Skip to main content

Global transcriptional response of Escherichia coli O157:H7 to growth transitions in glucose minimal medium



Global patterns of gene expression of Escherichia coli K-12 during growth transitions have been deeply investigated, however, comparable studies of E. coli O157:H7 have not been explored, particularly with respect to factors regulating virulence genes and genomic islands specific to this pathogen. To examine the impact of growth phase on the dynamics of the transcriptome, O157:H7 Sakai strain was cultured in MOPS minimal media (0.1% glucose), RNA harvested at 10 time points from early exponential to full stationary phase, and relative gene expression was measured by co-hybridization on high-density DNA microarrays. Expression levels of 14 genes, including those encoding Shiga toxins and other virulence factors associated with the locus of enterocyte effacement (LEE), were confirmed by Q-PCR.


Analysis of variance (R/MAANOVA, Fs test) identified 442 (36%) of 1239 O157-specific ORFs and 2110 (59%) of 3647 backbone ORFs that changed in expression significantly over time. QT cluster analysis placed 2468 of the 2552 significant ORFs into 12 groups; each group representing a distinct expression pattern. ORFs from the largest cluster (n = 1078) decreased in expression from late exponential to early stationary phase: most of these ORFs are involved in functions associated with steady state growth. Also represented in this cluster are ORFs of the TAI island, encoding tellurite resistance and urease activity, which decreased ~4-fold. Most ORFs of the LEE pathogenicity island also decreased ~2-fold by early stationary phase. The ORFs encoding proteins secreted via the LEE encoded type III secretion system, such as tccP and espJ, also decreased in expression from exponential to stationary phase. Three of the clusters (n = 154) comprised genes that are transiently upregulated at the transition into stationary phase and included genes involved in nutrient scavenging. Upregulated genes with an increase in mRNA levels from late exponential to early stationary phase belonged to one cluster (n = 923) which includes genes involved in stress responses (e.g. gadAB, osmBC, and dps). These transcript levels remained relatively high for > 3 h in stationary phase. The Shiga toxin genes (stx 1AB and stx 2B) were significantly induced after transition into stationary phase.


Expression of more than 300 O157-specific ORFs, many implicated in virulence of the O157 pathogen, was modulated in a growth dependent manner. These results provide a baseline transcriptional profile that can be compared to patterns of gene expression of this important foodborne pathogen under adverse environmental conditions.


Enterohemorrhagic Escherichia coli (EHEC), a food and water- borne pathogen of zoonotic origin, are an important cause of acute gastroenteritis in humans. O157:H7 is the predominant serotype of EHEC causing illness in the United States [1]. One of the hallmarks of EHEC pathogenesis is the formation of attaching and effacing lesions on intestinal epithelial cells [2, 3], resulting in intimate adherence of the bacterial cell to the intestinal mucosa. The ability to form A/E lesions is encoded by the Locus of Enterocyte Effacement (LEE), a 35 kb pathogenicity island [4]. Another principal virulence characteristic of EHEC is the production of Shiga toxins, two component cytotoxins that inhibits protein synthesis in eukaryotic cells [5]. The genes for production of Shiga toxin (stx1 and stx2) are located in intact or partial genomes of lambda prophages that are inserted into the chromosome [6, 7]. Together, the LEE and Shiga toxins are considered to be two of the principal virulence determinants that mediate development of hemorrhagic colitis and the life-threatening hemolytic uremic syndrome [8].

Regulation of these virulence factors is influenced by global regulators as well as regulators specific to the virulence factor. Many of the virulence factors in EHEC are carried on 'foreign' DNA, DNA which is not present in E. coli K-12, and their expression is typically regulated by transcriptional regulators also carried on these mobile elements. EHEC can utilize global regulators that are common to all E. coli as well as O157-specific regulators to control expression of virulence factors [8, 9]. For example, the expression of LEE is dependent on growth phase and responds to nutrient downshifts via ppGpp signaling through DksA to Ler (L EE e ncoded r egulator) and PchA [10]. Also, LEE expression can be upregulated by activation of Ler in the presence of bicarbonate ions [11], and by quorum sensing, in which autoinducer 3 activates LEE expression through Ler [12]. Shiga toxin expression is linked to expression of the phage lysis genes [13, 14] and can be controlled by the iron responsive regulator Fur [15]. In addition to the known virulence determinants, E. coli O157:H7 Sakai and EDL-933 contain ~1.3 Mb of sequence not found in E. coli K-12; these sequences occur as part of the genetic material of 18 lysogenic phages and 6 phage-like elements that have integrated into the Sakai genome [16] and as 177 O157 strain-specific islands (OI number) in the EDL-933 strain [17]. Recently it was identified that many effector proteins that are secreted via the LEE-encoded type three secretion system are specified by genes distributed throughout the genome on lambdoid phages [18].

In general, E. coli populations are ecologically and metabolically versatile and can adapt to growth under a wide range of conditions, adaptations essential to a bimodal lifestyle either in the primary habitat within animal hosts or in secondary habitats as free living cells in the natural environment [19]. Slowing metabolic activity during stationary phase reflects a survival mechanism in nutrient poor environments in which bacteria undergo a variety of morphological and physiological changes. In stationary phase, E. coli are more resistant to a number of stresses, including pH stress [20, 21] and osmotic stress [22], both of which occur during transit through the colonization of the host intestinal tract. Important to the transmission route of E. coli O157:H7 is the ability to persist in adverse environments until entering a new host. Hence studying growth transitions and how they relate to the lifestyle of E. coli O157:H7 is essential to understanding the persistence and spread of this pathogen from the bovine reservoir to foods and to humans.

Microarrays have been utilized to observe changes induced in the transcriptome of E. coli during growth transitions and in response to different environmental stresses. Global transcription profiling of E. coli K-12 indicates a coordinated response to growth arrest [23]. The sigma factor RpoS plays a crucial role in transcription of stationary phase and stress response genes [2426]. Microarray studies have been conducted to understand how E. coli K-12 responds to growth at low pH [27, 28], on rich and minimal media [29], during anaerobic growth [3032], and growth transitions in rich medium [33]. Transcriptional responses of O157:H7 during growth on plasma membranes [34] and upon exposure to norfloxacin [35], have been investigated, and microarrays have been used to identify genes regulated by luxS in E. coli O157:H7 [36], but a full picture of global gene expression during growth transitions, although well-studied in E. coli K-12, has not been investigated in the food borne pathogen E. coli O157:H7.

A key component to understanding the adaptations and ecology of E. coli O157:H7 is to determine how the O157-specific genomic elements respond to one of the critical stages of growth, stationary phase. It has been documented that some virulence factors are regulated by environmental cues that signal the presence of the host environment, but what is less understood is how these virulence factors are expressed in a non-host environment. Here we address the question of how O157:H7 responds to growth transitions by determining the changes in global gene expression patterns in O157:H7 during the transition from exponential to stationary phase in minimal medium, with particular interest in expression patterns of O157-specific genes. The dynamics of the transcriptome of E. coli O157:H7 Sakai was investigated by comparative microarray hybridizations at 10 different time points during exponential and stationary phase and compared to obtain a temporal pattern of gene expression. Gene expression patterns were also verified by Q-PCR of 14 ORFs including those encoding principal virulence factors (e.g. LEE-encoded proteins and Shiga toxins).


Analysis of expression ratios

The results are based on O157:H7 Sakai cultures collected from 10 time points of growth in minimal media with 4 replicates (i.e. 4 independent cultures that were each sampled 10 times) for a total of 40 samples. The resulting array data have been deposited in the NCBI Gene Expression Omnibus, accession GSE7477. The samples collected from the mid-exponential phase 3-hour time point (Fig. 1) were used as a common reference for hybridization and analysis for all subsequent time points in a replicated reference design where all the samples are biologically replicated, including the reference [37]. This design allowed for indirect comparison between all time points via the 3-hour samples [37, 38]. We incorporated the dye-swaps among the four biological replicates which can confound dye effects and biological replicate effects. However, if there is significant variation in the rate of dye incorporation from one labeling reaction to another, this would result in large dye-effects compared to biological replicate effect [39], an effect not found in our experiments. In general, the variance estimates for the biological replicate effects were small (median 0.0008), compared to the variance estimates (<0.5%) for the array effects (median 0.16). Most ORFs (4362/5886, 90%) had similar signal intensities for the mid- and late- exponential phase samples, indicating that there were few overall differences in expression between these time points (Fig. 2A and 2B). In contrast, there were many differences in expression found between mid-exponential phase samples and samples from the transition point (Fig. 2C) and early stationary phase (Fig. 2D).

Figure 1

Average growth of E. coli O157:H7 Sakai in MOPS minimal medium. Increase in cell density is measured at OD600 at 10 time periods (hr) of growth into stationary phase. Error bars represent the standard deviation of four culture replicates. Samples were taken from the culture at the time points plotted and RNA was extracted. The 5 growth phases, separated by dotted lines, are defined as exponential growth, late exponential growth, transition to stationary phase, early stationary phase, and stationary phase.

Figure 2

Average signal intensity plots for four of the time points of growth. The normalized signals of 4,886 ORFs for E. coli O157:H7 Sakai were averaged from four hybridizations, representing four biological replicates: A. Average signal intensity for exponential phase (3 h) vs late exponential phase (4 h). B. Average signal intensity for exponential phase (3 h) vs late exponential phase (4.6 h). C. Average signal intensity for exponential phase (3 h) vs transition to stationary phase (5 h). D. Average signal intensity for exponential phase (3 h) vs. stationary phase (6 h. The Lowess line is plotted for each graph in red.

Summary of significant changes in gene expression over time

Significant changes in expression through time were determined using ANOVA and the Fs-test [40] for the 4,886 O157:H7 Sakai ORFs targeted on the microarray. ANOVA does not capture how time affects gene expression, but how expression changes over time. The p values were adjusted to correct for type I error using the Benjamini-Hochberg (B-H) false discovery rate approach with the linear step-up correction implemented in R/MAANOVA. The B-H correction for multiple testing applies a step-wise correction to the p values, and allows selection of a suitable p value cutoff based on practical considerations. The 2,552 ORFs with the lowest adjusted p values (< 0.0000001) were considered for further analysis (Additional File 1). Fold change in expression between time points was determined by calculating the difference in log2 expression between the time points of interest (Additional Files 1 and 2).

Trends in gene expression identified by significant contrasts between subsequent time points and QT clustering

Expression profiles of the 2,552 ORFs with significant differences were clustered into 12 groups using Quality Threshold (QT) clustering (Fig. 3), which groups similar expression profiles based on jackknife correlations [41]. The majority of ORFs were classified into 12 groups; only 84 were not placed into a cluster. ORFs were also classified based on the time intervals at which a significant change in gene expression occurred. A subset of 524 ORFs had differences in expression from mid- to late-exponential phase (3 h to 4 h). The majority of ORFs, 1,668 of 4,886 (34%), had significantly changed in transcript level at the stationary phase transition point (4.6 to 5 h). Additionally, 1,045 of 4,886 (20%) ORFs significantly changed during the transition point to early stationary phase (5 to 5.5 h).

Figure 3

Quality Threshold (QT) clusters of the 2,552 ORFs with significant changes in gene expression. The number of ORFs in each cluster is listed at the top of each plot. Average expression profiles were determined for the ORFs in each cluster and plotted with the standard deviation for each time point.

Significant changes in exponential phase – response to limiting oxygen

There were 524 ORFs with significant differences in transcript level during exponential phase – a total of 113 genes changed only in exponential phase, whereas the remainder had altered levels of expression in exponential as well as stationary phase. ORFs that decreased in expression included those encoding enzymes of the TCA cycle, such as sucA, and sdhABCD (Table 1, Fig. 3; QTC 10). Many of the ORFs (94/524, 18%) that changed in transcript level are regulated by the transcription factor FNR [31, 32], whereas a smaller set of ORFs (29/524, 5%) are known to be regulated by ArcA [42]. Genes that increased included those encoding anaerobic electron acceptors, fumarate reductase, frdAB, and DMSO reductase, dmsA and dmsC, as well as a DMSO reductase paralog, ynfE (Table 1, Fig. 3; QTC 6). Genes encoding the cytochrome c biogenesis system (ccmABCDEFG) and the cytochrome c-like protein (napB and napC) also had a significant increase from 3 h to 4 h. The operon encoding the low affinity cytochrome oxidase bo (cyoABCD) decreased 9 to15- fold in expression from 3 h to 4 h (Table 1). Concomitantly, expression of the high affinity cytochrome oxidase bd, cydAB, increased 4–5 fold. Expression of the high and low affinity cytochrome oxidases is directly influenced by the level of available oxygen in the media [43, 44]. Gene set enrichment analysis (GSEA) found that genes involved in anaerobic energy metabolism were significantly enriched in the 4 h sample compared to the 3 h sample, and that genes involved in aerobic energy metabolism and the TCA cycle were significantly enriched in the 3 h compared to the 4 h culture (Additional File 3). The altered expression of the cytochrome oxidases as well as the increase in FNR-controlled, anaerobic associated genes and decrease in TCA cycle enzymes is consistent with the hypothesis that oxygen became limiting under the culture conditions that were used. Subsequent measurement of the residual dissolved oxygen tension in the culture during growth indicates that residual dissolved O2 levels decreased almost to zero (95% decrease in O2) during exponential phase (Fig. 4). As cultures entered stationary phase, the O2 levels began to increase, as the demand for O2 by the culture decreased.

Table 1 Significant ORFs (p value < 1 × 107) with greater than 4-fold significant change between 3 h and 4 h.
Figure 4

Residual dissolved oxygen (O 2 ) in MOPS minimal medium. The mean rDOT was determined for three independent cultures of E. coli O157:H7 Sakai in glucose limited MOPS minimal medium and plotted with the error bars indicating the standard deviation.

Significant change in O157-specific genes during exponential phase

In contrast to the backbone genes, only a small proportion (6%) of the O157-specific ORFs changed significantly in expression measured during exponential phase. While statistically significant, most of the differences in expression level for these 61 genes were not substantial, ranging from 1.2- to 2- fold (Additional File 2). Many of these ORFs are of unknown function, but of the 45 O157-specific ORFs that increased from 3–4 h, 9 of them were from a single Sakai phage (Sp) 15 region which encodes the Shiga toxin 1 phage, and contains 5 ORFs from Sp 5 which encodes the Shiga toxin 2 phage. Interestingly, ECs2381 a locus whose expression has been shown to contribute to colonization in the bovine gastrointestinal tract [45], was also significantly induced 1.8 fold from 3 to 4 h. There were 4 additional O157-specific ORFs that had a greater than 4-fold change in expression from 3–4 h (Table 1). These 4 ORFs were not associated with any of the Sakai phages, and have unknown or putative functions.

Genes with significant changes in expression at the stationary phase transition point

The majority of significant changes in transcript level occurred at the stationary phase transition point. There were 176 ORFs with a 4-fold or greater change in expression at the stationary phase transition point (Table 2). A total of 363 ORFs that significantly increased or decreased during the transition to stationary phase (4.6–5 h) also increased or decreased in early stationary phase (5–5.5 h) (Table 3). A reduced level of expression of ribosomal genes and genes involved in nucleotide and amino acid synthesis was observed from 4.6–5 h and 5–5.5 h (Tables 2 and 3, Fig. 3; QTC 1). Many of the genes that were significantly down-regulated upon entry into stationary phase have been also observed in E. coli K-12 [23], including decreased expression of the molecular chaperones encoded by groEL, htpG, and dnaK (Tables 2 and 3, Fig. 3; QTC 1, Additional File 1).

Table 2 Significant ORFs (p value < 1 × 107) with greater than 4-fold significant changes in expression at the stationary phase transition point (4.6 to 5 h).
Table 3 Significant ORFs (p value < 1 × 107) with greater than 4-fold significant changes in expression from late exponential to early stationary phase (4.6 to 5.5 h)

Expression of rpoS, encoding the stationary phase sigma factor, increased 7 fold during the transition point and the RpoS activator, dksA, increased 4.3 fold at the same interval (Table 2). Increased expression of genes known to be regulated by RpoS, such as bolA, dps, osmE, osmC, adhP and ugpB occurred at the stationary phase transition point (Tables 2 and 3, Fig. 3; QTC 2). Gene set enrichment analysis also indicated that these genes (members of the adaptation to atypical conditions group) were significantly enriched in the stationary phase time points compared to exponential phase (Additional File 4). ORFs encoding toxin-antitoxin systems, such as dinJ-yafQ (Table 3, Fig. 3; QTC 2) and chpB-chpS (Fig. 3; QTC 3), were upregulated at the stationary phase transition point. The greatest increase in transcriptional level for an upregulated gene (45-fold) during the stationary phase transition (4.6 to 5 h interval) was acs, encoding acetyl CoA synthetase (Table 2). Acs expression is controlled in part by RpoS [46]. Also induced at a very high level during the transition point into stationary phase were aceB (40 fold) and aceA (32 fold), which encode components of the glyoxalate shunt. Increased expression of acetyl CoA synthetase as well as genes of the glyoxalate shunt is consistent with the hypothesis that cells are scavenging acetate produced during rapid growth on glucose in exponential phase.

There are a variety of O157-specific genes that were downregulated in the transition to stationary phase include those encoding the O157 LPS antigen (ECs2839, ECs2840, and ECs2847) with a greater than 4 fold decrease in expression (Table 3, Fig. 3; QTC 1) as well as ECs2835, ECs2836, ECs2838, ECs2841, ECs2844, and ECs2845 (1.6 to 3 fold decrease) from that genomic region. ECs2113, part of the F9 fimbrial operon [47], decreased in expression from exponential to stationary phase (Table 3). Other ORFs in this operon, ECs2107-2112, also decreased significantly (~1.5 fold) upon entry into stationary phase. O157-specific ORFs that increased in stationary phase included ECs2182, ECs2737, and ECs5067, which encode putative transcriptional regulators. ECs2737 (pchC) encodes a perC-like homolog reported to regulate LEE transcription [48]. The recently identified chromosomal CcdAB toxin-antitoxin system [49], encoded by ECs0052 and ECs0053, increased 5.3-fold from the transition point to early stationary phase (Table 3, Fig. 3; QTC 2). ECs0415, afuA, encoding a periplasmic ferric iron binding protein, had the greatest increase in expression of the O157-specific ORFs, with a 37-fold increase during the transition point (Table 3, Fig. 3; QTC 2).

Genes with transient expression at the stationary phase transition point

A subset of 120 ORFs were expressed transiently from late exponential to early stationary phase. These ORFs changed significantly during the 4.6 to 5 h time interval, and then again during the 5 to 5.5 h interval, but in the opposite direction (Fig. 3; QTC 3, 11, and 12). Genes with this pattern of transcription included many involved in nutrient scavenging and turnover. Genes encoding both the anaerobic (glpABC) and aerobic (glpD) sn-glycerol 3 phosphate dehydrogenase increased >4 fold during the transition into stationary phase, and then decreased significantly in early stationary phase (Table 4). This pattern is consistent with the general model that the glp metabolic system serves as a salvage pathway for glycerol derived from degradation of phospholipids and triacylglycerol. The murPQ operon was expressed transiently, with a 5.3 fold increase followed by a 2 fold decrease in early stationary phase. The products of murPQ are involved in peptidoglycan turnover [50] suggesting that membrane components are being recycled and utilized for energy. A dicarboxylate transporter, dctA, was induced 24-fold upon entry into stationary phase, then decreased 4-fold in early stationary phase (Table 4). Genes encoding transport and binding proteins for carbohydrates, sugar alcohols and acids were significantly enriched at the stationary phase transition point compared to mid-exponential phase, also indicating the release of glucose-dependent catabolite repression (Additional File 4). Genes encoding salvage and scavenging pathways are also induced during glucose-limited growth [51, 52] (Additional File 1).

Table 4 Significant ORFs (p value < 1 × 107) with transient expression at the stationary phase transition point with expression changes greater than 4-fold.

Only 12 O157-specifc ORFs were expressed transiently upon entry into stationary phase. ECs1772, a putative intestinal colonization factor, decreased upon entry into stationary phase, then significantly increased in expression (>4 fold increase from 5 to 5.5 h) (Table 4).

Genes with a significant change in expression in early stationary phase

The interval with the second largest number of significantly modulated ORFs was in early stationary phase, during the 5 to 5.5 h interval. A number of these ORFs that had significant increases in expression levels during this interval are transcriptional regulators, such as the AraC-like regulators gadW and gadX, involved in regulation of the glutamate decarboxylase system of acid resistance (Table 5). Genes encoding DNA binding proteins, such as hha and cbpA, also increased in transcription in early stationary phase. Components of the glycerol metabolism pathway were activated in early stationary phase, including ugpC and ugpE, components of the glycerol 3 phosphate transporter, and ugpQ, the cytosolic glycerophosphoryl diester phosphodiesterase (Table 5).

Table 5 Significant ORFs (p value < 1 × 107) with greater than 4-fold expression changes in early stationary phase.

Expression of LEE island genes upon entry to stationary phase

Of the 41 LEE island ORFs, 23 had a significant (p < 0.00001) differences in expression in growth into stationary phase under minimal media (Fig. 5). Most of these ORFs were downregulated in transition to stationary phase (between 4.6 and 5 h). Three of these ORFs, espB, espD, and espZ, also had a significant change in expression in early stationary phase (5 h to 5.5 h). The greatest reduction in expression was observed in espB and espD, which both decreased 5.5- fold from late log to early stationary phase. Expression of espA also decreased > 5- fold over this interval, but the error variance for the espA probe was large and thus was not deemed significant. While the overall trend for the LEE island genes was a decline in expression during the transition to stationary phase, the exception was espZ which had a 3.2 fold increase in expression from late log to early stationary phase, indicating that this ORF may be activated by stationary phase associated regulators. The LEE ORFs as a group were found to be significantly enriched in the 3 h log-phase sample compared to the 5 h, 5.5 h, and 6 h stationary phase samples (Additional File 4).

Figure 5

Heatmap of gene expression over time for the ORFs of the LEE Pathogenecity Island. Expression values determined from the ANOVA analysis are represented colorimetrically, with dark red representing expression = 1.8 and dark green representing expression = -1.8 on a log2 scale. ORFs marked with an asterisk denotes that a statistically significant change in expression levels of that ORF was detected for at least one time point.

Based on the array data, a significant change in expression could not be detected for espA or eae as a result of the high variability in signal for probes targeting those ORFs. Because espB, espD and tir significantly decreased in expression, we used Q-PCR to determine if expression levels changed for espA and eae. Q-PCR confirmed the array data for espB, and detected a 32-fold decrease in espA and a 16-fold decrease in eae transcript levels from 4.6 to 5.5 h (Fig. 6A). Overall, the Q-PCR data detected greater amounts of decrease in espA, espB (22- to 35-fold), tir, and eae (7- to 16-fold) expression compared to the array data (3.2-fold decrease for tir and eae, 5- to 7-fold decrease for espA and espB) for the 4.6 to 5.5 h interval. Transcript levels for espZ were estimated to increase 3.2-fold for the 4.6 to 5.5 h interval from the array data. Q-PCR for espZ found only a slight change (~1.2-fold) in expression over the time intervals tested (Fig. 6A). Whereas the Q-PCR measurements did not confirm the increase in expression determined from the array data for espZ, it is clear that espZ expression does not follow the same pattern as other genes on the LEE island.

Figure 6

Relative changes in expression for virulence-associated genes from the LEE Pathogenecity Island. A. Expression differences of 5 genes transported through or adherence components of the LEE island as determined by Q-PCR. B. Expression differences of 3 genes encoding effectors (Esp) molecules transported through the LEE Type III Secretion System (TTSS) as determined by Q-PCR results. Average changes in expression for 2 independent replicate cultures were determined relative to the transcript level of each ORF at the 3 h mid-exponential phase time point.

Expression changes of LEE effectors

More than 60 putative effectors that are translocated through the LEE-encoded TTSS were recently identified [53] and of these TTSS effectors, 33 were found here to have significant (p < 0.00001) changes in expression over time. A total of 7 TTSS effector genes (nleA, nleG, espM1, espM2, nleG2-2, espZ, and nleG2-3) amplified in expression between mid- and late- exponential phase (3 h to 4 h). At the stationary phase transition point, 11 ORFs increased in expression and 7 ORFs had a significant reduction in expression (espJ, tccP, tir, espF, espB, espM2, and nleD). In early stationary phase, eleven ORFs were upregulated, with the exception of espB (2.5 fold decrease). From late exponential to early stationary phase (4.6 h to 5.5 h), 8 ORFs had an increase in expression > 2- fold, with espY1 increasing 9 fold. Over this same time interval, 5 TTSS effector ORFs had a 2-fold or greater decrease in expression, with espB decreasing 5.3 fold (Table 6). Array data for the TTSS effectors espY1 and espM1 indicated an increase in expression from exponential to stationary phase, which was reflected in the Q-PCR data for these ORFs (Fig. 6B). Expression of the TTSS espJ decreased in stationary phase and a 21 -fold decrease in expression during the transition from log to stationary phase (4.6 to 5 h) was determined by Q-PCR (Fig. 6B).

Table 6 Significant changes in transcript levels of Type III secretion system (TTSS) effectors.

Significant increase of genes from the acid fitness region in stationary phase

A cluster of 12 genes located in the E. coli K-12 genome at position 3652706 to 3665603 bp has recently been termed the acid fitness island (AFI) [54]. In E. coli O157:H7 Sakai genome, the homologous region contains a 9 kb insertion between yhiF (ECs4378) and yhiD (ECs4388). This AFI genomic region contains multiple transcriptional regulators that control expression of the glutamate dependent acid resistance (GDAR) system, such as GadE, GadX, and GadW, that are also known to influence LEE expression [55, 56]. The microarray comparisons demonstrate that the expression of the AFI genes increased significantly from exponential to stationary phase, and included some of the greatest increases in transcript level among the whole genome (Fig. 7). Transcript levels of gadA increased 111 fold and levels of gadE increased 85 fold from mid-exponential (3 h) to stationary phase (6 h). Over half of the AFI genes (7/12) had a significant increase in expression from mid- to late- exponential phase (Fig. 7). Transcript levels of gadA, gadE, and slp also increased significantly from 4.3 to 4.6 h, which is different from the expression profiles of the other known stationary phase activated genes that increase from 4.6–5 h. GSEA found the AFI ORFs as a group to be significantly enriched in all of the time points subsequent to the mid-exponential phase sample (Additional File 4).

Figure 7

Expression heat map of 12 genes from the acid fitness region (AFI). Expression values determined from the ANOVA analysis are represented colorimetrically, with dark red representing expression = 3 and dark green representing expression = -4 on a log2 scale. Over half of the AFI genes (7/12) had a significant increase in expression from mid- to late- exponential phase

Increased expression of Shiga toxin genes upon entry into stationary phase

More than 50% of Sp 15 ORFs and 60% of Sp 5 ORFs had significant (p < 0.00001) changes in expression over time into stationary growth phase. The genes encoding Shiga toxin 1, stx1A and stx1B, as well as those encoding Shiga toxin 2, stx2A and stx2B, had a significant difference in expression between mid- and late- log phase (3 to 4 h). Stx1 genes decreased ~1.8 fold over this interval, while Stx2 genes increased ~1.7 fold. Stx1 genes increased significantly in early stationary phase, with a ~4- fold increase from 5 to 6 h. A significant 2.5- fold increase for stx2B occurred at the transition point into stationary phase, but a concomitant increase was not observed for the stx2A ORF on the microarray. Q-PCR was used to monitor expression of stx1A, stx2A, and stx2B over time (Fig. 8). Expression of stx1A decreased 4-fold from mid-exponential phase to the stationary phase transition point, and then increased 2-fold during early stationary phase (5 to 6 h interval), which is similar to the array data. A 2-fold increase in expression from mid- to late- log phase was observed for stx2B using Q-PCR. While stx2B transcript levels remained higher in stationary phase compared to exponential phase, a subsequent increase during the stationary phase transition was not observed. Transcript levels of stx2A increased 1.5-fold from mid- to late- exponential phase, and decreased 1.5-fold from late exponential to stationary phase (Fig. 8). The ORFs encoding the phage replication factors O and P, as well as the CII regulatory protein and the Q antiterminator protein, increased in expression upon entry into stationary phase, but it is difficult to attribute these ORFs to a specific phage, as there are multiple copies of these ORFs throughout the genome.

Figure 8

Relative changes in expression of Shiga toxin genes. Changes in expression of the phage-encoded toxin genes stx1A, stx2A, and stx2B as determined by Q-PCR. Average changes in expression for 2 independent replicate cultures were determined relative to the transcript level of each ORF at the 3 h mid-exponential phase time point.

Expression changes in tellurite resistance genes and urease genes

Proteins involved in tellurite resistance are encoded on a pathogenicity island (TAI) along with genes encoding urease and are present in a single copy in E. coli O157:H7 Sakai [16]. Significant changes in expression over time were detected only for terZ, terD, and terE (Fig. 9A). Probes targeting terB and terC are not present on the array. Expression of these three genes increased significantly between mid- and late- exponential phase (~1.5 fold) and then decreased 5.9- to 7.4- fold during the transition into stationary phase and in early stationary phase (4.6 h to 5.5 h interval). These three ORFs were grouped into QT cluster 1. The ureD, ureA, ureC, ureE, and ureG ORFs had a significant reduction in expression from late exponential to early stationary phase (1.5 to 3 fold) and were grouped into QT cluster 1 (Fig. 9B). Also located on this genomic island is the adherence factor iha [57], but there is no significant change in iha mRNA over time. Quantitative real-time PCR targeting terZ and ureD confirmed the reduced expression of these ORFs from exponential to stationary phase, detecting an 8-fold decrease for terZ and ureD (Fig. 9C).

Figure 9

Expression profiles of 12 ORFs encoded on the tellurite resistance island (TAI). Expression values determined from the ANOVA analysis are plotted over time for the A. tellurite resistance and B. the urease-encoding ORFs. C. Average relative expression of terZ and ureD was determined by Q-PCR of two independent replicate cultures.


By monitoring global transcript levels over time, we have measured the numerous alterations in gene expression patterns of E. coli O157:H7 when growing exponentially and then entering stationary phase in minimal medium. E. coli O157:H7 responded to low O2 by increasing expression of genes encoding anaerobic electron acceptors and decreasing expression of genes encoding the TCA cycle. O157-specific ORFs, including nleA and nleG, which encode TTS effectors, were upregulated during this interval. As cells entered stationary phase, ~50% of the ORFs in the genome responded with significant changes in transcript level. The global trends include down regulation of some of virulence factors, such as the components of the TTSS and some of the TTSS effectors, in stationary phase. Other TTSS effectors and the genes encoding Shiga toxin increased in transcript level upon entry to stationary phase. Entry into stationary phase was characterized by the increase in expression of known stress and survival related genes, such as the osmotic stress proteins encoded by osmB and osmC, and the glutamate dependent acid resistance system encoded by gadA and gadBC.

As E. coli O157:H7 cells grew in the minimal medium, O2 levels in the medium decreased and were not replenished from the atmosphere (Fig. 4). The decrease in available oxygen was reflected in the transcriptome, with significant increases in genes encoding anaerobic electron acceptors, decrease in transcript levels of TCA cycle genes, and increase in expression of pyruvate formate lyase and concomitant decrease in expression of the pyruvate dehydrogenase complex. Decreased expression of cyo and sdh operons, with concurrent increased expression of the cyd operon was also observed during anaerobic growth of E. coli K-12 at pH 4.7 compared to aerobic growth at acidic pH [28] (Additional File 1). 5% (73/1239) of O157-specific ORFs also had significant changes in expression over the same time interval, which leads one to speculate that these O157-specific ORFs may be regulated by transcription factors that influence expression of respiratory genes, such as FNR and ArcA. Virulence factor expression in enteric pathogens, such as Salmonella enterica serovar Typhimurium and Vibrio cholerae, can be controlled by the redox state regulators FNR and ArcA [5860]. A search for the FNR and ArcA binding sequences 500 bp upstream of all ORFs in the Sakai genome (using Pattern Search on coliBASE [61]) did not identify any of the ORFs that were significantly expressed or repressed in exponential phase. It may be that these O157-specific ORFs are regulated indirectly through FNR or ArcA, or are modulated in a completely different manner. Recently Ando et al. [62] demonstrated that anaerobic growth in the presence of the electron acceptors nitrate or TMAO accelerates maturation of the TTSS, independently of new protein synthesis. They found that the narGHIJ operon must be present for the maturation of the TTSS. Our findings indicate that the expression of this operon increases significantly in late exponential phase, which is consistent with the narGHIJ operon contribution to the maturation of the TTSS in this environment.

Much is known about regulation of LEE expression; multiple circuits of activation and repression have been identified. LEE expression can be induced in response to a number of environmental stimuli, and while common transcriptional regulators and signaling systems can activate LEE expression, they all typically act on Ler or the recently described GrlA and GrlR regulators located on the LEE island [63], which then alter expression of other LEE genes. As LEE expression is thought to be maximal under conditions similar to the intestinal environment [11], we would expect to see increased expression of LEE under these conditions compared to the minimal medium used here. We did observe similar growth-phase dependent changes in expression in the LEE genes that has also been observed for O157:H7 grown in rich medium. LEE expression is known to be maximal in late exponential phase [12, 64]. Nakanishi et al. determined that LEE can be stimulated during nutrient downshifts through activation of ppGpp and DksA [10]. The pattern of espB expression observed during induction of LEE with ppGpp is similar to our observation, in that espB and espA expression increases from mid- to late-exponential phase, and then decreases in stationary phase (Fig. 5 and Fig. 6A). LEE transcription can also be stimulated by quorum sensing, via the auto-inducer 3 signaling system that activates ler [12]. The quorum sensing activated regulator QseA which activates LEE transcription [65] had a decrease in transcript levels from exponential to stationary phase (Fig. 3; QTC 1). Many of the common E. coli regulators that have been shown to repress LEE expression, such as Hha [66], SdiA [67], GadE [56], YhiF [56], and IHF [68, 69], all increased in mRNA levels from exponential to stationary phase (Table 2, Fig. 3; QTC 2). Other LEE activators, such as PchC [48], had a significant increase in mRNA in stationary phase.

The decrease in expression of LEE genes from exponential to stationary phase that we observed is similar to the expression changes of LEE genes when O157:H7 is attached to red blood cells [34]. In comparing our dataset to the genes that were found to be significantly up- or down-regulated in O157:H7 attached to red blood cells, we found that of the 299 ORFs reported to be down regulated in attached cells, 185 had significant changes in expression over time, and 73% of these ORFs were classified in QT cluster 1 (Fig. 3), which includes the LEE genes. Of the 105 ORFs reported to be upregulated in attached cells, 89 had significant changes in expression over time, and 56% of these significant ORFs were placed in QT cluster 2 (Fig. 3), suggesting that many of the ORFs expressed or repressed during attachment to plasma membranes are also regulated by growth phase.

Not as much known about regulation of non-LEE encoded TTSS effector genes. The presence of some of the TTSS effectors in the genome is variable among O157 strains [70]. The TTSS effectors tccP and espJ are located on Sakai phage 14 (EDL-933 OI#79); tccP (espFu) is necessary for actin recruitment and A/E lesion formation in E. coli O157:H7, while espJ influences colonization dynamics [7173]. Reading and colleagues recently determined that the AI-3, epinephrine, and norepinephrine response regulator QseF (yfhA) is an indirect transcriptional activator of tccP [74]. Our data indicate that as Sakai O157:H7 entered stationary phase, qseF mRNA levels decrease significantly (2 fold) (QTC 1) and tccP-espJ levels decreased significantly as well. This decrease in tccP and espJ transcript levels in stationary phase is in contrast to a previous study, where the activity of the tccP-espJ promoter was found to be the same in exponential vs. stationary phase cultures of EDL-933 in Dulbecco's Modified Eagle Medium as monitored by a gfp fusion [75]. Some of the TTSS effectors appear to respond to low O2 conditions, as 7 effectors increased significantly from mid- to late-exponential phase when oxygen levels decreased. Of the TTSS effectors with significant changes in expression over time, ~50% have decreased expression from exponential to stationary phase, while the other half have increased expression in stationary phase. Several of these TTSS effectors are predicted to be pseudogenes; one example is espL3, an ORF that was found to have a significant increase in expression at the stationary phase transition point. Many of the TTSS effectors with increased expression levels in stationary phase are members of the NleG family (Table 6), one of the largest families of TTS effectors that contain many duplications [53]. The utility of the increased expression of some of the TTS effectors in stationary phase is unclear and warrants further investigation.

The ability to resist low pH is a crucial adaptation and critical component of the low infectious dose of E. coli O157:H7. The GDAR system provides the greatest protection against low pH, allowing survival at levels of gastric acidify (e.g. pH 2.0) as long as glutamate is present in the environment [76]. Genes encoding the glutamate decarboxylase isozymes and the glutamate-gamma-aminobutyric acid antiporter are under tight control and can be activated via a number of environmental cues and transcriptional regulators [77], and many components of the system are located in the AFI region [54]. Also located on the AFI are hdeA, slp, and yhiF, which encode proteins involved in protecting cells from their metabolic by-products when the medium is acidified, and yhiD and hdeD, which are required for density-dependent acid resistance [78]. ORFs of the AFI are induced by a variety of stressful environments, including growth in the presence of acetate [79], anaerobic, low pH growth [28], during attachment to red blood cells [34], and in stationary phase [23]. Induction of the AFI ORFs as a response to multiple types of stresses highlights the importance of acid stress survival to the route of transmission for E. coli. Transcription factors located on the AFI, such as GadE and GadX, regulate a large number (>40) of ORFs outside of the GAD system in E. coli K-12 [54, 80], and it is known that these factors serve as repressors of LEE transcription [55, 56], but it is unknown what the effect of these regulators are on a genome-wide scale in E. coli O157:H7.

Chromosomal Toxin-antitoxin (TA) loci encode TA systems which are thought to induce reversible stasis, allowing survival during episodes of extreme nutritional stress [81, 82]. Increased expression of TA loci in E. coli K-12 during growth transitions led to the development of a model integrating inhibition of all macromolecular synthesis within the stringent response model [23], while recent findings suggest that the chromosomal TA systems do not significantly influence bacterial fitness during nutrient limitation [83]. E. coli O157:H7 has some of the same TA loci as E. coli K-12; dinJ-yafQ [84], chpAR, chpBS [85], as well as the recently described chromosomal ccdAB system that is present in E. coli O157:H7 and not in E. coli K-12 [49]. CcdB encodes a DNA gyrase toxin [49], chpB encodes a translation inhibitor [86], and yafQ encodes an mRNA endoribonuclease [84]. Here we observed increased expression of the dinJ-yafQ, chpBS, and ccdAB TA loci upon entry to stationary phase, indicating that these systems are actively transcribed in stationary phase, but their utility remains to be elucidated.

Important questions remain about the factors regulating the Shiga toxins which are encoded by lambda-like bacteriophages [7]. These phages can be transferred between many E. coli strains in nature and were the principal virulence factors acquired as a key step in the evolution of E. coli O157:H7 [87]. How they become influenced in expression and regulated when acquired into a new genome is not fully understood. Early work described two types of Shiga toxin [88] encoded by Stx1 and Stx2, both located within the late operons of the Stx-encoding phages [89, 90]. Stx1 expression is regulated in part by the iron-dependent transcriptional repressor, Fur [15]. Exposure of E. coli O157:H7 to low iron conditions result in an increase of stx1 transcription [14]. Our growth experiments indicate that the stx1 genes decrease in expression from mid- to late-exponential phase, at the same interval when O2 levels decrease (Fig. 8). Iron becomes more soluble as O2 concentration decreases, which could lead to the decreased stx1 expression. Stx1 expression can be induced at 2 other promoters in addition to the Fur binding site, those that require the phage anti-terminator Q [91] and the phage anti-terminator N [14]. Stx2 transcription is initiated at the late phage promoter and is dependent on the anti-repressor Q protein [13]. The production of Stx2 is part of the lytic cycle of the bacteriophage, and can be induced by antimicrobials [92, 93]. Herold et al. found that phage genes were induced upon exposure of E. coli O157:H7 EDL-933 to norfloxacin; phage late genes and stx2 were induced 150-fold [35]. Levels of stx2 mRNA are higher in starved O157:H7 compared to exponentially growing cells [94], indicating an overall trend of stx2 expression induction by stress. Here we observed that stx1 expression increased in stationary phase and stx2 expression was higher in stationary phase compared to exponential phase (Fig. 8). Herold et al. also reported that stx2A is transcribed more efficiently than stx2B in O157:H7 EDL-933, and may be regulated at the posttranscriptional level [35]. This difference in transcription efficiency between stx2A and stx2B was not reflected in our data, as stx2A levels did not appear to increase as the stx2B levels did in stationary phase.

The genomic island termed the Tellurite and Adherence Island (TAI) contains genes encoding tellurite resistance, urease, and an adherence factor, iha [57]. The urease operon and the tellurite genes are thought to be ubiquitously distributed in EHEC strains [95, 96] but is not present in O157:NM or O55:H7 strains [97]. Urease activity has been detected in relatively few EHEC strains. In O157:H7 Sakai, the urease genes have been detected at the transcript level, but a functional protein is not produced due to a stop codon present in ureD [98]. Saridakis et al. surveyed 25 O157:H7 strains that were isolated from humans, cattle, and pigs for ureC and the ability to resist low pH using the urease enzyme. UreC was detected in all of the isolates but there was no evidence for urease-mediated acid resistance [99]. Here we see that the genes encoding urease are expressed during exponential growth and transcript levels decrease significantly as cells enter stationary phase. The genes encoding tellurite resistance are functional in EHEC, and can be separated into two classes – those that are expressed constitutively during growth (terDEZ) and those that are expressed only during growth in the presence of tellurite (terBCF) [96]. Our data show this as well, with terD, E, and Z all expressed during exponential growth, with a significant decrease in transcript levels upon entry to stationary phase, and no significant changes in transcript levels for terA and F.

As E. coli O157:H7 cells transitioned to stationary phase, we observed a transient increase in expression of ORFs encoding proteins involved in acquisition and utilization of alternative carbon sources. ORFs involved in acetate uptake and utilization increased in expression, and some of these genes, such as acs, are known to be controlled, in part, by RpoS [46]. RpoS is activated during slow growth on glucose [52], and growth on less desirable carbon sources [100]. RpoS is also activated during the switch from one carbon source to another, as observed in E. coli K-12 growing in a rich medium [33] and in minimal medium [23]. The glyoxalate shunt is thought to play an important role under glucose-limited conditions, where the absence of catabolite repression is an important factor for activity of the cycle [101]; here we observed that ORFs encoding glyoxalate shunt enzymes increased in expression as E. coli O157:H7 entered stationary phase (Tables 3 and 4). Over expression of high affinity sugar transporters occurs during the short-term adaptation of E. coli K-12 to carbon-limited growth [52, 102]. Sugar transporters and carbon uptake systems are also activated in E. coli K-12 during growth on less desirable carbon sources [100]. In E. coli O157:H7, we observed increased expression of ORFs encoding transport systems for carbohydrates, sugar alcohols and acids during the transition to stationary phase. These data suggest that as glucose was depleted, other carbon sources were sought, but were not available in the minimal medium. In a rich medium, we would expect cells to resume growth, albeit at a lower rate, once glucose was consumed.

Transition to stationary phase is characterized by complex physiological changes to the bacterial cell. Many of the changes in expression patterns observed for O157:H7 in this study have also been observed for E. coli K-12 entering stationary phase [23]. Growth under stressful conditions, such as in the presence of acetate [79] or at low pH under anaerobic conditions [28], also induce expression of stationary phase associated genes. The sigma factor RpoS plays a critical role in transcribing genes associated with stationary phase and stress response [2426], many of these genes, such as bolA, dksA, osmC, and osmB, have been identified in this study as well. We still do not know about most of the O157-specific ORFs, and what regulators control their expression. Here we have classified some of the well-known O157 genes, as well as O157 genes of unknown function, as being modulated during growth transitions. Global transcriptional profiling of isogenic mutants for a specific regulator have been conducted for a number of regulators in E. coli K-12. As we know more about the regulation of virulence loci, such as the LEE, by global regulators, it would be very useful to perform these profiling experiments in E. coli O157:H7 and other pathogenic E. coli.


The principal contribution of this study is the first complete description of global transcriptome profiling of the human pathogenic strain, E. coli O157:H7, during growth transitions. Many O157-specific ORFs responded to the growth transition with significant changes in gene expression including multiple genes that are encoded on pathogenicity islands or toxin-converting bacteriophages. Because a majority of the O157-specific ORFs have unknown functions, these findings will have practical use for comparing transcriptional responses to a variety of growth conditions and elucidating the function and factors regulating these unknown ORFs, some of which may contribute to disease processes. These results also generate new hypotheses and provide opportunities to identify ORFs of interest, based on expression profiles, and compare expression within E. coli O157:H7 populations or with other EHEC populations. It is useful to understand what transcriptional changes occur from exponential to stationary phase, as these transitions are important adaptations to the growth and survival of E. coli O157:H7 in natural conditions.


Growth conditions

E. coli O157:H7 RIMD0509952 (Sakai) involved in a radish sprout outbreak [103] was stored at -70°C in LB broth and 10% glycerol. The Sakai strain was inoculated into 10 mL of LB broth from freezer stocks and grown to OD600 = 0.1. This ~4 h period of growth in LB allowed cells to recover before transfer to MOPS minimal media. MOPS minimal media (10×) was prepared as described in Neidhardt et al. [104] with the addition of 20 mL of micronutrient solution instead of 10 mL. MOPS minimal media contained 100 mL of 10× MOPS minimal media, 10 mL of 0.132 M K2HPO4, and 5 mL of 20% D-glucose per liter. Cultures were inoculated into 50 mL of MOPS minimal media with 0.1% glucose at a ratio of 1:200 and grown to stationary phase at 37°C with shaking. Cultures in MOPS were transferred to 100 mL MOPS in 250 mL plastic culture flasks (Nalgene, Rochester, N.Y.) at a ratio of 1:75 and grown for 9 h before transfer again to 100 mL MOPS in 250 mL plastic culture flasks at a ratio of 1:30 and this culture was used to sample RNA during growth (Fig. 1). Four independent cultures were sampled at 3 h (OD600 ≈ 0.22), 4 h (OD600 ≈ 0.45), 4.33 h (OD600 ≈ 0.73), 4.66 h (OD600 ≈ 0.93), 5 h (OD600 ≈ 1), 5.5 h, 6 h, 7 h, 7.5 h, and 8 h. After 5 h, the culture density remained at OD600 ≈ 1. Residual dissolved oxygen tension was monitored using a dO2 probe and Consort C535 multimeter (Topac Instrumentation, Hingham, Mass.) for three independent cultures in MOPS minimal medium.

RNA isolation

At each time point, 4 mL of culture was mixed with 8 mL of RNAProtect (Qiagen, Valencia, Calif.), vortexed, and centrifuged at 4°C, 7500 rpm for 10 min to pellet cells. The supernatant was removed and cell pellets stored at -70°C for 1 week or less before RNA extraction. Cell pellets were suspended in 700 uL of 95C lysis buffer (20 mM sodium acetate pH 5.2, 2 mM EDTA pH 8.0, and 0.5% SDS), held at 95°C for 30s then mixed with 700 uL of 65°C Acid-Phenol: Chloroform, pH 4.5 (with IAA, 125:25:1) (Ambion, Austin, Tex.). Samples were held at 65°C with periodic shaking for at least 6 minutes before centrifuging at 12000 rpm for 10 min. Supernatant was extracted again with acid-phenol:chloroform and then with chloroform:isoamyl alcohol (24.1). RNA was precipitated for at least 1 hour at -20°C in 2.5 V 100% ethanol and 1/10 V 3 M sodium acetate pH 5.2. RNA was suspended in 1 mM sodium citrate pH 6.5 and stored at -70°C. RNA samples were purified and treated with Dnase using the Rneasy kit (Qiagen). RNA quality was assessed by electrophoresis on formaldehyde-agarose gel. RNA was checked for complete DNA digestion by PCR.

cDNA synthesis and hybridizations

Reverse transcription reactions contained 6 ug RNA, 2 ug random primers (Invitrogen, Carlsbad, Calif.), 1× first strand buffer (Invitrogen), 10 mM DTT, 400 U Superscript II (Invitrogen), 0.5 mM each dATP, dCTP, and dGTP, 0.3 mM dTTP, and 0.2 mM amino-allyl dUTP. 30 uL reactions were incubated overnight at 42°C. cDNA was purified using PCR cleanup columns (Qiagen) with Phosphate wash buffer (5 mM K2HPO4, pH 8.0, 80%EtOH) and phosphate elution buffer (4 mM K2HPO4, pH 8.5). Amino-allyl labeled cDNA was dried and suspended in 0.1 M sodium carbonate pH 9.3 and coupled with either Cy3 or Cy5 (Amersham Biosciences, Piscataway, N.J.). Uncoupled dye was removed by another purfication using the PCR cleanup kit. Concentration of cDNA and amount of incoporated dye was measured for each sample using a Nanodrop spectrophotometer (Ambion).

Oligo microarray

The E. coli oligo set version 1.0.1 (Operon) was printed onto Corning UltraGaps (Corning Incorporated, Acton, Mass.) coated slides at the Research Technology Support Facility at Michigan State University. The Qiagen oligo set contained 5,978 probes specific for three E. coli strains, K-12 (MG1655) [105], O157:H7 Sakai [16], and O157:H7 EDL-933 [17], in which 5,943 probes were 70-mer oligonucleotides, and 35 probes had lengths that ranged from 41–69 bp. All of these probes are spotted in duplicate on each array. The oligo set also contained 12 randomized negative control 70-mer oligonucleotides. There are a common set of 3807 probes that target 3807 ORFs present based on the genome sequences of E. coli K-12 MG1655, E. coli O157:H7 EDL-933, and O157:H7 Sakai, which we refer to as the backbone ORFs. There are also 1741 probes targeting ORFs that are specific to one or two of the genomes, typically to O157:H7 EDL-933 and Sakai. These probes are referred to as the O157-specific ORFs. Probes designed to target K-12-specific or EDL-933 specific ORFs were not considered after the data analysis stage because these genes are not present in the Sakai genome. We used the Qiagen annotation for the probe set, and analyzed only the 4886 probes that were assigned a ECs number.

Hybridization conditions

Arrays were cross-linked by exposure to 600 mJ UV before blocking in 1% SDS, 5× SSC, and 1 mg/mL BSA at 42°C for 1 hour. After blocking, arrays were washed 2× 5 min in 0.1× SSC and 2× 30 s in H2O. Dried arrays were placed into hybridization cassettes (TeleChem International, Sunnyvale, Calif.) and the cDNA samples were suspended in 10 mM EDTA, denatured at 95°C for 5 min and then mixed with 40 uL of SlideHyb buffer 1 (Ambion) and loaded under a coverslip onto the array. Hybridizations were carried out at 47°C for 16–18 h. After hybridization, arrays were washed in 2× SSC, 0.5% SDS 37C for 5 min, followed by 2× 5 min in 0.1× SSC, 0.1% SDS 37°C, and then 2 × 2.5 min in room temperature 0.1× SSC. Arrays were scanned using an Axon 4000b scanner (Molecular Devices, Sunnyvale, Calif) and images were analyzed using GenePix 6.0 (Molecular Devices). The 3 h sample served as a common reference and was hybridized with all subsequent samples. Array data have been deposited at the NCBI Gene Expression Omnibus (Accession GSE7477).

Quantitative real-time PCR

Expression levels of 14 ORFs determined to be differentially expressed (p < 0.00001) were verified by quantitative real-time PCR (Q-PCR). Primer pairs were designed based on the published reference genome sequence of E. coli O157:H7 strain Sakai using the Primer3 server, and secondary structure scrutinized using the Mfold algorithm [106, 107] (Additional File 4). cDNA was synthesized from 1 ug of total RNA using the iScript Select cDNA synthesis kit (BioRad, Hercules, Calif.) and random hexamers supplied with the kit. Template cDNA was diluted to 10-1 to 10-3 for use in Q-PCR. Q-PCR reactions contained 12.5 uL 2× iQ SYBR green supermix (BioRad), 0.63 uL of each primer (10 uM stock), 9.24 uL H2O, and 2 uL cDNA and were conducted under the following conditions: 2 min at 95C, followed by 40 cycles of 10 sec at 95C then 20 sec at the specific annealing temperature. Q-PCR reactions were performed in triplicate for each cDNA sample tested. Relative expression was determined using the method described by Pfaffl et al. [108] and the fluorescence data from the 16S rRNA target were used for normalization within samples. All samples were then compared to the expression levels of the mid-exponential (3 hour) samples. The average log2 expression and standard deviation from two independent RNA samples are reported for each time point tested.

Data analysis

Raw intensity values for all probes on each array were normalized using pin-tip LOWESS [109] in R v.2.2.1 [110] with the MAANOVA (v. 0.98–8) package [111]. Signals from two replicate probes on each array were averaged and log2 transformation applied. Differences in transcription levels over time were determined using a mixed model ANOVA in R/MAANOVA, where the log transformed intensity data, Y = A (array) + D (dye) + T (time) + S (sample = biological replicate) + E (error). The ANOVA modeling allows for consideration of appropriate error structures for experiments with multiple sources of variation in microarray measurements [112]. The random effects of the model were biological replicate and array effects, whereas the fixed effects were time point and dye effects [113]. The Fs statistic, a shrinkage estimator for gene-specific variance components that makes no assumptions about the distribution of variances across genes, was estimated [40]. Significant changes in expression over time were determined by calculating the p values for the Fs statistic for each gene using 1000 random permutations. The p-values were adjusted to correct for type I error with the Benjamini-Hochberg (B-H) linear step-up correction implemented in R/MAANOVA and a cutoff adjusted p-value of 0.0000001. Pair wise contrasts of time points were estimated by the t-test in R/MAANOVA. Contrast p-values were corrected for multiple testing by using the B-H step-up correction. Probes with significant changes in expression over time were grouped by the time points at which the significant change occurred, as determined by the contrast analysis (adjusted p value < 0.05). QT clustering of significant ORFs was conducted in MeV v. 3.1 [116] with a diameter of 0.4 and a minimum cluster size of 15.

Gene Set Enrichment Analysis (GSEA) [114] was conducted using GSEA v. 2.0 [115] on a ranked list of log2 expression ratios (determined from the ANOVA estimates) for 5 of the time points relative to the mid-exponential time point. The following parameters were used for the analysis: 1000 permutations and exclusion of gene sets with less than 5 members and greater than 700 members. Gene sets were defined based on the E. coli O157:H7 Sakai role categories obtained from the Comprehensive Microbial Resource at The Institute for Genomic Research [116]. The GSEA results are summarized in Additional File 4. All gene sets determined to be enriched with a FDR (q-value) less than 0.25 are reported.


  1. 1.

    Mead PS, Slutsker L, Dietz V, McCaig LF, Bresee JS, Shapiro C, Griffin PM, Tauxe RV: Food-related illness and death in the United States. Emerg Infect Dis. 1999, 5: 607-625.

    PubMed Central  CAS  PubMed  Google Scholar 

  2. 2.

    Frankel G, Phillips AD, Rosenshine I, Dougan G, Kaper JB, Knutton S: Enteropathogenic and enterohaemorrhagic Escherichia coli: more subversive elements. Mol Microbiol. 1998, 30: 911-921. 10.1046/j.1365-2958.1998.01144.x.

    CAS  PubMed  Google Scholar 

  3. 3.

    Vallance BA, Chan C, Robertson ML, Finlay BB: Enteropathogenic and enterohemorrhagic Escherichia coli infections: emerging themes in pathogenesis and prevention. Can J Gastroenterol. 2002, 16: 771-778.

    PubMed  Google Scholar 

  4. 4.

    McDaniel TK, Jarvis KG, Donnenberg MS, Kaper JB: A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens. Proc Natl Acad Sci U S A. 1995, 92: 1664-1668. 10.1073/pnas.92.5.1664.

    PubMed Central  CAS  PubMed  Google Scholar 

  5. 5.

    Paton JC, Paton AW: Pathogenesis and diagnosis of Shiga toxin-producing Escherichia coli infections. Clin Microbiol Rev. 1998, 11: 450-479.

    PubMed Central  CAS  PubMed  Google Scholar 

  6. 6.

    Herold S, Karch H, Schmidt H: Shiga toxin-encoding bacteriophages--genomes in motion. Int J Med Microbiol. 2004, 294: 115-121. 10.1016/j.ijmm.2004.06.023.

    CAS  PubMed  Google Scholar 

  7. 7.

    Schmidt H: Shiga-toxin-converting bacteriophages. Res Microbiol. 2001, 152: 687-695. 10.1016/S0923-2508(01)01249-9.

    CAS  PubMed  Google Scholar 

  8. 8.

    Kaper JB, Nataro JP, Mobley HL: Pathogenic Escherichia coli. Nat Rev Microbiol. 2004, 2: 123-140. 10.1038/nrmicro818.

    CAS  PubMed  Google Scholar 

  9. 9.

    Spears KJ, Roe AJ, Gally DL: A comparison of enteropathogenic and enterohaemorrhagic Escherichia coli pathogenesis. FEMS Microbiol Lett. 2006, 255: 187-202. 10.1111/j.1574-6968.2006.00119.x.

    CAS  PubMed  Google Scholar 

  10. 10.

    Nakanishi N, Abe H, Ogura Y, Hayashi T, Tashiro K, Kuhara S, Sugimoto N, Tobe T: ppGpp with DksA controls gene expression in the locus of enterocyte effacement (LEE) pathogenicity island of enterohaemorrhagic Escherichia coli through activation of two virulence regulatory genes. Mol Microbiol. 2006, 61: 194-205. 10.1111/j.1365-2958.2006.05217.x.

    CAS  PubMed  Google Scholar 

  11. 11.

    Abe H, Tatsuno I, Tobe T, Okutani A, Sasakawa C: Bicarbonate ion stimulates the expression of locus of enterocyte effacement-encoded genes in enterohemorrhagic Escherichia coli O157:H7. Infect Immun. 2002, 70: 3500-3509. 10.1128/IAI.70.7.3500-3509.2002.

    PubMed Central  CAS  PubMed  Google Scholar 

  12. 12.

    Sperandio V, Mellies JL, Nguyen W, Shin S, Kaper JB: Quorum sensing controls expression of the type III secretion gene transcription and protein secretion in enterohemorrhagic and enteropathogenic Escherichia coli. Proc Natl Acad Sci U S A. 1999, 96: 15196-15201. 10.1073/pnas.96.26.15196.

    PubMed Central  CAS  PubMed  Google Scholar 

  13. 13.

    Wagner PL, Neely MN, Zhang X, Acheson DW, Waldor MK, Friedman DI: Role for a phage promoter in Shiga toxin 2 expression from a pathogenic Escherichia coli strain. J Bacteriol. 2001, 183: 2081-2085. 10.1128/JB.183.6.2081-2085.2001.

    PubMed Central  CAS  PubMed  Google Scholar 

  14. 14.

    Wagner PL, Livny J, Neely MN, Acheson DW, Friedman DI, Waldor MK: Bacteriophage control of Shiga toxin 1 production and release by Escherichia coli. Mol Microbiol. 2002, 44: 957-970. 10.1046/j.1365-2958.2002.02950.x.

    CAS  PubMed  Google Scholar 

  15. 15.

    Calderwood SB, Mekalanos JJ: Iron regulation of Shiga-like toxin expression in Escherichia coli is mediated by the fur locus. J Bacteriol. 1987, 169: 4759-4764.

    PubMed Central  CAS  PubMed  Google Scholar 

  16. 16.

    Hayashi T, Makino K, Ohnishi M, Kurokawa K, Ishii K, Yokoyama K, Han CG, Ohtsubo E, Nakayama K, Murata T, Tanaka M, Tobe T, Iida T, Takami H, Honda T, Sasakawa C, Ogasawara N, Yasunaga T, Kuhara S, Shiba T, Hattori M, Shinagawa H: Complete genome sequence of enterohemorrhagic Escherichia coli O157:H7 and genomic comparison with a laboratory strain K-12. DNA Res. 2001, 8: 11-22. 10.1093/dnares/8.1.11.

    CAS  PubMed  Google Scholar 

  17. 17.

    Perna NT, Plunkett G, Burland V, Mau B, Glasner JD, Rose DJ, Mayhew GF, Evans PS, Gregor J, Kirkpatrick HA, Posfai G, Hackett J, Klink S, Boutin A, Shao Y, Miller L, Grotbeck EJ, Davis NW, Lim A, Dimalanta ET, Potamousis KD, Apodaca J, Anantharaman TS, Lin J, Yen G, Schwartz DC, Welch RA, Blattner FR: Genome sequence of enterohaemorrhagic Escherichia coli O157:H7. Nature. 2001, 409: 529-533. 10.1038/35054089.

    CAS  PubMed  Google Scholar 

  18. 18.

    Tobe T, Ando H, Ishikawa H, Abe H, Tashiro K, Hayashi T, Kuhara S, Sugimoto N: Dual regulatory pathways integrating the RcsC-RcsD-RcsB signalling system control enterohaemorrhagic Escherichia coli pathogenicity. Mol Microbiol. 2005, 58: 320-333. 10.1111/j.1365-2958.2005.04828.x.

    CAS  PubMed  Google Scholar 

  19. 19.

    Gordon DM, Bauer S, Johnson JR: The genetic structure of Escherichia coli populations in primary and secondary habitats. Microbiology. 2002, 148: 1513-1522.

    CAS  PubMed  Google Scholar 

  20. 20.

    Small P, Blankenhorn D, Welty D, Zinser E, Slonczewski JL: Acid and base resistance in Escherichia coli and Shigella flexneri: role of rpoS and growth pH. J Bacteriol. 1994, 176: 1729-1737.

    PubMed Central  CAS  PubMed  Google Scholar 

  21. 21.

    Gorden J, Small PL: Acid resistance in enteric bacteria. Infect Immun. 1993, 61: 364-367.

    PubMed Central  CAS  PubMed  Google Scholar 

  22. 22.

    Jenkins DE, Chaisson SA, Matin A: Starvation-induced cross protection against osmotic challenge in Escherichia coli. J Bacteriol. 1990, 172: 2779-2781.

    PubMed Central  CAS  PubMed  Google Scholar 

  23. 23.

    Chang DE, Smalley DJ, Conway T: Gene expression profiling of Escherichia coli growth transitions: an expanded stringent response model. Mol Microbiol. 2002, 45: 289-306. 10.1046/j.1365-2958.2002.03001.x.

    CAS  PubMed  Google Scholar 

  24. 24.

    Lacour S, Landini P: SigmaS-dependent gene expression at the onset of stationary phase in Escherichia coli: function of sigmaS-dependent genes and identification of their promoter sequences. J Bacteriol. 2004, 186: 7186-7195. 10.1128/JB.186.21.7186-7195.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  25. 25.

    Weber H, Polen T, Heuveling J, Wendisch VF, Hengge R: Genome-Wide Analysis of the General Stress Response Network in Escherichia coli: {sigma}S-Dependent Genes, Promoters, and Sigma Factor Selectivity. J Bacteriol. 2005, 187: 1591-1603. 10.1128/JB.187.5.1591-1603.2005.

    PubMed Central  CAS  PubMed  Google Scholar 

  26. 26.

    Patten CL, Kirchhof MG, Schertzberg MR, Morton RA, Schellhorn HE: Microarray analysis of RpoS-mediated gene expression in Escherichia coli K-12. Mol Genet Genomics. 2004, 272: 580-591. 10.1007/s00438-004-1089-2.

    CAS  PubMed  Google Scholar 

  27. 27.

    Tucker DL, Tucker N, Conway T: Gene expression profiling of the pH response in Escherichia coli. J Bacteriol. 2002, 184: 6551-6558. 10.1128/JB.184.23.6551-6558.2002.

    PubMed Central  CAS  PubMed  Google Scholar 

  28. 28.

    Hayes ET, Wilks JC, Sanfilippo P, Yohannes E, Tate DP, Jones BD, Radmacher MD, BonDurant SS, Slonczewski JL: Oxygen limitation modulates pH regulation of catabolism and hydrogenases, multidrug transporters, and envelope composition in Escherichia coli K-12. BMC Microbiol. 2006, 6: 89-10.1186/1471-2180-6-89.

    PubMed Central  PubMed  Google Scholar 

  29. 29.

    Tao H, Bausch C, Richmond C, Blattner FR, Conway T: Functional genomics: expression analysis of Escherichia coli growing on minimal and rich media. J Bacteriol. 1999, 181: 6425-6440.

    PubMed Central  CAS  PubMed  Google Scholar 

  30. 30.

    Salmon KA, Hung SP, Steffen NR, Krupp R, Baldi P, Hatfield GW, Gunsalus RP: Global gene expression profiling in Escherichia coli K12: effects of oxygen availability and ArcA. J Biol Chem. 2005, 280: 15084-15096. 10.1074/jbc.M414030200.

    CAS  PubMed  Google Scholar 

  31. 31.

    Kang Y, Weber KD, Qiu Y, Kiley PJ, Blattner FR: Genome-wide expression analysis indicates that FNR of Escherichia coli K-12 regulates a large number of genes of unknown function. J Bacteriol. 2005, 187: 1135-1160. 10.1128/JB.187.3.1135-1160.2005.

    PubMed Central  CAS  PubMed  Google Scholar 

  32. 32.

    Constantinidou C, Hobman JL, Griffiths L, Patel MD, Penn CW, Cole JA, Overton TW: A reassessment of the FNR regulon and transcriptomic analysis of the effects of nitrate, nitrite, NarXL, and NarQP as Escherichia coli K12 adapts from aerobic to anaerobic growth. J Biol Chem. 2006, 281: 4802-4815. 10.1074/jbc.M512312200.

    CAS  PubMed  Google Scholar 

  33. 33.

    Baev MV, Baev D, Radek AJ, Campbell JW: Growth of Escherichia coli MG1655 on LB medium: determining metabolic strategy with transcriptional microarrays. Appl Microbiol Biotechnol. 2006, 71: 323-328. 10.1007/s00253-006-0392-8.

    CAS  PubMed  Google Scholar 

  34. 34.

    Dahan S, Knutton S, Shaw RK, Crepin VF, Dougan G, Frankel G: Transcriptome of enterohemorrhagic Escherichia coli O157 adhering to eukaryotic plasma membranes. Infect Immun. 2004, 72: 5452-5459. 10.1128/IAI.72.9.5452-5459.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  35. 35.

    Herold S, Siebert J, Huber A, Schmidt H: Global expression of prophage genes in Escherichia coli O157:H7 strain EDL933 in response to norfloxacin. Antimicrob Agents Chemother. 2005, 49: 931-944. 10.1128/AAC.49.3.931-944.2005.

    PubMed Central  CAS  PubMed  Google Scholar 

  36. 36.

    Sperandio V, Torres AG, Giron JA, Kaper JB: Quorum sensing is a global regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7. J Bacteriol. 2001, 183: 5187-5197. 10.1128/JB.183.17.5187-5197.2001.

    PubMed Central  CAS  PubMed  Google Scholar 

  37. 37.

    Yang YH, Speed T: Design issues for cDNA microarray experiments. Nat Rev Genet. 2002, 3: 579-588.

    CAS  PubMed  Google Scholar 

  38. 38.

    Churchill GA: Fundamentals of experimental design for cDNA microarrays. Nat Genet. 2002, 32 Suppl: 490-495. 10.1038/ng1031.

    PubMed  Google Scholar 

  39. 39.

    Kerr MK, Martin M, Churchill GA: Analysis of variance for gene expression microarray data. J Comput Biol. 2000, 7: 819-837. 10.1089/10665270050514954.

    CAS  PubMed  Google Scholar 

  40. 40.

    Cui X, Hwang JT, Qiu J, Blades NJ, Churchill GA: Improved statistical tests for differential gene expression by shrinking variance components estimates. Biostatistics. 2005, 6: 59-75. 10.1093/biostatistics/kxh018.

    PubMed  Google Scholar 

  41. 41.

    Heyer LJ, Kruglyak S, Yooseph S: Exploring expression data: identification and analysis of coexpressed genes. Genome Res. 1999, 9: 1106-1115. 10.1101/gr.9.11.1106.

    PubMed Central  CAS  PubMed  Google Scholar 

  42. 42.

    Liu X, De Wulf P: Probing the ArcA-P modulon of Escherichia coli by whole genome transcriptional analysis and sequence recognition profiling. J Biol Chem. 2004, 279: 12588-12597. 10.1074/jbc.M313454200.

    CAS  PubMed  Google Scholar 

  43. 43.

    Alexeeva S, Hellingwerf KJ, Teixeira de Mattos MJ: Quantitative assessment of oxygen availability: perceived aerobiosis and its effect on flux distribution in the respiratory chain of Escherichia coli. J Bacteriol. 2002, 184: 1402-1406.

    PubMed Central  CAS  PubMed  Google Scholar 

  44. 44.

    Tseng CP, Albrecht J, Gunsalus RP: Effect of microaerophilic cell growth conditions on expression of the aerobic (cyoABCDE and cydAB) and anaerobic (narGHJI, frdABCD, and dmsABC) respiratory pathway genes in Escherichia coli. J Bacteriol. 1996, 178: 1094-1098.

    PubMed Central  CAS  PubMed  Google Scholar 

  45. 45.

    Dziva F, van Diemen PM, Stevens MP, Smith AJ, Wallis TS: Identification of Escherichia coli O157:H7 genes influencing colonization of the bovine gastrointestinal tract using signature-tagged mutagenesis. Microbiology. 2004, 150: 3631-3645. 10.1099/mic.0.27448-0.

    CAS  PubMed  Google Scholar 

  46. 46.

    Shin S, Song SG, Lee DS, Pan JG, Park C: Involvement of iclR and rpoS in the induction of acs, the gene for acetyl coenzyme A synthetase of Escherichia coli K-12. FEMS Microbiol Lett. 1997, 146: 103-108. 10.1111/j.1574-6968.1997.tb10178.x.

    CAS  PubMed  Google Scholar 

  47. 47.

    Low AS, Dziva F, Torres AG, Martinez JL, Rosser T, Naylor S, Spears K, Holden N, Mahajan A, Findlay J, Sales J, Smith DG, Low JC, Stevens MP, Gally DL: Cloning, expression, and characterization of fimbrial operon F9 from enterohemorrhagic Escherichia coli O157:H7. Infect Immun. 2006, 74: 2233-2244. 10.1128/IAI.74.4.2233-2244.2006.

    PubMed Central  CAS  PubMed  Google Scholar 

  48. 48.

    Iyoda S, Watanabe H: Positive effects of multiple pch genes on expression of the locus of enterocyte effacement genes and adherence of enterohaemorrhagic Escherichia coli O157 : H7 to HEp-2 cells. Microbiology. 2004, 150: 2357-2571. 10.1099/mic.0.27100-0.

    CAS  PubMed  Google Scholar 

  49. 49.

    Wilbaux M, Mine N, Guerout AM, Mazel D, Van Melderen L: Functional interactions between coexisting toxin-antitoxin systems of the ccd family in Escherichia coli O157:H7. J Bacteriol. 2007, 189: 2712-2719. 10.1128/JB.01679-06.

    PubMed Central  CAS  PubMed  Google Scholar 

  50. 50.

    Uehara T, Suefuji K, Jaeger T, Mayer C, Park JT: MurQ Etherase is required by Escherichia coli in order to metabolize anhydro-N-acetylmuramic acid obtained either from the environment or from its own cell wall. J Bacteriol. 2006, 188: 1660-1662. 10.1128/JB.188.4.1660-1662.2006.

    PubMed Central  CAS  PubMed  Google Scholar 

  51. 51.

    Wick LM, Quadroni M, Egli T: Short- and long-term changes in proteome composition and kinetic properties in a culture of Escherichia coli during transition from glucose-excess to glucose-limited growth conditions in continuous culture and vice versa. Environ Microbiol. 2001, 3: 588-599. 10.1046/j.1462-2920.2001.00231.x.

    CAS  PubMed  Google Scholar 

  52. 52.

    Franchini AG, Egli T: Global gene expression in Escherichia coli K-12 during short-term and long-term adaptation to glucose-limited continuous culture conditions. Microbiology. 2006, 152: 2111-2127. 10.1099/mic.0.28939-0.

    CAS  PubMed  Google Scholar 

  53. 53.

    Tobe T, Beatson SA, Taniguchi H, Abe H, Bailey CM, Fivian A, Younis R, Matthews S, Marches O, Frankel G, Hayashi T, Pallen MJ: An extensive repertoire of type III secretion effectors in Escherichia coli O157 and the role of lambdoid phages in their dissemination. Proc Natl Acad Sci U S A. 2006, 103: 14941-14946. 10.1073/pnas.0604891103.

    PubMed Central  CAS  PubMed  Google Scholar 

  54. 54.

    Hommais F, Krin E, Coppee JY, Lacroix C, Yeramian E, Danchin A, Bertin P: GadE (YhiE): a novel activator involved in the response to acid environment in Escherichia coli. Microbiology. 2004, 150: 61-72. 10.1099/mic.0.26659-0.

    CAS  PubMed  Google Scholar 

  55. 55.

    Shin S, Castanie-Cornet MP, Foster JW, Crawford JA, Brinkley C, Kaper JB: An activator of glutamate decarboxylase genes regulates the expression of enteropathogenic Escherichia coli virulence genes through control of the plasmid-encoded regulator, Per. Mol Microbiol. 2001, 41: 1133-1150. 10.1046/j.1365-2958.2001.02570.x.

    CAS  PubMed  Google Scholar 

  56. 56.

    Tatsuno I, Nagano K, Taguchi K, Rong L, Mori H, Sasakawa C: Increased adherence to Caco-2 cells caused by disruption of the yhiE and yhiF genes in enterohemorrhagic Escherichia coli O157:H7. Infect Immun. 2003, 71: 2598-2606. 10.1128/IAI.71.5.2598-2606.2003.

    PubMed Central  CAS  PubMed  Google Scholar 

  57. 57.

    Tarr PI, Bilge SS, Vary JC, Jelacic S, Habeeb RL, Ward TR, Baylor MR, Besser TE: Iha: a novel Escherichia coli O157:H7 adherence-conferring molecule encoded on a recently acquired chromosomal island of conserved structure. Infect Immun. 2000, 68: 1400-1407. 10.1128/IAI.68.3.1400-1407.2000.

    PubMed Central  CAS  PubMed  Google Scholar 

  58. 58.

    Krishnan HH, Ghosh A, Paul K, Chowdhury R: Effect of anaerobiosis on expression of virulence factors in Vibrio cholerae. Infect Immun. 2004, 72: 3961-3967. 10.1128/IAI.72.7.3961-3967.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  59. 59.

    Sengupta N, Paul K, Chowdhury R: The global regulator ArcA modulates expression of virulence factors in Vibrio cholerae. Infect Immun. 2003, 71: 5583-5589. 10.1128/IAI.71.10.5583-5589.2003.

    PubMed Central  CAS  PubMed  Google Scholar 

  60. 60.

    Fink RC, Evans MR, Porwollik S, Vazquez-Torres A, Jones-Carson J, Troxell B, Libby SJ, McClelland M, Hassan HM: FNR Is a Global Regulator of Virulence and Anaerobic Metabolism in Salmonella enterica Serovar Typhimurium (ATCC 14028s). J Bacteriol. 2007, 189: 2262-2273. 10.1128/JB.00726-06.

    PubMed Central  CAS  PubMed  Google Scholar 

  61. 61.

    Chaudhuri RR, Khan AM, Pallen MJ: coliBASE: an online database for Escherichia coli, Shigella and Salmonella comparative genomics. Nucleic Acids Res. 2004, 32: D296-9. 10.1093/nar/gkh031.

    PubMed Central  CAS  PubMed  Google Scholar 

  62. 62.

    Ando H, Abe H, Sugimoto N, Tobe T: Maturation of functional type III secretion machinery by activation of anaerobic respiration in enterohaemorrhagic Escherichia coli. Microbiology. 2007, 153: 464-473. 10.1099/mic.0.2006/000893-0.

    CAS  PubMed  Google Scholar 

  63. 63.

    Deng W, Puente JL, Gruenheid S, Li Y, Vallance BA, Vazquez A, Barba J, Ibarra JA, O'Donnell P, Metalnikov P, Ashman K, Lee S, Goode D, Pawson T, Finlay BB: Dissecting virulence: systematic and functional analyses of a pathogenicity island. Proc Natl Acad Sci U S A. 2004, 101: 3597-3602. 10.1073/pnas.0400326101.

    PubMed Central  CAS  PubMed  Google Scholar 

  64. 64.

    Yona-Nadler C, Umanski T, Aizawa S, Friedberg D, Rosenshine I: Integration host factor (IHF) mediates repression of flagella in enteropathogenic and enterohaemorrhagic Escherichia coli. Microbiology. 2003, 149: 877-884. 10.1099/mic.0.25970-0.

    CAS  PubMed  Google Scholar 

  65. 65.

    Sperandio V, Li CC, Kaper JB: Quorum-sensing Escherichia coli regulator A: a regulator of the LysR family involved in the regulation of the locus of enterocyte effacement pathogenicity island in enterohemorrhagic E. coli. Infect Immun. 2002, 70: 3085-3093. 10.1128/IAI.70.6.3085-3093.2002.

    PubMed Central  CAS  PubMed  Google Scholar 

  66. 66.

    Sharma VK, Zuerner RL: Role of hha and ler in transcriptional regulation of the esp operon of enterohemorrhagic Escherichia coli O157:H7. J Bacteriol. 2004, 186: 7290-7301. 10.1128/JB.186.21.7290-7301.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  67. 67.

    Kanamaru K, Tatsuno I, Tobe T, Sasakawa C: SdiA, an Escherichia coli homologue of quorum-sensing regulators, controls the expression of virulence factors in enterohaemorrhagic Escherichia coli O157:H7. Mol Microbiol. 2000, 38: 805-816. 10.1046/j.1365-2958.2000.02171.x.

    CAS  PubMed  Google Scholar 

  68. 68.

    Li M, Rosenshine I, Tung SL, Wang XH, Friedberg D, Hew CL, Leung KY: Comparative proteomic analysis of extracellular proteins of enterohemorrhagic and enteropathogenic Escherichia coli strains and their ihf and ler mutants. Appl Environ Microbiol. 2004, 70: 5274-5282. 10.1128/AEM.70.9.5274-5282.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  69. 69.

    Friedberg D, Umanski T, Fang Y, Rosenshine I: Hierarchy in the expression of the locus of enterocyte effacement genes of enteropathogenic Escherichia coli. Mol Microbiol. 1999, 34: 941-952. 10.1046/j.1365-2958.1999.01655.x.

    CAS  PubMed  Google Scholar 

  70. 70.

    Ogura Y, Kurokawa K, Ooka T, Tashiro K, Tobe T, Ohnishi M, Nakayama K, Morimoto T, Terajima J, Watanabe H, Kuhara S, Hayashi T: Complexity of the genomic diversity in enterohemorrhagic Escherichia coli O157 revealed by the combinational use of the O157 Sakai OligoDNA microarray and the Whole Genome PCR scanning. DNA Res. 2006, 13: 3-14. 10.1093/dnares/dsi026.

    CAS  PubMed  Google Scholar 

  71. 71.

    Garmendia J, Phillips AD, Carlier MF, Chong Y, Schuller S, Marches O, Dahan S, Oswald E, Shaw RK, Knutton S, Frankel G: TccP is an enterohaemorrhagic Escherichia coli O157:H7 type III effector protein that couples Tir to the actin-cytoskeleton. Cell Microbiol. 2004, 6: 1167-1183. 10.1111/j.1462-5822.2004.00459.x.

    CAS  PubMed  Google Scholar 

  72. 72.

    Campellone KG, Robbins D, Leong JM: EspFU is a translocated EHEC effector that interacts with Tir and N-WASP and promotes Nck-independent actin assembly. Dev Cell. 2004, 7: 217-228. 10.1016/j.devcel.2004.07.004.

    CAS  PubMed  Google Scholar 

  73. 73.

    Dahan S, Wiles S, La Ragione RM, Best A, Woodward MJ, Stevens MP, Shaw RK, Chong Y, Knutton S, Phillips A, Frankel G: EspJ is a prophage-carried type III effector protein of attaching and effacing pathogens that modulates infection dynamics. Infect Immun. 2005, 73: 679-686. 10.1128/IAI.73.2.679-686.2005.

    PubMed Central  CAS  PubMed  Google Scholar 

  74. 74.

    Reading NC, Torres AG, Kendall MM, Hughes DT, Yamamoto K, Sperandio V: A Novel Two-Component Signaling System That Activates Transcription of an Enterohemorrhagic Escherichia coli Effector Involved in Remodeling of Host Actin. J Bacteriol. 2007, 189: 2468-2476. 10.1128/JB.01848-06.

    PubMed Central  CAS  PubMed  Google Scholar 

  75. 75.

    Garmendia J, Frankel G: Operon structure and gene expression of the espJ--tccP locus of enterohaemorrhagic Escherichia coli O157:H7. FEMS Microbiol Lett. 2005, 247: 137-145. 10.1016/j.femsle.2005.04.035.

    CAS  PubMed  Google Scholar 

  76. 76.

    Lin J, Smith MP, Chapin KC, Baik HS, Bennett GN, Foster JW: Mechanisms of acid resistance in enterohemorrhagic Escherichia coli. Appl Environ Microbiol. 1996, 62: 3094-3100.

    PubMed Central  CAS  PubMed  Google Scholar 

  77. 77.

    Foster JW: Escherichia coli acid resistance: tales of an amateur acidophile. Nat Rev Microbiol. 2004, 2: 898-907. 10.1038/nrmicro1021.

    CAS  PubMed  Google Scholar 

  78. 78.

    Mates AK, Sayed AK, Foster JW: Products of the Escherichia coli acid fitness island attenuate metabolite stress at extremely low pH and mediate a cell density-dependent acid resistance. J Bacteriol. 2007, 189: 2759-2768. 10.1128/JB.01490-06.

    PubMed Central  CAS  PubMed  Google Scholar 

  79. 79.

    Arnold CN, McElhanon J, Lee A, Leonhart R, Siegele DA: Global analysis of Escherichia coli gene expression during the acetate- induced acid tolerance response. J Bacteriol. 2001, 183: 2178-2186. 10.1128/JB.183.7.2178-2186.2001.

    PubMed Central  CAS  PubMed  Google Scholar 

  80. 80.

    Ma Z, Masuda N, Foster JW: Characterization of EvgAS-YdeO-GadE branched regulatory circuit governing glutamate-dependent acid resistance in Escherichia coli. J Bacteriol. 2004, 186: 7378-7389. 10.1128/JB.186.21.7378-7389.2004.

    PubMed Central  CAS  PubMed  Google Scholar 

  81. 81.

    Buts L, Lah J, Dao-Thi MH, Wyns L, Loris R: Toxin-antitoxin modules as bacterial metabolic stress managers. Trends Biochem Sci. 2005, 30: 672-679.

    CAS  PubMed  Google Scholar 

  82. 82.

    Gerdes K, Christensen SK, Lobner-Olesen A: Prokaryotic toxin-antitoxin stress response loci. Nat Rev Microbiol. 2005, 3: 371-382. 10.1038/nrmicro1147.

    CAS  PubMed  Google Scholar 

  83. 83.

    Tsilibaris V, Maenhaut-Michel G, Mine N, Van Melderen L: What is the benefit for E. coli to have multiple toxin-antitoxin systems in their genomes?. J Bacteriol. 2007

    Google Scholar 

  84. 84.

    Motiejunaite R, Armalyte J, Markuckas A, Suziedeliene E: Escherichia coli dinJ-yafQ genes act as a toxin-antitoxin module. FEMS Microbiol Lett. 2007, 268: 112-119. 10.1111/j.1574-6968.2006.00563.x.

    CAS  PubMed  Google Scholar 

  85. 85.

    Masuda Y, Miyakawa K, Nishimura Y, Ohtsubo E: chpA and chpB, Escherichia coli chromosomal homologs of the pem locus responsible for stable maintenance of plasmid R100. J Bacteriol. 1993, 175: 6850-6856.

    PubMed Central  CAS  PubMed  Google Scholar 

  86. 86.

    Christensen SK, Pedersen K, Hansen FG, Gerdes K: Toxin-antitoxin loci as stress-response-elements: ChpAK/MazF and ChpBK cleave translated RNAs and are counteracted by tmRNA. J Mol Biol. 2003, 332: 809-819. 10.1016/S0022-2836(03)00922-7.

    CAS  PubMed  Google Scholar 

  87. 87.

    Reid SD, Herbelin CJ, Bumbaugh AC, Selander RK, Whittam TS: Parallel evolution of virulence in pathogenic Escherichia coli. Nature. 2000, 406: 64-67. 10.1038/35017546.

    CAS  PubMed  Google Scholar 

  88. 88.

    O'Brien AD, Newland JW, Miller SF, Holmes RK, Smith HW, Formal SB: Shiga-like toxin-converting phages from Escherichia coli strains that cause hemorrhagic colitis or infantile diarrhea. Science. 1984, 226: 694-696. 10.1126/science.6387911.

    PubMed  Google Scholar 

  89. 89.

    Neely MN, Friedman DI: Arrangement and functional identification of genes in the regulatory region of lambdoid phage H-19B, a carrier of a Shiga-like toxin. Gene. 1998, 223: 105-113. 10.1016/S0378-1119(98)00236-4.

    CAS  PubMed  Google Scholar 

  90. 90.

    Plunkett G, Rose DJ, Durfee TJ, Blattner FR: Sequence of Shiga toxin 2 phage 933W from Escherichia coli O157:H7: Shiga toxin as a phage late-gene product. J Bacteriol. 1999, 181: 1767-1778.

    PubMed Central  CAS  PubMed  Google Scholar 

  91. 91.

    Neely MN, Friedman DI: Functional and genetic analysis of regulatory regions of coliphage H-19B: location of shiga-like toxin and lysis genes suggest a role for phage functions in toxin release. Mol Microbiol. 1998, 28: 1255-1267. 10.1046/j.1365-2958.1998.00890.x.

    CAS  PubMed  Google Scholar 

  92. 92.

    Kimmitt PT, Harwood CR, Barer MR: Toxin gene expression by shiga toxin-producing Escherichia coli: the role of antibiotics and the bacterial SOS response. Emerg Infect Dis. 2000, 6: 458-465.

    PubMed Central  CAS  PubMed  Google Scholar 

  93. 93.

    Azizoglu RO, Drake M: Impact of antibiotic stress on acid and heat tolerance and virulence factor expression of Escherichia coli O157:H7. J Food Prot. 2007, 70: 194-199.

    CAS  PubMed  Google Scholar 

  94. 94.

    Leenanon B, Elhanafi D, Drake MA: Acid adaptation and starvation effects on Shiga toxin production by Escherichia coli O157:H7. J Food Prot. 2003, 66: 970-977.

    CAS  PubMed  Google Scholar 

  95. 95.

    Nakano M, Iida T, Ohnishi M, Kurokawa K, Takahashi A, Tsukamoto T, Yasunaga T, Hayashi T, Honda T: Association of the urease gene with enterohemorrhagic Escherichia coli strains irrespective of their serogroups. J Clin Microbiol. 2001, 39: 4541-4543. 10.1128/JCM.39.12.4541-4543.2001.

    PubMed Central  CAS  PubMed  Google Scholar 

  96. 96.

    Taylor DE, Rooker M, Keelan M, Ng LK, Martin I, Perna NT, Burland NT, Blattner FR: Genomic variability of O islands encoding tellurite resistance in enterohemorrhagic Escherichia coli O157:H7 isolates. J Bacteriol. 2002, 184: 4690-4698. 10.1128/JB.184.17.4690-4698.2002.

    PubMed Central  CAS  PubMed  Google Scholar 

  97. 97.

    Friedrich AW, Kock R, Bielaszewska M, Zhang W, Karch H, Mathys W: Distribution of the urease gene cluster among and urease activities of enterohemorrhagic Escherichia coli O157 isolates from humans. J Clin Microbiol. 2005, 43: 546-550. 10.1128/JCM.43.2.546-550.2005.

    PubMed Central  CAS  PubMed  Google Scholar 

  98. 98.

    Nakano M, Iida T, Honda T: Urease activity of enterohaemorrhagic Escherichia coli depends on a specific one-base substitution in ureD. Microbiology. 2004, 150: 3483-3489. 10.1099/mic.0.27280-0.

    CAS  PubMed  Google Scholar 

  99. 99.

    Saridakis CE, Johnson RP, Benson A, Ziebell K, Gyles CL: Influence of animal origin and lineage on survival of Escherichia coli O157:H7 strains in strong and weak acid challenges. J Food Prot. 2004, 67: 1591-1596.

    PubMed  Google Scholar 

  100. 100.

    Liu M, Durfee T, Cabrera JE, Zhao K, Jin DJ, Blattner FR: Global transcriptional programs reveal a carbon source foraging strategy by Escherichia coli. J Biol Chem. 2005, 280: 15921-15927. 10.1074/jbc.M414050200.

    CAS  PubMed  Google Scholar 

  101. 101.

    Fischer E, Sauer U: A novel metabolic cycle catalyzes glucose oxidation and anaplerosis in hungry Escherichia coli. J Biol Chem. 2003, 278: 46446-46451. 10.1074/jbc.M307968200.

    CAS  PubMed  Google Scholar 

  102. 102.

    Wick LM, Weilenmann H, Egli T: The apparent clock-like evolution of Escherichia coli in glucose-limited chemostats is reproducible at large but not at small population sizes and can be explained with Monod kinetics. Microbiology. 2002, 148: 2889-2902.

    CAS  PubMed  Google Scholar 

  103. 103.

    Michino H, Araki K, Minami S, Takaya S, Sakai N, Miyazaki M, Ono A, Yanagawa H: Massive outbreak of Escherichia coli O157:H7 infection in schoolchildren in Sakai City, Japan, associated with consumption of white radish sprouts. Am J Epidemiol. 1999, 150: 787-796.

    CAS  PubMed  Google Scholar 

  104. 104.

    Neidhardt FC, Bloch PL, Smith DF: Culture medium for enterobacteria. J Bacteriol. 1974, 119: 736-747.

    PubMed Central  CAS  PubMed  Google Scholar 

  105. 105.

    Blattner FR, Plunkett G, Bloch CA, Perna NT, Burland V, Riley M, Collado-Vides J, Glasner JD, Rode CK, Mayhew GF, Gregor J, Davis NW, Kirkpatrick HA, Goeden MA, Rose DJ, Mau B, Shao Y: The complete genome sequence of Escherichia coli K-12. Science. 1997, 277: 1453-1474. 10.1126/science.277.5331.1453.

    CAS  PubMed  Google Scholar 

  106. 106.

    Zuker M: Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 2003, 31: 3406-3415. 10.1093/nar/gkg595.

    PubMed Central  CAS  PubMed  Google Scholar 

  107. 107.

    Rozen S, Skaletsky H: Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol. 2000, 132: 365-386.

    CAS  PubMed  Google Scholar 

  108. 108.

    Pfaffl MW: A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001, 29: e45-10.1093/nar/29.9.e45.

    PubMed Central  CAS  PubMed  Google Scholar 

  109. 109.

    Quackenbush J: Microarray data normalization and transformation. Nat Genet. 2002, 32 Suppl: 496-501. 10.1038/ng1032.

    PubMed  Google Scholar 

  110. 110.

    Team RCD: R: A language and environment for statistical computing. 2005, Vienna, Austria, R Foundation for Statistical Computing

    Google Scholar 

  111. 111.

    R/MAANOVA: . []

  112. 112.

    Kerr MK: Linear models for microarray data analysis: hidden similarities and differences. J Comput Biol. 2003, 10: 891-901. 10.1089/106652703322756131.

    CAS  PubMed  Google Scholar 

  113. 113.

    Cui X, Churchill GA: Statistical tests for differential expression in cDNA microarray experiments. Genome Biol. 2003, 4: 210-10.1186/gb-2003-4-4-210.

    PubMed Central  PubMed  Google Scholar 

  114. 114.

    Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005, 102: 15545-15550. 10.1073/pnas.0506580102.

    PubMed Central  CAS  PubMed  Google Scholar 

  115. 115.

    Gene Set Enrichment Analysis (GSEA) v. 2.0. []

  116. 116.

    Comprehensive Microbial Resource at The Institute for Genomic Research. []

Download references


The authors thank Katherine Schaeffer for assistance with the dO2 measurements and Galeb Abu-Ali, Peter Bergholz, and Sivapriya Kailasan Vanaja for helpful scientific discussions and critical review of previous versions of the manuscript. This work was supported by Food Safety NRI #2005-35201-16362 from the United States Department of Agriculture to TSW.

Author information



Corresponding author

Correspondence to Thomas S Whittam.

Additional information

Authors' contributions

TMB designed and carried out all the microarray experiments, data analysis, and composed the manuscript. LMW and WQ assisted in the experimental design and statistical data analyses. JTR designed and optimized the Q-PCR assays and LMO conducted the experimental assays. TSW oversaw the project, contributed to the design and final analyses, and manuscript completion. All authors have read and approved the final version of the manuscript.

Electronic supplementary material


Additional File 1: Entire data set of O157 ORFs sorted by significance. Log2 expression indices for each time point for 4886 ORFs in the multigenome microarray (XLS 3 MB)

Additional File 2: Significant O157-specific ORFs. Log2 expression indices for each time point (XLS 377 KB)

Additional File 3: QPCR primers. PCR primers designed for QPCR analysis. (DOC 50 KB)


Additional File 4: GSEA results. Gene Set Enrichment Analysis (GSEA) conducted using GSEA v. 2.0 on a ranked list of log2 expression ratios (determined from the ANOVA estimates) for 5 of the time points relative to the mid-exponential time point. (XLS 25 KB)

Authors’ original submitted files for images

Rights and permissions

Open Access This article is published under license to BioMed Central Ltd. This is an Open Access article is distributed under the terms of the Creative Commons Attribution License ( ), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article

Cite this article

Bergholz, T.M., Wick, L.M., Qi, W. et al. Global transcriptional response of Escherichia coli O157:H7 to growth transitions in glucose minimal medium. BMC Microbiol 7, 97 (2007).

Download citation


  • Stationary Phase
  • Shiga Toxin
  • Early Stationary Phase
  • DMSO Reductase
  • Stx1 Gene