Permissive Fatty Acid Incorporation Promotes Staphylococcal Adaptation to FASII Antibiotics in Host Environments

in a broader, host-relevant context. We report that S. aureus rapidly adapts to FASII antibiotics without FASII mutations when exposed to host environments. FASII antibiotic administration upon signs of infection, rather than just after inoculation as commonly practiced, fails to eliminate S. aureus in a septicemia model. In vitro , serum lowers S. aureus membrane stress, leading to a greater retention of the substrates required for environmental fatty acid (eFA) utilization: eFAs and the acyl carrier protein. In this condition, eFA occupies both phospholipid positions, regardless of anti-FASII selection. Our results identify S. aureus membrane plasticity in host environments as a main limitation for using FASII antibiotics in mono-therapeutic treatments


RESULTS
A FASII Antibiotic Does Not Clear S. aureus Infection in a Septicemia Murine Model Results of animal tests are decisional checkpoints for antibiotic development.FASII antibiotic challenge tests to date administer treatments within minutes to a few hours post-infection (summarized in Morvan et al., 2017), i.e., prior to bacterial dissemination to host organs and before clinical symptoms would call for antibiotic treatment (Leekha et al., 2011;Surewaard et al., 2016).This consideration guided the design of the infection and treatment protocol used here (Figure 1A).MRSA strain USA300 was administered by the intravenous route.Antibiotic treatments were initiated 16 h (T16) post-infection, at which time animals exhibited signs of sickness (lethargy and ruffled fur).Group 1 received no treatment.As an antibiotic, group 2 received AFN-1252, a pipeline FASII inhibitor targeting FabI, an enoyl-acyl-carrier-protein-reductase, following recommended dosing (Kaplan et al., 2013b).Group 3 received vancomycin, which was used to validate that treatment starting at T16 was feasible.At T40 (i.e., 24 h post-antibiotic treatment), bacterial counts were significantly lower in the organs of both antibiotic-treated animals compared to the untreated group, attesting to AFN-1252 activity (Figure 1B).At T88, vancomycin-treated animals were essentially free of bacteria.However, organs from AFN-1252treated mice still contained S. aureus CFU (colony forming units); bacterial counts were increased 10-fold in kidneys (to 5 3 10 6 ; p % 0.01), decreased 10-fold in the liver (p % 0.05), and unchanged in the spleen.FASII antibiotic treatments thus failed to eliminate S. aureus in a septicemia model.
The fatty acid kinase FakA functions with FakB1 or FakB2, which orient specificity, for fatty acid incorporation in S. aureus (Parsons et al., 2014).The fakA and fakB1fakB2 mutants show attenuated virulence (Lopez et al., 2017), which could reflect the need for host fatty acid incorporation during infection.
Although single fakB1 or fakB2 mutants may display partial phenotypes (each still incorporates some fatty acids; Cuypers et al., 2019;Parsons et al., 2014), we asked whether such a mutant would be more sensitive than the parental strain to AFN-1252 treatment.This was tested by infecting mice with a 1:1 mix of USA300 and fakB2 and following the same antibiotic protocol as above.Erythromycin resistance of fakB2 (Fey et al., 2013) was used to determine the proportion of fakB2 present during infection.The USA300:fakB2 average ratios in untreated mice 88 h post-infection were 1.6, 9.9, and 3.3, respectively, in kidney, liver, and spleen in organs at 88 h (Figure 1C).In contrast, AFN-1252-treated mice showed average USA300:fakB2 ratios of 68.9, 19.2, and 6.8, respectively, in kidney, liver, and spleen.These results indicate that fakB2 is preferentially eliminated by AFN-1252.They support the proposal that fatty acid incorporation occurs during infection and contributes to anti-FASII treatment failure.The underlying mechanisms leading to FASII inhibitor escape in host-relevant conditions were investigated.

Host Constituents Promote Rapid Staphylococcal Adaptation to FASII Antibiotics
We reasoned that in septicemic infection, serum and other host constituents bind eFAs and may neutralize FASII inhibitors (Balemans et al., 2010;Hunt et al., 2016;Lacey and Lord, 1981;Litus et al., 2018).The effect of serum on FASII antibiotic activity was tested using triclosan, an extensively used biocide that also targets FabI (McMurry et al., 1998).S. aureus triclosan sensitivity was compared in a medium containing a 3-fatty-acid cocktail (FA, containing C14:0, C16:0, and C18:1cis, 0.17 mM each; with triclosan, FA-Tric) and the same medium supplemented with serum (SerFA-Tric) (Figure 2A).USA300 growth without serum was inhibited by triclosan, with the emergence of FASII mutants usually after 24-48 h of incubation (Morvan et al., 2016).However, serum supplementation markedly shortened latency compared to BHI (brain heart infusion medium) cultures-to 8 h-and was followed by near-normal growth (Figure 2A).Similar results were observed with the unrelated S. aureus Newman strain (Figure S1A).If S. aureus outgrowth were due to triclosan titration by serum, FASII would remain active so that bacterial fatty acid composition would be endogenous.However, the contrary occurred: bacterial fatty acid profiles during outgrowth in the SerFA-Tric medium were totally exogenous (Figures 2B and S1B).As expected, albumin, a major serum constituent, also resulted in FASII bypass (Figure S1C).Similarly, when USA300 was grown with liver or kidney extracts (without added fatty acids) and triclosan, outgrowth kinetics were similar to those of SerFA-Tric cultures, and cells bypassed the FASII block by incorporating organ-derived fatty acids (Figure S1D).Importantly, pre-incubation in SerFA prior to FASII antibiotic treatments shortened the time prior to S. aureus outgrowth.USA300 was challenged with the AFN-1252, which led to a longer (10 h) latency phase prior to outgrowth than did the triclosan.However, pre-incubation in serum shortened latency to about 6.5 h for both drugs, compared to that in the non-selective medium (Figure S2), indicating that bacterial pre-exposure to the lipid-rich host environment contributes to limiting FASII antibiotic efficacy.
Staphylococcus epidermidis, haemolyticus, and lugdunensis are emerging pathogens that, like S. aureus, synthesize branched-chain fatty acids.Representative strains were grown in SerFA and treated with AFN-1252 as above (Figure S3).All cultures grew after overnight incubation and displayed exogenous fatty acid profiles, indicating that these staphylococcal species also bypass FASII inhibitors.
These results show that in serum, S. aureus and other staphylococcal species escape anti-FASII inhibition and maintain robust growth by replacing endogenously synthesized fatty acids with eFAs.They indicate serum actually enhances, rather than prevents, eFA incorporation by S. aureus.
Curves are the average of three biological replicates.
Representative fatty acid profiles from over 10 determinations are shown.(C) Cell vitality and permeability were evaluated by fluorescence microscopy using, respectively, Syto9 TM and propidium iodide (PI) probes.USA300 was grown in FA-Tric and SerFA-Tric for 6 h; mid-exponential phase cultures in BHI, FA, and SerFA were used as references.Proportions of vital and permeable bacteria were determined on ~10,000 bacteria per condition, issued from three independent experiments giving comparable images (Figure S4).Proportions of vital (green), permeable (red), and mixed-cell clusters (yellow) are shown.(D) CFUs were determined in parallel on three independent cultures per condition: BHI starting culture (T0), FA-Tric, and SerFA-Tric samples (T6 = 6 h).(E) Fatty acid profiles were determined on four biological replicates grown as in (C).Proportions of eFAs compared to total fatty acids are shown after 6 h of selection.Note that fatty acid profiles are fully exogenous upon outgrowth.
no detectable genome mutations.The other clones carried SNPs corresponding to commonly found variants and are likely unrelated to FASII antibiotic adaptation (described in Table S1).The absence of FASII mutations distinguishes this adaptation mechanism from resistance due to FASII target or initiator gene mutations (Ciusa et al., 2012;Morvan et al., 2016).The S. aureus evasion of FASII antibiotics in serum identifies a unique strategy of condition-dependent adaptation.
Serum Lowers Fatty-Acid-Induced Bacterial Membrane Permeability and Improves Fitness Numerous fatty acids reportedly perturb bacterial membrane integrity and are a source of stress, while serum albumin neutralizes these effects (Lacey and Lord, 1981;Litus et al., 2018;Nicolaides, 1974;Parsons et al., 2012).Accordingly, serum abolished the eFA-provoked growth lag in the non-selective medium (Figure 2A).Serum effects on S. aureus vitality, permeability, and cell state were examined.Free fatty acids had  strong permeabilizing effects on cells from the FA and FA-Tric, as compared to BHI cultures, as evaluated by fluorescence microscopy; these effects were offset by serum in the SerFA and SerFA-Tric cultures (Figures 2C, S4A, and S4B).Plating efficiency was $10 3fold higher after 6 h of growth in SerFA-Tric compared to FA-Tric (Figure 2D).The accumulation of tetrads comprising mixed-stained cells in triclosan-treated cultures correlates with the observed latency prior to outgrowth (Figure S4C).Importantly, serum facilitates fatty acid incorporation in the latency period, as seen in the 6-h FA-Tric and SerFA-Tric cultures (Figure 2E).The bacterial stress state in FA-Tric supplemented or not supplemented with serum was also assessed by a proteomics approach using the USA300 spa strain (Figure S5).Differences in stress-related protein abundance between anti-FASIItreated and control cultures were, overall, more pronounced when the serum was absent.Serum therefore improves S. aureus fitness in fatty-acid-containing environments and contributes to FASII antibiotic adaptation via eFA incorporation.

S. aureus Grown in Serum Shows a Greater Retention of the ACP and a Reduced Capacity for eFA Efflux via FarE
We questioned how serum affects the availability of two key substrates required for FASII antibiotic adaptation: the acyl carrier protein (ACP) and eFAs.ACP is required for both de novo fatty acid synthesis via FASII and eFA incorporation in the phospholipid 2-position (Figure 3A) (Majerus et al., 1964;Morvan et al., 2016).eFAs induce membrane leakage that depletes S. aureus ACP pools and could limit eFA incorporation during FASII inhibition (Parsons et al., 2011(Parsons et al., , 2012)).S. aureus ACP pools were compared by immunoblotting using anti-ACP antibodies in total extracts from cultures grown without and with serum and triclosan (Figure 3B).ACP levels were lower in extracts from cells grown in eFA, compared to the BHI medium, as reported (Parsons et al., 2012).ACP was barely detected in FA-Tric-grown cells after 2 h, and it remained low at 4 h.In contrast, the addition of serum reversed this effect, leading to greater intracellular ACP availability.S. aureus USA300 dSer2FA pre-cultures and cultures were prepared with 10% delipidated serum (dSer) and C17:1tr and C18:1cis.AFN-1252, 0.5 mg/ml was added at OD 600 = 0.1 at the start of kinetics.(A) Fatty acid composition at indicated time points is shown as the proportion of endogenous branched-chain fatty acids (BCFAs C15, i15, ai15; purple), endogenous saturated (straight chain) fatty acids (SCFAs C18:0, C20:0; pink), and eFAs (C17:1tr and C18:1cis, green); corresponding profiles are shown in Figure S6B.Gray curve represents OD 600 readings at indicated times.Shown is the average of two independent experiments.(B) Kinetics of phosphatidylglycerol (PGly) profile modifications during FASII bypass in FASII antibiotics.Masses are represented: red, PGly species with one eFA and one endogenous fatty acid; green, PGly species with eFAs in both positions.Fatty acid composition of each major PGly mass is given in Table S3 (dSer2FA-AFN).Shown is representative result from two independent experiments.BHI and dSerFA cultures (Figure 4) were analyzed together with this experiment.The same experiment performed using triclosan in place of AFN-1252 (data not shown) gave comparable results.

Phosphatidylglycerol species
Intracellular eFA pools are limited by a recently characterized S. aureus fatty acid efflux pump (FarE), which is induced by unsaturated eFAs (Alnaseri et al., 2015(Alnaseri et al., , 2019)).Proteomics analysis (as above) confirmed high FarE expression in all eFA-containing media without serum.In contrast, nearly no FarE was detected in cells issued from serum-containing cultures, despite the presence of eFAs (Figure 3C).The greater S. aureus intracellular retention of ACP and reduced eFA efflux capacity in serum is consistent with more efficient eFA incorporation and FASII antibiotic adaptation (Figure 2E).

Exogenous Fatty Acids Can Occupy Both Phospholipid Positions in the Absence and Presence of FASII Antibiotics
The accepted rationale for developing S. aureus FASII inhibitors is the presumed stringent requirement for endogenous branched-chain fatty acid ai15 at the phospholipid 2-position, catalyzed by 1-acyl-sn-glycerol-3-phosphate acyltransferase PlsC (Albanesi et al., 2013;Balemans et al., 2010;Parsons et al., 2011;Schiebel et al., 2014;Zhang and Rock, 2008).The above evidence that FASII antibiotic adaptation leads to eFA incorporation at both phospholipid positions (Figures 2B and  S1B), and previous unexplained observations (Delekta et al., 2018), gave evidence against the generality of phospholipid stringency.However, eFA incorporation in both S. aureus phospholipid positions could be a last-resort choice when the preferred substrate ai15 is unavailable.Alternatively, the use of eFAs versus ai15 may simply depend on the intracellular substrate availability, which increases in serum.eFA incorporation was assessed in non-selective conditions to discriminate between these alternative hypotheses.To remove ambiguity in distinguishing the endogenous from the exogenous fatty acids in phospholipid identifications, a cocktail of unsaturated eFA 17:1trans (tr) and 18:1cis prepared in delipidated serum (dSer2FA medium) was used to supplement S. aureus USA300 growth.Although structurally distinct from S. aureus endogenous fatty acids, 17:1tr and 18:1cis did not interfere with growth in the serum-containing medium (optical density [OD] 600 = 6.4 for both BHI and dSer2FA cultures at 6 h).In this non-selective growth condition, 17:1tr and 18:1cis comprised about 60% of the total fatty acid content (Figures 4A and S6A) at the expense of straight-chain saturated fatty acids, which decreased from 50% in BHI to about 10% in the dSer2FA cultures.Importantly, eFAs occupied both positions in 20%-25% of phosphatidylglycerol (PGly) species after 6 h without the antibiotic (Figure 4B; Table S2).These results prove that wild-type S. aureus incorporates dissimilar fatty acids in both phospholipid positions without the need for FASII antibiotic selection and rules out the previously assumed fatty acid selectivity of phospholipid-synthesizing enzymes.

Serum
Membrane stress

aureus Adaptation to FASII Antibiotics
The membrane stress state impacts eFA and ACP intracellular pools and dictates FASII antibiotic adaptation as shown in the model.(A) High stress: free eFAs permeabilize membranes.Fatty acid efflux via FarE and ACP leakage (Alnaseri et al., 2015(Alnaseri et al., , 2019;;Parsons et al., 2012) leads to the depletion of FASII-bypass substrates.In this condition, eFA-PO 4 is used by PlsY to charge the phospholipid 1-position.FASII antibiotics would arrest growth, leading to cell death or the emergence of FASII initiation mutants that bypass FASII (Gloux et al., 2017;Morvan et al., 2016).(B) Low stress: serum (yellow disk) or other host components reduce eFA toxicity (Lacey and Lord, 1981; this report) and improve membrane integrity.Higher ACP and/or eFA-PO 4 pools drive PlsX directionality to eFA-ACP production.eFA-ACP and endogenous ACP (if not blocked by anti-FASII) compete for phospholipid synthesis at the 2-position via PlsC.eFAs occupy both phospholipid positions even without FASII antibiotics (Figure 4B).eFA (green) and ACP (blue) abundance is represented by font size; phospholipids, ''p'' form; thick arrows, favored reactions; thin ar-rows, reduced reactions.Dashed circle, permeable membrane; solid circle, intact membrane.PlsY mediates eFA incorporation in the phospholipid 1-position; PlsX and PlsC, catalyze fatty acid insertion in the phospholipid 2-position.Only eFA processing is presented.29, 3974-3982, December 17, 2019 3979 shown), in parallel with OD 600 (Figures 5 and S6B).For both FASII antibiotics, bacterial transition from mixed to exclusively exogenous fatty acids was concomitant with outgrowth from latency starting at 8 h post-treatment, and it was completed at 10 h (Figure 5A).For both FASII antibiotic treatments, the predominant phospholipid species at 10 h were totally exogenous (Figure 5B; Table S3).The sharp increase in exclusively eFA-containing phospholipids coincides with the exit from latency, as expected from the coordination between membrane phospholipid synthesis and cell growth (Vadia and Levin, 2015;Vadia et al., 2017).These results show that when S. aureus membrane integrity is maintained, as by the host serum, eFA incorporation is not stringent, and they reflect competition between fatty acids synthesized by S. aureus and those available from the environment.eFAs are incorporated in both phospholipid positions in the absence of selection and completely replace endogenous fatty acids in the presence of FASII antibiotics.

DISCUSSION
S. aureus is shown here to adapt to FASII antibiotics in host environments and without FASII mutations.A FASII antibiotic was ineffective in a septicemia model in which the treatment protocol respected the interval between primo-infection and treatment time.The design of antibiotic challenge tests based on realistic intervals between infection and treatment should improve the predictive value of animal studies, which are decisive for scale-up to clinical trials.Based on this study, AFN-1252 and likely any FASII inhibitor would be ineffective as stand-alone treatments of S. aureus deep infection.
S. aureus responses to FASII antibiotics are schematized in a model (Figure 6): greater bacterial membrane integrity and intracellular ACP and eFA pools, as shown here in serum-containing host environments, lead to complete eFA replacement in phospholipids.The rapid kinetics, absence of FASII mutations during antibiotic treatment, and eFA-replete phospholipids in the absence of selection indicate that adaptation impacts a general, non-FASII-mutated, S. aureus population.Serum and other host components promote low stress and FASII adaptation, which links this model to anti-FASII treatment failure.In contrast, skin surfaces producing fatty acids would correspond to highstress environments unfavorable to FASII bypass except by FASII mutation (Figure 6A; Morvan et al., 2016;Pishchany et al., 2018).Previous in vitro and in vivo studies reporting FASII antibiotic efficacy were based on S. aureus strain 8325 and derivatives (Parsons et al., 2013(Parsons et al., , 2011)).This bacterial lineage bears an atypical 161/288 amino acid deletion in fatty acid kinase subunit FakB1, which may impact saturated eFA entry and FASII antibiotic adaptation (Parsons et al., 2014;data not shown).FASII inhibitors were recently suggested for the elimination of Clostridium difficile and Listeria monocytogenes (Marreddy et al., 2019;Yao et al., 2016).Like S. aureus, these pathogens might show relaxed phospholipid stringency in host environments; niche-dependent antibiotic adaptation in these cases remains to be tested.
Bacterial metabolism and environmental stress play unquestionable roles in the outcome of antimicrobial treatments.For example, reduced metabolism in bacterial persisters is recog-nized as a major means of escape from antibiotic killing (Chuard et al., 1997;Lewis, 2007).In contrast, FASII antibiotic adaptation involves robust bacterial proliferation concomitant with the compensatory incorporation of host fatty acids.Such metabolic rescue is likely frequent, as bacteria commonly salvage metabolites from their environments.This consideration may set a logical limit when selecting targets for antimicrobial drug development.
As shown here, S. aureus and other firmicute pathogens may have reduced genetic requirements in host biotopes, including a non-essentiality of FASII.Interestingly, a wall-less L-form S. aureus multiplies in the presence of cell wall antibiotics and survives by the oversynthesis of fatty acids (Kawai et al., 2018 and references therein).An intriguing possibility is that such ''primitive bacteria'' can use the lipid supply of the host, taking their lifestyle one step further toward antibiotic adaptation and parasitism.
corresponded to fakB2 mutants.This number was subtracted from the total number of CFUs to determine the number of wild-type USA300 bacteria.

Determination of S. aureus fatty acid profiles
Aliquots of S. aureus cultures (routinely an OD 600 equivalent R 1 was used) were centrifuged and washed once in 0.9% NaCl containing 0.02% Triton X-100, followed by two washes in 0.9% NaCl.Whole cell esterified fatty acid determinations were then done as described (Yamamoto et al., 2006).Briefly, cell pellets were treated with 0.5 mL of 1N sodium methoxide in methanol.Heptane (200 ml) was then added, together with methyl-10-undecenoate (Sigma-Aldrich) as internal standard, vortexed for 1 min, and centrifuged.Fatty acid methyl esters were recovered in the heptane phase.Analyses were performed in a split-splitless injection mode on an AutoSystem XL Gas Chromatograph (Perkin-Elmer) equipped with a ZB-Wax capillary column (30 m x 0.25 mm x 0.25 mm; Phenomenex, France).Data were recorded and analyzed by TotalChrom Workstation (Perkin-Elmer).S. aureus fatty acid peaks were detected between 12 and 32 min of elution, and identified with retention times of purified esterified fatty acid standards.

Fluorescence microscopy
Bacteria were grown in BHI, FA, Ser-FA to OD 600 = 0.5-1, or FA-Tric and SerFA-Tric for 6 hours.Cells (OD 600 equivalent of 10-20) were centrifuged and washed once in phosphate buffered saline.Syto9 TM (0.5 ml sample of a 5 mM solution in DMSO; Thermo Fischer Scientific) and propidium iodide (3 ml of 1.5 mM solution in water; Sigma, France) were added to 30 ml samples for respectively viable and permeabilized cell visualization.Cells were observed 10 min post-staining by fluorescent microscopy using a Zeiss AxioObserver Z1 inverted fluorescence microscope equipped with a Zeiss AxioCam MRm digital camera and Zeiss fluorescence filters.Images were processed with the Zeiss ZEN software package using a 38 HE Green Fluorescent Protein filter (excitation wavelength 450/ 490 nm; beam splitter, 495 nm; emission, 500/550 nm) and 45 Texas Red filter (excitation: 540/580 nm; beam splitter, 585 nm; emission, 595-668 nm).The numbers of live (green) and permeabilized (red) cells were counted manually.Dividing cells and tetrads were counted as single entities.Tetrads and clusters containing both viable and permeabilized cells (in FA-Tric and SerFA-Tric samples) were classified in a separate ''mixed'' category.The proportion of tetrads among total cells was determined by manual counting; 15 micrographs from 3 independent cultures per condition were evaluated, based on a total of $10 000 cell clusters per condition.
Genome sequencing S. aureus USA300 and Newman strains were harvested after exit from latency phase for cultures grown in SerFA-Tric (12 h cultures; 3 independent samples for each strain), or in SerFA-AFN (15 h cultures; 3 independent samples of USA300).Cultures in BHI (one sample for each strain), and in SerFA (for USA300) were sequenced as references.DNA extractions were performed using the QIAGEN ''DNeasy â Blood & Tissue Kit,'' following manufacturer's protocol, except that cell pellets were first resuspended in 0.1 mg lysostaphin/ml Tris 10 mM (AMBI, USA) and incubated 30 minutes at 37 C. Genomic DNA sequencing by Illumina HiSeq next generation sequencing was outsourced (GATC-Biotech, Konstanz, Germany).Coverage was estimated to be at least 70-fold for USA300 SerFA-Tric samples, and at least 400-fold for all other samples.The 2 3 150 paired-end reads were analyzed using ''Variation Analysis'' method provided by Patricbrc.org.Bowtie2 (Patricbrc.org) was used to align sequences and SAMtools to identify SNPs (Wattam et al., 2017).SNPs that differed in non-antibiotic-treated USA300 and Newman strains from those in the reference sequence (GenBank Nucleotide accession codes NC_007793.1 and NC_009641 respectively) were subtracted prior to variant screening.Variants were identified as representing at least 80% of reads in sequences for which there were at least 10 reads.

ACP assessment by immunoblotting
The USA300 spa::Tn strain (SAUSA300_0113; Fey et al., 2013) was used for immunoblotting to avoid IgG titration by Protein A; this strain was confirmed to behave like its parent with respect to FASII antibiotics.An overnight BHI USA300 spa preculture was used to inoculate BHI, FA, SerFA, FA-Tric, or SerFA-Tric media at OD 600 = 0.1.Cultures were harvested at OD 600 = $1 for BHI, FA, and SerFA, and after 2 h or 4 h for FA-Tric or SerFA-Tric.All samples were adjusted to equivalent OD 600 values, and washed twice in TE-protease inhibitor (cOmplete Tablets, Mini EASYpack Roche, Germany, as per supplier's instructions), prior to lysis with Fastprep.Samples (20 mg per well as quantified by the Bradford Protein Assay kit (BioRad)) were treated for 3 min at 95 C and then loaded on 12.5% SDS-PAGE gels run at 150 V for 2 h.Gels were then electro-transferred to PVDF membranes (0.2 mm; BioRad; 75mA) for 3 h on a semi-dry transfer unit (Hoefer TE 70).Western blotting and exposure used an ECL kit (Perkin-Elmer) as per supplier's instructions.Rabbit anti-S.aureus ACP antibodies (Morvan et al., 2016) were used at 1:1,300 dilution.
Proteomic analyses of USA300 responses to anti-FASII Four independent overnight BHI precultures of USA300 spa::Tn strain were used to inoculate BHI, FA, SerFA, FA-Tric, or SerFA-Tric media at OD600 = 0.1.Culture extracts were prepared as described for western blotting, and after verification of protein concentration and quality, 10 mg of each protein sample was short-run on SDS-PAGE.Further sample treatment by LC-MS/MS, and bioinformatics and statistical analyses of data are as described (Pe ´rez-Pascual et al., 2017).The reference genome GenBank Nucleotide accession code NC_007793.1 was used for protein annotation.The complete list of proteins expressed in the five growth conditions is available on the Mendeley database (https://doi.org/10.17632/9292c75797.2;https://data.mendeley.com/datasets/9292c75797/draft?a=bb34343b-6314-4421-89bd-b25e2bbdb0df).
Cell Reports 29, 3974-3982.e1-e4,December 17, 2019 e3 Extraction of polar membrane lipids Lipid extractions were performed as described with modifications (Bligh and Dyer, 1959;Thedieck et al., 2006).Briefly, freeze-dried cell material (100 mg) was extracted with 9.5 mL of chloroform-methanol 0.3% NaCl (1:2:0.8v/v/v) at 80 C for 15 min.All following steps were done at room temperature.Extracts were vortexed for 1 h and centrifuged for 15 min at 4000 rpm.Supernatants were collected and cell debris was re-extracted with 9.5 mL of the same mixture, vortexed 30 min, and centrifuged.Supernatants were then pooled and 5 mL each of chloroform and 0.3% NaCl was added and mixed.Phase separation was achieved by centrifugation at 4000 rpm for 15 min.The upper phase was discarded and the collected chloroform phase was evaporated to dryness under a nitrogen stream and stored at À20 C.

Phosphatidylglycerol identification
The polar membrane lipid samples were injected in chloroform in a chromatographic system (ThermoFisher Scientific) including a Dionex U-3000 quaternary RSLC, a WPS-3000RS autosampler and a column oven.Lipid separation was carried out by liquid chromatography on a PVA-Sil column (150 3 2.1 mm I.D., 120 A) (YMC Europe GmbH) with a 10x2 mm guard column packed with the same material.Column temperature was thermostatically controlled at 35 C. The chromatographic method separates phospholipids according to class, and was performed as described (Imbert et al., 2012;Moulin et al., 2015).
For phosphatidylglycerol species identification, this system was coupled to an LTQ-Orbitrap Velos Pro (ThermoFisher) equipped with an H-ESI II probe.Spray voltage was set at 3.3 kV.Heater temperature of the probe was set at 200 C. Sheath gas, auxiliary gas and sweep gas flow rates were set at 20, 8 and 0 (arbitrary unit) respectively.Capillary temperature was set at 325 C and S-lens RF level at 60%.Analysis was performed in negative mode to obtain structural information on phosphatidylglycerol fatty chains.The mass spectrometer is equipped with two analyzers: a double linear ion trap (LTQ Velos Pro) for fragmentation at low resolution and an orbital trap (Orbitrapª) for high resolution detection.Detection was performed in full MS Scan with 100,000 resolution and data dependent MS 2 and MS 3 with collision induced dissociation (CID; collision energy set at 35).Chromatographic retention time was used for polar head identification by comparing to commercial standards.Phospholipid identification was performed using high resolution mass full scan to obtain the formula of the entire PG species (Pulfer and Murphy, 2003), and MS 2 /MS 3 fragmentation to obtain structural information about fatty acid chain composition of each species.

QUANTIFICATION AND STATISTICAL ANALYSIS
Means and standard errors of replicate growth curves and proportions of fatty acids were determined (Excel, Microsoft, USA).For phospholipid determinations, the average of two experiments with the range of duplicates is presented.Data from animal experiments using USA300 only were statistically analyzed by the non-paired non-parametric Mann-Whitney test with GraphPad Prism 5.0 (GraphPad Software, San Diego, California).The same test was used in USA300:fakB2 competition studies to compare results in non-treated and antibiotic-treated animals.In the competition study, the paired and non-parametric Wilcoxon test was used to determine p values in separate organs.

Figure 2 .
Figure 2. Positive Effects of Serum on S. aureus Adaptation to a FASII Antibiotic Figure 6.Conditional S. aureus Adaptation to FASII Antibiotics Figure 3. Impact of Serum on S. aureus Intracellular Retention of ACP and eFAs Expression of fatty acid efflux pump FarE.Proteomic analyses were performed on S. aureus USA300 spa treated as in (B).Samples are in the same order as in (B).Mean values and standard deviation are presented.Black, BHI; pink, FA; light green, SerFA; red, FA-Tric 2 h; middle green, SerFA 2 h; burgundy, FA-Tric 4 h; dark green, SerFA 4 h.Quadruplicate independent samples were used for immunoblots and proteomic analyses.The full proteomic study is available at https://data.mendeley.com/datasets//9292c75797/2.