribosomal structure and biogenesis.

Differential regulation of metabolic and stress-related proteins and virulence factors

During biofilm formation, the bacterial protein synthesis machinery was observed to decrease (Tables 1–3). This change was represented by COG ‘E,’ which represents amino acid transport and metabolism, which showed decreased abundance as a class of proteins. As biofilms have limited availability to energy resources, it was not surprising that COG ‘C—Energy production,’ which indicated metabolic activities and energy production, were lowered during biofilm formation (Tables 1–3). Bacteria in biofilms are adapting to a slower growth rate as exhibited by an overall lower energy demand through the tricarboxylic acid (TCA) cycle and ATP synthesis. The downregulated metabolic genes in Burkholderia biofilms observed here may explain the observed decrease in antibiotic sensitivities as also reported previously (Meylan et al.2017). Stress related proteins such as alkyl hydroperoxide reductase/thiol specific antioxidant (AhpC/Tsa) and heat shock protein HslVU exhibited increased abundance, suggesting bacterial adaptation to biofilm-associated stress. Proteins passing the ANOVA significance test and proteins whose relative abundance changed significantly across different stages of planktonic and biofilm forms as mention earlier are provided in Tables 1–3.

Table 1.

Level of protein abundance (biofilms versus early- and mid-log planktonic forms) obtained at different time points.

				Log2 fold change
Locus tag	Protein name	Protein description (COGs)	COG categories	Early biofilm	Maturing biofilm	Late biofilm	Early biofilm	Maturing biofilm	Late biofilm	References
				Versus early log phase			Versus mid log phase
Increased Abundance
BTH_I3014		glutamate synthase (NADH) large subunit (EC 1.4.1.14)	E	2.38	2.38	1.95	2.30	2.29	1.86
BTH_I1634	serC	phosphoserine aminotransferase apoenzyme (EC 2.6.1.52)	E	2.16	2.44	1.98	1.93	2.21	1.75	(Rodrigues et al. 2006)
BTH_I1550	glk	glucokinase (EC 2.7.1.2)/transcriptional regulator, RpiR family	G	2.05	2.55	2.63	1.52	2.02	2.10	(Kumar et al. 2014)
BTH_I2904	paaF	phenylacetate-CoA ligase (EC 6.2.1.30)	H	4.71	4.65	2.94	3.26	3.20	1.49
BTH_I0768	coaBC	Phosphopantothenate-cysteine ligase (EC 6.3.2.5)	H	1.34	1.19	1.53	1.28	1.12	1.47
BTH_I1994	aceB	malate synthase (EC 2.3.3.9)	C	3.29	3.30	3.21	2.67	2.68	2.58
BTH_II2188		2-methylcitrate synthase (EC 2.3.3.5)	C	3.93	3.55	4.45	3.33	2.95	3.84
BTH_I2916		PDZ domain protein	R	2.04	2.13	2.42	2.24	2.33	2.61
BTH_I1017		oxidoreductase, short chain dehydrogenase	R	1.33	1.74	2.80	1.50	1.91	2.97
BTH_I1373	tolB	tolB protein	U	1.91	2.43	2.88	1.25	1.77	2.22	(Imperi et al.2009; Lo Sciuto et al.2014)
BTH_I2962		unnamed protein product; Similar to Hcp protein	U	2.32	2.31	2.61	1.07	1.06	1.36	(Lim et al. 2015)
BTH_II2017	acsA	acetyl-coenzyme A synthetase (EC 6.2.1.1)	I	1.41	1.87	1.78	1.13	1.59	1.50	(Li et al.2016)
BTH_I2608		3-oxoadipate CoA-succinyl transferase alpha subunit	I	4.84	5.52	5.45	3.74	4.41	4.34
BTH_I3015		Protein of unknown function (DUF1568) domain protein	X	2.70	5.06	5.40	1.04	3.41	3.75
BTH_I0164	hslU	heat shock protein HslVU, ATPase subunit HslU	O	3.55	4.18	2.95	2.61	3.23	2.00	(Rohrwild et al. 1996)
BTH_I0763	clpA	ATP-dependent Clp protease ATP-binding subunit ClpA	O	2.87	3.01	2.70	1.38	1.52	1.21
BTH_I0498	uvrA	excinuclease ABC, A subunit	L	1.41	1.79	1.31	1.21	1.58	1.10
BTH_I2418	dltA	peptide synthetase-like protein	Q	2.95	6.90	5.61	3.52	7.47	6.18
BTH_I2638		DNA-binding response regulator	T	1.64	2.20	2.05	1.21	1.77	1.62
BTH_I2469		putative serine protein kinase, PrkA	T	4.61	5.06	4.55	1.75	2.21	1.69
BTH_I2217	rho	transcription termination factor Rho	K	1.13	1.42	1.05	1.43	1.72	1.36	(Carrano et al. 1998; Italiani and Marques2005; Skordalakes et al.2005)
BTH_I1234	rpsI	SSU ribosomal protein S9P	J	1.26	1.04	1.38	1.37	1.15	1.48
BTH_I0780	rpmB	LSU ribosomal protein L28P	J	1.11	1.21	1.40	1.19	1.29	1.48
BTH_II0701		hypothetical protein	S	2.07	2.45	2.77	1.43	1.81	2.12
Decreased Abundance
BTH_I2994	hisG	ATP phosphoribosyltransferase (homohexameric) (EC 2.4.2.17)	E	−2.37	−2.14	−2.17	−1.68	−1.45	−1.48
BTH_I1047	ilvC	ketol-acid reductoisomerase (EC 1.1.1.86)	E	−2.13	−1.82	−2.34	−2.26	−1.96	−2.48
BTH_I2199	thrC	threonine synthase	E	−1.74	−2.10	−1.87	−1.05	−1.41	−1.18
BTH_I0817	cysD	sulfate adenylyltransferase subunit 2 (EC 2.7.7.4)	E	−3.53	−4.49	−3.95	−2.38	−3.33	−2.79
BTH_I1775	argD	succinylornithine aminotransferase apoenzyme (EC 2.6.1.81)	E	−1.24	−1.84	−1.86	−2.00	−2.60	−2.61
BTH_I0951		aminotransferase family protein	E	−1.93	−2.62	−2.99	−2.43	−3.13	−3.50
BTH_I0889		amino acid ABC transporter membrane protein 1	E	−2.33	−2.84	−3.57	−1.19	−1.69	−2.43
BTH_I0887		amino acid ABC transporter ATP-binding protein	E	−2.55	−3.38	−2.55	−1.79	−2.62	−1.79
BTH_I0861		Orn/Lys/Arg decarboxylase	E	−1.51	−1.90	−1.22	−1.34	−1.73	−1.05
BTH_I0791	argD	acetylornithine aminotransferase apoenzyme (EC 2.6.1.11)	E	−2.43	−2.22	−2.00	−1.91	−1.70	−1.47
BTH_II2037		aromatic amino acid aminotransferase apoenzyme (EC 2.6.1.57)	E	−2.09	−1.93	−2.01	−1.79	−1.63	−1.71
BTH_I1894	eno	enolase (EC 4.2.1.11)	G	−1.87	−2.22	−1.78	−2.00	−2.35	−1.90	(Al-Maleki et al.2015)
BTH_I1489		phosphomannomutase (EC 5.4.2.8)	G	−2.56	−1.44	−2.02	−2.56	−1.43	−2.01
BTH_II1908	aepX	phosphoenolpyruvate mutase (EC 5.4.2.9)	G	−2.20	−2.21	−2.16	−2.47	−2.48	−2.42
BTH_I0354		sporulation-related repeat protein	D	−2.27	−2.30	−2.24	−1.91	−1.94	−1.88
BTH_I0011	gspG	type II secretion system protein G (GspG)	N	−1.28	−1.35	−1.41	−1.61	−1.68	−1.74
BTH_I3235	yidC	protein translocase subunit yidC	M	−3.90	−3.07	−3.66	−2.93	−2.09	−2.68	(Sachelaru et al.2017)
BTH_I1119	murC	UDP-N-acetylmuramate–L-alanine ligase (EC 6.3.2.8)	M	−1.41	−1.78	−1.51	−1.15	−1.52	−1.24
BTH_I2225		lipoprotein NlpD, putative	M	−2.43	−2.75	−4.21	−1.24	−1.57	−3.03	(Roland and Regine2018)
BTH_I2082		penicillin-binding protein, 1A family	M	−3.58	−3.73	−3.26	−2.84	−2.99	−2.53	(Spratt1975)
BTH_I2035	bamA	Beta-barrel assembly machine subunit BamA	M	−2.70	−2.81	−2.21	−2.30	−2.41	−1.81	(Anwari et al.2012)
BTH_I1483	wbpM	capsular polysaccharide biosynthesis	M	−2.12	−2.57	−1.68	−1.79	−2.24	−1.35
BTH_I0380		penicillin-binding protein 6	M	−1.81	−3.05	−1.94	−1.17	−2.41	−1.30
BTH_I1561	FtsK	cell division protein FtsK	M	−1.95	−1.55	−1.72	−1.41	−1.01	−1.18
Decreased Abundance
BTH_I0386	lipA	lipoic acid synthetase	H	−1.63	−1.78	−1.45	−1.68	−1.83	−1.49
BTH_II0614	dxs	1-deoxy-D-xylulose−5-phosphate synthase (EC 2.2.1.7)	H	−3.24	−3.51	−4.98	−2.78	−3.04	−4.51
BTH_I1139		octaprenyl-diphosphate synthase	H	−2.67	−2.83	−2.63	−1.91	−2.07	−1.86	(Okada et al.1997)
BTH_I0556	ubiB	2-octaprenylphenol hydroxylase (EC 1.14.13.-)	H	−2.41	−2.42	−2.27	−1.77	−1.79	−1.63
BTH_I3312	atpF	ATP synthase F0 subcomplex B subunit	C	−1.52	−2.59	−1.73	−1.25	−2.33	−1.46
BTH_I3311	atpH	ATP synthase F1 subcomplex delta subunit	C	−3.47	−3.58	−3.70	−3.36	−3.48	−3.59
BTH_I1069	nuoI	NADH dehydrogenase subunit I (EC 1.6.5.3)	C	−1.71	−2.04	−1.69	−1.42	−1.75	−1.39
BTH_I1064	nuoD	NADH dehydrogenase subunit D (EC 1.6.5.3)	C	−1.77	−2.52	−2.36	−1.61	−2.36	−2.20
BTH_I1063	nuoC	NADH dehydrogenase subunit C (EC 1.6.5.3)	C	−2.54	−3.08	−2.98	−2.32	−2.85	−2.75
BTH_I0646	sucC	succinyl-CoA synthase, beta subunit	C	−1.64	−1.21	−1.72	−1.76	−1.34	−1.85
BTH_I3110	maeB	allosteric NADP-dependent malic enzyme (EC 1.1.1.40)	C	−2.47	−4.12	−3.63	−2.45	−4.10	−3.61
BTH_I1654		electron transfer flavoprotein alpha subunit apoprotein	C	−1.79	−2.50	−2.68	−1.64	−2.35	−2.53
BTH_I1072		NADH dehydrogenase subunit L (EC 1.6.5.3)	C	−1.72	−2.33	−1.32	−1.61	−2.23	−1.22
BTH_I0608		cytochrome c family protein	C	−4.98	−3.96	−3.80	−3.90	−2.89	−2.72
BTH_I0600		homodimeric glycerol 3-phosphate dehydrogenase (quinone)	C	−3.18	−2.54	−2.23	−3.41	−2.78	−2.46
BTH_I0550		oxidoreductase, FAD-binding	C	−3.87	−4.13	−3.62	−2.53	−2.79	−2.28
BTH_I0427	ctaD	cytochrome c oxidase, subunit I	C	−2.47	−3.89	−3.03	−1.84	−3.26	−2.40
BTH_I0426	coxB	cytochrome c oxidase, subunit II	C	−2.53	−3.14	−2.37	−1.22	−1.83	−1.06
BTH_I0150		Domain of unknown function (DUF344) superfamily	C	−3.42	−2.65	−2.50	−3.18	−2.41	−2.26
BTH_II0708	fdxH	formate dehydrogenase (quinone-dependent) iron-sulfur subunit	C	−2.05	−1.96	−2.09	−1.59	−1.50	−1.63
BTH_II0663		succinate dehydrogenase subunit B (EC 1.3.5.1)	C	−1.57	−2.37	−2.24	−1.68	−2.48	−2.36
BTH_I2184		Protein of unknown function subfamily	S	−2.06	−2.54	−2.08	−1.41	−1.89	−1.43
BTH_I1907		Beta-barrel assembly machine subunit BamC	S	−1.49	−1.40	−1.56	−1.21	−1.13	−1.28	(Anwari et al.2012)
BTH_II1072	betA	choline dehydrogenase (EC 1.1.99.1)	R	−3.42	−1.31	−2.09	−3.43	−1.32	−2.10
Decreased Abundance
BTH_I1767		radical SAM domain protein	R	−2.28	−1.29	−2.52	−2.38	−1.39	−2.62
BTH_II2244		radical SAM domain protein	R	−4.77	−3.79	−4.73	−4.90	−3.91	−4.85
BTH_I2765		exopolyphosphatase	P	−2.82	−4.43	−4.17	−3.31	−4.91	−4.66	(Kornberg, Rao and Ault-Riche 1999; Malde et al. 2014; Chuang et al. 2015)
BTH_I0818	cysN	sulfate adenylyltransferase subunit 1 (EC 2.7.7.4)	P	−2.50	−2.27	−2.49	−2.32	−2.09	−2.30
BTH_I0439		D-methionine ABC transporter	P	−1.80	−1.77	−1.73	−1.91	−1.88	−1.84
BTH_I3048	secY	protein translocase subunit secY/sec61 alpha	U	−3.59	−2.94	−3.82	−2.12	−1.47	−2.35	(Sachelaru et al. 2017)
BTH_I1729	lepB	signal peptidase I (EC:3.4.21.89). Serine peptidase.	U	−3.57	−4.13	−3.48	−1.89	−2.45	−1.81	(Rahman et al. 2003)
BTH_I1276	secD	protein-export membrane protein SecD	U	−2.35	−2.65	−2.63	−1.77	−2.06	−2.04
BTH_I0973	phaZ	intracellular PHB depolymerase	I	−2.55	−1.85	−1.38	−4.02	−3.32	−2.85
BTH_I0783	ispH	4-hydroxy-3-methylbut-2-enyl diphosphate reductase	I	−1.31	−2.65	−2.29	−1.66	−2.99	−2.63
BTH_II2359	shc	squalene-hopene cyclase	I	−3.45	−3.70	−3.20	−2.64	−2.88	−2.39	(Schmerk, Bernards and Valvano 2011)
BTH_I2231	ndk	nucleoside diphosphate kinase (EC 2.7.4.6)	F	−2.26	−2.44	−2.96	−3.12	−3.30	−3.82	(Al-Maleki et al.2015; Yu, Rao and Zhang 2017)
BTH_I1250	purH	bifunctional purine biosynthesis protein PurH	F	−2.59	−2.11	−2.15	−3.02	−2.55	−2.59
BTH_I0739	adk	Adenylate kinase (EC 2.7.4.3)	F	−1.99	−1.82	−1.32	−2.87	−2.71	−2.21	(Markaryan et al.2001)
BTH_I0666	purC	phosphoribosylaminoimidazole-succinocarboxamide synthase	F	−1.62	−2.83	−2.19	−1.12	−2.33	−1.69
BTH_I2118	tig	trigger factor	O	−1.59	−2.71	−3.03	−1.96	−3.09	−3.40	(Hesterkamp et al. 1996)
BTH_I1306	grpE	co-chaperone GrpE	O	−3.12	−2.77	−2.98	−4.36	−4.01	−4.22	(Liberek et al.1991; Dubern et al. 2005)
BTH_I1876	iscU	FeS assembly scaffold apoprotein IscU	O	−1.83	−2.01	−1.95	−1.12	−1.31	−1.25	(Bonomi et al. 2005)
BTH_I1457	groES	chaperonin, 10 kDa	O	−2.16	−2.16	−2.64	−3.51	−3.50	−3.99
BTH_I0415		carboxy-terminal protease	O	−2.05	−1.22	−1.78	−2.15	−1.31	−1.88	(Bandara et al. 2008)
BTH_I1756		ABC transporter, permease/ATP-binding protein	Q	−2.40	−2.44	−2.13	−1.76	−1.80	−1.50	(Garmory and Titball 2004)
BTH_I2081		phage shock protein A (PspA) family protein	T	−3.76	−3.66	−3.96	−1.25	−1.15	−1.46	(Adrien et al.2017)
BTH_I2565	nusA	NusA antitermination factor	K	−1.38	−1.09	−1.85	−1.68	−1.39	−2.15	(Wells et al. 2016)
Decreased Abundance
BTH_I2248	tex	transcription accessory protein, TEX	K	−3.24	−3.35	−2.60	−3.23	−3.34	−2.59	(He et al.2006; Moule et al.2015)
BTH_I3071	fusA2	translation elongation factor 2 (EF-2/EF-G)	J	−1.01	−1.81	−1.16	−1.55	−2.36	−1.71
BTH_I2030	frr	ribosome recycling factor	J	−2.91	−2.22	−2.88	−3.81	−3.11	−3.78
BTH_I1056	pnp	polyribonucleotide nucleotidyltransferase	J	−1.41	−1.66	−2.40	−2.07	−2.32	−3.07
BTH_I0779	rpmG	LSU ribosomal protein L33P	J	−3.09	−3.03	−3.12	−3.64	−3.59	−3.67
BTH_I0584	glyS	glycyl-tRNA synthetase beta chain (EC 6.1.1.14)	J	−2.21	−3.12	−2.73	−1.09	−2.00	−1.61
BTH_I0366		hypothetical protein	S	−2.47	−1.96	−2.61	−1.91	−1.39	−2.05

Table 2.

Level of protein abundance (biofilms versus early- and mid-stationary planktonic forms) obtained at different time points.

				Log2 fold change
Locus tag	Protein name	Protein description (COGs)	COG categories	Early biofilm	Maturing biofilm	Late biofilm	Early biofilm	Maturing biofilm	Late biofilm	References
				Versus early stationary phase			Versus mid stationary phase
Increased Abundance
BTH_II0679	trpB	tryptophan synthase, beta chain (EC 4.2.1.20)	E	1.69	1.47	1.50	1.29	1.78	1.57
BTH_I1550	glk	glucokinase (EC 2.7.1.2)/transcriptional regulator, RpiR family	G	1.16	1.40	1.66	1.91	1.74	1.99	(Kumar et al.2014)
BTH_I1553		carbohydrate ABC transporter substrate-binding protein	G	2.62	1.14	3.92	2.45	4.33	2.86	(Higgins 1992)
BTH_II1908	aepX	phosphoenolpyruvate mutase (EC 5.4.2.9)	G	1.26	1.52	1.25	1.52	1.31	1.57
BTH_I2608		3-oxoadipate CoA-succinyl transferase alpha subunit	I	1.88	2.06	2.56	2.74	2.48	2.66
BTH_I2245	purA	Adenylosuccinate synthetase (EC 6.3.4.4)	F	1.16	1.23	2.02	2.09	1.42	1.49
BTH_I0979	purT	formate-dependent phosphoribosylglycinamide formyltransferase	F	1.67	1.13	2.53	2.00	2.48	1.94
BTH_I3111	pyrE	orotate phosphoribosyltransferase (EC 2.4.2.10)	F	2.49	2.21	2.82	2.53	2.61	2.33
BTH_I0164	hslU	heat shock protein HslVU, ATPase subunit HslU	O	3.30	3.62	3.92	4.24	2.69	3.01
BTH_I2122	lon	ATP-dependent proteinase	O	1.35	1.58	1.48	1.71	1.90	2.13	(Rogers et al.2016)
BTH_I2092		antioxidant, AhpC/Tsa family	Q	1.39	2.13	2.51	3.25	2.03	2.77	(Chae et al.1994; Loprasert et al. 2003)
BTH_I2248	tex	transcription accessory protein, TEX	K	1.99	2.22	1.88	2.11	2.63	2.86	(He et al.2006; Moule et al.2015)
BTH_I3195	rpsU1	SSU ribosomal protein S21P	J	3.07	3.76	3.62	4.32	2.39	3.09
BTH_I3072	rpsG	SSU ribosomal protein S7P	J	1.66	2.02	1.71	2.07	1.43	1.79
BTH_I3064	rpsS	SSU ribosomal protein S19P	J	2.03	2.35	2.25	2.56	1.70	2.01
BTH_I3063	rplV	LSU ribosomal protein L22P	J	1.58	1.84	1.39	1.65	1.32	1.58
BTH_I3061	rplP	LSU ribosomal protein L16P	J	1.71	1.98	1.98	2.25	1.32	1.59
BTH_I3055	rpsN	SSU ribosomal protein S14P	J	2.72	2.20	3.66	3.14	3.65	3.13
BTH_I3052	rplR	LSU ribosomal protein L18P	J	1.64	1.71	1.75	1.82	1.38	1.44
BTH_I3047	infA	bacterial translation initiation factor 1 (bIF−1)	J	1.26	1.03	1.67	1.44	1.48	1.25
BTH_I3041	rplQ	LSU ribosomal protein L17P	J	2.72	2.69	3.09	3.06	2.55	2.52
BTH_I2592	rplT	LSU ribosomal protein L20P	J	1.51	1.81	1.04	1.33	1.26	1.56
BTH_I2181	rpsR	SSU ribosomal protein S18P	J	2.40	1.70	2.64	1.95	2.46	1.76
BTH_I1661	rpsP	SSU ribosomal protein S16P	J	1.70	1.95	1.97	2.22	1.98	2.23
Increased Abundance
BTH_I1234	rpsI	SSU ribosomal protein S9P	J	1.32	1.39	1.10	1.17	1.44	1.50
BTH_I1233	rplM	LSU ribosomal protein L13P	J	1.75	1.90	1.72	1.87	1.39	1.54
BTH_I0780	rpmB	LSU ribosomal protein L28P	J	2.39	2.39	2.49	2.49	2.68	2.68
BTH_I2965		hypothetical protein	S	3.71	4.29	2.43	3.01	3.35	3.93
Decreased Abundance
BTH_I3253	gcvP	glycine dehydrogenase (decarboxylating) alpha subunit	E	−2.07	−1.76	−1.72	−1.41	−2.76	−2.46
BTH_II0673	leuD	3-isopropylmalate dehydratase, small subunit (EC 4.2.1.33)	E	−1.62	−1.46	−3.06	−2.90	−2.68	−2.52
BTH_I3023	aroB	3-dehydroquinate synthase (EC 4.2.3.4)	E	−2.18	−1.71	−1.76	−1.29	−1.59	−1.11
BTH_I1617		glutamine amidotransferase, class I	E	−2.29	−3.44	−1.35	−2.50	−2.03	−3.18
BTH_I1050	leuA	2-isopropylmalate synthase (EC 2.3.3.13)	E	−2.95	−2.47	−1.96	−1.48	−2.18	−1.69
BTH_I0861		Orn/Lys/Arg decarboxylase	E	−1.58	−1.96	−1.97	−2.35	−1.29	−1.67
BTH_I0710	kynU	Kynureninase (EC 3.7.1.3)	E	−2.31	−2.88	−1.65	−2.22	−1.66	−2.23
BTH_I1894	eno	enolase (EC 4.2.1.11)	G	−1.22	−1.71	−1.57	−2.07	−1.12	−1.62	(Al-Maleki et al. 2015)
BTH_I0962	tal	transaldolase (EC 2.2.1.2)	G	−1.31	−1.47	−1.37	−1.52	−1.90	−2.06
BTH_II0631	gnd	6-phosphogluconate dehydrogenase (decarboxylating)	G	−2.49	−2.39	−1.84	−1.75	−2.24	−2.14
BTH_I0011	gspG	type II secretion system protein G (GspG)	N	−1.22	−1.99	−1.29	−2.05	−1.34	−2.11
BTH_I2082		penicillin-binding protein, 1A family	M	−2.02	−2.38	−2.17	−2.53	−1.71	−2.07
BTH_I2035	bamA	Beta-barrel assembly machine subunit BamA	M	−1.76	−1.76	−1.86	−1.86	−1.27	−1.27	(Anwari et al. 2012)
BTH_I0857		ompA family protein	M	−2.33	−2.44	−2.38	−2.48	−2.03	−2.14	(Smani et al.2014)
BTH_I0556	ubiB	2-octaprenylphenol hydroxylase (EC 1.14.13.-)	H	−2.00	−2.63	−2.01	−2.64	−1.86	−2.49
BTH_I3311	atpH	ATP synthase F1 subcomplex delta subunit	C	−1.44	−2.52	−1.55	−2.63	−1.66	−2.75
BTH_I1069	nuoI	NADH dehydrogenase subunit I (EC 1.6.5.3)	C	−1.93	−1.78	−2.26	−2.11	−1.91	−1.75
Decreased Abundance
BTH_I1068	nuoH	NADH dehydrogenase subunit H (EC 1.6.5.3)	C	−1.81	−1.85	−2.69	−2.73	−1.95	−1.99
BTH_I1064	nuoD	NADH dehydrogenase subunit D (EC 1.6.5.3)	C	−1.15	−1.63	−1.90	−2.39	−1.74	−2.23
BTH_I1063	nuoC	NADH dehydrogenase subunit C (EC 1.6.5.3)	C	−2.16	−2.33	−2.69	−2.86	−2.59	−2.76
BTH_II0658	mdh1	malate dehydrogenase (NAD) (EC 1.1.1.37)	C	−1.71	−2.48	−1.00	−1.77	−1.33	−2.09
BTH_I3101	paaK	phenylacetate-CoA oxygenase/reductase, PaaK subunit	C	−2.02	−1.82	−2.95	−2.75	−4.59	−4.40	(Grishin and Cygler 2015)
BTH_I2808	pckG	Phosphoenolpyruvate carboxykinase	C	−4.19	−3.69	−6.32	−5.83	−4.63	−4.14
BTH_I2554	lpdA	dihydrolipoamide dehydrogenase	C	−2.27	−2.25	−2.70	−2.67	−3.09	−3.07
BTH_I1654		electron transfer flavoprotein alpha subunit apoprotein	C	−1.30	−1.17	−2.00	−1.88	−2.18	−2.06
BTH_I1072		NADH dehydrogenase subunit L (EC 1.6.5.3)	C	−2.51	−2.85	−3.13	−3.47	−2.11	−2.45
BTH_I1067	nuoG	NADH dehydrogenase subunit G (EC 1.6.5.3)	C	−2.08	−1.64	−2.84	−2.40	−2.86	−2.42
BTH_I0608		cytochrome c family protein	C	−2.32	−2.68	−1.30	−1.66	−1.14	−1.50
BTH_II2187	acnD	aconitase (EC 4.2.1.3)	C	−4.15	−4.54	−6.23	−6.62	−4.64	−5.04
BTH_II2020		fumarase, class I, homodimeric (EC 4.2.1.2)	C	−1.85	−1.86	−3.41	−3.42	−3.78	−3.79
BTH_II0663		succinate dehydrogenase subunit B (EC 1.3.5.1)	C	−1.31	−1.38	−2.11	−2.18	−1.99	−2.05
BTH_II0662	sdhA	succinate dehydrogenase subunit A (EC 1.3.5.1)	C	−2.17	−2.28	−3.38	−3.49	−2.24	−2.35
BTH_II0654	acnA	aconitase (EC 4.2.1.3)	C	−1.82	−1.95	−2.31	−2.45	−1.94	−2.07
BTH_II0362		succinate dehydrogenase subunit B (EC 1.3.5.1)	C	−3.19	−3.76	−3.86	−4.44	−2.88	−3.46
BTH_I2916		PDZ domain protein	R	−1.97	−1.37	−1.88	−1.28	−1.59	−1.00
BTH_I1673		microcin-processing peptidase 1	R	−3.91	−3.49	−5.64	−5.21	−4.54	−4.12
BTH_I1296		microcin-processing peptidase 2	R	−4.03	−3.18	−2.69	−1.84	−2.50	−1.65
BTH_I2765		exopolyphosphatase	P	−1.53	−2.16	−3.14	−3.77	−2.89	−3.51	(Kornberg, Rao and Ault-Riche 1999; Malde et al. 2014; Chuang et al. 2015)
BTH_I1276	secD	protein-export membrane protein SecD	U	−1.18	−1.09	−1.48	−1.38	−1.46	−1.36	(Sachelaru et al. 2017)
BTH_I2255		poly(R)-hydroxyalkanoic acid synthase, class I	I	−3.22	−2.30	−3.58	−2.67	−2.36	−1.45
BTH_I0973	phaZ	intracellular PHB depolymerase	I	−3.47	−3.54	−2.76	−2.84	−2.30	−2.37
Decreased Abundance
BTH_I1250	purH	bifunctional purine biosynthesis protein PurH	F	−2.17	−3.13	−1.70	−2.66	−1.74	−2.70
BTH_I0739	adk	Adenylate kinase (EC 2.7.4.3)	F	−2.97	−3.57	−2.81	−3.40	−2.31	−2.91	(Markaryan et al.2001)
BTH_I2048		SPFH domain, Band 7 family protein	O	−2.79	−2.88	−1.95	−2.04	−1.67	−1.76
BTH_I0415		carboxy-terminal protease	O	−2.05	−2.51	−1.22	−1.68	−1.78	−2.24	(Bandara et al. 2008)
BTH_I2088	mfd	transcription-repair coupling factor	L	−1.46	−1.38	−1.46	−1.37	−1.48	−1.40	(Roberts and Park 2004)
BTH_I1388		1-Cys peroxiredoxin (EC 1.11.1.15)	Q	−2.60	−3.32	−2.31	−3.03	−1.54	−2.27	(Kaihami et al. 2014)
BTH_I0756	fusA1	translation elongation factor 2 (EF-2/EF-G)	J	−1.32	−1.76	−1.59	−2.03	−1.12	−1.56
BTH_I0485	raiA	SSU ribosomal protein S30P/sigma 54 modulation protein	J	−2.06	−2.19	−2.13	−2.26	−2.09	−2.22
BTH_I2218		hypothetical protein	S	−1.99	−2.20	−1.74	−1.95	−1.61	−1.82
BTH_I1867		hypothetical protein	S	−3.96	−4.14	−1.47	−1.66	−2.18	−2.37
BTH_I1491		hypothetical protein	S	−5.27	−5.53	−5.38	−5.64	−4.64	−4.90
BTH_I1191		hypothetical protein	S	−2.91	−2.73	−2.53	−2.36	−2.33	−2.15
BTH_I1155		hypothetical protein	S	−2.14	−2.06	−1.70	−1.62	−1.79	−1.71
BTH_I0605		hypothetical protein	S	−1.75	−1.95	−1.88	−2.08	−1.67	−1.86
BTH_I0454	cydA	cytochrome d ubiquinol oxidase subunit I	C	−3.90	−4.28	−5.61	−5.98	−4.19	−4.56

Table 3.

Level of protein abundance (biofilms versus early-/mid-log and stationary planktonic forms) obtained at different time points.

				Log2 fold change
Locus tag	Protein name	Protein description	COG categories	Early biofilm	Maturing biofilm	Late biofilm	Early biofilm	Maturing biofilm	Late biofilm	Early biofilm	Maturing biofilm	Late biofilm	Early biofilm	Maturing biofilm	Late biofilm	References
				Versus early log phase			Versus mid log phase			Versus early stationary phase			Versus mid stationary phase
Increased Abundance
BTH_I1550	glk	glucokinase (EC 2.7.1.2)/transcriptional regulator	G	2.05	2.55	2.63	1.52	2.02	2.10	1.16	1.66	1.74	1.40	1.91	1.99	(Kumar et al.2014)
BTH_I2608		3-oxoadipate CoA-succinyl transferase alpha subunit	I	4.84	5.52	5.45	3.74	4.41	4.34	1.88	2.56	2.48	2.06	2.74	2.66
BTH_I0164	hslU	heat shock protein HslVU, ATPase subunit HslU	O	3.55	4.18	2.95	2.61	3.23	2.00	3.30	3.92	2.69	3.62	4.24	3.01
BTH_I1234	rpsI	SSU ribosomal protein S9P	J	1.26	1.04	1.38	1.37	1.15	1.48	1.32	1.10	1.44	1.39	1.17	1.50
BTH_I0780	rpmB	LSU ribosomal protein L28P	J	1.11	1.21	1.40	1.19	1.29	1.48	2.39	2.49	2.68	2.39	2.49	2.68
Decreased Abundance
BTH_I0861		Orn/Lys/Arg decarboxylase	E	−1.51	−1.90	−1.22	−1.34	−1.73	−1.05	−1.58	−1.97	−1.29	−1.96	−2.35	−1.67
BTH_I1894	eno	enolase (EC 4.2.1.11)	G	−1.87	−2.22	−1.78	−2.00	−2.35	−1.90	−1.22	−1.57	−1.12	−1.71	−2.07	−1.62	(Al-Maleki et al. 2015)
BTH_I0011	gspG	type II secretion system protein G (GspG)	N	−1.28	−1.35	−1.41	−1.61	−1.68	−1.74	−1.22	−1.29	−1.34	−1.99	−2.05	−2.11
BTH_I2082		penicillin-binding protein, 1A family	M	−3.58	−3.73	−3.26	−2.84	−2.99	−2.53	−2.02	−2.17	−1.71	−2.38	−2.53	−2.07
BTH_I2035	bamA	Beta-barrel assembly machine subunit BamA	M	−2.70	−2.81	−2.21	−2.30	−2.41	−1.81	−1.76	−1.86	−1.27	−1.76	−1.86	−1.27	(Anwari et al. 2012)
BTH_I0556	ubiB	2-octaprenylphenol hydroxylase (EC 1.14.13.-)	H	−2.41	−2.42	−2.27	−1.77	−1.79	−1.63	−2.00	−2.01	−1.86	−2.63	−2.64	−2.49
BTH_I3311	atpH	ATP synthase F1 subcomplex delta subunit	C	−3.47	−3.58	−3.70	−3.36	−3.48	−3.59	−1.44	−1.55	−1.66	−2.52	−2.63	−2.75
BTH_I1069	nuoI	NADH dehydrogenase subunit I (EC 1.6.5.3)	C	−1.71	−2.04	−1.69	−1.42	−1.75	−1.39	−1.93	−2.26	−1.91	−1.78	−2.11	−1.75
BTH_I1064	nuoD	NADH dehydrogenase subunit D (EC 1.6.5.3)	C	−1.77	−2.52	−2.36	−1.61	−2.36	−2.20	−1.15	−1.90	−1.74	−1.63	−2.39	−2.23
BTH_I1063	nuoC	NADH dehydrogenase subunit C (EC 1.6.5.3)	C	−2.54	−3.08	−2.98	−2.32	−2.85	−2.75	−2.16	−2.69	−2.59	−2.33	−2.86	−2.76
BTH_I1654		electron transfer flavoprotein alpha subunit apoprotein	C	−1.79	−2.50	−2.68	−1.64	−2.35	−2.53	−1.30	−2.00	−2.18	−1.17	−1.88	−2.06
BTH_I1072		NADH dehydrogenase subunit L (EC 1.6.5.3)	C	−1.72	−2.33	−1.32	−1.61	−2.23	−1.22	−2.51	−3.13	−2.11	−2.85	−3.47	−2.45
BTH_I0608		cytochrome c family protein	C	−4.98	−3.96	−3.80	−3.90	−2.89	−2.72	−2.32	−1.30	−1.14	−2.68	−1.66	−1.50
BTH_II0663		succinate dehydrogenase subunit B (EC 1.3.5.1)	C	−1.57	−2.37	−2.24	−1.68	−2.48	−2.36	−1.31	−2.11	−1.99	−1.38	−2.18	−2.05
BTH_I2765		exopolyphosphatase	P	−2.82	−4.43	−4.17	−3.31	−4.91	−4.66	−1.53	−3.14	−2.89	−2.16	−3.77	−3.51	(Kornberg, Rao and Ault-Riche 1999; Malde et al. 2014; Chuang et al. 2015)
BTH_I1276	secD	protein-export membrane protein SecD	U	−2.35	−2.65	−2.63	−1.77	−2.06	−2.04	−1.18	−1.48	−1.46	−1.09	−1.38	−1.36
BTH_I0973	phaZ	intracellular PHB depolymerase	I	−2.55	−1.85	−1.38	−4.02	−3.32	−2.85	−3.47	−2.76	−2.30	−3.54	−2.84	−2.37
BTH_I1250	purH	bifunctional purine biosynthesis protein PurH	F	−2.59	−2.11	−2.15	−3.02	−2.55	−2.59	−2.17	−1.70	−1.74	−3.13	−2.66	−2.70
BTH_I0739	adk	Adenylate kinase (EC 2.7.4.3)	F	−1.99	−1.82	−1.32	−2.87	−2.71	−2.21	−2.97	−2.81	−2.31	−3.57	−3.40	−2.91	(Markaryan et al.2001)
BTH_I0415		carboxy-terminal protease	O	−2.05	−1.22	−1.78	−2.15	−1.31	−1.88	−2.05	−1.22	−1.78	−2.51	−1.68	−2.24	(Bandara et al. 2008)

Temporal proteomic changes in planktonic versus biofilm states

Table 3 displays proteins whose abundance in biofilms changed versus early-/mid-log and stationary planktonic forms of bacteria. Antibiotic resistance in biofilms is a highly complex phenomenon that may involve multiple proteins, their regulation and metabolic phenotypic alterations in bacteria. In our study, we demonstrated that changes in abundance of protein(s) might be contributing to the decreased antibiotic sensitivity as several detected proteins play roles in cellular growth/arrest (e.g. the reduction of ATP-consuming pathway [BTH_I3312, BTH_I3311, BTH_I3310, BTH_I3309, BTH_I3308, BTH_I1069, BTH_I1068, BTH_I1064, BTH_I1063, BTH_I1062 BTH_I1072, BTH_I0427, BTH_I0426, BTH_II0419], the TCA cycle depression [BTH_II0658, BTH_I0647, BTH_II2187, BTH_II0663, BTH_II0662, BTH_II0362, BTH_II2020], the shift of metabolic bypass via glyoxylic shunt [BTH_I1994]), decreased metabolic activity (e.g. protein synthesis and DNA synthesis), pore diffusion and efflux (e.g. BTH_I2578, BTH_I1276 BTH_I3048, BTH_I2035, BTH_I2252, BTH_I0857, BTH_I1907, BTH_I3235, BTH_I0681, BTH_I0682, BTH_I1373), cell division (e.g. reduction of cell wall division [BTH_I1123, BTH_I1122, BTH_I1117, BTH_I1119, BTH_I2082, BTH_I3029, BTH_I0380, BTH_I1561]. Table S3 (Supporting Information) details B. thailandensis proteins that putatively play roles in biofilm-specific tolerance as identified in this study. Notable is that the transcription accessory protein TEX, which is involved in pathogen fitness (He et al.2006;Moule et al.2015), showed decreased abundance in biofilm samples compared to early-/mid-log planktonic states. However, TEX exhibited increased abundance at later biofilms time points suggesting temporal roles of TEX in bacterial fitness and long-term survival. We curated virulence factors and stress-related proteins from the published literature as well as different online databases as described in the materials and methods section. We manually curated 121 putative virulence factors and stress-related proteins to allow comparison of their changes in overabundance between planktonic and biofilm samples (Table S2, Supporting Information; Fig. 4). Increased protein abundance in biofilms as shown in cluster (3) is comprised primarily of several putative stress-related response proteins such as 3-oxoadipate CoA-succinyl transferase alpha subunit (BTH_I2608), heat shock protein HslVU, ATPase subunit HslU (BTH_I0164), excinuclease ABC, A subunit (BTH_I0498), ATP-dependent proteinase, a serine peptidase of the MEROPS family S16 (BTH_I2122), ATP-dependent Clp protease ATP-binding subunit ClpA (BTH_I0763), peptide synthetase-like protein (BTH_I2418) and antioxidant, AhpC/Tsa family (BTH_I2092) (Fig. 4). A number of virulence factors show distinct temporal abundance patterns in planktonic (early log, mid log, early stationary and mid stationary) and biofilm (early, maturing and late) growth forms. Several virulence factors showed increased abundance throughout the growth of biofilms in early-maturing and late-maturing stages (Fig. 4, Table S2, Supporting Information). Cluster (1) constitutes proteins whose abundance is generally increased in early- and mid-log bacteria while their abundance level changes were mixed between stationary and biofilm samples. However, cluster (2) comprises proteins whose abundance is mostly upregulated in early- and mid-stationary phases and later exhibited decreased abundance in biofilms (Fig. 4). Cluster (3) also includes virulence factors type VI secretion system protein TssH-1 (BTH_I2958) (Shalom, Shaw and Thomas2007), RNA-binding protein Hfq (BTH_I2239), type VI secretion system protein TssD-1 (BTH_I2962) (Shalom, Shaw and Thomas2007), and phosphoserine aminotransferase apoenzyme (EC 2.6.1.52) (BTH_I1634). In addition, UDP-N-acetylglucosamine 1-carboxyvinyltransferase (EC 2.5.1.7) (BTH_II1407), which is an antibiotic target in susceptible species, is differentially abundant in biofilms. UDP-N-acetylglucosamine 1-carboxyvinyltransferase's abundance initially found to be decreased abundance in early- and mid-log phase, however, early-stationary phase exhibit an over of protein abundance. In mid-stationary, early and maturing biofilms its abundance level is decreased, which suggested the protein is dynamically abundant as bacteria temporally passage from one phase to another (Fig. 4).

Figure 4.

Protein abundance (z-scored) profiles of Burkholderia virulence factors during planktonic and biofilm growth. About 121 manually curated virulence factors and stress-related proteins are compared for their temporal abundance between planktonic and biofilm samples. Cluster (1), (2) and (3) show proteins, whose abundance is increased in Early-/Mid-log, Early/Mid-stationary and Early/Maturing/Late biofilm samples, respectively. Cluster (1) and (2) have majority of differentially regulated Burkholderia virulence factors and cluster (3) contains stress-related response proteins (BTH_I2608, BTH_I0164, BTH_I0498, BTH_I2122, BTH_I0763, BTH_I2418 and BTH_I2092) (Table S2, Supporting Information).

DISCUSSION

Biofilm formation is a universal bacterial attribute wherein free-floating planktonic bacteria adapt to form sessile colonies, which in their later stages shed new planktonic bacteria to restart the cycle between free-floating and stationary biofilm states. Biofilms provide bacteria with survival advantages such as a protective barrier ‘extracellular/exopolymer matrix’ that also contribute to diminished antibiotic sensitivities as well as horizontal gene transfer among biofilm dwellers (Aguila-Arcos et al.2017). Not surprisingly, transcriptomics does not capture all the changes that occur at the proteome level because only a subset of proteins is concordantly identified in the two datasets (Table S4, Supporting Infomation; (Chattagul et al.2019)). The transition from planktonic cells to sessile biofilm communities occurs through reshaping the proteome, which is not apparent from the temporally synchronous transcriptome, which is reported to correlate poorly with their respective proteomes (Whiteley et al. 2001; Freiberg et al. 2016). The present global proteomic work revealed temporal changes in proteomes that support Burkholderia biofilm growth such as adaptation to environmental challenges, including stress, nutrient limiting conditions and antibiotic response (Fig.5). In addition, bacteria grown as biofilms in vitro and in vivo have limited oxygen supplies, while those grown in vivo are exposed to cell envelope, oxidative, nitrosative, DNA damage and protein structural denaturation (unfolding/misfolding) sources of stress.

Figure 5.

Proteome rewiring supports bacterial biofilm.

In the present study, we observed that most MIC and P-MIC values from multiple antimicrobial classes were within expected ranges of susceptibility except for TMP, which produced a P-MIC value close to the MBEC value in B. thailandensis (Fig. 1). An earlier study demonstrated that MIC and P-MIC values were within ranges of susceptibility, but some isolates gave dissimilar results possibly due to the use of different methods of bacterial seeding for the MIC and P-MIC assays (Sawasdidoln et al. 2010). Moreover, the participation of specific functions in multidrug efflux pumps may have influenced elimination of antimicrobial agents (Biot et al. 2011), especially trimethoprim as previously reported (Webb et al. 2017). The non-specific nature of trimethoprim resistance within the planktonic shedding biofilm and biofilm cells is still unclear, nevertheless, it has been reported that the trimethoprim resistance mechanism is linked to an increased level of dihydrofolate reductase in planktonic resistant cells (Fleming, Datta and Gruneberg 1972).

While B. thailandensis is rarely a human pathogen, there have been a few reported deaths (Glass et al. 2006; Chang et al. 2017). Notably, B. thailandensis shares closely related virulence factors and has key elements in genomic similarity with B. pseudomallei, a causative agent of melioidosis (Ngamdee et al. 2015). Our aim here was to study the proteomes of B. thailandensis, as a surrogate model system for the human pathogen B. pseudomallei. Although, the broad similarities between B. thailandensis and B. pseudomallei have been previously described (Holden et al. 2004;Kespichayawattana et al.2004; Yu et al. 2006), the absence of human virulence in B. thailandensis means caution should be used in interpreting our results relative to pathogenicity of B. pseudomallei in certain aspects.

Although bacteria can genetically encode resistance elements that help them avoid eradication by antibiotics, the ability of bacteria to phenotypically adapt in response to environmental conditions is their most universal method of avoiding elimination by stresses like those presented by antibiotics. Metabolically dormant bacteria present in biofilms, although typically capable of possessing genetic susceptibilities to antibiotics, reduce the efficacy of antimicrobial therapy through acquired phenotypic tolerance, such as diminished drug uptake (Meylan et al.2017). These phenotypic changes are more readily apparent when studying the proteome, but may not be reflected in the transcriptome. The metabolic modulations perhaps provide bacteria with an advantage in avoiding antibiotic sensitivities to different classes of drugs in bacterial biofilm forms. Interestingly, biofilm forms of Burkholderia also block antibiotic response by avoiding production of reactive oxygen species (Van Acker et al.2013). Slow replication and metabolic rates give adaptive advantages to biofilms while stress-associated proteins help control the core proteome as well as nucleic acids to maintain cellular homeostasis. Earlier work has shown that biofilm bacteria inhibit cellular respiration, such as via a diminished TCA cycle through glyoxylate shunt (Meylan et al.2017). In conformity with the earlier work that presented diminished metabolic activities providing bacteria with adaptive advantages, we also found TCA cycle depression (decreased abundance of malate dehydrogenase, succinyl-CoA synthase, aconitase, succinate dehydrogenase, fumarase) in Burkholderia biofilms (Tables 1–3). Additionally, an overabundance of malate synthase aceB suggests a glyoxylate bypass while TCA cycle associated enzymes showed decreased abundance (Table 1). Given that the glyoxylate shunt serves as an alternative to the TCA cycle that is induced during stress, roles of malate synthase aceB are well established in the glyoxylate shunt (Cortay et al. 1989). Under conditions of biofilm growth, levels of propionate accumulation are reduced through its catabolism while 2-methylcitrate synthase is induced (Gerike et al.1998). Here we found that 2-methylcitrate synthase showed increased abundance in the biofilm samples (Tables 1–3). Increased levels of glyoxylate correlate with diminished bacterial metabolism, as it inhibits TCA cycle activity and antibiotic sensitivities (Meylan et al.2017). Glyoxylate shunts carbon away from the TCA cycle (Quayle, Keech and Taylor 1961; Caspi et al. 2012) and protects bacteria against tobramycin antibiotic lethality (Meylan et al.2017). Interestingly, we also observed decreased abundance of TCA cycle enzymes such as malate dehydrogenase, succinyl-CoA synthase, aconitase, succinate dehydrogenase and fumarase (Tables 1–3). In addition, TCA cycle attenuation aids in formation of the exopolysaccharide matrix and consequently decreases antibiotic response in bacterial biofilms (Zhu et al.2009). Among other hexokinases, glucokinase has a lower affinity for glucose while it is also involved in the first step of the biosynthetic pathway that produces precursors for biofilm exopolysaccharide synthesis (Mongkolrob, Taweechaisupapong and Tungpradabkul2015). Glucokinase levels were observed to be upregulated in Burkholderia in biofilm stimulating conditions, indicating that it may support biofilms (Tables 1–3).

Exopolyphosphatases hydrolyze a key regulator of stress responses and virulence known as long-chain polyphosphate (poly(P)) that can alter bacterial cell membrane permeability and antibiotic tolerance (Kornberg, Rao and Ault-Riche 1999; Malde et al.2014; Chuang et al.2015). For example, the M. tuberculosis strain deficient in the exopolyphosphatase gene exhibited elevated levels of poly(P), a reduced bacterial growth rate, reduced antibiotic susceptibility, increased cell wall thickness and a metabolic shift in carbon source utilization (Chuang et al.2015). Growing Burkholderia in biofilm-stimulating conditions showed decreased abundance of exopolyphosphatase (Tables 1–3), suggesting poly(P) accumulation may assist in metabolic shift and adaptation during biofilm formation. In bacteria, transcription termination factor Rho-controlled transcription correlates with biofilm formation, cell motility, growth, metabolism and sporulation; absence of Rho suppresses biofilm formation (Bidnenko et al.2017) and several antibiotics target Rho factor (Carrano et al. 1998; Italiani and Marques2005; Skordalakes et al.2005). In Burkholderia biofilms, the upregulation of Rho we detected might facilitate biofilm formation (Table 1). Additionally, Gram-negative bacterial outer membrane stability factor TolB, a periplasmic protein that is also a potential drug target, which plays roles in intracellular traffiking and secretion (Imperi et al.2009; Lo Sciuto et al.2014), is upregulated in Burkholderia biofilms (Table 1).

In response to environmental stress present during biofilm formation, bacteria produce stress-associated proteins that help maintain protein and nucleic acid integrity and/or remove non-required or damaged proteins (Rohrwild et al.1996). In Burkholderia biofilms, heat shock protein HslVU showed increased abundance; ATPase subunit HslU proteins usually exhibited increased protein abundance in response to stress (Tables 1–3) as they degrade damaged proteins. Additionally, alkyl hydroperoxide reductase/thiol specific antioxidant (AhpC/Tsa) family proteins play various roles in stress responses. For example, AhpC directly reduces organic hyperoxides whereas TSA protects against sulfur-containing radicals. In comparison to their planktonic forms, Burkholderia biofilms showed upregulated AhpC/Tsa proteins (Table 2). Most of the ANOVA significant proteins detected in the biofilm state data were found to be concordant with earlier published studies of bacterial biofilms (Tables 1–3). However, some of the protein abundance were observed to be discordant, for example chaperonin/co-chaperone (groES, groS, GrpE) abundance. Interestingly, a transcriptomics study of biofilms also reported downregulation of groES and GrpE (Scherr et al. 2013). Additionally, Freiberg et al. have explained the poor correlation of transcriptome and proteome in GAS biofilm as referred to temporospatial heterogeneity of the complexity in biofilm population (Freiberg et al. 2016). Although, these two sets of data were represented to be discordant, some of translational levels were strongly correlated to the transcriptional levels as shown in Table S4 (Supporting Information). All in all, comparative proteome profiles of biofilm and planktonic bacteria showed proteomic rewiring that supports biofilm formation wherein decreased abundance of metabolic proteins and increased abundance of stress-associated proteins is observed (Fig. 5). As proposed earlier, potentiating TCA cycle with a combination of antibiotics may serve as unique strategy to inhibit biofilm-forming bacteria (Meylan et al.2017). The present temporal profiling of planktonic and biofilm forming Burkholderia provides a unique opportunity to understand how bacterial proteome correlates with an overall adaptive response.

Conflict of interest

None declared.