Polyamine Transport Is Required for Stress Responses and Capsule Production in Streptococcus Pneumoniae

Infections due to Streptococcus pneumoniae, a commensal in the nasopharynx, still claim a signicant number of lives worldwide. Genetic plasticity, antibiotic resistance, limited serotype coverage of the available polysaccharide-based conjugate vaccines confounds therapeutic interventions. Pathogenic systems that allow successful adaption and persistence in the host could be potential innovative targets for mediations. Polyamines are ubiquitous polycationic molecules and regulate many cellular processes. We previously reported that deletion of potABCD, an operon that encodes a putrescine/spermidine transporter ( ∆ potABCD), resulted in an un-encapsulated attenuated phenotype. Here we characterize the transcriptome, metabolome, and stress responses of S. pneumoniae that is dependent on the polyamine transporter. Expression of genes involved in oxidative stress responses and the central metabolism was reduced while that of genes involved in the Leloir, tagatose, and pentose phosphate pathways was increased in ΔpotABCD. Downregulation of genes of the central metabolism will reduce production of precursors of capsule polysaccharides. Metabolomics results show reduced glutathione and pyruvate levels in the mutant. We also show that the potABCD operon protects pneumococci against hydrogen peroxide and nitrosative stress. These results show the importance of the potABCD operon and polyamine transport in pneumococcal physiology and tness that represents a novel target for therapeutic interventions. There was no signicant difference in the amount of NADPH produced by ΔpotABCD (3.8 ± 0.4 µg/mg) compared to the WT (4.0 ± 0.6 µg/mg). We observed no signicant difference in the amount of H 2 O 2 generated by ΔpotABCD (1 mM ± 0.05) and WT (1 mM ± 0.01). However, we observed a signicant difference between the GSH/GSSG ratio of the WT (1.3 ± 0.1) and ΔpotABCD (1.7 ± 0.1, p ≤ 0.05). A high GSH/GSSG ratio indicates increased GSH production, for which one stimulus is oxidative stress. These results show that intracellular levels of NADPH, H 2 O 2, and intracellular pH i are not dependent on polyamine transport.


Introduction
Despite years of intensive research, infections due to Streptococcus pneumoniae (pneumococcus) still claim countless lives across the entire globe [1]. Pneumococci account for up to 15% of pneumonia cases in the USA and 27% worldwide [2]. Following colonization of the nasopharynx, pneumococci can translocate to sterile sites and cause infections such as community-acquired pneumonia, meningitis, septicemia and otitis media [3]. Well-coordinated metabolic networks for e cient exploitation of the host micro-nutrients and immune response evasion strategies are crucial for pneumococcal pathogenesis. Serotype diversity, limited serotype coverage of the available vaccines, serotype replacement, and increase in multidrug-resistant strains confound intervention strategies to limit the spread of pneumococci [4][5][6]. A better understanding of pneumococcal physiology and survival mechanisms in the host can identify novel therapeutic targets.
Polyamines are polycationic molecules that interact with RNA, DNA, and phospholipids and modulate cellular processes such as cell division, transcription, and translation [7,8]. Putrescine, spermidine and cadaverine are the principal bacterial polyamines and their intracellular concentrations are tightly regulated by transport, biosynthesis, and degradation [8]. In pathogenic bacteria, polyamines are known to regulate virulence, bio lm formation, stress responses, in vivo tness, and host-pathogen interactions [9]. Therefore, failure to sustain intracellular polyamine levels could alter regulatory homeostasis which could interfere with in vivo survival and pathogenesis. Our previous work has shown that the polyamine transport operon potABCD is essential for virulence in murine models of pneumococcal infections [10]. We have shown that in a murine model of pneumonia, ΔpotABCD is more invasive than the wild type strain (WT) but is more susceptible to opsonophagocytosis. Uptake of ∆potABCD by neutrophils does not require antibody opsonization [11]. PotD, either alone or in combination with other proteins, has been shown to elicit protection against pneumococcal colonization, pneumonia, and sepsis in mice [12][13][14].
We have recently shown that deletion of the potABCD operon resulted in reduced intracellular concentrations of putrescine and spermidine and an un-encapsulated phenotype [10,15]. Since pneumococcal capsular polysaccharide (CPS) is a determinant of virulence, reduced CPS could explain reported in vivo attenuation of ΔpotABCD. Initial characterization of the ΔpotABCD proteome identi ed altered expression of over 100 proteins including virulence factors such as pneumolysin [10]. However, the limited proteome coverage did not allow for the identi cation of speci c mechanisms of metabolic regulation that could explain the observed phenotype. To determine polyamine dependent metabolic regulation that is at the intersection of pneumococcal virulence, we compared the ∆potABCD and WT transcriptome and metabolome using RNA-Seq and metabolomics. Given the role of polyamines in the defense against reactive radicals, and the diversity of stress conditions encountered by pneumococci, reduced intracellular polyamine levels are expected to adversely affect stress responses that are critical for in vivo tness. We therefore, examined the role of polyamine transport on the susceptibility of pneumococci to oxidative and nitrosative stress. We determined the impact of impaired polyamine transport on the intracellular pH (pH i ), GSH/GSSG ratio, and production of Nicotinamide adenine dinucleotide phosphate (NADPH), and hydrogen peroxide (H 2 O 2 ). Our results show that polyamine transport is required for the regulation of stress responses and central metabolism that adversely affect capsular polysaccharide synthesis in pneumococci and present an attractive target for developing novel therapeutics.

Results
Polyamine transport modulates pneumococcal gene expression.
Comparison of the transcriptome pro les of TIGR4 and ΔpotABCD identi ed regulatory mechanisms responsive to polyamine transport. We identi ed a total of 1333 genes whose expression varied signi cantly in ΔpotABCD. Expression of 651 and 682 genes was downregulated and upregulated in ∆potABCD, respectively (Tables 1-3 and Supp Table 2). Gene Ontology analysis of the differentially expressed genes (DEGs) identi ed signi cant enrichment of ve categories which included: phosphotransferase systems (PTS), galactose metabolism, ABC transporters, amino and nucleotide sugars as well as fructose and mannose metabolism. Major biological functions and pathways represented by the DEGs with a fold change of ≥ 2 and metabolites are discussed in the following sections and shown in Tables 1-3 and Table 5, respectively.
Polyamine transporter and pneumococcal stress responses.
Polyamines protect cells against reactive oxygen species (ROS) via regulation of stress response genes or by directly scavenging reactive free radicals [8,16]. Therefore, de ciency of polyamines could impact the redox status and render ΔpotABCD susceptible to oxidative stress. Downregulation of genes that encode Page 4/22 treR, a scavenger of H 2 O 2, molecular chaperones and their regulator, hrcA (Table 3), indicate impaired redox and repair systems which could compromise in vivo tness of ΔpotABCD [17,18]. Reduced expression of genes that encode several tRNAs (Supp. Table 2) could represent cellular adaptation during stress to meet the demand for redox homeostasis [19]. Increased expression of some genes that encode regulators implicated in pneumococcal stress responses, the arginine deiminase system (ADS) and glutamine transporters could be in response to oxidative stress in the mutant [20,21] (Table 1). Elevated glutamine in ux and ADS could be in response to meet the increased demand for energy (ATP) and to restore the buffering capacity (ammonia). Increased expression of ABC transporters for the import of iron, manganese, and phosphate (Tables 1) could result in cationic imbalance and negatively impact cellular functions and redox homeostasis [22]. These results show that de ciency of the polyamine transporter impairs pneumococcal stress systems, renders the mutant susceptible to oxidative stress and triggers polyamine independent redox systems to combat the stress. The signature of oxidative stress was further revealed by metabolic shift towards the PPP in the mutant which is usually in response to oxidative stress to produce NADH/NADPH which are cofactors in antioxidant systems [23]. Expression of genes that encode enzymes of the Leloir pathway involved in galactose catabolism and generation of UDP-galactose (UDP-Gal) was upregulated in the mutant ( Table 2). Expression of lacF-2 and the lacDCBA operon involved in the import and interconversion of galactose via the tagatose pathway to fructose6-phosphate and glyceraldehyde3-phosphate (G3P) was upregulated ( Table 2). Upregulated Leloir and tagatose pathways could be in response to high demand for PPP precursors in the mutant. Increased PPP and G3P may contribute to upregulation of tktC and tktN which encode a transketolase that catalyzes the interconversion of sugar-phosphates in the pathway ( Table 2). There was an increase in the expression of genes that encode enzymes that degrade fucose ( Table 2). Fucose degradation by triosephosphate isomerase yields G3P, an intermediate of glycolysis and PPP [24]. Increased expression of a putative PTS system involved in the import and interconversion of L-ascorbate to xylulose5-phosphate, an intermediate of PPP (Table 2) further suggests increased activity of the PPP. Genes involved in ascorbate utilization are co-transcribed upstream with a transcriptional regulator (bglG) whose expression was upregulated in ΔpotABCD ( Table 2). Shunting carbons to the PPP may be in response to increased demand of NADPH due to oxidative stress in the polyamine transporter-de cient mutant.
Glycolytic pathway and production of precursors for the pneumococcal capsule.
During stress, organisms modulate their gene expression to limit energy consuming processes such as CPS synthesis to preserve energy for redox systems. Gene expression pro le in ΔpotABCD indicate a limited ow of carbohydrates through the main glycolytic pathway which could be in response to stress. Expression of a sucrose operon regulator and genes involved in sucrose uptake was reduced. Expression of malP, which encodes an enzyme that breaks down glycogen to glucose1-phosphate and the malXCD operon, that encodes a maltose/maltodextrin transporter, was downregulated (Table 3). Reduced in ux of sucrose, maltose, and reduced breakdown of glucose in glycolysis will deplete glycolytic intermediates in ΔpotABCD. UDP-N-acetylglucosamine (UDP-GlcNAc) is a precursor of the three pneumococcal CPS sugar repeat units (UDP-N-acetylmannosamine (UDP-ManNAc), UDP-N-acetylgalactosamine (UDP-GalNAc) and UDP-N-acetylfucosamine (UDP-FucNAc). UDP-GlcNAc levels are expected to be low due to the elevated levels of nagA and nagB, involved in its breakdown to fructose6-phosphate (Table 3) and subsequently reduce the levels of precursors for CPS synthesis. Impaired CPS production was further apparent with the reduced expression of genes involved in the transport of N-acetylgalactosamine (Tables 3). Moreover, the pyr operon involved in the de novo synthesis of pyrimidine nucleotides and pyrR, the regulator of this operon, were repressed ( Table 3). Repression of pyrimidine biosynthetic genes will reduce UTP levels, a precursor for UDP required for the activation of UDP-sugars for CPS synthesis [24].
Expression of asd and dap involved in the synthesis of lysine, a constituent of the pneumococcal peptidoglycan layer (PG), was reduced ( Table 3). The overall effect of the above gene expression changes is reduced carbon ow through the central metabolism possibly due to stress and this will impact the production of precursors for CPS repeat unit sugars and the PG layer in the mutant. To validate RNA-Seq results, we measured the expression of selected genes using qRT-PCR (Table 4). Genes that encode for tktC, tktN, the ascorbate regulator (bglG), and the choline-binding protein pcpA, were upregulated, which is consistent with RNA-Seq (Table 4). Although not differentially expressed in the RNA-Seq results, polyamine synthesis genes speE and speA were upregulated in the qRT-PCR results, suggesting that polyamine synthesis could compensate for the loss of polyamine transport. Bacteria adapt to changing microenvironments by modifying metabolite levels to maintain redox homoeostasis. To gain further insight into the role of polyamine transport on the pneumococcal stress signature observed at the transcriptome level, intracellular levels of NADPH, endogenous H 2 O 2, pH i , and GSH/GSSG ratio between WT and ΔpotABCD were compared. Our results show that deletion of the potABCD operon does not impact pH i, as the pH i of TIGR4 (7.5) and that of the mutant (7.2) were both within physiological range (Fig. 1).
There was no signi cant difference in the amount of NADPH produced by ΔpotABCD (3.8 ± 0.4 µg/mg) compared to the WT (4.0 ± 0.6 µg/mg). We observed no signi cant difference in the amount of H 2 O 2 generated by ΔpotABCD (1 mM ± 0.05) and WT (1 mM ± 0.01). However, we observed a signi cant difference between the GSH/GSSG ratio of the WT (1.3 ± 0.1) and ΔpotABCD (1.7 ± 0.1, p ≤ 0.05). A high GSH/GSSG ratio indicates increased GSH production, for which one stimulus is oxidative stress. These results show that intracellular levels of NADPH, H 2 O 2, and intracellular pH i are not dependent on polyamine transport.
Metabolic pro le of polyamine transporter de cient S. pneumoniae.
Our metabolomics results identi ed signi cant differences in the levels of several metabolites in response to potABCD deletion ( Table 5). Levels of N-acetylglucosamine (GlcNAc) and pyruvate were reduced which could impact CPS production in ΔpotABCD. Increased activity of the PPP was evident by the higher levels of sedoheptulose1, 7-bisphosphate, a precursor for either erythrose4-phosphate or ribose5-phosphate.
Increased galactose metabolism was shown by high levels of UDP-glucose, a precursor for glucose1phosphate, an intermediate of glucose6-phosphate that can be channeled to PPP. The metabolome also showed reduced concentration of trehalose6-phosphate which could impact oxidative stress responses in ΔpotABCD. The concentration of glutathione was higher, which could be in response to oxidative stress in ΔpotABCD. Metabolomics results are concordant with our RNA-Seq results.
PotABCD is required for combating hydrogen peroxide and nitrosative stress. cadaverine supplementation (Fig. 2). Polyamine supplementation was done on ΔpotABCD cultured in chemically de ned growth medium (CDM) which lacks synthesized polyamines.
Compared to WT, ΔpotABCD was signi cantly more susceptible to S-nitrosoglutathione (GSNO) that generates RNS. Exposure to GSNO for 60 min at a concentration of 2.5 mM resulted in a signi cant reduction in the percentage survival of ΔpotABCD that ranged between 32% at 15-min and 73% at 60 min post exposure compared to the WT (0% at 15 min and 10% at 60 min post exposure). There was no signi cant difference in the survival of the complement pABG5-potABCD strain or supplementation with polyamines in the presence of GSNO (Figure 3). These results suggest that ΔpotABCD is more susceptible to H 2 O 2 and nitrosative stress than WT. Data with pABG5-potABCD complement or polyamine supplementation indicates that impact of impaired polyamine transport could be speci c to the type of stress.

Discussion
Findings in this study show that de ciency of polyamine transport impairs pneumococcal stress responses and disrupts CPS and PG synthesis (Fig. 4). This study also shows that the potABCD operon is involved in hydrogen peroxide and nitrosative stress responses in pneumococci. Polyamines contribute to cellular homeostasis by scavenging reactive radicals or balance intracellular pH through the consumption of a proton during their synthesis via the decarboxylation of amino acids [25,26]. Therefore, reduced intracellular polyamine levels reported in Ayoola et al. [15] and decreased expression of some genes involved in redox balance in the current study, could render ΔpotABCD susceptible to stress. In addition, polyamines regulate transcription of several genes including those involved in oxidative stress responses. Downregulation of expression of treR that impact the levels of trehalose a scavenger of ROS and the impaired DNA repair system could make the mutant sensitive to stress. The damage caused by H 2 O 2 can be ampli ed via the Fenton reaction with the generation of hydroxyl radicals, the primary cause of damage to biomolecules such as DNA [27]. Therefore, increased expression of the iron transporter could aggravate the effects of oxidative stress and impair pneumococcal survival in vivo, as it moves through different host niches.
In response to the stress caused by polyamine transport de ciency, ΔpotABCD triggered other known stress response systems to restore the redox status. Acquisition of P i and manganese is essential for normal growth, virulence, and oxidative stress responses in many pathogenic bacteria [17,22,28]. The PPP is often upregulated as a quick cellular response to meet NADPH demands to combat oxidative stress [23]. The ADS produces ammonia (NH 3 ) that protect cells against acid stress and a molecule of ATP for basal cellular functions during stress [29,30]. Reduced glutathione and glutathione metabolism play a signi cant role during oxidative stress in many organisms [31][32][33]. Therefore, increased expression of genes involved in cation in ux, PPP, the ADS, and a high GSH/GSSG ratio could be in response to stress in ΔpotABCD. [17,34]. Upregulation of the ADS could explain the observation that ∆potABCD can maintain pH i in the physiological range. The polyamine modulon which represent genes whose expression is regulated by polyamines is well described in Escherichia coli and includes genes involved in cell proliferation, bio lm formation, and detoxi cation of ROS [35]. Our results show that de ciency of the polyamine transporter affected the transcript level of several regulators, and thus indirectly altering genes controlled by these regulators. Interestingly, we observed that some polyamine transporter responsive genes are also part of the CcpA regulatory network, for example genes involved in stress responses, amino acid synthesis, and carbohydrate metabolism [36,37]. These results suggest that the intersection of polyamine metabolism and CcpA regulatory network is needed for in vivo tness. These results warrant future studies to identify mechanisms of co-regulation of polyamine metabolism and other transcriptional regulators that impact pneumococcal virulence.
Furthermore, upregulation of the Leloir and the tagatose pathways that produce PPP intermediates, and transketolase, a key enzyme in PPP, further suggest decreased carbon ow through glycolysis, supported by decreased pyruvate levels in ΔpotABCD. Decreased glycolysis could result in reduced acetyl-CoA, a precursor for UDP-GlcNAc, and a donor for the three N-acetylated sugars at the intersection of CPS and PG repeat unit biosynthesis [38,39]. However, upregulation of genes involved in the Leloir pathway in ΔpotABCD is contrary to what was observed in ΔspeA [15]. Deletion of arginine decarboxylase (speA) resulted in downregulation of genes of the Leloir pathway, implicating a role for polyamine synthesis in the regulation of this pathway. Arginine decarboxylase in ∆potABCD could result in the upregulation of the Leloir pathway which warrants further investigation. The PPP generates ribulose5-phosphate which can either be used for nucleotide synthesis or converted to sedoheptulose1, 7-bisphosphate, a precursor for erythrose4-phosphate and G3P. High levels of sedoheptulose1, 7-bisphosphate suggest that ∆potABCD promotes PPP activity possibly to increase production of NADPH levels required in redox systems. Downregulation of the pyr operon, responsible for the interconversion of uracil and uridine monophosphate (UMP), could impact UDP production necessary for UDP-sugar repeat unit synthesis and further impair CPS biosynthesis. Results from this study and our previous report [15] demonstrate that polyamine mediated CPS and PG regulation is dependent on both polyamine transport and synthesis.
Reduction of CPS precursors could explain the un-encapsulated phenotype and attenuation of ΔpotABCD in murine models of colonization, pneumonia, and sepsis [10,15].
Hydrogen peroxide produced by the pneumococcus is essential for its pathogenesis and protection against other common respiratory tract inhabitants. H 2 O 2 is cytotoxic to host cells, causes apoptosis in respiratory epithelial cells, and promotes colonization of the upper respiratory tract [40]. Therefore, pneumococci must adapt to survive the high levels of H 2 O 2 produced via the pyruvate oxidase system.
Despite a high GSH/GSSG ratio, maintenance of NADPH levels and upregulation of several genes involved in oxidative stress responses, ∆potABCD was more susceptible to hydrogen peroxide and nitrosative stress comparable to the WT. The increased susceptibility to H 2 O 2 cannot be attributed to increased intracellular H 2 O 2 levels in ∆potABCD since they were comparable to that produced by WT.
These results suggest that polyamines could be essential for the regulation of up/down stream functions of pneumococcal oxidative stress responses which warrants further investigation. These results are consistent with the known roles of polyamines in other bacterial pathogens [8]. Putrescine and spermidine protect cells from ROS by increasing expression of genes that code for free radical scavengers [41,42]. Polyamine synthesis [43] or transport genes are upregulated [9] in response to H 2 O 2 stress. Cadaverine protects Salmonella typhimurium and E. coli against nitrosative and acid stress [44,45]. Pneumococci increase transcription of the substrate binding protein (potD) of the polyamine transporter in response to oxidative and thermal stress and during murine bacteremia [9]. We also reported that speA, a gene that encodes an arginine decarboxylase, regulates pneumococcal nitrosative, H 2 O 2 and superoxide stress responses (manuscript under review). These results show that polyamine uptake from the environment may potentially help pneumococci to survive various host microenvironments. However, only supplementation with 2.5 mM cadaverine restored mutant viability in the presence of H 2 O 2 but not GSNO. These results indicate that providing the transport function in trans (pABG5-potABCD complement) or by exogenous polyamine supplementation has varying effects on the viability of ∆potABCD that is dependent on the type of stress and the type and concentration of polyamines.
In summary, polyamine transport is essential for pneumococcal stress responses and a dysregulation of these responses impacts the synthesis of CPS and PG. Through horizontal gene transfer, a non-virulent pneumococci can become encapsulated and virulent during coinfections in the host [46] which confounds management strategies. Therefore, gaining insights into virulence mechanisms that are capsule independent but modulate capsule production, like polyamine metabolism, present a novel avenue for exploring new vaccine and therapeutic interventions against this deadly pathogen. Polyamine transport systems are conserved across pneumococcal serotypes and are a promising therapeutic avenue due to their immunogenic potential [12,14]. Future studies focusing on uncovering the interconnected network of pneumococcal polyamine redox homeostasis, CPS and PG synthesis will contribute towards deconvolution of the complex regulatory networks that impact stress responses and pneumococcal adaptation in vivo.

Materials And Methods
Bacterial strains and growth conditions. S. pneumoniae serotype 4 strain TIGR4 [47], ΔpotABCD [11], and the complement strain (pABG5-potABCD) [15] were used in this study. All strains were grown in either chemically de ned medium (CDM) [48] or Todd-Hewitt broth supplemented with 0.5% yeast extract (THY) or on 5% sheep blood agar plates (BAP) in 5% CO 2 at 37°C. All assays were performed in triplicate in three independent experiments. RNA Sequencing.
Total RNA was isolated and puri ed from mid-log phase cultures (OD 600 0.4) of TIGR4 and ΔpotABCD (n = 4) grown in THY (a complete medium that mimics nutrients in the host milieu) using the RNeasy® Mini Kit (Qiagen, Valencia, CA, USA). RNA quality was checked with an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). RNA-Seq analysis was performed as described earlier [15]. Removal of failed reads, mapping of the short sequence reads to the TIGR4 reference genome, and identi cation of differentially expressed genes were performed with CLC Genomic Workbench 20.0.3 (Qiagen, Valencia, CA, USA).
Paired end reads of both WT and ΔpotABCD were mapped to the TIGR4 genome using CLC proprietary read mapper and read counts were estimated by EM estimation algorithm [49] and DEGs were identi ed based on the fold change generated by the edgeR algorithm. Changes in gene expression with a fold change of ± 1.3 at a false discovery rate (FDR) of ≤ 0.05 were considered signi cant. Functions and pathways represented by DEGs were identi ed utilizing multiple bioinformatics resources such as MetaCyc [50], Gene Ontology [51], KEGG [52], UniProt [53], and STRING [54]. RNA-Seq raw data and metadata are available at NCBI GEO with the accession number XXXXXXX.
Quantitative real time PCR.
To validate RNA-Seq results, we measured expression of selected genes by quantitative reverse transcription PCR (qRT-PCR). The primers used for qRT-PCR are listed in supplementary material (Supp. Table 1). All primers were validated by performing a melt curve analysis with SYBR Green (Thermo Fisher Scienti c Waltham, MA, USA). In brief, total RNA was puri ed from mid-log phase cultures (OD 600 0.4) of TIGR4 and ΔpotABCD grown in THY (n = 3). Puri ed RNA (7.5 ng/reaction) was reverse-transcribed into cDNA and PCR was performed using the SuperScript III Platinum SYBR Green One-Step qRT-PCR Kit (Thermo Fisher Scienti c, Waltham, MA, USA) as previously described [55]. Relative quanti cation of gene expression was determined by the Stratagene Mx3005P qPCR System (Agilent, Santa Clara, CA, USA). Expression of selected genes was normalized to the expression of gapDH and fold changes determined by the comparative C T method.
Measurement of intracellular pH.
The intracellular pH (pH i ) was determined based on the method described by Clementi et al., [56] with detected using a plate reader for 5 min. 10 µM carbonyl cyanide 3-chlorophenylhydrazone (CCCP) was added to one well (to serve as control) and the reading taken for another 5 min. CCCP is a protonophore that uncouples proton motive force and causes a rapid decrease in pH i (Millipore-Sigma, St. Louis, MO, USA). Nigericin was added to both CCCP-treated control and untreated sample (to a nal concentration of 1 mM) to equilibrate the pH i to the pH of the buffer and uorescence was read for an additional 5 min.
Fluorescence was calculated and the pH i was interpolated from the calibration curve.
Measurement of intracellular NADPH.  [57]. Samples were run with a spray voltage of 3 kV. The nitrogen sheath gas was set to a ow rate of 10 psi with a capillary temperature of 320°C. Automatic gain control target was set to 3e6.
The samples were analyzed with a resolution of 140,000 and a scan window of 85-800 m/z from 0 to 9 min and 110-1000 m/z from 9 to 25 min. Files generated by Xcalibur (RAW) were converted to the open source mzML format [49] via the open source msconvert software as part of the ProteoWizard package [49]. Maven (mzroll) software, Princeton University [58,59] was used to automatically correct the total ion chromatograms based on the retention times for each sample [58,59]. Metabolites were manually identi ed and integrated using known masses (± 5 ppm mass tolerance) and retention times (1 ≤ 1.5 min). Unknown peaks were automatically selected via Maven's automated peak detection algorithms. A database of 275 metabolites veri ed using exact m/z and known retention times, expanded from the original database [57] was used. The statistical analysis on metabolite peak intensity post CFU normalization was done by MetaboAnalyst 4.0 [60]. Quantile normalization which is highly e cient in normalizing metabolite variations from mass spectrometry [61] was used to normalize the data.
Signi cant differences in metabolite peak intensity between ΔpotABCD and TIGR4 were identi ed by a Ttest at an adjusted FDR of ≤ 0.05. .
Mid-log phase cultures of TIGR4, ΔpotABCD and the complement pABG5-potABCD strain grown in CDM were centrifuged at 10,000 x g for 2 min and cells suspended in PBS. The cells (10 7 CFU/mL) in 100 µl were supplemented with a nal concentration of 2.5 mM GSNO (Sigma-Aldrich, Israel), a nitric oxide producer, and incubated at 37°C in 5% CO 2 for 60 min. In addition, ΔpotABCD challenged with 2.5 mM GSNO was supplemented with cadaverine, putrescine or spermidine (½MIC, ¼MIC, MIC). Control reactions had untreated bacteria, and CFUs were determined by serial dilution in PBS and plating on BAP every after 15 min. Results from three independent experiments were expressed as percent survival of treated bacteria relative to the untreated bacteria.
Statistical analysis. Con icts of interest.
The authors declare no con ict of interest in this work.
Funding information.
Grant: #P20GM103646 (Center for Biomedical Research Excellence in Pathogen Host Interactions) from the National Institute for General Medical Sciences.