Enzymes Catalyzing the TCA- and Urea Cycle Influence the Matrix Composition of Biofilms Formed by Methicillin-Resistant Staphylococcus aureus USA300

In methicillin-sensitive Staphylococcus aureus (MSSA), the tricarboxylic acid (TCA) cycle is known to negatively regulate production of the major biofilm-matrix exopolysaccharide, PIA/PNAG. However, methicillin-resistant S. aureus (MRSA) produce a primarily proteinaceous biofilm matrix, and contribution of the TCA-cycle therein remains unclear. Utilizing USA300-JE2 Tn-mutants (NARSA) in genes encoding TCA- and urea cycle enzymes for transduction into a prolific biofilm-forming USA300 strain (UAS391-Erys), we studied the contribution of the TCA- and urea cycle and of proteins, eDNA and PIA/PNAG, to the matrix. Genes targeted in the urea cycle encoded argininosuccinate lyase and arginase (argH::Tn and rocF::Tn), and in the TCA-cycle encoded succinyl-CoA synthetase, succinate dehydrogenase, aconitase, isocitrate dehydrogenase, fumarate hydratase class II, and citrate synthase II (sucC::Tn, sdhA/B::Tn, acnA::Tn, icd::Tn, fumC::Tn and gltA::Tn). Biofilm formation was significantly decreased under no flow and flow conditions by argH::Tn, fumC::Tn, and sdhA/B::Tn (range OD492 0.374−0.667; integrated densities 2.065−4.875) compared to UAS391-EryS (OD492 0.814; integrated density 10.676) (p ≤ 0.008). Cellular and matrix stains, enzymatic treatment (Proteinase K, DNase I), and reverse-transcriptase PCR-based gene-expression analysis of fibronectin-binding proteins (fnbA/B) and the staphylococcal accessory regulator (sarA) on pre-formed UAS391-Erys and Tn-mutant biofilms showed: (i) < 1% PIA/PNAG in the proteinaceous/eDNA matrix; (ii) increased proteins under no flow and flow in the matrix of Tn mutant biofilms (on average 50 and 51 (±11)%) compared to UAS391-Erys (on average 22 and 25 (±4)%) (p < 0.001); and (iii) down- and up-regulation of fnbA/B and sarA, respectively, in Tn-mutants compared to UAS391-EryS (0.62-, 0.57-, and 2.23-fold on average). In conclusion, we show that the biofilm matrix of MRSA-USA300 and the corresponding Tn mutants is PIA/PNAG-independent and are mainly composed of proteins and eDNA. The primary impact of TCA-cycle inactivation was on the protein component of the biofilm matrix of MRSA-USA300.


Introduction
Nosocomial and community-acquired infections caused by Staphylococcus aureus range from superficial to life-threatening [1]. The pathogenic ability of S. aureus is greatly facilitated by its capacity to form biofilms, sessile microbial communities that remain embedded in an extracellular polymeric glycocalyx (matrix) or slime layer [2]. Interestingly, recent studies have highlighted differences in biofilm formation between methicillin-sensitive S. aureus (MSSA) and their (multi-) drug resistant counterpart, methicillin-resistant S. aureus (MRSA) [3,4]. In MSSA, the primary polysaccharide that forms the biofilm matrix is encoded by the icaADBC operon and is known as polysaccharide intercellular adhesin PIA or poly-N-acetylglucosamine PNAG [5]. On the other hand, MRSA exhibits a primarily proteinaceous biofilm matrix, with very little contribution of PIA/PNAG [6], that is mediated by adhesins such as the fibronectin binding proteins FnbpA/B [7]. In addition, recent reports also show an important contribution of extracellular DNA (eDNA) to the MRSA biofilm matrix [8]. eDNA in S. aureus is released by cell lysis, which has been shown to be dependent on autolysins such as the major autolysin atl [9,10], and on the holin/antiholin system cidA/lrgA [10,11].
The tricarboxylic acid (TCA) cycle is a central metabolic pathway that generates energy (ATP) and precursors for biosynthesis of macromolecules like 2-oxoglutarate [12]. Its role in regulating PIA/PNAG production in staphylococcal species has been well-studied. In S. epidermidis, environmental changes that inhibited TCA-cycle activity also resulted in a massive derepression of PIA biosynthetic genes and increased PIA production [13,14]. This inverse correlation was also confirmed for MSSA in a rabbit catheter model of biofilm infection [15]. However, the contribution of the TCA-cycle, if any, to biofilm formation by MRSA remains unclear, given the primarily protein-based matrix and the lack of studies on TCA-cycle inhibition using fluorocitrate or transposon (Tn) mutants.
In this study, utilizing Tn mutants, biofilm models, and various stains and enzymes, we studied the importance of the TCA-and urea cycle for biofilm formation by MRSA-USA300 and the net contribution of proteins, eDNA and PIA/PNAG, to the matrix.

Bacterial Strains and Growth Conditions
The strains used in this study are shown in Table 1. Bursa aurealis transposon (Tn) insertion mutations encoding functionally non-redundant TCA-and urea cycle enzymes ( Figure 1) in USA300-JE2 were obtained from the Nebraska Transposon Mutant Library (NTML, www.beiresources.org) [16]. Parental strains UAS391, UAS391-Ery S (erythromycin resistance cured UAS391), and JE2 are all MRSA belonging to the highly virulent and widespread clonal lineage, USA300. These, as well as Tn insertion mutants, were routinely grown on Brain-Heart infusion (BHI; Becton, Dickinson and Company, Franklin Lakes, NJ, USA) supplemented with 0.1% D(+)-glucose monohydrate (Merck Millipore, Billerica, MA, USA) and BHI Bacto™ agar (Becton, Dickinson and Company, USA) for biofilm, transduction and complementation experiments. Lysogeny broth (LB; Becton, Dickinson and Company, USA) was used for Escherichia coli. For the Tn-carrying S. aureus transductants with the erythromycin resistance marker ermB, 5 or 10 µg/mL erythromycin (Sigma-Aldrich ® , Merck KGaA, St. Louis, MO, USA) was supplemented to the growth medium. Figure 1. The urea and tricarboxylic acid (TCA)-cycles. Arginine is synthesized via the urea cycle. Carbamoyl phosphate reacts with ornithine to generate citrulline. Addition of aspartate to citrulline creates L-argininosuccinate. ATP is cleaved to AMP and pyrophosphate to drive this reaction forward. Arginine is cleaved off of Largininosuccinate by the enzyme encoded by argH and can be used for protein synthesis. Hydrolysis of arginine generates ornithine and urea. Fumarate is the other product of the ArgH-catalyzed reaction and can be used in the TCA-cycle. Acetyl-CoA derived from pyruvate and other catabolic pathways enters the TCA-cycle. The acetyl group condenses with four-carbon oxaloacetate to produce citrate. Citrate rearranges to isocitrate, which is decarboxylated and forms NADH + H + by transferring 2H + + 2e − . 2-Oxoglutarate is decarboxylated and transfers 2H + + 2e − to form NADH + H + , while incorporating CoA to form succinyl-CoA. Succinate forms fumarate by transferring 2H + + 2e − resulting in FADH2. Water is incorporated, and oxaloacetate is formed when 2H + + 2e − are transferred to form NADH + H + . The pathway marked in green highlights the proposed model for NADH reoxidation (Arnon-Buchanan cycle). Addition of aspartate to citrulline creates L-argininosuccinate. ATP is cleaved to AMP and pyrophosphate to drive this reaction forward. Arginine is cleaved off of L-argininosuccinate by the enzyme encoded by argH and can be used for protein synthesis. Hydrolysis of arginine generates ornithine and urea. Fumarate is the other product of the ArgH-catalyzed reaction and can be used in the TCA-cycle. Acetyl-CoA derived from pyruvate and other catabolic pathways enters the TCA-cycle. The acetyl group condenses with four-carbon oxaloacetate to produce citrate. Citrate rearranges to isocitrate, which is decarboxylated and forms NADH + H + by transferring 2H + + 2e − . 2-Oxoglutarate is decarboxylated and transfers 2H + + 2e − to form NADH + H + , while incorporating CoA to form succinyl-CoA. Succinate forms fumarate by transferring 2H + + 2e − resulting in FADH 2 . Water is incorporated, and oxaloacetate is formed when 2H + + 2e − are transferred to form NADH + H + . The pathway marked in green highlights the proposed model for NADH reoxidation (Arnon-Buchanan cycle).

Growth Rate Analysis
To exclude the possibility of changes in biofilm mass due to a pleiotropic effect on the bacterial growth rate, an overnight grown culture of the Tn mutants or UAS391-Ery S was diluted until a concentration of 0.5 McFarland and 20 µL was added to 180 µL fresh BHI-medium in a 96-well microtiter plate (CELLSTAR ® 96 Well Plate Flat Bottom (polystyrene), Greiner Bio-One, Austria). The optical density of each well was measured with a spectrophotometer (MultiSkan™ GO Microplate Spectrophotometer, Thermo Fisher Scientific Inc., Waltham, MA, USA) using SkanIt™ software during a course of 24 h at 37 • C (measurements were taken every 15 min, at 600 nm with shaking at 5 Hz and an amplitude of 15 mm). Growth rates were calculated based on the exponential portion of the curve, the maximum culture density, and the duration of the growth lag phase using GrowthRates software [23]. In total, 96 measurements were made and the growth pattern of each mutant or UAS391-Ery S was measured in 8 different wells.

Quantitative Biofilm Assay under Static (No Flow) Conditions
UAS391-Ery S , Tn and complemented mutants were studied as 24 h-old biofilms under flow or no flow conditions as described [18] with one modification; washing to remove planktonic bacteria was performed by gently submerging the plate in a tub of 1× PBS (Thermo Fisher Scientific Inc., Waltham, MA, USA). OD values were measured at 492 nm (Multiskan FC photometer, Thermo Fisher Scientific Inc., Waltham, MA, USA), normalized to the blank and compared to simultaneously run MRSA ATCC reference strains 6538 and 5374 (positive and negative control, respectively), as well as UAS391-Ery S . The assay was performed on three distinct days, each time on three different plates and control strains, UAS391-Ery S as well as Tn and complementation mutants on the same plate were added in 6 different wells.

Quantitative Biofilm Assay under Flow (Dynamic) Conditions
All Tn and complemented mutants as well as UAS391-Ery S were also tested for biofilm formation under dynamic conditions in the Bioflux™ system using glass 48-well plates (Fluxion Biosciences Inc., Alameda, CA, USA), as described by Reference [17]. Imaging was performed with a high-end fluorescence Carl Zeiss™ microscope (Axio Observer ® with Cell Observer SD, ApoTome.2, LSM710, Göttingen, Germany) using ZEN pro 2012 software (Zeiss Efficient Navigation ® , Göttingen, Germany). Actual fluorescence quantification of the obtained images was performed using the program ImageJ (Image Processing and Analysis in Java), which measured integrated density (http://imagej.nih.gov/). The assay was performed on two distinct days. Control strains, UAS391-Ery S , as well as Tn and complemented mutants on the same plate were added in duplicate.

Analysis of Biofilm Matrix Composition
Additionally, UAS391-Ery S , Tn and complemented mutants grown under flow and no flow conditions were studied for differences in cell densities and viability, as well as matrix composition using LIVE/DEAD™ (BacLight™ Bacterial Viability Kit), SYPRO ® Ruby (FilmTracer™ SYPRO ® Ruby Biofilm Matrix Stain), or wheat germ agglutinin (WGA) (Wheat Germ Agglutinin, Texas Red™-X Conjugate) fluorescent stains (Thermo Fisher Scientific Inc., Waltham, MA, USA). LIVE/DEAD™ stain consists of SYTO™ 9 which stains the entire cell mass green followed by propidium iodide which will only stain the dead or dying cells with a compromised membrane (red). WGA Texas Red™-X Conjugate binds to sialic acid and N-acetylglucosaminyl residues of PIA/PNAG, and FilmTracer™ SYPRO ® Ruby Biofilm Matrix Stain labels most classes of proteins, such as glycoproteins, phosphoproteins, lipoproteins, calcium binding proteins, and fibrillar proteins. Briefly, after 24 (no flow) or 17 h (flow) growth and rinsing with either 1× PBS (no flow) or 0.9% sodium chloride (flow) to remove planktonic cells, biofilms were stained and microscopically visualized using the ImageJ program for data measurements, as explained before. Under no flow conditions (used for quantification), the assay was performed on three distinct days, each time on three different plates and control strains, UAS391-Ery S as well as mutants on the same plate were added in 6 different wells. Under flow conditions (used for visualization), the assay was performed on two distinct days, with control strains, UAS391-Ery S as well as Tn and complemented mutants added in duplo on the same plate. Percentages compared to UAS391-Ery S were calculated as µm 2 area covered. In order to quantify the proportion of protein and eDNA in the biofilm matrix, pre-formed biofilms, grown under flow or no flow conditions, were rinsed once with either 1X PBS (no flow) or 0.9% sodium chloride (flow) incubated for 2 (no flow) or 5h (flow) at 37 • C with Proteinase K (Sigma-Aldrich ® , Merck KGaA, St. Louis, MO, USA) (100 µg/mL in culture medium with 10 mM Tris-HCl, pH 7.5) or DNaseI (100 U/mL in culture medium) (Sigma-Aldrich ® , Merck KGaA, St. Louis, MO, USA). Control wells were treated with the appropriate buffer. Afterwards, the wells were washed and stained, as described before. Under no flow conditions (used for quantification), the assay was performed on three distinct days, each time on three different plates and control strains, UAS391-Ery S as well as mutants on the same plate were added in 6 different wells. Under flow conditions (used for visualization), the assay was performed on two distinct days, with control strains, UAS391-Ery S as well as Tn, and complemented mutants added in duplo on the same plate.

Relative Gene Expression Analysis
To measure the impact of the Tn insertion in the target gene on the expression of the global regulator sarA and on the fibronectin-binding proteins encoded by fnbA/B, 24h-old no flow biofilms of Tn mutants and of UAS391-Ery S were mechanically disrupted using bead beating (FastPrep ® -24 classic homogenization instrument, MP Biomedicals, Irvine, CA, USA). Total RNA was isolated (Masterpure™ Complete DNA and RNA Purification kit, Epicentre ® , Madison, WI, USA), 1 µg RNA was purified (Turbo DNA-free™, Ambion ® , Thermo Fisher Scientific Inc., Waltham, MA, USA) and first-strand cDNA was synthesized using random primers (Reverse Transcription System, Promega Corporation, Madison, WI, USA). Reverse transcriptase-PCR (RT-PCR) was performed (StepOnePlus™ system, Applied Biosystems ® , Thermo Fisher Scientific Inc., Waltham, MA, USA) with Power SYBR™ Green PCR Master Mix (Thermo Fisher Scientific Inc., Waltham, MA, USA). Gene-specific primers are listed in Supplementary Table S1. For data normalization, housekeeping gene gyrB (SAUSA300_0005) was used as an internal reference and the fold change in gene expression was calculated using the comparative C t method (2 −∆∆Ct ).

Complementation Experiments
To confirm that the changed biofilm phenotype of the Tn mutants was caused by inactivation of the target gene and not because of secondary mutations in the genome, complementation by a cloned wild type copy of the target genes was performed. Briefly, total genomic DNA of UAS391-Ery S was purified and the argH, acnA, icd, gltA, fumC, sucC, sdhA, sdhB, and rocF genes with a 25-26 bp. overlap corresponding to the nucleotide sequences flanking the EcoRI site of the shuttle vector pALC2073, were amplified, as described in Reference [17]. For genes in an operon, the distal genes were also included in the PCR fragments that were used for complementation. Primers are listed in Supplementary Table S1. PCR-fragments were cloned (2× Gibson Assembly ® Master Mix, New England BioLabs ® Inc., Ipswich, MA, USA) in the EcoRI-linearized (New England BioLabs ® Inc., Ipswich, MA, USA) pALC2073 vector, transformed into recombination-impaired CaCl 2 -competent E. coli DH5α and transformants were selected on LB supplemented with carbenicillin (100 µg/mL). Transformants were checked by Sanger sequencing using pALC2073 vector primers TetR2, pALC-2, and internal gene sequence primers of interest (Supplementary Table S1). Constructs were first introduced by electroporation into the restriction-deficient S. aureus host RN4220 to adapt plasmid DNA from E. coli to S. aureus modifications. Transformants were selected on LB agar plates supplemented with 10 µg/mL chloramphenicol (Sigma-Aldrich ® , Merck KGaA, St. Louis, MO, USA). Subsequently, plasmid DNAs isolated from this strain were introduced by electroporation into the corresponding UAS391-Ery S Tn mutants and the transformants were again selected on LB medium supplemented with 10 µg/mL chloramphenicol. The expression of the cloned genes was induced by adding 0.1 µg/mL anhydrotetracycline (Sigma-Aldrich ® , Merck KGaA, St. Louis, MO, USA) to the growth media.

Statistical Analysis
Biomass quantification in the dynamic flow and no flow assays, as well as growth rate analysis was performed using the R Project software (version 3.1.2.) (R foundation, Vienna, Austria). A Welch two-sample t-test or a Wilcoxon Rank Sum test was used when data was either distributed normally or not, based on a Shapiro-Wilk Normality test. p-values < 0.05 were considered significant. Comparison of average growth rates of argH::Tn and rocF::Tn mutants with UAS391-Ery S showed no decrease in growth rates (p = 0.158 and p = 0.207, respectively) ( Table 2 and Supplementary Figure S1). Next, no flow and flow biofilms formed by the argH::Tn and rocF::Tn mutants were quantified compared to UAS391-Ery S . The argH::Tn mutant showed a significant decrease in biofilm formation both in the no flow (p < 0.001) and in the flow model (p = 0.008), compared to UAS391-Ery S (Table 2 and Figure 2). However, a similarly significant decrease was not observed with the rocF::Tn mutant (p ≥ 0.103) ( Table 2 and Figure 2). The biomass under flow and no flow conditions increased for argH::Tn upon complementation with pGV5990 to quantities comparable to UAS391-Ery S (p ≤ 0.045), while the complemented rocF::Tn (with pGV6003) showed no change in biofilm formation, as compared to rocF::Tn mutant or to UAS391-Ery S (p ≥ 0.057) ( Table 2). Table 2. Overview of results. Quantification of formed biofilm mass (optical density, no flow assay; integrated density, dynamic assay), growth rate (growth curve assay), ratio live:dead cells (Syto™ 9 Green Fluorescent Acid and propidium iodide stain), protein component (Filmtracer™ SYPRO™ Ruby Biofilm Matrix stain), and PIA/PNAG component (WGA Texas Red™-X Conjugate stain). Standard deviations are mentioned next to each value (±) and the percentage value compared to UAS391-Ery S is mentioned between brackets. NT refers to not tested.
In all TCA-and urea cycle Tn mutants, sarA showed a distinct upregulation (1.35 to 3.63-fold), with the exception of sucC::Tn, which showed a 0.92-fold downregulation (Supplementary Figure S2). Overall, a knockout mutation in either the TCA-or urea cycle was associated with a decrease in fnbA and fnbB expression (0.53-to 0.78-fold and 0.51 to 0.77-fold, respectively) (Supplementary Figure S2).  Fluorescence microscopy observations of no flow biofilm matrix structure obtained from UAS391-Ery S , argininosuccinate lyase (argH), fumarate hydratase class II (fumC), succinate dehydrogenase (flavoprotein subunit) (sdhA), succinate dehydrogenase iron-sulfur protein (sdhB), and citrate synthase II (gltA) Tn mutants. Since microscopy images of gltA::Tn, acnA::Tn, icd::Tn, sucC::Tn, and rocF::Tn, as well as corresponding complemented mutant biofilms were comparable to each other, the matrix formed by gltA::Tn serves as an example picture for all. The top row shows total cells stained with SYTO™ 9 and PI. The middle row shows PNAG stained with WGA Texas Red™-X Conjugate, and is combined with Bright-field imaging. The bottom row shows the protein component stained with FilmTracer™ SYPRO™ Ruby Biofilm Matrix.

Discussion
The biofilm matrix of MRSA-USA300 and corresponding Tn mutants is PIA/PNAG-independent and mainly composed of proteins and eDNA.
Using enzymatic digest and biofilm-matrix staining experiments to assess the contribution of PIA/PNAG, protein, and eDNA to the biofilm matrix of MRSA-USA300, we firstly showed that the biofilm matrix of USA300 UAS391-Ery S and Tn mutants was primarily composed of proteinaceous material and eDNA with < 1% contribution of PIA/PNAG. Interestingly, and in contrast to MSSA, inactivation of the TCA-cycle in MRSA Tn mutants did not result in any increase of PIA/PNAG in the biofilm matrix. In MSSA, decreased TCA-cycle activity was reported to shunt metabolites toward PIA/PNAG production [14]. We have previously whole genome sequenced the UAS391 strain and found an intact functional icaADBC operon [19]. These data fully support the results of Pozzi et al. that showed that high level expression of PBP2a-the product of the methicillin resistance gene, mecA, harboured on the SCCmec element that differentiates MRSA and MSSA-blocks icaADBC-dependent polysaccharide biofilm development and promotes formation of proteinaceous biofilms [5]. Of note, S. aureus also produces a capsular polysaccharide (type 5 and 8), which has been implicated in biofilm formation [24]. The role of the TCA-cycle in capsular polysaccharide production was demonstrated by Sadykov et al. who showed that in the absence of glucose, the capsular sugar precursor fructose 6-phoshate is synthesized by gluconeogenesis from the TCA-cycle intermediate oxaloacetate [25]. However, the USA300 clonal lineage, including the UAS391 strain, harbors conserved mutations in the cap5 locus, and does not produce a capsular polysaccharide [26], which also made it easier to exclude the contribution of PIA/PNAG to the biofilm matrix of UAS391 and its Tn mutants.

TCA-Cycle Inactivation Impacts the Protein Component of the Biofilm Matrix of MRSA-USA300
Upon comparison of the net protein contribution to the matrix under no flow and flow conditions, it was clear that the net contribution of proteins to the entire Tn mutant biomass (on average 50% and 51%, respectively) was significantly higher than for UAS391-Ery S (on average 22% and 25%, respectively).
Several studies have reported a role for proteins and eDNA in the ica-independent MRSA biofilm phenotype [8,9,27]. Houston et al. demonstrated an important role for eDNA during the primary attachment and early stages of MRSA biofilm formation by employing a ∆atl knockout mutant in MRSA isolate BH1CC [9]. These authors also reported that DNaseI impaired biofilm development by MRSA isolates from clonal complex 5 (CC5), CC22 and sequence type 239 (ST239). Moreover, treatment of USA300 biofilms after 6 h and 22 h of growth demonstrated both a significant impact of DNaseI and proteinase K, with the latter having the largest impact on the total biofilm mass [27]. Our results also indicate the possibility of the USA300 biofilm matrix containing other components, as protein and eDNA only accounted for on average 23.5% and 48.5% of the biomass. These might have been teichoic acids associated with the cell wall (cell wall teichoic acid, WTA) or cell membrane (lipoteichoic acid, LTA) [28].
A prior study on an ica-knockout MSSA strain RN6390 has shown that, in the presence of citrate, the fibronectin-binding proteins, FnBPA and FnBPB, stimulate biofilm formation by promoting both cell-to-surface and cell-to-cell interactions, which is part of a larger network of virulence factors that are controlled by the staphylococcal accessory regulator, SarA [29]. All urea and TCA-cycle Tn mutants in our study showed a significant downregulation in fnbA and fnbB gene expression compared to UAS391-Ery S , whereas sarA was upregulated for all Tn mutants except sucC::Tn. SarA has been reported to work synergistically with the two-component saeRS system to repress extracellular proteases that would otherwise reduce the accumulation of critical proteins that contribute to the biofilm matrix [30]. Downregulation of fnbA/B would potentially lead to a decreased protein biofilm matrix, but upregulation of sarA might neutralize and counteract this effect in the Tn mutants.

Inactivation of Specific TCA-Cycle Genes Is Associated with a High Metabolic Fitness Cost
Using live-dead staining on flow biofilms, we found a significantly higher amount of dead cells in the biofilms formed by argH::Tn, fumC::Tn, sdhA::Tn, sdhB::Tn, sucC::Tn, and rocF::Tn mutants. However, these mutants neither demonstrated attenuated or slower growth on growth curve assays nor were the total number of cells in their respective biofilms significantly different from those in UAS-391-Ery s biofilms. However, Halsey et al. have shown that Tn mutants with mutations in TCA-cycle genes past the 2-oxoglutarate node (fumC, sdhA, and sucC) did not grow at all in their planktonic S. aureus growth assay [31]. The fact that these defects were not detected in our corresponding TCA-cycle Tn mutants might be due to differences in growth media.
A higher number of dead cells found in the biofilms of Tn mutants might be indicative of a high metabolic cost for the bacterium. It is important to note that propidium iodide does not only stain cells with a compromised membrane, but also eDNA. However, the proportion of eDNA in the biofilm matrix, observed by DNase I digestion, did not differ significantly between the Tn mutants (on average 52% and 55% under flow and no flow conditions, respectively) and UAS391-Ery S (on average 44% and 53% under flow and no flow conditions, respectively). Thus, based on our data, inactivation of the TCA-cycle is likely not associated with any change in eDNA biofilm matrix composition.
In conclusion, we identified an important role of the TCA-cycle in mediating biofilm formation, specifically by influencing the matrix composition, in MRSA USA300 biofilms. These metabolic pathway hits require further screening of MRSA of different clonal lineages to confirm commonality of the target mechanisms and eventually yield interesting therapeutic targets.