Plague, caused by Y ersinia pestis, is one of the most dangerous diseases that impressed a horror onto human consciousness that persists to this day. Cases of plague can be normally controlled by timely antibiotic administration. Streptomycin is the first-line antibiotic for plague treatment. In this study, a DNA microarray was used to investigate the changes in the gene expression profile of Y. pestis upon exposure to streptomycin. A total of 345 genes were identified to be differentially regulated, 144 of which were up-regulated, and 201 down-regulated. Streptomycin-induced transcriptional changes occurred in genes responsible for heat shock response, drug/analogue sensitivity, biosynthesis of the branched-chain amino acids, chemotaxis and mobility and broad regulatory functions. A wide set of genes involved in energy metabolism, biosynthesis of small macromolecules, synthesis and modification of macromoclecules and degradation of small and macro molecules were among those down-regulated. The results reveal general changes in gene expression that are consistent with known mechanisms of action of streptomycin and many new genes that are likely to play important roles in the response to streptomycin treatment, providing useful candidates for investigating the specific mechanisms of streptomycin action.
Plague is a zoonotic disease caused by the gram-negative bacterium Yersinia pestis. It has caused social devastation on a scale unmatched by any other infectious agent. Plague is ever present in endemic areas, circulating in various mammalian species. There are hundreds of cases of human plague reported annually, and plague has been recognized as a re-emerging disease by the World Health Organization . Y. pestis has recently become of public interest because of its potential as an agent of bioterrorism. Since its first development in 1944, streptomycin has been the antibiotic of choice for the treatment of most forms of plague . When administered in the early phase of the disease, it can effectively reduce the overall human mortality to 5–14%, whereas untreated, the mortality rate is between 50% and 90%. Streptomycin is a water soluble aminoglycoside that is marketed as the sulfate salt of streptomycin. Streptomycin interferes with several steps of protein synthesis, its most conspicuous effects being the stimulation of translational errors and a slowing down of translocation resulting in the production of faulty proteins. Streptomycin binds to the decoding center of bacterial 16S rRNA in the absence of ribosomal proteins, and protects a set of bases in the decoding region against dimethyl sulfate attack .
Although the development of antibiotics has significantly lowered the mortality rates, there still exists the serious threat to public health posed by pathogenic bacteria that have developed resistance to the traditional antibiotics, which is also the case for plague.
In 1995, a Y. pestis strain, resistant to streptomycin, ampicillin, chloramphenicol, kanamycin, sulfonamides, tetracycline and minocycline, was isolated in Madagascar from a 16-year-old boy . Therefore, it is important to extensively understand the molecular mechanisms of action of the traditional antibiotics, and attempts should be conducted as well for the development of new antibiotics. The availability of the genome sequences [7,,9] and the subsequent development of DNA microarrays to profile the transcriptome [10,,,13] have opened a window for monitoring global changes in gene expression patterns in Y. pestis.
Here we used the whole-genome DNA microarray to investigate the global transcriptional response of Y. pestis triggered by the treatment of streptomycin, giving an overall picture of the molecular mechanisms of action of the antibiotic in vanquishing this deadly pathogen.
Materials and methods
Bacterial strain, medium and MIC determination
Yersinia pestis strain 201 was used in this study. It was isolated from Microtus brandti in Inner Mongolia, China. It has major phenotypes as F1+ (able to produce fraction 1 antigen or the capsule), VW+ (presence of V antigen), Pst+ (able to produce pesticin) and Pgm+ (pigmentation on Congo-red media). Strain 201 has an LD50 of less than 100 cells for mice by subcutaneous challenge. Strain 201 belongs to a newly established Y. pestis biovar, microtus . Biovar microtus strains are supposed to be avirulent to humans, although they are highly lethal to mice . A chemically defined TMH medium  was used for cultivating the bacteria.
Minimal inhibitory concentration (MIC) was determined by the broth dilution method . The tests were performed in sterile glass tubes with an initial inoculum of approximately 105 CFU of Y. pestis cells (overnight cultures) in 1 ml of TMH medium per tube. Streptomycin sulfate (Amresco) was dissolved in each tube and dilution series were made to give a final concentration range from 0.1 to 128 μg/ml. Bacteria were then cultured at 37 °C for 20 h. The MIC was defined as the lowest concentration that prevents the development of visible growth.
Bacterial growth and RNA isolation
Strain 201 was grown at 26 °C to the middle exponential growth phase (an A620 of about 0.6) in the TMH medium. The cell cultures were 1:20 diluted in fresh TMH medium and the cells experienced at least 10 generations in the medium prior to reaching to the middle exponential growth phase. Bacteria were then transferred to grow at 37 °C for 1 h to be ready for antibiotic treatment. Cells were incubated at 37 °C for 30 min under the treatment of 10× MIC (80 μg/ml) of streptomycin; the control culture was allowed to continue growing at 37 °C for the same period of time with adding the same volume of distilled water. Immediately before harvesting for RNA isolation, bacterial cells were mixed with RNA protect Bacteria Reagent (Qiagen) to minimize RNA degradation. Total RNA was isolated by using the MasterPure™ RNA Purification kits (Epicenter). RNA quality was monitored by agarose gel electrophoresis and RNA quantity was measured by spectrophotometer. Two independent bacterial cultures for each test or control condition were prepared as biological replicates for RNA isolation.
Probe synthesis and mciroarray hybridization
Fifteen to 20 micrograms of RNA were used to synthesize cDNA in the presence of aminoallyl-dUTP, genome directed primers (GDPs) and random hexamer primers with the Superscript II system (Invitrogen). The reverse transcription of bacterial RNA by the mixture of GDPs and random hexamers has been proven to be more effective and reliable than with either GDPs or random hexamers only . The aminoallyl-modified cDNA was then labeled by Cy5 or Cy3 monofunctional dye (Amersham) according to the manufacturer’ instruction. Three separated labeled probes were made for each RNA preparation as technical replicates. Pairwise comparisons were made using dye swaps to avoid labeling bias.
Glass slides spotted with PCR amplicons representing about 95% of non-redundant annotated genes or ORFs of Y. pestis CO92 and 91001 were used for hybridization . Briefly, based on the genomic sequences of Y. pestis CO92 and 91001, a total number of 4015 annotated open reading frames (ORFs) were selected after the exclusion of ORFs encoding IS protein, integrase, and transposase. Specific primer pairs were designed to amplify nearly the full length of each gene. The purified 4005 successful amplifications were spotted on the CSS-1000 silylated glass slides (CEL) by using a SpotArray72 Microarray Printing System (Perkin–Elmer Life Sciences) to construct the DNA microarrays. The spotted slides were crosslinked by using a UV Stratalinker (Hoefer). NaBH4 was used to block the free aldehyde groups on the slide surface. The slides were prehybridized in a buffer containing 5× SSC, 0.1% SDS and 0.1% BSA, and then washed and blown to dry. The two differentially labeled cDNA samples were dried and then resuspended in hybridization solution (50% deionized formamide, 5× SSC, 0.1% SDS, 5× Denhardt's solution, and 0.5 μg/μl of sheared salmon sperm DNA). The labeled DNA samples hybridized with the slides at 42 °C for 18–20 h. After hybridization, the slides were washed in 1× SSC with 0.06% SDS for 2 min, then in 0.06× SSC for 2 min and finally in ethanol for 2 min. The slides were blown to dry and then were scanned by using a GenePix Personal 4100A Microarray Scanner (Axon Instruments).
The scanning images were processed and the data were further analyzed by using GenePix Pro 4.1 software (Axon Instruments) combined with Microsoft Excel software. Spots were analyzed by adaptive quantitation, and the local background was subsequently subtracted. Spots with background-corrected signal intensity (median) in both channels less than twofold of background intensity (median) were rejected from further analysis. Data normalization was performed on the remaining spots by total intensity normalization methods. The normalized log2 ratio of test/reference signal for each spot was recorded. Genes with less than three data points were considered unreliable, and their data points were discarded as well. The averaged log2 ratio for each remaining gene on the six replicate slides was ultimately calculated. Significant changes of gene expression were identified with SAM software  using one class mode (Δ= 1.01444) [the measurement is the log (red/green) ratio from two labelled samples hydridized to a cDNA chip, with green denoting before treatment and red, after treatment]. Significance analysis of microarrays (SAM) assigns a score to each gene on the basis of change in gene expression relative to the standard deviation of repeated measurements. For genes with scores greater than an adjustable threshold, SAM uses permutations of the repeated measurements to estimate the percentage of genes identified by chance, the false discovery rate (FDR).
Results and discussion
General comments on microarray profiling
In our present work, DNA microarray was used as a tool to survey the global effects mediated by the addition of increasing concentration of the bactericidal agent streptomycin to logarithmically growing cultures of Y. pestis strain 201. The MIC value of Y. pestis strain 201 versus streptomycin sulfate was determined as 8 μg/ml. In order to furthest understanding the transcriptional response of Y. pestis under the treatment with streptomycin, we magnify the inhibitory drug concentration to 10× MIC, meanwhile only the short-term (30 min) changes was examined to avoid confounding secondary drug effects.
A wide set of genes or operons was affected, both positively and negatively, upon the exposure to streptomycin. Of all the available 2380 genes in the final microarray dataset (data points of the remaining 1625 genes were missed in the data filtering process, indicating that these genes were not expressed at all in the experimental conditions or had data with low quality), 345 were identified to be differentially regulated; among them, 144 genes were up-regulated, 201 down-regulated (see Supplementary Tables 1 and 2 for details). Fig. 1 provides a summary of the differentially regulated genes grouped by functional categories, giving an overall picture of the alteration of the global gene transcription pattern of Y. pestis. Several general trends in gene transcriptional response can be elicited on the basis of the differentially regulated genes with proven or annotated functions. The microarray data provide a clear cascades that streptomycin disturbs the metabolic pathways and the structuring of cell envelope in Y. pestis, which is consistent with the proven or suggested mechanisms of action of streptomycin. Some other interesting observations, such as the induction of stress response proteins, the up-regulation of genes responsible for drug/analogue sensitivity and the induced biosynthesis of selective amino acids, show that Y. pestis takes several responsive mechanisms in confronting with the addition of streptomycin. In addition, a portion of genes whose transcription was altered encode proteins of unknown or unassigned functions. The detected changes in gene transcription provides seful candidates for investigating the specific mechanisms of action of this drug which is highly efficient against plague infection. Noticeably, the up-regulation of many genes, especially those with unknown functions, will provide novel insights into the unknown mechanisms of action of streptomycin against Y. pestis and can be useful starting points in the identification of new targets for antimicrobial chemotherapy.
Genes up-regulated by treatment of streptomycin
Stress response proteins
The category of genes with the distinct response to treatment with streptomycin was the heat shock group. Bacterial heat shock response appears to a global regulatory system for effective adaptation to various environmental changes including elevation in temperature, the addition of ethanol and heavy metals, high osmolarity, and starvation [20,21]. It involves the induction of the major heat shock proteins (MHSPs), including chaperones, proteases, transcriptional regulators and other structural proteins, that act by repairing and preventing damages caused by an accumulation of unfolded proteins . Previous studies on Escherichia coli have shown that the addition of streptomycin or kanamycin (another aminoglycoside antibiotics) results in the induction of the MHSPs [22,23]. The microarray analysis showed that streptomycin induces the heat shock regulon in Streptococcus pneumoniae. The heat shock response to aminoglycosides may be a common stress response for bacteria. In this study five heat shock genes (rpoH, rpoE, ibpA, ibpB, and htrA) were induced by the treatment of streptomycin. HtrA is a protease that digests abnormal proteins in the periplasm . Mutation in the htrA gene leads to decreased survival of Y. pestis in mice and/or macrophages . The heat shock genes ibpA and ibpB, encoding chaperones, play roles in assisting the folding of newly synthesized proteins, preventing aggregation of proteins, and repairing proteins that have been damaged or misfolded by heat shock and other stresses . The heat shock response in Gram-negative bacteria is positively controlled at transcriptional level by two sigma factors, σ32 (RpoH) and σE (RpoE). In E. coli, σ32 controls the expression of many MHSPs including FtsH, DnaKJ, GroELS, Lon, ClpAP, HslVU and GrpE, while σE positively controls the expression of the σ32 regulon, HtrA and FkpA . Our previous study showed that the sudden upshift of temperature from 37 to 45 °C for 10 min (heat shock) significantly enhanced the transcription of rpoH in Y. pestis, but rpoE was not differentially regulated . In this study, both rpoH and rpoE were up-regulated by the addition of streptomycin.
The psp FABCD locus encoding the phage shock proteins, initially identified in E. coli, was found to help to ensure survival of the bacterium in late stationary phase at alkaline pH, and protect the cell against dissipation of its proton-motive force against challenge . Until now, the precise biochemical and physiological functions of the psp genes are not understood. In this study, transcription of pspABC (pspD was not represented on the microarray) was up-regulated by 4- to 7-fold after the treatment of streptomycin. Interestingly, transcription of Y. pestis pspA was also up-regulated as well upon shift from 26 to 37 °C during steady-state vegetative growth . The psp genes likely play some roles in the accommodation of Y. pestis to environmental perturbations.
In E. coli, the plasmid-borne terZABCDE operon encoding tellurium resistance proteins was associated with resistance to tellurium, bacteriophage and colicins . These genes are part of a pathogenicity island that contains additional integrase, phage and urease genes and appear to be up-regulated in the presence of tellurite that has a long history as an antimicrobial agent and is often employed in selective media for the isolation of a wide range of pathogens . Four genes (terZABC) were up-regulated in this study, suggesting that the ter genes are associated with protection from other forms of stress or agents causing cellular damages. Two genes (YP0976-0977 according to the genome annotation of Y. pestis 91001 ) encoding an ATPase and an permeases of the ABC-type multidrug transport system and the acrB/acrE gene encoding a multidrug efflux protein were down-regulated by streptomycin treatment, indicating that these genes do not play an important role in confronting with streptomycin.
Induced biosynthesis of selective amino acids
Transcription of three (thrA, dapB and dapF) and six (YPO1687 and ilvADEMG) genes, responsible for biosynthesis of amino acids of aspartate and pyruvate families respectively, was greatly elevated in response to streptomycin treatment. Noticeably, the ilvADEMG operon encodes four of the enzymes for the biosynthesis of the branched-chain amino acids, isoleucine, leucine and valine in E coli; when the bacterium grows in a chemically defined medium, limitation of one of these three branched-chain amino acids caused an increase in expression of the enzymes, while the concomitant addition of all three amion acids cause a decrease in expression [30,31]. Streptomycin dramatically disturbs the cell envelope structure and trans-membrane transporting functions (see below). Therefore, it can be hypothesized that transport of some amino acids is disturbed, thus introducing a deprivation condition, which will induce the expression of the corresponding biosynthesis genes.
Broad regulatory functions
Transcriptional change of a gene encoding global regulatory protein may have pleiotropic effects. Nine genes involved in broad regulatory functions were up-regulated by treatment with streptomycin (see Supplementary Table 1). The most notable one is the Lrp gene encoding the leucine-responsive regulatory protein. In E. coli, Lrp regulates, both positively and negatively, the expression of more than 40 genes and proteins involved in amino acid biosynthesis, pili biosynthesis, ammonia assimilation, amino acid catabolism, and peptide transport [32,33]. A substantial fraction of operons regulated by Lrp are also regulated by leucine, and the effect of leucine on expression of these operons requires a functional Lrp protein . Therefore, the induced transcription of the Y. pestis lrp gene in response to streptomycin is consistent with the elevated expression of the ilv genes responsible for the biosynthesis of isoleucine and leucine. Together with the results above, we can postulate that some amino acids will play an important role in facing the presence of streptomycin by unknown mechanisms.
Chemotaxis and mobility
Motile bacteria can swim toward or away from specific environmental perturbations, such as nutrient-limiting or stress conditions; this behavior, called chemotaxis, is mediated by the bacteria changing direction by briefly reversing the direction of rotation of the flagellar motors . Surprisingly, four genes (YPO0704, flhB, fliR/mopE and motB/lafU) involved in chemotaxis and flagellar assembly were up-regulated by 2.0- to 4.0-fold in this study, although Y. pestis is uniformly nonmotile. We also found that seven flagellar genes were greatly up-regulated after heat shock (sudden temperature upshift from 37 to 45 °C) . Genome annotation of Y. pestis CO92 revealed a total of 86 chemotaxis- or flagellar-related genes that constitute several distinct clusters in chromosome, and mutations exist in most of these gene clusters, but it was still suggested that some form of motility might be possible in Y. pestis under atypical conditions .
Streptomycin-repressed cellular pathways
The addition of streptomycin repressed the transcription of a large number of genes involved in energy metabolism, biosynthesis of small macromolecules, synthesis and modification of macromolecules, degradation of small and macro molecules and transport/binding functions, suggesting that most of the metabolic pathways in Y. pestis were retarded after the drug treatment. This general repression of metabolic processes is likely to be a common response reflecting growth arrest, rather than organism- or drug-specific.
The highest number of genes which are regulated by streptomycin is found in the energy metabolism functional category, and that this regulation is always a decrease of the transcription rate. Fifty-two (39%) of the totally 132 energy-metabolism-related genes on the microarray were down-regulated, accounting for the overall slowing down of cellular energy generation. These genes or operons fall into the functional categories of aerobic respiration (16 genes), ATP-proton motive force (8), electron transport (9), fermentation (2), glycolysis (3), non-oxidative branch (1), pyruvate dehydrogenase (3) and tricarboxylic acid cycle (10), showing the wide inhibition of energy metabolisms in Y. pestis by streptomycin treatment.
Biosynthesis of small macromolecules
The proteins encoded by tauABC constitute a system for uptaking taurine as a source of sulfur and the product of tauD is involved in the oxygenolytic release of sulfite from taurine . The cys genes involved in uptake of sulfite, reduction of sulfite to sulfide and synthesis of cysteine from serine, and these three processes are classified into the cysteine biosynthetic pathway . Addition of the cysteine thiols to the vinyl groups of heme occurs in the oxidative environment of the bacterial periplasm . This heme-binding process is crucial for the maturation pathway of cytochrome c that is the indispensable component of electron transport. In this study, the transcription of three tau genes (tauACD) and nine cys genes (cysA, cysG, cysH, cysI, cysJ, cysK/cysZ, cysP, cysT and cysW) was repressed. It is speculated that the reduced rate of energy metabolism (electron transport) may result in a deceased requirement of dissociative cysteine that acts as a signal inhibiting the transcription of these genes involved in organosulfur utilization and cysteine biosynthesis
Some genes involved in biosynthesis of pyrimidine ribonucleotides (nadE/adgA) and purine ribonucleotides (purA, purB, purF, purL, purI and guaA) and also salvage of nucleosides and nucleotides (upp/uraP) were down-regulated, indicating a generally reduced cellular pool of nucleotides resulted from the retardance of cell growth.
Synthesis and modification of macromoclecules
Among the down-regulated genes is another group genes involved in synthesis and modification of macromoclecules. These genes could be mainly assigned into the functional categories of DNA replication, restriction/modification and repair (7 genes), RNA and aminoacyl tRNA synthesis (10), and biosynthesis of cytoplasmic glycogen (4), lipopolysaccharide (9) and peptidoglycan (7). It was not surprising that there was the general retardance of DNA and RNA synthesis. Our interest was primarily focused on the reduced biosynthesis of cytoplasmic and surface polysaccharides, and of the cell wall component of peptidoglycan (see below). The glgPACXB operon responsible for glycogen synthesis was down-regulated; the reduced accumulation of cytoplasmic glycogen should destroy the original volume of the Y. pestis cytoplasm.
Degradation of small and macro molecules
The RecBCD enzyme, encoded by recB, recC, and recD genes, is a potent large ATP-dependent exodeoxyribonuclease enzyme that is active on either ds or ss DNA and a weak ATP-stimulated endonuclease activity that acts only on ss DNA . The reduced transcription of the three rec genes in this study indicated a restrained metabolic activity in the degradation of DNAs. In addition, transcription of 13 genes involved in amino acids metabolism, carbon compounds metabolism, degradation of RNAs, or degradation of proteins and peptides was repressed.
Action of streptomycin on cell envelope
A bacterial cell has the four major structural components: a cytoplasmic region, a cell membrane, a cell wall, and some sorts of surface structure; the later three coin a descriptive term “cell envelope”. The cytoplasm is surrounded by the plasma membrane, while the cell wall contribute to prevent damage to the underlying protoplast. Outside the cell wall, there is the surface structure composed of polysaccharides or proteins. In this study, at least 25 genes encoding membrane components (membrane proteins, lipoproteins, porins and exported proteins) were up-regulated by the streptomycin treatment. The induced expression of selective membrane components indicates the remodeling of cell membrane of Y. pestis upon exposure to streptomycin.
The cell wall of Gram-negative bacterium consists of a thin peptidoglycan sheet between the plasma (inner) membrane and an outer membrane. Peptidoglycan is a giant macromolecule of periodic structure that forms the polymer to constitute the shape-maintaining structure of cell wall . The Y. pestis mur operon responsible for peptidoglycan synthesis was down-regulated by the streptomycin treatment. Lipopolysaccharide (LPS) is a major component of the outer membrane. Y. pestis expresses rough type of LPS without O-antigen polysaccharide due to the inactivation of the O-antigen gene cluster . Our microarray analysis identified nine genes, encoding enzymes for LPS synthesis, whose transcription was down-regulated.
An essential feature of living organisms is the ability to accumulate nutrients against a concentration gradient and to excrete the various end products of metabolism. The main cellular structure responsible for nutrient transport is the plasma membrane and the outer membrane in Gram-negative bacteria. Under streptomycin treatment, at least 29 Y. pestis genes encoding transport/binding proteins were differently regulated; among them, 12 genes were up-regulated and 17 down-regulated. Noticeably, the down-regulated genes mainly fall into the functional categories of transport of amino acids and amines, cations, and carbohydrates, organic acids and alcohols. This showed that streptomycin could slowdown the ability of Y. pestis to acquire amino, cation and carbohydrate through affecting the relevant transporters.
As a member of the aminoglycoside class of antibacterial agents, streptomycin irreversibly binds to the 30S ribosomal subunit to inhibit protein synthesis and cause misreading of mRNA . However, additional hypotheses have been proposed to suggest the ‘lethal mechanism’ of aminoglycosides, including changes in the cell membrane, mistranslation of critical proteins, and other uncharacterized secondary effects . In this study, the detecting transcriptional changes in the genes responsible for cell envelope function categories supports the notion that streptomycin has the profound detrimental effects on the normal structuring of Y. pestis cell envelope.
Differentially regulated plasmid-borne genes
Yersinia pestis normally carries three prototypical virulence plasmids: a 9.5-kb plasmid pPCP1, a 110-kb pMT1 and a 70-kb pCD1. At least 23 plasmid genes were differentially regulated after streptomycin treatment. Y. pestis has acquired two unique plasmids, pPCP1 and pMT1; both of them contribute to the extreme pathogenicity of Y. pestis. All the three pathogenic yersiniae, Y. pestis, Y. pseudotuberculosis, and Y. enterocolitica, carry the pCD1 plasmid that encode a type III secretion system (T3SS). The T3SS allows extracellular Yersinia that is docked at the surface of cells of the immune system to deliver Yop effectors into the cytosol of these cells. These Yops could prevent phagocytosis of the bacteria and inhibit the host immune response. The T3SS composes of four elements: (i) type III secretion machinery called Ysc; (ii) a set of proteins required to translocate the effector proteins inside the eukaryotic cells; (iii) a control system, and (iv) Yop effector proteins . The up-regulated T3SS genes after streptomycin treatment contain lcrF encoding thermoregulatory protein, yopT encoding cytotoxic effector protein, two syc genes sycE and sycT, and four ysc genes yscT, yscN, yscB and yscC. On the other hand, four genes encoding type III secretion substrates YopB, YopD, YopH and YopP, and two ysc genes yscQ and yscR were down-regulated.
Financial supports for this work came from the National Natural Science Foundation of China (No. 30430620).
Appendix A Supplementary data
Supplementary data associated with this article can be found, in the online version at doi:10.1016/j.femsle.2005.01.018.