Introduction

Polyamines are ubiquitous polycationic molecules essential for cell growth from bacteria to mammalian cells. A variety of biogenic polyamines has been discovered from different organisms: putrescine (Put) and spermidine (Spd) predominate in prokaryotic cells, whereas Spd and Spm are mainly found in eukaryotic cells [7]. The distinct distribution of Spm suggests that this tetra-amine may be associated with some new activities or specific regulations during evolution. Of interest, increased Spm levels have been demonstrated at inflammatory sites of infection and in cancer cells, suggesting the significance of Spm in cell proliferation and immune functions in human body [2, 48]. In contrast, many studies showed that exogenous Spm may become toxic to bacteria with accumulation, the inhibition of which is especially strong in S. aureus [15, 22]. Moreover, we also found that at low concentrations, Spm by itself has negligible effects while is able to sensitize bacteria to β-lactam antibiotics; such synergy is most potent in S. aureus [23]. However, the molecular target of Spm toxicity in S. aureus remains largely unknown.

In some bacteria, Spm toxicity can be alleviated by catabolic pathways. There are two common mechanisms that have been reported to initiate Spm degradation: glutamylation by γ-glutamylpolyamine synthetases (GPS) in P. aeruginosa [46] and acetylation by Spd/Spm acetyltransferase (SSAT) in E. coli and B. subtilis. In most S. aureus strains, there are no such Spm catabolic enzymes; the only exception is the USA300 lineage, which expresses speG encoded SSAT to convert toxic Spm to non-toxic N-1 acetyl-Spm through acetylation, thus affording complete resistance to Spm. But for non-USA300 clones, S. aureus can be rapidly killed by physiological Spm concentrations [22].

To elucidate the molecular mechanism of Spm toxicity in S. aureus, we applied in this study the transcriptomic approach to investigate the cellular response to Spm stress. The results suggest that Spm may interfere with iron homeostasis and sigB regulon, and consequently led to the growth inhibition. Furthermore, in depth characterization of tetM, the gene that displayed the most significant fold change in Spm transcriptome analysis, demonstrated specific induction of tetM by Spm and Tetracycline, but not by other polyamines or similar ribosome-targeted antibiotics. Despite not contributing to Spm toxicity, these tetM derived data may have a potential impact on the current model of transcriptional attenuation and the dissemination of antimicrobial resistance determinant carried by conjugative transposon.

Materials and Methods

Bacterial Strains, Plasmids, and Growth Conditions

Staphylococcus aureus Mu50 (ATCC 700699), RN4220 (kindly provided by R. P. Novick), and Escherichia coli DH5α (Bethesda Research Laboratories) were employed in this study. Plasmids pCN38 and pCN56 (obtained from NARSA) are shuttle vectors of E. coli and S. aureus for gene cloning and promoter-gfp fusion, respectively [6]. Both E. coli and S. aureus strains were routinely cultured and maintained in the Luria–Bertani (LB) medium. When required, the LB medium was buffered with 20 mM Tris–HCl at the indicated pH. Antibiotics were added to the medium as necessary at the following concentrations: ampicillin, 100 μg ml−1 for E. coli; kanamycin, 25 μg ml−1 for E. coli, 50 μg ml−1 for S. aureus; erythromycin, 10 μg ml−1 for S. aureus; chloramphenicol, 10 μg ml−1 for S. aureus.

Transcriptional Profiling Conditions and Microarrays

Staphylococcus aureus Mu50 was grown in the buffered LB broth (pH8.0) at 37 °C to log-phase (OD600 of 0.5), and exogenous Spm was added to a final concentration of 10 mM. RNA samples were extracted with phenol (Fisher), digested with RNase-free DNase I (Pierce) to remove genomic DNA, and purified with RNeasy mini-columns (Qiagen). Synthesis, fragmentation, and terminal labeling of cDNA were carried out according to the protocols of the manufacturer (Affymetrix). Labeled cDNA was hybridized to the GeneChip S. aureus genome array. After scanning, the images were processed with GCOS 1.4 software (Affymetrix). Processed microarray data files have been deposited in the Gene Expression Omnibus (GEO) database under accession number GSE43230.

qRT-PCR

Total RNA was prepared as described above for microarrays after the cell cultures were treated with compounds as indicated in the text. Purified RNA was subjected to VeriQuest SYBR Green One-Step qRT-PCR (Affymetrix) using ABI 7500 Fast System (Applied Biosystems). The relative quantification of mRNA was determined by change in the expression level of the target transcripts relative to gyrB transcript (housekeeping gene) in accordance with the manufacturer’s protocol. Data obtained are expressed as the mean fold change in transcript during stress condition relative to the compound-free control sample.

Construction of tetM-Promoter::gfp Fusion

A DNA fragment covering 705-bp upstream and 1-bp downstream of the tetM start codon was amplified by PCR from S. aureus Mu50 by primers 5′-CGC GGA TCC CGG ACT ACC AAC CCA AAC AGC TAG-3′and 5′-CCG GAA TTC TCA TGT GAT TTT CCT CCA TTC AA-3′. The PCR product was cloned in the BamHI-EcoRI sites upstream of a promoterless gfp gene of the shuttle vector pCN56-CM (replace ermC from pCN56 with cat194 from pCN38 to confer chloramphenicol resistance as selection marker). The correct construct pCNPtetM::gfp was introduced into RN4220 by electroporation and subsequently into Mu50.

Measurements of Promoter Activities

Bacteria carrying pCNPtetM::gfp were grown in LB (pH 8.0) with shaking at 37 °C to the exponential phase before exposure to polyamines (Spm, Spd or Put; 1 mM) or antibiotics translation inhibitors (tetracycline, chloramphenicol, erythromycin; 2 ng µl−1). After 2 h, the cells were collected by centrifugation and suspended into the same volume of Tris–Cl buffer (pH8.0). The green fluorescence was then measured with a spectrofluorometer (EnSpire® Multimode Plate Readers, PerkinElmer) by excitation at a wavelength of 485 nm and detection of emission at 515 nm, and the growth curve was monitored by measuring the optical density at wavelength 600 nm (OD600). The promoter activity was expressed as the ratio of relative fluorescence (RFU) over OD600.

Cloning of the tetM Gene

A DNA fragment covering the tetM structural gene as well as 705-bp upstream and 668-bp downstream of tetM was amplified from genomic DNA of S. aureus Mu50 by the primer pair 5′-CGC GGA TCC CGG ACT ACC AAC CCA AAC AGC TAG-3′) and 5′-GGA AGA TCT TTA GAG CCG TTG GTT TAG CGA TTA-3′. The PCR product was digested by BamHI and BglII, ligated to the BamHI site of pUC18. One positive clone was confirmed by enzyme digestion and named as pUCtetM. The insert in pUCtetM was purified after BamHI and EcoRI digestion and cloned into the same restriction sites of E. coli–S. aureus shuttle vector pCN38. The resulting plasmid pCNtetM was confirmed by sequencing and introduced into S. aureus strain RN4220 by electroporation.

MIC Measurements

The broth micro-dilution method in 96-well titer plates was employed to determine MICs of antibiotics or spermine according to the CLSI guideline. Serial two-fold dilutions of the tested compound were prepared in the LB broth. When required, the LB medium was buffered with 20 mM Tris–HCl of the indicated pH. Fresh overnight cultures of each bacterial strain were diluted and inoculated with approximately 5 × 105 CFU per well. The plates were properly wrapped and incubated without shaking at 37 °C for 16 to 18 h. The MIC was defined as the lowest concentration of the agent that inhibited the growth of the bacteria as detected by unaided eyes. The MICs were determined in triplicate.

Results

Transcriptome Analysis of S. aureus in Response to High Dose Spm

For most S. aureus strains which do not possess SpeG, Spm can inhibit the bacterial growth in a dose-dependent manner [23]. To obtain a comprehensive view of the transcriptional changes by Spm, we conducted a series of independent microarray experiments in the absence (control group) and the presence (experimental group) of 10 mM Spm upon 15, 30, and 60 min exposures. To further identify genes of changes with statistical significance, we applied the following criteria to each of the control-experimental microarray data sets: (1) P value ≤0.05, (2) absolute fold change in transcript level ≥2.5, and (3) for up-regulation, genes in the experimental group of over 200 in signal intensity; for down-regulation, genes in the control group of over 200 in signal intensity. Of the over 3,000 ORFs in the genome of S. aureus, a list of 581 genes showing significant changes upon Spm treatment was compiled by this approach. To examine how these genes were distributed with regard to their physiological and biochemical functions, they were further classified according to COG functional categories (Table S1). By comparing with the profiles before Spm treatment, genes showing significant changes upon exposure to Spm predominantly participate in iron transport and metabolism, SigB regulon, and ATP production. Transcriptional profiles of representative genes are shown in Fig. 1, and below described some major findings.

Fig. 1
figure 1

Differential expression of iron-regulation and sigB regulon following Spm exposure. Microarray results are presented as the fold change of experimental group (samples collected at different time points after compared to control (t = 0, before Spm). Candidate genes were categorized according to functional class. Common name of the encoded protein is provided, and gene names are indicated in parentheses

Full size image

Up-regulation

Iron Uptake and Metal-Dependent Transcriptional Regulators

As shown in Fig. 1, a number of genes encoding metal transporters (especially for iron) showed increased expression, indicative of possible dysregulation of metal homeostasis by Spm challenge. Iron acquisition by S. aureus is mainly mediated by siderophore or heme [17]. It has been reported that FhuABGD (ferric hydroxymate uptake) [38], HtsABC (heme transport system) [14], and SirABC (staphylococcal iron regulated transporter) [9] participate in siderophore import, while heme uptake go through the Isd system and the HtsBC permease [29, 43]. In specific, for siderophore import, staphyloferrin A is mediated by HtsABC and FhuA, staphyloferrin B by SirABC and FhuA, while xeno-siderophore by FhuABGD. Here, the fhuABG, htsABC, and isdF all showed 3–6-fold up-regulation by Spm. In addition, the norA gene encoding an iron-responsive efflux pump in S. aureus [12], also displayed over 2.5- fold increase. Moreover, four major metal-mediated repressors, Fur, PerR, Zur (three Fur family members) [19, 20, 24, 25], and MntR (DtxR-like protein) [21], all displayed induced expression. Although responding to different metal ions, these repressors share a similar mode of action: binding of metal ions to the regulators is essential for their DNA-binding activities in gene repression. Upon iron limitation, dissociation of F2+-Fur complex derepresses the Fur regulon for siderophore synthesis, transport, and regulation (fur itself) [11]. Meanwhile, PerR, Zur, or MntR also can modulate the Fur regulon directly or indirectly, further expanding the regulatory capability on iron uptake and storage. In addition to iron-dependence, the PerR regulon (including fur and perR itself) can be activated by peroxide in order to modulate resistance of oxidative stress [19]. Collectively, these observations revealed a strong link of Spm toxicity to iron-restricted stress.

Other Spm-Induced Gene Expression

Genes related to metabolism and transport of certain amino acids, particularly those for arginine biosynthesis, showed significant induction by Spm. Up-regulation of these genes infers an increasing need of specific amino acids during Spm stress. Another notable finding was the induction of potABCD encoding components of a putative ABC transporter for polyamine uptake. Investigation of this operon was recently reported [47]. Besides, some Spm-inducible genes may likely contribute to osmotic tolerance, such as those encoding glycine betaine transporter (opuD), choline dehydrogenase (betA), glycine betaine/carnitine/choline ABC transporter (opuCA/CB/CD), and glycine aldehyde dehydrogenase (gbsA). Of interest, the tetM, gene (SAV0398) for tetracycline resistance, and the kdpCBADE operon (two copies in Mu50 genome, SAV0074-0070 and SAV2075-2079) for potassium uptake were also significantly induced by Spm (Table S1).

Down-Regulation

The SigB Regulon

One very striking event upon Spm stress was the down-regulation of Sigma B (SigB, σB) regulon (Fig. 1). A total of 251 genes were reported in the σB regulon of S. aureus [5], and 96 genes of this regulon were found responsive to high dose Spm. Among these genes 79 ORFs positively controlled by σB exhibited reduced expression, while 17 genes which are repressed by σB display an upregulated pattern in response to Spm stress (Table S1, Fig. 1).

ATP Synthesis and Carbohydrate Metabolism/Transport

The addition of Spm to S. aurues decreased to different extents the expression of genes in major central metabolic pathways of energy production, including fermentation, glycolysis, gluconeogenesis, pyruvate conversion, pentose phosphate cycle, and TCA cycle. We also noticed (Table S1) a down-regulation of the membrane-bound ATP synthases (F0F1-ATPases). These data indicate an overall decreased activity or demand of energy production during Spm stress.

Many genes in carbohydrate metabolism and transport were down-regulated. In specific, expression of ptsIH for glucose uptake [34] was decreased, consistent with the reduced expression of genes in glycolysis and pentose phosphate cycle (the alternative glucose degradation pathway), as described above. Moreover, expression of the lacABCDFEG operon for lactose utilization was also decreased. The assimilation of lactose or galactose by S. aureus accumulates galactose 6-phosphate. This phosphorylated carbohydrate is the actual intracellular signal molecule for induction of the lactose (lac) genes, which can be further metabolized to triose phosphates used in the glycolysis [30]. Taken together, reduced expression of genes for sugar uptake and utilization was consistent with the pattern of down-regulation in ATP synthesis.

Spermine-Specific Induction of tetM

Among genes that were induced by 10 mM Spm, tetM exhibited the most significant fold change and high signals (over 8-fold). In S. aureus Mu50, tetM was identified on Tn5801 (Fig. 2a), which contains highly conserved ORFs in four functional modules: conjugation, regulation, recombination, and accessory genes such as tetM [10]. As shown in Fig. 2b, tetM was highly induced by exogenous Spm at either 1 mM or 10 mM. In comparison, Spm exerted no significant effect on the expression of other genes in Tn5801.

Fig. 2
figure 2

Schematic representation of Tn5801 from S. aureus strain Mu50 and the expression profiling of tetM with the flanking genes in response to Spm. a The ORFs within Tn5801 of S. aureus Mu50 are numbered SAV0392 to SAV0415 according to the original designations (GenBank accession no. BA000017). tetM is represented as black arrow. b Gene expression signals were from transcriptome analyses with Mu50 treated with high dose Spm (10 mM) for 15 min or low dose Spm (1 mM) for 120 min. GeneChip raw data are mean values from two independent sets of cultures. Expression change (n-fold) was the signal intensity ratio between Spm treated and untreated conditions

Full size image

To substantiate the finding of Spm on tetM expression, and to test whether tetM was responsive to other polyamines, the transcript levels of tetM were measured by qRT-PCR. The cultures of Mu50 in the exponential phase were exposed to various polyamines (Spm, Spd, and Put; 1 mM), and the cells exposed to low dose Tc (2 ng µl−1=4.16 µM) served as positive control of tetM induction. As shown in Fig. 3a, tetM expression was specifically induced by Spm and Tc to a comparable level but not by Spd or Put.

Fig. 3
figure 3

tetM responses to polyamines and protein inhibitors. a qRT-PCR analysis of tetM in response to different polyamines. The fold change represents the difference between Ct of drug-free condition (control) and Ct of the compound-supplement condition. b Promoter activities of PtetM. S. aureus strain Mu50 harboring pCNPtetM::gfp was grown for 2 h in the presence or absence of polyamine (1 mM of Put, Spd, Spm) or antibiotics (2 ng µl−1 of Tc, Em or Cm). Tc tetracycline, Em erythromycin, Cm chloramphenicol. Promoter activities were expressed as relative fluorescence versus OD600 of the culture. Standard deviations are derived from at least two independent cultures

Full size image

As described in Materials and Methods, a recombinant plasmid pCNPtetM::gfp was constructed to measure the promoter activity of tetM with gfp as the reporter gene. This plasmid was introduced into Mu50, and the effects of different polyamines and antibiotics were evaluated as shown in Fig. 3b. For polyamines, only Spm exerted an induction effect on tetM expression, consistent to the results of DNA microarrays and qRT-PCR. For ribosome-targeted antibiotics, only Tc was able to induce the tetM promoter; chloramphenicol and erythromycin showed no effect.

It is of interest to understand why Spm can induce tetM transcription and whether the expression of TetM is able to confer resistance to Spm. First, to test whether the TetM is indeed responsible for Tc resistance, the tetM gene including its regulatory region from Mu50 was cloned and transformed into E. coli DH5α as well as S. aureus RN4220 where tetM was absent. In the subsequent drug-susceptibility test, MIC of Tc in RN4220 harboring tetM increased more than 60- fold in comparison to its host, reaching to a level comparable to that in Mu50; similarly, tetM was able to raise the tolerance of E. coli to Tc, but to a lesser extent (Table 1).

Table 1 Effects of the tetM gene on MICs of tetracycline (Tc) and spermine (Spm)
Full size table

Since Spm toxicity also applies to strains of S. aureus without the tetM gene, it is unlikely that TetM is the molecular target of Spm toxicity. However, it was possible that TetM expression might promote Spm resistance. To test this hypothesis, Spm MIC was determined in tetM expressing strains of S. aureus (chromosome-borne as in Mu50 and plasmid-borne as in RN4220/tetM carrying a recombinant tetM-bearing plasmid). As revealed in Table 1, both RN4220/tetM and Mu50 showed no difference on the Spm sensitivity in spite of tetM expression, indicating that the up-regulation of tetM by Spm is not a defensive mechanism to Spm toxicity.

Although tetM may not play any role in Spm toxicity (Table 1), such specific induction by Spm is exceptionally interesting. Given that Tc acts by binding to the 30S subunit of ribosomes, we investigated the tetM response to other drugs against translation by measurements of the tetM promoter activities. As shown in Fig. 2b, unlike Tc, erythromycin and chloramphenicol were unable to induce tetM, suggesting that tetM exclusively responds to Tc and Spm, but not other ribosome-targeted drugs or other polyamines.

Increase of Spm Tolerance by KCl and NaCl

Another interesting observation revealed from our transcriptome data is the induction of kdpABCDE locus upon Spm exposure. The kdp genes, which are widely distributed among bacteria, originally were considered for potassium uptake and regulation [13], while recent lines of evidence also showed that the Kdp system is involved in virulence of some bacteria including S. aureus [33, 45]. In E. coli, the KdpABC is a high-affinity K+ uptake ATPase, and its expression is regulated by the KdpDE two-component system in response to K+ limitation or turgor pressure [18].

The connection between Spm and K+ homeostasis was originally described in eukaryotic systems, where Spm enters and occludes the transmembrane pore domain of inwardly rectifying potassium channel (Kir) [37, 44]. In bacteria, despite of less complexity, many potassium transport systems (like Kdp-ATPase, Trk) have been shown to behave similarly to the eukaryotic countparts. Given this, we proposed that in S. aureus Spm may block the K+ uptake in a similar manner as that in the eukaryotic potassium channel, and therefore supplement with a high concentration of K+ may increase Spm resistance by counteracting Spm obstruction on the uptake channel. To test this, Spm MIC in Mu50 was measured in LB containing 300 mM KCl. Interestingly the addition of KCl indeed increased Spm MIC from 1 mM to 32 mM.

It should be noted that aside from K+ limitation, high osmolarity also has been proposed to serves as the signal for induction of the kdp operon [3]. To test whether the induction of kdp by Spm was due to osmolarity upshift, we also measured Spm MIC in the presence of other osmotic shock agents - 300 mM NaCl and 600 mM sucrose. Surprisingly, like KCl, NaCl also enhanced the tolerance of Mu50 to Spm with MIC of 32 mM; nonetheless, osmotic shock with 600 mM sucrose showed no effect.

Discussion

Spm Toxicity and Iron Limitation

In this study, we analyzed changes in the transcriptome of S. aureus Mu50 upon Spm stress. The stimulation of iron acquisition, including iron uptake genes (fhuABG 4-fold, htsABC 12- fold) and metal-specific regulators (Fur, PerR, Zur, MntH 3-fold) was particularly of note, indicative of a specific connection between Spm and iron-depletion. Similarly, early reports have indicated the effects of low iron stress and siderophore production on spermine biosynthesis [4]. In fact, Spm has been known to have iron chelating properties by forming ternary complexes between Fe2+ ions and phosphate groups [26–28], which can contribute to polyamine anti-oxidant activity through inhibiting Fe2+ auto-oxidation and in consequence hydroxyl radical generation [31, 40, 41]. Therefore, the interesting changes seen with the up-regulation of metallo-regulators Fur, Zur, and MntR could directly result from metal chelation. In addition, the antioxidant properties of Spm would explain the lack of regulation on PerR-dependent antioxidant components (like katA) since more iron is being transported, alongside the metal chelation reducing Fenton chemistry. Taken together, we hypothesize that Spm may induce stress through iron starvation in S. aureus.

Iron is essential for bacterial growth due to its involvement in multiple metabolic processes, including respiration [heme-containing cytochromes, (Fe–S)-containing ferredoxins] and key enzymes in central metabolism [(Fe–S)-containing enzymes in TCA, like SdhA, CitB, CitG] [16, 17]. Thus, the potential sequestration of iron by Spm may lead to deficiency on electron transport and energy production. Of note, a recent report shows that a number of spontaneous spermine-resistant mutants of S. aureus are small colony variants with deficiency on menadione synthesis. Mutations affecting the generation of menadione and associated electron transport may compromise Spm toxicity [22]. From our transcriptome data, genes for iron acquisition and for the biosynthesis of cytochromes showed significant induction under Spm stress, inferring a correlation between heme/iron and Spm. Hence, both heme/iron and menaquinine, two indispensable components for electron transport chain, seem to have connections with Spm. It also suggests the possible effect of Spm on cellular respiration and energy production through these connections of unknown nature.

Collectively, we propose a model for Spm toxicity as described below. Spm may generate iron-depleted conditions or change the intracellular iron status by iron chelating properties. Such iron starvation signals are sensed by metal-specific regulators (Fur, PerR, MntR, Zur), which initiate derepression of iron uptake systems. Even though low dose Spm is sufficient to induce iron uptake, the delivered impact may be sequestered by redirecting central metabolism, thus explaining why low dose Spm has no inhibition on S. aureus. Upon high dose Spm, the potential detrimental effect resulting from iron limitation might be overwhelming, which leads to the instant inhibition on cell growth. Furthermore, the fact that high dose Spm is able to turn off the SigB regulon may cause the cells even more difficult to survive under Spm stress.

Down-Regulation of SigB Regulon by Spm

The alternative sigma factor σB of S. aureus is in charge of the general stress response, expression of virulence determinants, and modulation of antibiotic resistance. It controls a large regulon, either directly by recognizing conserved σB promoter sequences (e.g., asp23, clpL) or indirectly via σB-dependent elements (e.g., RNAIII, agrA) [32, 36]. In specific, σB has been shown to influence the expression of several global regulators, including up-regulation of SarA, SarS (SarH1) as well as down-regulation of RNAIII and agr regulatory system, all of which exhibited an altered pattern of expression by Spm in our study. Of note, the function of σB in virulence factor production is exactly the opposite from that of RNAIII/agr quorum sensing circuits. Therefore, the increased expression of exoprotein and toxin genes under Spm stress might be a consequence of lower σB and/or higher RNA III transcription.

From the results of transcriptome analysis, it was very clear that the σB regulon was turned off under Spm stress. Conserved in most bacteria, the major physiological function of the σB regulon was to protect the cells when under stress. However, it was surprising that Spm stress in fact shuts down the σB regulon. We considered this effect could be very detrimental to the cells. This novel finding warrants future investigation on the molecular mechanism of Spm effects on σB.

Potential Mechanisms of tetM Induction by Tetracycline and Spm

Tetracycline (Tc) functions by binding to a specific region in the 16 s rRNA of the 30S ribosomal subunit and preventing access of aminoacyl-tARNA (aa-tRNA) to the ribosomal A site [8]. The ribosomal protection protein TetM can release Tc binding and rescue the ribosome to return to the elongation cycle [42]. The mechanism for tetM regulation has been proposed through transcriptional attenuation [39]. A "leader peptide" immediately upstream of tetM, contains rare amino acids; therefore, the ribosome pauses at these positions due to the shortage of rare aminoacyl-tRNAs (aa-tRNAs), leading to retarded translation. The non-synchronism of RNA polymerase and ribosome makes the transcription terminated at the stem-loop termination sites which are upstream of tetM start codon, resulting in no expression of tetM when Tc is absent. When cells are exposed to Tc which retards the translation, there is a backup of charged tRNAs, including the rare ones needed for the tetM leader peptide. The increased availability of these aa-tRNAs allows for coupled translation and transcription, thus enables disruption of the step-loop terminator. Transcription then proceeds into tetM, and more TetM were expressed to protect ribosome from Tc attack.

Spm also can bind to 16S rRNA in the 30S subunit [1]. This determined Spm-binding site partially overlaps with the Tc binding site. The effects of Spm on ribosomal functions were further characterized in many aspects: (1) Spm can enhance the binding of aa-tRNA to the A site to ribosome; (2) Spm cross-linking to rRNA reduces the translational accuracy and therefore increases the probability of using “wrong aa-tRNA” for translation in the lack of right cognate aa-tRNA; (3) Spm had no impact on the peptide bond formation [1]. Based on these, along with the transcriptional attenuation model for Tc regulation on tetM, as well as the specific induction of tetM by Spm and Tc, it was reasonable to propose that a) Spm may behave like Tc by inhibiting translation to induce tetM; or b) Spm may facilitate aa-tRNA incorporation to 30S binding site thus bypass the transcriptional attenuation of tetM and enhance its expression. As a small molecule, Tc or/and Spm may use riboswitch to achieve the regulatory action. Considering the fact that Spm does not restrain the aa-tRNA binding as mentioned above, the former mechanism is less favorable. Alternatively, given the complicated structure of tetM mRNA and Tc/Spm small molecules, it is interesting to speculate riboswitch as potential mechanism of tetM regulation by Tc and/or Spm instead of the current model of coupled transcription-translation as described above.

Protection of Spermine Stress by KCl/NaCl

In this study, expression of the kdpCBAFDE operon was found inducible by exogenous Spm. Since Spm was reported to block the Kdp-like channel of potassium uptake [37, 44], we hypothesized that growth inhibition of S. aureus by Spm might be related to potassium uptake, and that induction of the kdp operon might be the result of low potassium inside the cells. Although addition of potassium (300 mM KCl) indeed increased Spm MIC up to 32-fold, a similar fold of protection against Spm can also be observed in the presence of 300 mM NaCl but not by 600 mM of sucrose. Since high osmolarity by NaCl or sucrose was reported to induce high levels of kdp expression [35], it was apparent that induction of the kdp operon alone cannot protect the cells from Spm stress, and the mechanism of NaCl/KCl protection against Spm stress remained to be answered.