Methionine Limitation Impairs Pathogen Expansion and Biofilm Formation Capacity

Methionine auxotrophs need high concentrations of free methionine to overcome growth defects.

We chose three different model organisms, namely, E. coli, S. aureus, and P. aeruginosa, for analysis of physiological effects associated with a lack of methionine biosynthesis. Each organism uses distinct methionine biosynthesis pathways and enzyme classes to perform the reactions (Table 1). E. coli performs the acetylation step using succinyl-CoA and a MetA-class enzyme. In contrast, S. aureus and P. aeruginosa use acetyl-CoA and a MetX-class enzyme. E. coli and S. aureus employ the transsulfurylation pathway to create homocysteine, while P. aeruginosa relies on the direct sulfurylation pathway. Finally, all model organisms perform the methylation of homocysteine using MetE and MetH-class enzymes.

Isogenic mutants defective in key enzymes of the methionine biosynthesis pathways in each organism were obtained. Markerless deletion mutants in S. aureus (strains Newman, SH1000, and USA300 LAC) were created by allelic replacement. Transposon insertion mutants of P. aeruginosa PA14 as well as E. coli BW25113 strains with met genes replaced by a kanamycin marker were acquired from mutant libraries (https://cgsc2.biology.yale.edu/, http://pa14.mgh.harvard.edu/cgi-bin/pa14/home.cgi). Each organism carries the two redundant enzyme classes MetE and MetH. Mutants defective in either one alone would not be expected to be auxotrophic for methionine. Indeed, P. aeruginosa mutants deficient in either MetE or MetH showed wild-type (WT) levels of growth yield in methionine-dependent growth assays (see Fig. S1 in the supplemental material). Therefore, we focused on the other steps of the pathway.

All mutant strains grew in a similar fashion to the parental strain when cultivated in complex medium (data not shown) but were unable to grow when inoculated into defined medium without methionine, confirming methionine-auxotrophic phenotypes (Fig. 2). Only E. coli BW25113 ΔmetC showed slow growth in the absence of methionine and reached an optical density (OD) of ∼0.5 after 50 h of incubation (see Fig. S2 in the supplemental material). However, in our standard experiments (∼12 to 15 h of growth), the increase in optical density was hardly detectable. Furthermore, the results showed that the metY gene of P. aeruginosa was dispensable, while metZ was essential for proliferation without external methionine. This result was unexpected since both enzymes are supposedly able to perform the direct sulfurization reaction.

All of the mutants needed large amounts of methionine to reach the same density as the WT. Mutants of S. aureus strains Newman, USA300 LAC, and SH1000 needed a concentration of 50 to 100 µM methionine to overcome auxotrophy (Fig. 2; see Fig. S3 in the supplemental material). In contrast, E. coli and P. aeruginosa needed even higher concentrations (100 to 200 µM). Furthermore, we observed that growth yields of S. aureus Newman and P. aeruginosa PA14 wild-type strains did not increase when additional methionine was available. This result suggests that autotrophic strains are actively synthesizing endogenous methionine rather than acquiring it form exogenous sources.

The individual methionine auxotrophs of each species showed very similar phenotypic characteristics. This finding makes it unlikely that secondary site mutations or polar effects might be associated with mutagenesis. However, we thought to confirm this in one model organism and complemented the E. coli ΔmetA strain by expression of the metA gene from the IPTG inducible vector pME6032. This abolished the observed differences in final bacterial density (see Fig. S4 in the supplemental material).

These experiments indicate that blockage of methionine biosynthesis cannot be easily overcome by using environmentally available methionine and that auxotrophic strains have growth disadvantages at concentrations reported to be available in human serum (25 to 48 µM).

Human serum contains an insufficient source of methionine.

The addition of 30 µM methionine to defined medium was insufficient to restore a WT phenotype to methionine auxotrophs of all three organisms. However, it remained unclear whether this finding would apply in vivo as human serum is a complex mixture of free amino acids, proteins, and peptides that might be used as a source of methionine by invading microorganisms. Therefore, we investigated the potential of heat-inactivated pooled human serum to permit the growth of methionine auxotrophic mutants. For each species, we chose a representative mutant defective in the first step of methionine biosynthesis (acetylation). Therefore, free methionine as well as all methionine precursors downstream of the acetylation reaction (compare in Fig. 1) should be able to serve to bypass the blockage. Since all strains grew poorly in pooled human serum (data not shown), we supplemented human serum with methionine-free defined medium and compared the number of CFUs after 24 h of incubation. As shown in Fig. 3, the mutants retained a distinct growth disadvantage over the entire range of serum concentrations tested and even 50% pooled human serum did not allow them to reach WT levels of growth yield. Importantly, the addition of free methionine to human serum abolished the differences between the WT and mutant strains (Fig. 3D), indicating that methionine limitation was responsible for the observed defects.

These data indicate that the combined sources of methionine present in human serum are insufficient to restore a WT phenotype in mutant strains, suggesting that methionine biosynthesis might be a valid target for new antimicrobials.

Methionine auxotrophy reduces protein-dependent biofilm formation by staphylococci.

Both staphylococci and pseudomonads form clinically relevant biofilm, for instance, in catheter- and joint replacement-associated infections (19). S. aureus SH1000 biofilm relies predominantly on proteins (20). Therefore, we speculated that the capacity to form biofilm might depend on the ability to synthesize methionine and investigated this using the methionine biosynthesis inhibitor propargylglycine (PG). PG specifically inhibits cystathionine-γ-synthase (21) which is well conserved among staphylococcal species. Indeed, in the absence of free methionine, PG reduced the growth yields of S. aureus SH1000 (Fig. 4A). PG concentrations reducing growth yields only partially had profound effects on biofilm formation of the strain (Fig. 4A and B). To exclude that this effect reflected the decreased final growth yields in the presence of PG, we analyzed the protein levels in the cell wall of SH1000. Cell wall-anchored proteins, such as the fibronectin-binding proteins (22) or SasG (23), are associated with biofilm formation in different staphylococcal lineages. Therefore, we assumed that cell wall-associated proteins might also contribute to biofilm formation under our experimental conditions. We investigated this using cell wall extracts of SH1000 grown under methionine-limiting conditions in the presence or absence of PG. The extracts were analyzed by infrared quantification of proteins separated by SDS-PAGE and stained with Coomassie blue. PG treatment led to a decrease in cell wall protein content (Fig. 4C), supporting the idea that reduced protein amounts in the cell wall decreased the capacity to form proteinaceous biofilms.

We also created a ΔmetC deletion mutant in the SH1000 background and tested the mutant for biofilm formation. Similar to S. aureus Newman mutants described above, SH1000 ΔmetC was auxotrophic for methionine and needed 50 µM methionine to reach the same growth yields as the WT (Fig. S3 in the supplemental material). The auxotrophic mutant also showed a reduced biofilm formation, and WT levels of biofilm were restored by the addition of 30 to 50 µM methionine (Fig. 4D).

Next, we investigated whether methionine limitation also affects the formation of polysaccharide-dependent biofilms. S. epidermidis RP62A produces a biofilm dependent on the polysaccharide intracellular adhesion (PIA) (24). Similar to the inhibition of SH1000, PG reduced growth yields of S. epidermidis RP62A in methionine-deficient medium (Fig. 5A). However, the biomass of RP62A biofilms remained unaltered (Fig. 5B), suggesting that biosynthesis of the PIA matrix was not dependent on methionine biosynthesis. To validate this finding, we isolated the biofilm matrix of S. epidermidis RP62A grown in methionine-deficient medium in the presence and absence of PG and assessed PIA semiquantitatively. Indeed, we found that the amount of PIA was not reduced by PG. In contrast, the amount of PIA seemed slightly increased in the presence of PG (Fig. 5C). Of note, no PIA was detected when an icaA-deficient mutant was used, confirming the specificity of this assay (data not shown).

In conclusion, these data indicate that in contrast to polysaccharide biofilms, proteinaceous staphylococcal biofilm formation depends on the availability of sufficient amounts of methionine.

Methionine auxotrophy leads to the dispersal of P. aeruginosa biofilms.

The capacity to form biofilms is key for P. aeruginosa to cause ventilator-associated pneumonia, catheter-associated urinary tract infections, and chronic lung infections in cystic fibrosis (CF) patients (9–11). Its biofilm consists of exopolysaccharides, such as galactose and mannose-rich polysaccharide Psl (25), the glucose-rich polysaccharide (Pel) (26), and alginate (27). In addition, extracellular DNA (eDNA) and proteins are present (28). Biofilm formation by the P. aeruginosa PA14 WT strain and the isogenic methionine auxotrophic mutants was compared. Defined medium was supplemented with 100 µM methionine, and biofilm formation was measured at the liquid/air interface using the crystal violet microtiter plate method (29) which allows staining and quantification of biofilm mass comprising embedded cells and surrounding matrix. Only a minor difference in growth yield between the WT and the mutant strains was observed. (Fig. 6A). However, all mutants produced a significantly reduced biofilm mass under these conditions. Also, the P. aeruginosa ΔmetY mutant, which did not show any growth deficit compared with the parental strain, exhibited impaired biofilm formation (Fig. 6B).

Given the strong effects of methionine auxotrophy on P. aeruginosa biofilm formation, we asked whether this reflects differences in the production of the extracellular matrix or within the number of bacterial cells. To investigate this question, the Calgary biofilm device was used consisting of a 96-well plate filled with defined medium and a lid containing pegs that reach into the medium (30). The biofilm formed on the pegs was dissociated to determine viable counts of bacteria located within the biofilm (31). In media containing 100 µM methionine, biofilms of methionine auxotrophs and the autotrophic parental strain harbored similar numbers of CFU (Fig. 6C, top), suggesting that differences between WT and mutants in the crystal violet assay were due to different quantities of extracellular matrix. This result opened the question as to whether methionine biosynthesis is relevant for the initial formation of the biofilm or whether it is also important for biofilm growth and maintenance. To investigate this question, P. aeruginosa biofilm was allowed to form on pegs in the presence of 100 µM methionine. After 48 h, the pegs were transferred to fresh medium containing different concentrations of methionine and incubated for 24 to 48 h, after which CFUs were enumerated. The WT and the ΔmetY mutant were able to maintain a constant number of CFUs from 48 to 96 h in medium lacking methionine. In contrast, the viable counts of ΔmetX and ΔmetZ mutants decreased by 4 to 5 orders of magnitude in the same time frame. In half of the experiments, no viable bacteria were recovered from biofilms of the ΔmetX and ΔmetZ mutants (Fig. 6C). This effect was abolished when extracellular methionine (10 to 50 µM) was present in the medium. However, significantly fewer ΔmetX bacteria were recovered from the biofilm than from the WT even in the presence of 30 to 50 µM methionine.

The biofilm matrix of P. aeruginosa PA14 binds Congo red, giving WT strain colonies a distinct red color in the presence of this dye. In contrast, Δpel mutants that are deficient in the carbohydrate-processing Pel proteins fail to stain with Congo red (26). P. aeruginosa WT and the ΔmetX mutants were grown on defined M9 agar plates containing 30 to 100 µM methionine and Congo red. All strains formed comparably sized colonies. The PA14 WT strain showed an intense red color on all concentrations of methionine, indicating the formation of a glycocalyx. In contrast, the color of ΔmetX colonies was dependent on the methionine concentration. The strain remained virtually colorless in the presence of 30 to 50 µM methionine, strongly resembling the reported phenotype of Pel-deficient mutants. This phenotype was reversed by increasing the concentration of methionine (Fig. 6D).

These data suggest that the availability of sufficient levels of methionine is critical for both the development and maintenance of the biofilm, as well as for glycocalyx formation.

Bacterial strains and growth conditions.

All bacterial strains used in this study are listed in Table 2. If not specified otherwise, S. aureus strains were grown at 37°C in tryptic soy broth (TSB; Oxoid) or on tryptic soy agar (TSA). E. coli and P. aeruginosa strains were grown at 37°C in lysogeny broth (LB) or on lysogeny agar (LA) (Oxoid).

All liquid overnight cultures for the experiments described below were grown in 20-ml volumes in 100-ml flasks containing one baffle. Cultures were incubated at 37°C with 140 rpm agitation in an Innova 44 incubator (New Brunswick Scientific). CO₂ content was not controlled.

Creation of markerless deletion mutants in S. aureus.

Targeted mutagenesis of S. aureus was performed using the thermosensitive plasmid pIMAY and allelic exchange. The procedure is described in detail elsewhere (59). In brief, 500-bp DNA fragments upstream and downstream of the genes of interest were amplified by PCR (oligonucleotides are summarized in Table 3). A sequence overlap was integrated into the fragments to allow fusion and created an ATG-TAA scar in the mutant allele. The 1-kb deletion fragments were created using spliced extension overlap PCR and cloned into pIMAY. The deletion plasmids were used to transform S. aureus target strains using a standard procedure at 30°C. Integration of the plasmid was selected at 37°C, and excision was promoted by growing the strains in liquid culture overnight at 30°C. Loss of the plasmid was selected on plates containing 1 µg/ml anhydrotetracycline, and the sensitivity of strains toward chloramphenicol was confirmed. Strains carrying the mutant allele were identified by colony PCR.

Construction of pME6032::Meta.

The metA gene was amplified from the chromosome of E. coli BW25113 by PCR (Table 3). The PCR product was purified and treated with EcoRI and BglII. After purification, the DNA fragment was cloned into pME3032 treated with the same endonucleases as mentioned previously. The recombinant plasmid was used to transform E. coli XL1-Blue using standard techniques. The plasmid was validated by Sanger sequencing and transferred to E. coli BW25113ΔmetA.

Bacterial growth in methionine-limited media.

For overnight cultivation, S. aureus was grown in synthetic nasal medium 3 (SNM3) (15) containing 30 µM methionine. E. coli and P. aeruginosa strains were grown in M9 medium containing 100 µM methionine. Cells were harvested by centrifugation (10 min, 5,000 × g), washed with phosphate-buffered saline (PBS), and adjusted to an OD at 600 nm (OD₆₀₀) of 1. In a 48-well plate, (Nunclon Delta surface) 500-µl volumes of defined medium (SNM20 for S. aureus and M9 for E. coli and P. aeruginosa) containing different methionine concentrations were inoculated to an OD₆₀₀ of 0.01. The plates were sealed with adhesive covers (Excel Scientific) and incubated at 37°C in an Epoch 2 microplate reader (BioTek) for 20 h while agitating (807 rpm). The OD₆₀₀ was measured every 15 min.

For complementation of E. coli ΔmetA, the overnight cultures were grown and harvested as described above. For the growth assay in the microtiter plate, M9 medium without methionine was supplemented with 1 mM IPTG. The plate was incubated as described above.

Growth in human serum.

For overnight cultivation, staphylococci were grown in synthetic nasal medium 3 (SNM3) (15) containing 30 µM methionine. E. coli and P. aeruginosa strains were grown in synthetic M9 medium containing 100 µM methionine. Cells were harvested by centrifugation (10 min, 5,000 × g), washed with PBS, and adjusted to an OD₆₀₀ of 0.1. Heat-inactivated pooled human serum derived from healthy males (Sigma-Aldrich) was diluted in either SNM20 or M9 media to a final concentration of 50%, 25%, or 10%, and 200-µl volumes were transferred to a 96-well plate (Greiner). Each well was inoculated with 15 µl of cells (OD₆₀₀, 0.1; final concentration, ∼5 × 10⁶ CFU/ml), sealed with adhesive covers, and incubated at 37°C with shaking (140 rpm) for 24 h in an Innova 44 incubator. To enumerate CFUs, samples were taken and serial dilutions were plated on TSB agar plates.

Crystal violet assay for quantification of bacterial biofilms.

S. aureus was grown overnight in SNM3 (15) containing 30 µM methionine. S. epidermidis was grown in modified Tris minimal succinate (mTMS) medium containing 30 µM methionine. mTMS was prepared according to a previous report (60), with the exception that casamino acids were replaced with proteinogenic amino acids (A, 990 µM; R, 660 µM; C, 66 µM; E, 660 µM; G, 990 µM; H, 330 µM; L, 1,980 µM; K, 990 µM; F, 990 µM; P, 990 µM; S, 792 µM; T, 1,320 µM; W, 132 µM; and V, 660 µM) and ornithine-HCl (660 µM). After overnight growth, cells were harvested by centrifugation (10 min, 5,000 × g) and adjusted to OD₆₀₀ of 1 in SNM3/mTMS containing 1% glucose. This suspension was diluted 1:200 in SNM3/mTMS containing 1% glucose. For S. aureus experiments, 96-well Nunclon Δ surface microtiter plates (Thermo Fisher) were coated overnight at 4°C with 100 µl of 0.045 µg/ml fibrinogen (Calbiochem) to increase biofilm formation. After the plates were washed three times with PBS, 200 µl of the inoculated defined medium was added to the wells. For inhibition experiments, a final concentration of 100 µg/ml propargylglycine was added. The plate was incubated statically for 24 h at 37°C. The plates were washed thrice with PBS, stained with 100 µl of 2.3% crystal violet (Sigma-Aldrich), and washed five times with 300 µl PBS. Retained color was extracted in 100 µl 5% acetic acid. OD at 570 nm (OD₅₇₀) was quantified using a CLARIOstar device (BMG Labtech).

For P. aeruginosa biofilms, bacteria were grown overnight in M9 medium containing 100 µM methionine. Overnight cultures were harvested, adjusted to an OD₆₀₀ of 1 in M9 medium (100 µM methionine), and used to inoculate (1:200) M9 medium with appropriate concentrations of methionine. A total of 200 µl was transferred to 96-well plates and grown statically at 37°C for 48 h. The 96-well plate was washed and stained with crystal violet as described above.

Isolation of S. aureus cell wall extracts.

S. aureus was grown overnight in SNM3 containing 30 µM methionine. Cells were harvested by centrifugation (10 min, 5,000 × g), and washed with SNM3 without methionine. A 20-ml volume of SNM3 was inoculated to an OD₆₀₀ of 0.1, and PG (0 to 100 µg/ml) was added. The culture was incubated with agitation for 24 h. Cells were collected (10 min, 5,000 × g) and washed once with wash buffer (WB; 10 mM Tris-HCl [pH 7] and 10 mM MgCl₂). A total of 1 ml WB was adjusted to an OD₆₀₀ of 2. Cells were collected and resuspended in 1 ml digestion buffer (10 mM Tris-HCl [pH 7], 10 mM MgCl, 500 mM sucrose, 0.3 mg/ml lysostaphin, 250 U/ml mutanolysin, 30 µl protease inhibitor cocktail [Roche; 1 complete mini tablet dissolved in 1 ml H₂O], and 1 mM phenylmethylsulfonyl fluoride [PMSF]) and incubated at 37°C for 1 h. The protoplasts were removed by centrifugation (8,000 × g for 10 min at 4°C). The supernatant (containing the cell wall fraction) was collected, and 15 µl was used for SDS-PAGE analysis using standard techniques. Next, 7.5% SDS gels were stained with Coomassie brilliant blue G 250 (Serva). As Coomassie blue protein stain is a strong 700-nm fluorophore, the protein amount within individual bands was quantified using an Odyssey CLx device (LI-COR) which allows detection and accurate quantification of fluorescent signals in the near infrared spectrum.

Isolation of S. epidermidis biofilm matrix.

A 10-ml bacterial culture (mTMS, 30 µM methionine) was grown overnight in a 50-ml flask containing one baffle at 37°C. Bacteria were harvested and washed once with PBS. Next, 2.5 ml methionine-free mTMS was inoculated to an OD₆₀₀ of 0.01, and 0 to 100 µg/ml PG was added. A total of 200-µl volumes of the inoculated medium were dispersed to 12 wells of a 96-well plate (Nunclon Delta surface). The plate was incubated at 37°C for 24 h. The plate was incubated for 30 min in an ultrasonic bath (Bandelin SONOREX RK 100). The biofilm was scratched off the wells and centrifuged at 3,000 × g for 10 min. The pellet was resuspended in 200 µl of 0.1 M EDTA (pH 8). The suspension was incubated for 30 min in an ultrasonic bath and centrifuged at 8,000 × g for 10 min. The supernatant was used for immunoblot analysis. Serial dilutions were prepared, and 5 µl was spotted on a nitrocellulose membrane (LI-COR). The membrane was blocked for 20 min with blocking buffer (Thermo Scientific) and incubated for 20 min with 130 ng/ml wheat germ agglutinin coupled to horseradish peroxidase (Biotium). The blot was washed twice with Tris-buffered saline with Tween 20 (TBST) and once with Tris-buffered saline (TBS). Detection was carried out using the WesternSure premium chemiluminescent substrate (LI-COR) and the Bio-Rad ChemiDoc XRS imaging system.

Peg plate assay of biofilm formation and detachment of P. aeruginosa PA14.

P. aeruginosa strains were grown overnight in M9 medium with 100 µM methionine. Overnight cultures were diluted to an OD₆₀₀ of 0.1 in M9 medium (with 100 µM methionine), and 200 µl was transferred to a 96-well plate (Thermo Scientific) with a transferable 96-peg solid-phase (TSP) plate on top (Thermo Scientific). The pegs of the TSP plate reached into the inoculated medium and allowed initial bacterial adhesion. The plates were incubated statically for 2 h at 37°C. The TSP plate was transferred to a new 96-well plate filled with 200 µl of sterile M9 medium containing 100 µM methionine and incubated statically for 48 h at 37°C. If needed, the TSP was subsequently transferred to a new microtiter plate containing sterile M9 medium with various concentrations of methionine and incubated for an additional 24 to 48 h. For enumeration of living cells within the biofilms, TSP plates were transferred onto a 96-well plate containing 200 µl 0.1 M EDTA and 0.1% CHAPS (3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate) and incubated for 1 h on a rocking table. CFUs were enumerated by plating serial dilutions of the bacterial suspensions onto LB agar plates.

Congo red assay to detect P. aeruginosa glycocalyx.

P. aeruginosa strains were grown overnight in M9 medium containing 100-µM methionine. Cells were harvested, washed with methionine-free M9 medium and adjusted to an OD₆₀₀ of 0.025 in the same medium. A total of 5 µl of the bacterial suspension was spotted on defined M9 agar plates containing 30 to 100 µM methionine as well as 40 µg/ml Congo red and 20 µg/ml Coomassie brilliant blue G-250. The color of arising colonies was assessed after incubation at 37°C for 24 h followed by further incubation at room temperature for 24 h.

Statistics.

Statistical analysis was performed by using GraphPad Prism.