High-Density Microarray-Mediated Gene Expression Profiling of Escherichia coli

Abstract

A nearly complete collection of 4,290 Escherichia coli open reading frames was amplified and arrayed in high density on glass slides. To exploit this reagent, conditions for RNA isolation from E. coli cells, cDNA production with attendant fluorescent dye incorporation, DNA-DNA hybridization, and hybrid quantitation have been established. A brief isopropyl-β-d-thiogalactopyranoside (IPTG) treatment elevated lacZ, lacY, and lacA transcript content about 30-fold; in contrast, most other transcript titers remained unchanged. Distinct RNA expression patterns between E. coli cultures in the exponential and transitional phases of growth were catalogued, as were differences associated with culturing in minimal and rich media. The relative abundance of each transcript was estimated by using hybridization of a genomic DNA-derived, fluorescently labeled probe as a correction factor. This inventory provided a quantitative view of the steady-state level of each mRNA species. Genes the expression of which was detected by this method were enumerated, and results were compared with the current understanding of E. coli physiology.

Escherichia coli K-12 has been exhaustively studied for over 50 years. Early experiments measured the molecular fluxes from small compounds into macromolecular constituents (33). These studies were followed by others in which small molecule pools of central metabolic building blocks (21), nucleotides (3), and amino acids were enumerated. The levels of several macromolecular components, including individual species of proteins (26), have been measured. Such measurements of the steady state provide a census of the cellular content, while changes upon imposition of a stress catalogue the cell's fight for survival. This response to an insulting or adverse condition can take many forms, from relieving end product inhibition to derepressing transcription (20).

In E. coli, experiments to define stress-related, global regulatory responses have often relied upon either the isolation of operon fusions induced by a particular stress (16) or proteomic measures in which the protein fractions from stressed and unstressed cultures are separated by a two-dimensional method prior to comparison (37). Each method has an inherent technological hurdle; the map location of responsive gene fusions must be ascertained precisely, while induced or repressed proteins excised from the two-dimensional gels must be correctly identified.

Alternatively, mRNA measurements utilizing techniques such as hybridization to DNA and primer extension have allowed the monitoring of individual gene's expression profiles. Recently, expression profiling of most yeast genes has been reported (8, 40); such measurements were facilitated by high-density arrays of individual genes and specific labeling of cDNA copies of eukaryotic mRNA by using poly(A) tail-specific primers. Thus, the lack of a poly(A) tail and the extremely short bacterial mRNA half-life represent hurdles for the application of DNA microarray technology to prokaryotic research. Nonetheless, early attempts at comprehensive expression profiling using large DNA fragments from an ordered λ library of E. coli genomic fragments as a capture reagent and radiolabeled cDNA as a probe suggested that these problems were not insurmountable (6).

Here we present a means to successfully perform microarray-based comprehensive gene expression profile analyses with E. coli. We show that such experimentation can be informative by examination of (i) differences in gene expression profiles caused by growth of E. coli in either minimal or rich medium, (ii) changes in gene expression associated with the transition from exponential-phase to stationary-phase growth in minimal medium, and (iii) the specificity of induction mediated by isopropyl-β-d-thiogalactopyranoside (IPTG), the classic lac operon inducer. Moreover, a method for determining the relative abundance of each transcript was developed and used to provide a census of the mRNA composition of E. coli under each of the growth conditions mentioned above.

MATERIALS AND METHODS

Microbiological methods.

E. coliMG1655 (1) was cultured with aeration in either the minimal medium, M9 (23), supplemented with 0.4% glucose or in the rich medium, Luria-Bertani (LB) (23), at 37°C. The overnight culture was diluted 250-fold into fresh medium and aerated at 37°C. Samples of the minimal medium culture were harvested at A₆₀₀s of 0.40 (exponential phase; just under five generations) and 1.6 (transition to stationary phase, just less than seven generations) prior to RNA isolation. An IPTG induction (23) was performed to examine the specificity with which it affects gene expression. The LB medium-grown culture was split when it achieved an appropriate density (A₆₀₀ of 0.40). To one portion was added IPTG to a final concentration of 1 mM; the untreated sample served as a control. Incubation of both samples was continued with aeration at 37°C for another 15 min (A₆₀₀ of 0.45 for both cultures) before RNA isolation was initiated.

RNA isolation.

Shaved ice was added to 50-ml samples which were pelleted immediately in a refrigerated centrifuge by spinning at 10,410 × g for 2 min. Each resultant pellet was resuspended in a mixture containing 100 μl of Tris-HCl (10 mM, pH 8.0) and 350 μl of β-mercaptoethanol-supplemented RLT buffer (Qiagen RNeasy Mini kit; Valencia, Calif.) that was kept on ice. The cell suspension was added to a chilled 2-ml microcentrifuge tube containing 100 μl of 0.1-mm-diameter zirconia-silica beads (Blospec Products Inc., Bartlesville, Okla.). The cells were broken by agitation at room temperature for 25 s with a Mini-Beadbeater (Biospec Products, Inc.). Debris was pelleted by centrifugation for 3 min at 16,000 × g and 4°C; the resultant supernatant was mixed with 250 μl of ethanol. This mixture was loaded onto Qiagen RNeasy columns from the Qiagen RNeasy Mini kit. RNA isolation was completed by using the protocol supplied with this kit. Incubation for 1 h at 37°C in 40 mM Tris (pH 8.0), 10 mM NaCl, 6 mM MgCl₂ with RNase-free RQ1 DNase (1 U/μl; Promega, Madison, Wis.) digested any genomic DNA contaminating the RNA preparation. The digestion products were purified by a second passage through an RNeasy column (Qiagen). The product was eluted from the column in 50 μl of RNAse-free water prior to determining sample concentration by an A₂₆₀ reading. RNA preparations were stored frozen at −20°C until use.

Synthesis of fluorescent cDNA from total RNA.

To a volume brought to 22 μl with double-distilled water (ddH₂O) were added 6 μg of total RNA template and 12 μg of random hexamer primers (Operon Technologies, Inc., Alameda, Calif.). Annealing was accomplished by incubation for 10 min at 70°C followed by 10 min at room temperature. cDNA probes were synthesized with SuperScript II reverse transcriptase (10 U/μl; Life Technologies, Inc., Gaithersburg, Md.) in the presence of deoxynucleoside triphosphates (dNTPs) (dATP, dGTP, and dTTP, each at 0.1 mM; dCTP at 50 μM) and Cy3- or Cy5-dCTP at 25 μM. In order were added 8 μl of 5× SuperScript II reaction buffer (Life Technologies, Inc.), 4 μl of 0.1 M dithiothreitol, 2 μl of the dNTP mix (2 mM dATP, 2 mM dGTP, 2 mM TTP, 1 mM dCTP), 2 μl of 0.5 mM Cy3- or Cy5-labeled dCTP (Amersham Pharmacia Biotech, Arlington Heights, Ill.), and 2 μl of SuperScript II reverse transcriptase. cDNA synthesis proceeded at 42°C for 2.5 h before the reaction was terminated by heating at 94°C for 5 min. The RNA templates were hydrolyzed with 0.25 M NaOH. The reaction was then neutralized by adding HCl and Tris-HCl (pH 6.8). The labeled cDNA was purified with a PCR purification kit (Qiagen), dried, and stored at −20°C. Labeling efficiency was calculated by using the A₂₆₀ and either A₅₅₀ for Cy3 incorporation or A₆₅₀ for Cy5 labeling measurements.

Fluorescent copying of genomic DNA.

Genomic DNA was isolated from strain MG1655 by a standard procedure (38). Genomic DNA, sheared with a nebulizer to approximately 2-kbp fragments, was used to prepare labeled DNA. Three micrograms of this DNA was mixed with 6 μg of random hexamer primers (Operon Technologies, Inc.) in 33 μl of ddH₂O. DNA was denatured by heating at 94°C prior to annealing on ice for 10 min. Fluorescent copying of the genomic DNA was accomplished with the Klenow fragment of DNA polymerase I (5 U/μl; Promega, Madison, Wis.). To the DNA mixture was added 6 μl of 10× Klenow buffer (supplied with the enzyme), 3 μl of the dNTP mix described above, 12 μl of ddH₂O, 3 μl of 0.5 mM Cy3-dCTP (Amersham Pharmacia Biotech), and 3 μl of the Klenow fragment of DNA polymerase I. After a static, 2.5-h incubation at room temperature, the labeled DNA probe was purified with a PCR purification kit (Qiagen) before being dried in a speed vacuum.

Amplification of 4,290 E. coli genes.

Our amplification method was based on a previously described protocol (31). Specific primer pairs (Sigma Genosys, The Woodlands, Tex.) for each protein-specifying gene of E. coli were used in two consecutive PCR amplifications. Two amplifications were performed to prevent contaminating genomic DNA within the initial PCR product from being spotted to the microarray. Any such carried-over material was eliminated by the “dilution” associated with the second amplification reaction. Genomic DNA (30 ng) was used as the template in the first round of PCR amplification, and 500-fold-diluted PCR products served as templates for PCR reamplification. Duplicate 50-μl scale reactions were performed in the reamplification. The PCRs were catalyzed with ExTaq polymerase (Panvera, Madison, Wis.) with the four dNTPs (Amersham Pharmacia Biotech) present at 0.2 mM and the primers at 0.5 μM. Twenty-five cycles of denaturation at 95°C for 15 s, annealing at 64°C for 15 s, and polymerization at 72°C for 1 min were conducted. A 2-μl aliquot of each PCR product was sized by electrophoresis through agarose gels. More than 95% of these resultant second PCR products displayed visible bands of the correct size. Second-round PCR mixtures, devoid of templates and primers, were saved to be spotted onto slides to serve as negative controls for hybridization experiments. Each second-round PCR mixture was purified with 96-well PCR purification kits (Qiagen). The eluant was dried with a vacuum centrifuge.

Arraying amplified genes.

Twenty microliters of 6 M Na₂SCN or 50% dimethyl sulfoxide was added to each dried DNA sample (≥0.1-ng/nl final concentration). A generation II DNA spotter (Molecular Dynamics, Sunnyvale, Calif.) was used to array the samples onto coated glass slides (Amersham Pharmacia Biotech). Two aliquots of approximately 1 nl from 1,536 resuspended PCR products were arrayed in duplicate on each slide; three slides were used to order all amplified E. coli genes. To serve as controls, 76 specific E. coli PCR products, 8 amplified genes of Klebsiella pnuemoniae, and 12 plant cDNA clones were also spotted onto each slide. Arrayed glass slides, after baking at 80°C for 2 h, were stored in a desiccator at room temperature under vacuum.

Hybridization and washing.

Arrayed slides were placed in isopropanol for 10 min, boiled in ddH₂O for 5 min, and dried by passage of ultraclean N₂ gas prior to prehybridization. The prehybridization solution (PHS) was 3.5× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) (BRL, Life Technologies Inc., Gaithersberg, Md.), 0.2% sodium dodecyl sulfate (SDS; BRL, Life Technologies, Inc.), 1% bovine serum albumin (fraction V; Sigma, St. Louis, Mo.). The hybridization solution (HS) contained 4 μl of ddH₂O, 7.5 μl of 20× SSC, 2.5 μl of 1% SDS (BRL, Life Technologies Inc.), 1 μl of 10 mg of salmon sperm DNA per ml (Sigma), and 15 μl of formamide (Sigma). The slides were incubated at 60°C for 20 min in PHS to block nonspecific binding of probe. The slides were next rinsed five times in ddH₂O at room temperature and twice in isopropanol before being dried by the passage of nitrogen. The dried probe was resuspended in the HS and denatured by heating at 94°C for 5 min. Thirty microliters of the probe containing HS was applied to a dried, prehybridized slide, covered with a coverslip (Corning, Corning, N.Y.), and put into a sealed hybridization chamber containing a small reservoir of water to maintain moisture. Hybridization occurred for approximately 14 h at 35°C. Coverslips were removed in washing buffer I (2× SSC–0.1% SDS) warmed to 35°C prior to incubation for 5 min. Next, the slides were washed sequentially for 5 min in 1× SSC–0.1% SDS and 0.1× SSC–0.1% SDS. Slides were then passed through three baths, each passage lasting 2 min, of 0.1× SSC. The slides were dried with a nitrogen gas flow.

Data collection and analysis.

Hybridization to each slide was quantified with a confocal laser microscope (Molecular Dynamics, Sunnyvale, Calif.) the photomultiplier tube of which was set to 700 and 800 V for the Cy-3 and Cy-5 signals, respectively. The images obtained were analyzed with ArrayVision 4.0 software (Imaging Research, Inc., Ontario, Canada). The fluorescent intensity (I_i) associated with each spotted gene (i) was reduced by subtracting the fluorescence (N_i) of an adjoining, nonspotted region of the slide. These readings (R_i = I_i − N_i) were exported to a spreadsheet for further data manipulation. The four “no-DNA” spots derived from PCR mixtures devoid of template were controls used to determine the noise (background signal) level.

The 96 genes present on each slide were used as internal controls (C⁹⁶) to derive equivalent readings (ER_i = R_i/C⁹⁶_j) among the three slides (j) of an individual whole genome array set. Hybridization to this 96-gene set allowed correction for any difference in the hybridization to the three slides within a set. This accounted for slide-to-slide differences in signal acquisition.

For the IPTG induction experiment, it was presumed that the overall transcriptional pattern did not change significantly. Thus, the equivalent reading of each gene was summed (Σ_iⁿ ER = ER_i); normalization by multiplication with a correction factor (CF) of the summed values and the underlying equivalent readings was performed to equalize the summed readings of the control and treated samples (Σ_iⁿ ER_control = CF × Σ_iⁿ ER_treated). This allowed calculation of fold induction of each gene's expression by comparison of each gene's normalized equivalent reading, norm ER, from a pair of conditions. The fold induction of any gene's transcript by a chemical treatment is norm ER_treated/ norm ER_control, where norm ER_control = ER_control and norm ER_treated = CF × ER_treated.

RNA abundance.

To convert normalized equivalent readings into measures of transcript abundance (AB), a further correction was needed. That correction required the hybridization signal arising from an equimolar concentration of all transcripts. The surrogate for this correction factor was the fluorescent intensities arising from hybridization with the fluorescent copy of genomic DNA. Thus, the fluorescent intensities from hybridization with RNA-derived probes were corrected by using fluorescent intensities arising from genomic DNA-derived probes. The abundance of each gene's transcription product(s) was determined by dividing the normalized equivalent reading of a genomic DNA-derived sample into the normalized equivalent reading from the RNA-derived sample (AB = norm ER_transcripts/ norm ER_genome). The convention of Riley and Labedan (32) was followed in grouping genes into functional sets.

β-Galactosidase content.

β-Galactosidase content was measured by the method of Miller using the combined action of chloroform and SDS to disrupt the cell envelope, thus allowing entry of the substrate (23). Two independent cultures of MG1655 were grown in LB medium at 37°C to densities (A₆₀₀s) of 0.43 and 0.47, split, treated with 1 mM IPTG for 15 min, and assayed. The data from these two independent experiments performed on different days were averaged.

RESULTS

Array quality.

Preliminary experiments (data not shown) indicated that the fluorescent signals obtained with labeled cDNA derived from random priming of E. coli genomic DNA (3 μg) were not saturating and were well within the linear range of the instrumentation. The noise level was determined by averaging the readings of the four control spots derived from PCR mixtures lacking a DNA template after subtracting the background signal derived from an adjacent area that had not been spotted.

Using a probe derived from genomic DNA, hybridization signals exceeding noise by a factor of 2 were observed for 4,228 (99.5%) of the 4,290 arrayed genes. Sixty two genes (see Appendix) either failed in PCR amplifications or did not bind sufficient signal to be detected when present at a presumably equimolar ratio with the other genes in the probe sample. Interestingly, 4 (ilvL, leuL, rhoL, and tnaL) of the 27 known genes of this class encode short (<300 bp), attenuation leader polypeptides (19). This suggests that short open reading frames ORFs may not be as readily detected with this method as longer genes.

IPTG induction.

The effect of 1 mM IPTG upon expression of the arrayed genes was investigated. Duplicate RNA preparations of the control and induced cells were each labeled with Cy-3 and Cy-5 by cDNA synthesis. Averaging of measurements was essential for optimal signal detection (Fig. 1). lacZYA induction above the background was detected when the results of a single hybridization experiment in which Cy3-labeled cDNAs derived from treated and control cells were separately hybridized to individual slide sets as viewed in a log-log plot (Fig. 1A). Variation in the measurement of other transcripts was also significant, as indicated by the width of the spread in the data points falling along the diagonal of this scatter plot. Thus, dual-labeling experiments were performed. Improvements were observed by labeling the control sample with Cy-5 and the induced sample with Cy-3 before hybridizing to a single set of three slides (Fig. 1B). However, there was a skewing of the data away from the abscissa (x axis) and towards the ordinate (y axis).

FIG. 1.

Analysis of IPTG induction. Basal expression levels expressed as normalized equivalent readings were plotted on the ordinate, and induced levels, also normalized equivalent readings, were plotted on the abscissa. (A) Results obtained when two Cy-3-labeled probes were hybridized to duplicate whole genome array sets. (B) Experiment in which the Cy-5-labeled cDNA copy of control DNA and the Cy-3-labeled copy of induced RNA were coannealed to a single slide set. The RNAs used to generate the results in panel B were each labeled with the other dye to allow a “reciprocal” hybridization. The resulting data were averaged with the data presented in panel B to yield the scatter plot depicted in panel C. A second independent set of RNA samples were isolated, their cDNAs were labeled with both dyes, and the products were hybridized in both possible combinations to generate the results depicted in panel D. (E) Averaged results of the two independent experiments depicted in panels C and D.

This suggested that dye bias could influence the results. Thus, averaging of these results with others obtained by using differentially labeled cDNA samples (induced cDNA labeled with Cy-5 and control cDNA labeled with Cy-3) from the same set of two RNA preparations resulted in a decreased variation between the treated and control samples. Such “label swapping” and/or repetition, which averaged four normalized equivalent readings of each transcript derived from each RNA sample, lessened the skewing and decreased the scatter (Fig. 1C).

The experiment depicted in Fig. 1C was replicated; fresh cultures were induced and nucleic acids were processed to yield the data depicted in Fig. 1D. Each data point of the experiments shown in Fig. 1C and D represents four measurements of individual transcript abundance; this repetition when averaged yielded the tight constellation shown in Fig. 1E, which combined the data used to generate Fig. 1C and D.

Examination of the extent of hybridization to any individual gene revealed a wide dynamic range with more than a thousand-fold variation in signal intensity between genes (Fig. 1). The expression of only eight genes increased by a factor of more than 2 after exposure to 1 mM IPTG for 15 min (Fig. 1E), while repression by more than a factor of 2 was not observed. These induced genes are listed in Table 1. As expected, the most highly induced RNA ratios (10- to 40-fold) (Table 1) corresponded to the lac operon structural genes, as has been recently reported (31). Measurement of the β-galactosidase content of MG1655 cultures treated with 1 mM IPTG for 15 min resulted in 540 ± 60 Miller units (n = 2), while untreated cultures (n = 2) yielded 17 ± 9 Miller units. Thus, lac transcription and translation increased in parallel.

TABLE 1.

Genes with elevated expression after IPTG treatment

Gene	Function	Level of expression ina:
		LB medium			Minimal medium
					Exponential phase		Transition phase
		Uninduced ppm	Uninduced rank	IPTG induction (fold)	ppm	Rank	ppm	Rank
lacA	Thiogalactoside acyltransferase	40	3,747	36.0	0.25	4,244	21	3,816
lacZ	β-Galactosidase	89	2,420	29.0	2	3,879	20	3,849
lacY	Galactoside permease	61	3,125	14.0	6	4,202	16	3,975
b2324	Peptidase?	280	621	5.3	73	2,639	54	2,717
uxaA	Altronate hydrolase	290	575	4.0	78	2,530	90	1,990
b1783/yeaG	?	370	401	3.6	320	576	1,000	136
melA	α-Galactosidase	41	3,729	2.9	14	4,050	17	3,966
b0956/ycbG	Hydrogenase?	260	678	2.5	140	1,573	130	1,529

^appm, fraction of particular transcript/summed transcripts hybridizing to all ORFs on the microarrays expressed in ppm; rank, genes ranked in order of expression, with 1 being the most highly expressed gene for each of the experiments. 

A commonality among some of the other induced genes was intriguing. b0956, encoding a putative dehydrogenase and preceded by a catabolite activator protein binding site, may have a catabolic function that parallels those of the lac operon and two other induced genes, melA and uxaA. melA encodes an α-galactosidase, while uxaA specifies an enzyme of hexuronate catabolism (2). Interestingly, potential melA induction by IPTG was suggested by hybridization of cDNA to an ordered set of λ clones carrying inserts of the E. coli chromosome (6) and in another microarray experiment (31).

The function of the other induced genes is even more speculative; upstream of b1783 is a ς⁵⁴ binding site, while peptidase function is hypothesized for b2324. Induction of these latter genes was not observed by Richmond et al. (31).

Estimate of steady-state transcript levels.

The percentage of RNA that programs protein synthesis has been determined under a wide variety of growth regimens (4). Here we estimated the fraction of those protein-specifying transcripts devoted to each arrayed gene. Hybridization signals arising from annealing of RNA-derived Cy-3 labeled cDNA populations were normalized by dividing by the signal generated with Cy-3 fluorescent cDNA arising from copying of sheared E. coli genomic DNA as a probe. Each spot's corrected signal from RNA-derived cDNA hybridization reflected the amount of RNA in the sample. Three RNA samples were thus measured; they were isolated from cells growing exponentially in rich medium, growing exponentially in minimal medium, and cells in minimal medium making a transition from the exponential phase to the stationary phase (for culture conditions, see Materials and Methods). RNAs from certain central metabolic (gapA and ptsH), defense (ahpC and cspC), DNA metabolic (hns), surface structure (acpP, ompACFT, and lpp), translation (rplBCKLMPWX, rpmBCl, rpsACDHJNS, trmD, fusA, infC, and tufAB), transcription (rpoAB), and unassigned (b4243) genes (32) were abundant (>0.1%, among the top 100 transcripts) in all three samples.

Transcripts of exponential-phase cells cultured in rich medium.

High-density microarrays were used to measure the transcriptional content of cells growing in rich medium. A total of 1,776 genes were apparently not expressed; their hybridization signals did not exceed the noise by a factor of 2. Of these nonexpressed genes, function has been ascribed to 465. They are listed in the Appendix. Each gene having an assigned function was expressed in at least one of the two tested stages of cultivation in minimal medium.

The other genes, each representing between 0.0007 and 1% of the summed hybridizing signal, were expressed in LB broth-grown cells. The distribution of genes as a function of expression level is plotted in Fig. 2, while Fig. 3 depicts fractional expression as a function of summed genes. The percentage of transcripts was plotted as a function of genes summed in Fig. 3. The order in which genes were summed was based upon expression level, with the most highly expressed gene in each condition summed first.

FIG. 2.

Distribution of expressed genes. The histogram plots number of genes as a function of expression range. Diagonally striped, solid, and horizontally striped bars reflect distributions observed in RNAs derived from cells growing exponentially in minimal medium, cells transitioning to the stationary phase in minimal medium, and cells growing exponentially in rich medium, respectively. Expression of 766, 1,030, and 1,776 genes was not detected under the three respective conditions.

FIG. 3.

Fractional expression. The extent of ORF transcripts is plotted as a function of genes summed. The order in which genes were summed was based upon expression level, with the most highly expressed gene summed first.

Fewer genes were expressed in LB medium than in minimal medium (Fig. 2); the fraction of rare transcripts was smaller in rich medium (Fig. 3). The 50 most highly expressed genes in LB broth-grown cells are listed in left-most columns of Table 2; 29 of these intensely transcribed genes encode proteins involved in translation and protein folding (2, 22, 32).

TABLE 2.

Highly expressed genes under three different culture conditions

Genea	ppm (100) in LB medium (exponential phase)b	Genea	ppm (100) in minimal medium (exponential phase)b	Genea	ppm (100) in minimal (transition phase)b
infC (TA)	70	cspA (SR)	54	hdeA (SS)	160
rplK (TA)	68	metE (BS)	50	hdeB (SS)	99
rplL (TA)	66	tufB (TA)	48	rmf (TA)	83
rplA (TA)	48	ompA	46	dps (SS)	65
hemK	47	ilvC (BS)	42	lpp	63
rpmI (TA)	46	rmf (TA)	38	ompC	59
rplW (TA)	44	tufA (TA)	38	icdA (CM)	59
rplJ (TA)	43	ompT	37	metE (BS)	53
acpP	42	infC (TA)	36	gapA (CM)	49
lpp	40	ahpC (SR)	34	tufA (TA)	49
glpQ	39	rplM (TA)	31	ompA	44
fusA (TA)	39	ptsH (CM)	31	infC (TA)	44
gatB	38	aceB (CM)	30	uspA (SR)	40
rpsF (TA)	38	lpp	29	tufB (TA)	39
tufB (TA)	37	rpsJ (TA)	28	ilvC (BS)	37
ompC	37	cirA	28	rpsN	36
mopB (PF)	35	gapA (CM)	26	eno (CM)	36
atpF	35	rpmI (TA)	26	ahpC (SR)	35
hns	35	yjjS	26	ompT	33
rpmB (TA)	34	rpmC (TA)	24	gadA (SS)	33
ompA	33	fusA (TA)	24	aceB (CM)	33
tnaA	33	b2745	24	rplX (TA)	32
rpoA	33	ompF	23	fusA (TA)	32
trmD (TA)	32	cspC (SR)	22	ptsH (CM)	31
rplI (TA)	31	aceA (CM)	21	rpsD (TA)	29
gapA (CM)	30	pyrB (BS)	21	b3512	29
rplM (TA)	28	rplK (TA)	21	gpmA (CM)	28
rpmG (TA)	28	rpsD (TA)	21	metK (BS)	28
rpsC (TA)	27	cysK (BS)	20	rpmC (TA)	27
rplT (TA)	27	ptsI (CM)	20	gadB (SS)	27
rplX (TA)	27	b1452	19	rpsV (TA)	27
priB	26	rpsS (TA)	19	cysK (BS)	26
ompF	25	fepA	19	rpsJ (TA)	25
hupA	25	pyrI (BS)	18	rpsH (TA)	25
rpsJ (TA)	25	aroF (BS)	18	rplE (TA)	25
rplB (TA)	25	rpsH (TA)	17	aceA (CM)	25
rplU (TA)	25	rpsN (TA)	17	b2266	23
tig (PF)	24	b0805	17	rplM (TA)	23
tufA (TA)	24	ompC	17	rpsS (TA)	23
rplD (TA)	24	rpsA (TA)	17	nlpD	22
rplC (TA)	24	thrL (BS)	17	acpP	22
gatA	24	rplX (TA)	16	rpmI (TA)	22
rpsA (TA)	24	rplL (TA)	16	rpoS (SS)	21
gatY	23	rpsM (TA)	16	rpoA	20
rpsS	23	b4243	16	hns	20
ppa	22	rplB (TA)	16	b4253	20
gatZ	22	b0817	16	rplB (TA)	20
cspE (SR)	21	folE (BS)	15	b1452	19
cspC (SR)	21	icdA (CM)	15	b0817	19
mopA (PF)	21	rplW (TA)	15	b1003	19

^aPF, protein-folding genes; SR, stress-responsive genes; CM, central metabolic enzyme-specifying genes; BS, biosynthetic genes; TA, translation-associated genes; SS, rpoS (sigma S)-controlled genes. 

^bFraction of transcripts hybridizing to specified gene/summed transcripts hybridizing to all ORFs on the microarrays (expressed in ppm). 

Table 3 summarizes the steady-state mRNA levels obtained for several ribosomal protein-specifying operons. Sixty-nine determinations are reported, of which 63 (greater than 90%) exceeded 500 ppm (parts per million) (0.05%). Of those measurements not meeting this threshold, two (rpsL and yceD) were associated with poor-quality PCR products (see Table 3, footnote). The other four (secY, rpmJ, dnaG, and rpoD) represented hybridization to the penultimate or final genes of operons. Moreover, gradients of transcript levels were observed for the Spc, S10, L10(β), S15, L35, S21 (ς), and L13 operons. The bases of such gradients were not investigated.

TABLE 3.

Expression of operons encoding ribosomal proteins in an exponential-phase culture in LB medium

Operon (min)	Genea	Proteinb	ppm (100)	Rank
Spc (73)	rplN	L14	8.6	122
	rplX	L24	27.1	31
	(TR)
	rplE	L5	10.0	102
	rpsN	S14	17.2	59
	rpsH	S8	13.7	79
	rplF	L6	6.9	167
	rplR	L18	14.9	70
	rpsE	S5	11.6	87
	rpmD	L30	5.4	227
	rplO	L15	5.8	206
	secY	SecY	1.9	1053
	rpmJ	L36	4.0	359
S10 (73)	(TR)
	rpsJ	S10	24.6	35
	rplC	L3	23.9	41
	rplD	L4	24.0	40
	rplW	L23	43.9	7
	rplB	L2	24.6	36
	rpsS	S19	22.8	45
	rplV	L22	15.7	66
	rpsC	S3	27.4	29
	rplP	L16	19.1	53
	rpmC	L29	13.6	80
	rpsQ	S17	7.8	142
α (73)	(TR)
	rpsM	S13	8.1	136
	rpsK	S11	8.4	128
	rpsD	S4	15.3	67
	rpoA	α	32.6	23
	rplQ	L17	7.1	163
Str (73)	rpsLc	S12	0.9	2291
	(TR)
	rpsG	S7	10.0	99
	fusA	EF-G	38.5	12
	tufA	EF-Tu	24.1	31
L11 (90)	(TR)
	rplK	L11	68.0	2
	rplA	L1	48.0	4
L10 or β (90)	(TR)
	rplJ	L10	43.4	8
	rplL	L7/12	66.4	3
	(TT)
	rpoB	β	11.4	89
	rpoC	β′	5.0	258
S15 (69)	(TR)
	rpsO	S15	16.0	63
	(TT)
	pnp	Pnp	5.7	215
S2 (4)	rpsB	S2	16.9	60
	tsf	EF-Ts	14.2	76
L34 (83)	rpmH	L34	6.9	166
	mpA	RNAse P	5.3	237
L28 (82)	rpmB	L28	33.5	20
	rpmG	L33	27.5	28
L35 (38)	infC	IF3	70.2	1
	(TR)
	rpmI	L35	46.2	6
	rplT	L20	27.2	30
S20 (0)	rpsT	S20	5.3	235
S1 (21)	rpsA	S1	23.8	43
S6 (95)	rpsF	S6	38.3	14
	priB	n	26.4	32
	rpsR	S18	19.0	54
	rplI	L9	30.6	25
S21 or ς (67)	rpsU	S21	17.3	58
	dnaG	DnaG	2.1	875
	rpoD	ς70	4.2	876
L25 (48)	rplY	L25	16.7	61
L28 (82)	rpmB	L28	33.5	20
	rpmG	L33	27.5	28
L32 (24)	yceDd	ORF	1.4	1463
	rpmF	L32	9.7	107
L13 (70)	rplM	L13	28.0	27
	rpsl	S9	8.0	140
trmD (57)	rpsP	S16	11.2	90
	b2608	ORF	10.1	98
	trmD	TrmD	32.3	24
	rplS	L19	6.3	185

^aGenes within operons are listed in the order in which they are transcribed. TR, sites of translational repression; TT, sites of transcriptional termination (15). 

^bRibosomal proteins are indicated by their standard alphanumeric designations (15). Other proteins are involved in translation (EF-G, EF-Tu, EF-Ts, and IF3), secretion (SecY), transcription (α, β, β′, and ς⁷⁰), RNA maturation and turnover (RNase P, TrmD, and Pnp), or replication (n and DnaG) (32). 

^cMultiple PCR products were derived from the rpsL-specific amplification. 

^dThe yceD product was not of the expected size. 

Transcripts of exponential-phase cells cultured in defined minimal medium.

The gene expression pattern of cells growing exponentially in minimal medium was also examined. At this cell density, the pH remained at 7.0. Expression levels varied from 0.001 to 0.7%. Apparently, as illustrated in Fig. 2, biosynthetic requirements mandated that a significantly greater fraction of the genome was expressed in minimal medium. Only 776 genes were expressed negligibly (signal/noise ratio of <2); 149 of these genes have presumed or demonstrated function. They are listed in the Appendix.

The 50 genes most highly expressed in logarithmically growing cells cultured in minimal medium with glucose as a carbon or energy source are enumerated in middle columns of Table 2. The distribution of genes as a function of expression level (Fig. 2) and the fractional expression as a function of summed genes with genes ranked by expression level (Fig. 3) were also plotted. Such broad-distribution analyses readily revealed the significant differences observed in expression of E. coli when grown in defined and rich media. In minimal media, many more genes were transcribed over a somewhat broader range.

Eight biosynthetic genes became highly expressed (Table 2). Notable among them were metE, encoding the aerobic methionine synthase, and ilvC, an isoleucine-valine biosynthetic gene subject to feed-forward transcriptional activation (35) by its substrates. Both the ilvC (27, 39)- and metE (10)-encoded enzymes are sluggish catalysts. The metE product accounts for about 5% of E. coli protein when cells are cultured in minimal medium with glucose as a carbon or energy source (36). Other highly expressed biosynthetic genes included folE and cysK;the folE product, GTP cyclohydrolase I, catalyzes both cleavage of the 5-membered ring of guanine and the rearrangement of the ribose moiety of the substrate, GTP (9). cysK, encoding o-acetylserine (thiol)-lyase isozyme A, is responsible for more than 90% of sulfur fixation under aerobic conditions (17). Transcripts of the pyrBI operon encoding aspartate transcarbamylase also were highly expressed during exponential growth in minimal medium relative to an LB broth-grown culture. This expression level is a characteristic signature of strain MG 1655, whose aspartate transcarbamylase content is elevated more than 100-fold when grown in the absence of uracil due to an rph mutation that is polar on pyrE (13). The other highly expressed transcripts, thrL and aroF, encoded, respectively, the threonine leader polypeptide (19) and the phenylalanine-inhibited first enzyme of the common aromatic pathway. The aroF product, one of three isozymes, is estimated to account for more than 80% of the activity catalyzing the first common step of aromatic amino acid synthesis (28).

Expression of several genes catalyzing fueling reactions was also elevated. As expected, ptsHI transcripts encoding phosphotransferase sugar transport common components (29) accumulated to a very high titer in glucose-minimal medium. Surprisingly, aceAB, encoding the glyoxylate shunt enzymes malate synthase and isocitrate lyase (7), was highly expressed. Perhaps the tricarboxylic acid (TCA) cycle functions in its branched state during this phase of growth requiring the glyoxylate shunt for anapleurotic replenishment (24). Alternatively, at this culture density, the cells have started to use the accumulated acetate.

These data were further analyzed by examining individual transcript levels within the context of operon structure (Table 4). Results with selected amino acid biosynthetic operons are presented in conjunction with the operon responsible for phosphoenolpyruvate: carbohydrate phosphotransferase system (PTS)-mediated sugar uptake. These biosynthetic operons are predominantly controlled by attenuation. Detection of attenuator transcripts was difficult (see above and the Appendix) this difficulty was compounded by the multiple PCR products obtained after amplification of hisL (Table 4). The thr leader was found to be more highly expressed than the cognate structural genes; this was not observed when the trp or his operons were analyzed. The most poorly expressed of the listed genes was ilvY, which encodes a specific DNA binding protein that activates ilvC transcription while repressing its own synthesis (35). Conversely, ilvC was the most highly expressed of this group, as was indicated in Table 2. Measurement of pts operon expression ranged from 1,200 to 3,100 ppm; a gradient of transcript levels was observed for pts. Multiple promoters and termination sites, a hallmark of this operon (29), might provide a basis for this observation.

TABLE 4.

Expression of some amino acid biosynthetic operons and the PTS operon in an exponential-phase culture in minimal medium

Operon	Genea	ppm (100)	Rank
Histidine	hisLb	3.3	556
	hisG	5.6	265
	hisD	2.5	818
	hisC	3.7	490
	hisB	3.3	555
	hisH	2.1	1,000
	hisA	1.8	1,185
	hisF	2.5	824
	hisI	4.2	417
Isoleucine-valine	ilvG	7.5	171
	ilvM	2.5	827
	ilvE	2.9	668
	ilvD	1.6	1,360
	ilvA	0.6	3,038
IRc regulator	ilvY∗∗	0.4	3,373
IR	ilvC∗	41.7	5
ALSd I	ilvB	1.3	1,765
	ilvN	1.7	1,295
ALS III	ilvIb	2.9	662
	ilvH	3.0	632
Leucine	leuA	11.3	82
	leuB	10.9	93
	leuC	6.3	215
	leuD	9.6	115
Threonine	thrL	16.8	41
	thrA	4.0	442
	thrB	5.4	276
	thrC	7.9	160
Tryptophan	trpL	6.6	199
	trpE	6.8	189
	trpD	4.3	396
	trpC	10.4	102
	trpB	11.6	77
	trpA	13.2	64
Proline AB	proA	2.5	835
	proB	2.4	863
Proline C	proC	1.1	1,994
Arginine E	argE∗∗	1.6	1,353
Arginine CBH	argC∗	2.0	1,088
	argB∗	2.9	684
	argH∗	2.2	967
PTS	ptsH	31.0	12
	ptsI	19.6	30
	crr	11.9	74

^aGenes within operons are listed in the order in which they are transcribed (2), except for the divergent clusters, where the sites of initiation lie between the genes labeled with one and two asterisks. 

^bMultiple PCR products were derived from the ilvI- and hisL-specific amplifications. 

^cIR, isomeroreductase. 

^dALS, acetolactate synthase. 

Comparable expression of genes within an operon was observed; the proA and proB mRNA titers were 240 to 250 ppm, while those for ilvl and ilvH were 290 and 300 ppm. Similarly, the ilvBN transcription level was found to be 130 ppm when ilvB was immobilized on the microarray and 170 ppm when ilvN was the capture reagent in the hybridization. Measures of argC, argB, and argH mRNA quantities differed by less than 50%, ranging from 200 to 300 ppm. Levels of mRNA from the leu operon were within a factor of 2, as were those of the thr structural genes. Measures of trp and his expression were within a factor of 3. Determinations of ilvGMEDA operon transcript levels were more variable; they ranged from 750 ppm for ilvG to 60 ppm for ilvA. There might be a biochemical basis for this variation, since transcript level paralleled gene order within the operon.

Transcripts of cells making a transition from the exponential phase to the stationary phase in defined minimal medium.

During this transition, at a point where the pH had dropped slightly to 6.7, significant changes in gene expression were expected and observed. Expressed gene levels were from 0.0023 to 1.6%. A total of 1,030 genes, of which 110 have a defined role (Appendix) did not appear to be expressed at this transitional phase of growth.

The 50 genes most highly expressed genes during this transition are listed in the rightmost columns of Table 2. Significantly, several rpoS-regulated genes, including hdeA (10-fold transcript elevation in comparison to the exponential-phase content) (12), hdeB (9-fold) (12), dps (4-fold) (12), gadA (8-fold) (5), and gadB (10-fold) (5), as well as rpoS (3-fold) (12) itself, became highly expressed. Despite this remodeling of transcription, the overall patterns of gene number as a function of expression level (Fig. 2) and fractional expression as a function of ranked gene (Fig. 3) were not as distinct as one might imagine in comparison to the patterns from exponentially growing cells.

Compilation.

The observed expression patterns are summarized in Table 5, where gene products were grouped by metabolic function according to an established classification scheme (32). Exponential growth in minimal medium elevated the amount of pyrimidine and amino acid biosynthetic transcripts with respect to growth in the rich broth, LB. In contrast, cofactor and purine transcripts did not appear to accumulate relative to growth in LB broth. Expression of glyoxylate shunt and other glucose metabolism-related transcripts was also elevated in minimal medium; the seven-fold elevation of glyoxylate shunt transcripts exceeded the average of that observed for amino acid biosynthetic mRNAs. Expression of genes involved in sulfur fixation was also elevated during growth in minimal medium.

TABLE 5.

Summary of three E. coli expression profiles

Characteristic (no. of genes summed)	Abundance (100 ppm) in:
	Minimal medium		LB medium (exponential phase)
	Exponential phase	Transition phase
Cell processes
Cell division (26)	110	100	100
Chemotaxis and motility (12)	14	7	11
Folding and ushering proteins (7)	32	61	110
Transport of large molecules: protein, peptide secretion (32)	82	101	100
Transport of small molecules
Amino acids, amines (49)	91	81	68
Anions (20)	29	28	23
Carbohydrates, organic acids, alcohols (82)	200	160	340
Cations (52)	120	98	76
Nucleosides, purines, pyrimidines (6)	10	9	17
Other (12)	21	27	12
Elements of external origin
Laterally acquired	240	170	230
Phage-related functions/prophages (27)	55	42	65
Plasmid-related functions (1)	2	6	9
Transposon-related functions (34)	58	35	38
Global functions
Energy transfer, ATP proton motive force (9)	77	54	150
Global regulatory functions (51)	176	290	180
Macromolecule metabolism
Basic protein synthesis, and modification (6)	47	48	74
Degradation of:
DNA (23)	38	30	31
RNA (11)	29	15	22
Polysaccharides (3)	6	3	4
Proteins and peptides (61)	84	93	110
Synthesis and modification of:
DNA (89)	230	190	310
Lipoprotein (11)	41	50	37
Phospholipids (11)	20	15	21
Polysaccharides (cytoplasmic) (6)	15	16	6
Protein translation and modification (34)	290	300	430
RNA synthesis (27)	100	100	150
Macromolecules
Glycoprotein
Lipopolysaccharide (13)	15	12	18
tRNA modification and charging (40)	130	130	210
Metabolism of small molecules
Amino acid biosynthesis (110)	120	93	33
Biosynthesis of cofactors, carriers (115)	720	690	640
Central intermediary metabolism
2′-Deoxyribonucleotide metabolism (12)	34	32	32
Amino sugars (10)	12	11	15
Entner-Douderoff (3)	4	3	6
Gluconeogenesis (4)	9	12	21
Glyoxylate bypass (5)	76	75	12
Other metabolism of glucose (3)	9	5	4
Nonoxidative branch, pentose pathway (8)	26	43	43
Nucleotide hydrolysis (2)	1	1	3
Nucleotide interconversions (13)	41	39	20
Phosphorus compounds (17)	32	30	22
Polyamine blosynthesis (8)	16	13	13
Nucleoside and nucleotide salvage (18)	37	38	54
Sugar-nucleotide synthesis, conversion (18)	42	34	48
Sulfur metabolism (10)	39	25	10
Pool, interconversion (46)	120	190	210
Degradation of small molecules
Amines (9)	16	10	25
Amino acids (17)	22	14	72
Carbon compounds (90)	140	110	250
Fatty acids (10)	20	17	30
Others (8)	15	21	8
Energy metabolism, carbon
Aerobic respiration (27)	77	58	120
Anaerobic respiration (80)	75	57	110
Electron transport (24)	32	24	48
Fermentation (21)	40	50	44
Glycolysis (18)	130	240	150
Oxidative branch, pentose pathway (2)
Pyruvate dehydrogenase (6)	46	41	40
TCA cycle (18)	89	120	93
Fatty acid biosynthesis: fatty acid and phosphatidic acid synthesis (23)	73	94	150
Nucleotide synthesis	190	130	100
Purine ribonucleotide synthesis (22)	110	77	83
Pyrimidine ribonucleotide biosynthesis (10)	79	49	18
Miscellaneous: not classified (109)	220	230	250
ORFs: unknown proteins (1,324)	3,500	3,400	2,600
Processes
Adaptation
Atypical conditions (16)	120	83	74
Osmotic (14)	38	63	26
Protection responses
Cell killing (3)	5	3	3
Detoxification (11)	80	83	97
Drug or analog sensitivity (32)	42	31	38
Structural elements
Cell envelope
Murein sacculus, peptidoglycan (34)	95	120	130
Outer membrane constituents (17)	230	260	200
Cell exterior constituents (16)	37	39	62
Surface polysaccharides and antigens: surface structures (57)	75	51	52
Ribosome constituents
Ribosomal and stable RNAs (3)
Ribosomal proteins (64)	790	860	1,500
Ribosomal maturation and modification (6)	56	110	7
ORFs not listed (102)

The rapid growth observed in the LB broth was reflected in the gene expression profile, as was the difference in carbon or energy source between glucose and amino acids. LB broth-grown cultures displayed elevated expression of genes specifying glucogenic enzymes and of genes whose products degrade small molecules. Expression of the ATP and proton-motive force-generating machinery, elevated by a factor of about 2, paralleled increased ribosomal protein, aminoacayl-tRNA synthetase and protein folding-associated expression.

Changes observed upon entering the transitional period between exponential and stationary phase growth in glucose minimal medium were less dramatic. Nonetheless, elevation of mRNAs specifying gluconeogenic, glycogenic, and TCA cycle enzymes was observed as was an increase in transcripts encoding enzymes responsible for metabolic pool interconversions and the nonoxidative branch of the hexose monophosphate shunt. Perhaps, as a prelude to the stationary phase, the cell needs to increase its ability to capture energy or convert specific, biosynthetic end products into other, alternative small molecules, such as trehalose. This sugar, whose accumulation occurs during stationary phase in a ς^S-dependent fashion, serves as an osmoprotectant (11). The cell also displayed an increased titer of protein folding and global regulatory function transcripts while making a transition between growth phases.

DISCUSSION

Comprehensive expression profiling has been performed previously with the yeast Saccharomyces cerevisiae (8). Here we report that such profiling can also be accomplished and refined by using a dye-swapping, high-density microarray technology when the prokaryote E. coli is used as an experimental system. Adaptation of RNA isolation and labeling protocols from eukaryotes to prokaryotes is not straightforward, because eukaryotic mRNA manipulations often exploit the specific 3′-polyadenylation of this molecular species. The short half-life of bacterial mRNA is another obstacle. We chose to isolate RNA by a standard procedure that included a centrifugation step and to reverse transcribe bulk prokaryotic RNA to prepare our hybridization probe. Thus, the reported measurements are subject to possible systematic errors, including differential mRNA stabilities (18). Despite the large amount of stable RNA in the sample, hybridization to protein-encoding genes was readily detected. Recently, independent studies of E. coli (31, 34) successfully applied nylon-based medium-density DNA array or glass-based high-density DNA array technologies to assess gene expression changes in response to growth medium and heat shock.

Nonetheless, errors could be introduced in the many steps from RNA purification to analyses of hybridization signals. As shown in Fig. 1, conditions have been optimized to yield highly reproducible data. The scatter plot of the optimized protocol (Fig. 1E) illustrated that measurements of gene expression were still subject to considerable variation when the signal was in the lowest part of the detectable range. It was found that expression of only eight genes was effected by IPTG treatment; all were induced. It was reassuring that the expected lacZYA induction was observed; the significance of the weaker inductions awaits confirmation by complementary techniques, perhaps based upon enzyme assay of gene fusions or analysis of an isogenic pair of strains differing in lacI. Such a correlation has been provided for IPTG induction of melA expression (31). Thus, such comprehensive gene expression profiling generates hypotheses requiring further study for verification.

Having developed confidence in the technology, it was applied to monitoring expression as a function of growth stage and medium. For these experiments, normalization of signal intensity was essential. Probe, derived from replication of genomic DNA and used as a replica of equimolar transcription of the entire genome, allowed calculation of mRNA inventories. Thus, we have provided measures of steady-state transcript levels under prescribed sets of conditions rather than the fold change in mRNA titers that represents the difference in expression between two conditions. These inventories were satisfying in several ways. First, the most highly transcribed genes in actively growing cells cultured in LB medium often encoded proteins involved in translation. In contrast, cultures at a similar growth stage in glucose minimal medium expressed to a very high level several small molecule biosynthetic genes and the means to utilize glucose.

Thus, agreement between this molecular analysis and the accumulated understanding of E. coli physiology (24, 25) was observed. This agreement was underscored in the analysis of cells making a transition from the exponential growth phase to the stationary phase; the elevated expression of several rpoS-controlled genes corresponded to expectations. Nonetheless, some caution is necessary; potential effects of differential mRNA stability (18) have yet to be considered.

It is most unlikely that the technology is limited to highly expressed genes. First, reproducible expression measurements were obtained over a wide dynamic range (Fig. 1E). Second, the data from Fig. 3 and Table 1 illustrate that lac operon expression, although low before IPTG induction, was detected, suggesting that most transcripts can be readily measured by the techniques described. The lower limits of expression that can be observed may be defined by the analyses of well-characterized “promoter-down” mutants (30) or “spiking” experiments. In the latter, the templates for cDNA synthesis would contain constant amounts of total RNA derived from a deletion mutant to which various quantities of a corresponding transcript synthesized in vitro has been added.

Compiling of data (Table 5) into functional groups (32) has become one method for the analysis of gene expression profiles. Work such as that presented here indicates that these categories do not respond as a bloc; rather subsets act differently, as has been observed for amino acid biosynthesis during a study of yeast (14). Nonetheless, such analyses allow global trends to be observed and the integration of gene expression patterns with the cell's overall physiological status. The histogram (Fig. 2) allows one to appreciate the quantity of transcripts that falls within each expression range. Figure 3 provides an indication of how transcriptional capacity is distributed under the three distinct conditions that were examined.

Moreover, unexpected results worthy of further study are found within the compilation presented in Table 5. Unlike transcripts dealing with energy transfer, ribosomal proteins, translation and aminoacyl-tRNA formation which were elevated in LB broth-grown cells, the sum of mRNAs specifying cell division proteins did not vary under the three conditions that were investigated. In a similar vein, hybridization to genes involved in the synthesis of the cell envelope was not increased when the probe was derived from cells cultured in LB broth.

In such global analyses, the reliability of the data obtained is an issue. The method described here makes four measurements of the transcript level present in each RNA sample. Moreover, the organization of genes into operons provides internal benchmarking of the measurements. Analyses of transcripts from ribosomal protein, biosynthetic, and PTS operons (Tables 3 and 4) suggests that the data are of high quality, since the range of the measurements for each operon is rather small. Such an analysis is consistent with the methodological improvements illustrated in Fig. 1.

Thus, an initial mRNA inventory was compiled. We believe that our analyses are subject to errors in measures of transcripts smaller than about 300 nucleotides and failures in PCR amplification. The compilation illustrated several physiological points. The protein biosynthetic demand during growth in rich medium was noted, accounting for about 15% of the polypeptide-specifying transcripts; what may be as significant is that more than one-quarter of all protein-specifying transcripts under any of the measured conditions lack a functional assignment. Consequently, gene expression profiling provides a further impetus to the continued study of E. coli.

Appendix

E. coli genes not giving a detectable signal when hybridized with genome-derived DNA were as follows: acrB, agaC, arcB, cydA, dacB, dnaT, entC, entF, exo, fruL, ilvL, leuL, lytB, pheL, phnA, potH, potl, putA, rbsD, rfaB, rhlB, rhoL, sdhC, selB, tdcA, tnaL, b0177, b0250, b0269, b0271, b0291, b0322, b0574, b1437, b1595, b1824, b1978, b2067, b2086, b2088, b2097, b2270, b2292, b2630, b2641, b2851, b2878, b3194, b3596, b3597, b3672, b3678, b3696, b3697, b3705, b4002, b4253, b4280, b4404, and b4405

Genes having a known function but not expressed when cells were cultured in rich medium were as follows: aas, acpDS, acrF, adiY, agaABDIRSVW, ais, alkB, alpA, apaGH, appY, aqpZ, araBEH, argACT, aroDE, arp, arsCR, artIMQ, asr, betIT, bglBG, bioCH, blc, bolA, cadAC, caiBF, cbl, cchB, ccmABCD, cdd, celAC, chaB, cheBRWYZ, chpAR, cmtAB, cof, cpsG, creB, criR, csgABDFG, cspBF, cvpA, cybC, cynST, cysHJUW, dam, dedA, deoR, dgkA, dicBC, dinIJ, dmsBC, dniR, dppBC, dsbE, dsdCX, dsrB, eaeH, ebgC, ecpD, emrY, endA, entDE, envRY, erfK, evgAS, exbB, farR, fdnI, fecIR, feoA, fepBDE, fes, fic, fimD_1, fimFGZ, fixX, flgACGHLMN, flhAD, fliACEFGJLMOPQRSTZ, folP, frdC, frvARX, frwD, ftsKL, fucAKORU, fumC, gabP, gadB, galKMPR, gapC, gatR, gcl, gcvA, gefL, gidB, gip, glcCDG, glgS, glnBK, glpEG, gltBDF, glvBCG, gntKUV, grxA, gusC, gutM, hdeAB, hdhA, hemH, hha, hhoB, hlpB, hlsMQ, hnr, hofBDFGH, holBE, hrpA, hslJS, htrCE, hyaBCDE, hybBDEFG, hycABDFGHI, hydN, hypACD, ibk, icc, ilvMNY, insA_2, insA_3, insA_4, katE, kch, kdgT, hdpE, kdtB, kduDI, kefC, lacA, lar, ldcC, leuDO, lhr, lit, livK, lrp, lysAR, marBR, may, mcrA, mdlB, mepA, metACR, mhpBE, moaD, moeB, molR, mreD, mscL, msyB, mukB, mutHY, nac, napBCFH, narIJZ, nel, nhaR, nikBDE, nirCD, nlp, nlpC, nrdEFG, nrfBFG, nupG, ogrK, osmBCE, pabA, panF, pfkB, pgpB, pheM, phnBCDFGHJKLOQ, phoBH, phpB, phrB, pinO, pitB, pnuC, potFG, ppdABCD, ppiC, prfH, priC, prkB, prmA, proBVW, prsA, pshM, psiF, pspBC, pssR, pstA, pth, ptrB, ptsO, purE, pyrIL, racC, rarD, rcsA, recDT, relBEF, rem, rfaHKYZ, rhaDRSCD, rimL, rmf, rna, rnb, rnhB, rnk, rpiBR, rpmHJ, rpsV, rspAB, sanA, sapBC, sdiA, sfa, sieB, slp, smf, smg, smpA, sms, sodC, sohA, soxRS, speBC, sprT, srlB, sspB, sugE, sulA, surE, syd, talC, tap, tdcCR, tdk, tehB, tesB, thiEH, thrLS, torADRT, tpr, treR, trkG, trpL, ubiX, ugpE, uhpT, uidA, umuD, ung, usg, uxaBC, vsr, wcaB, xapR, xasA, xerD and xylFH. Several operons have component genes that were apparently expressed and others that were not. Such inconsistencies (found in the dms, dpp, lac, nik, sap, and tdc operons) were not resolved. Genes without a known role are not listed.

The “well-defined, though quiescent” genes in cells growing exponentially in minimal medium were as follows: acpS, adiY, agaBDIV, alpA, aapZ, arsC, aslB, cadBC, caiBF, cchAB, cdsA, chpABRS, cmtAB, criR, csrA, cynST, deoR, ebgC, ecpD, emrB, envR, eutEJ, feoA, fimD_1, fimZ, fixAX, figAFH, frdC, frvARX, frwCD, fucAIKOU, galK, gcl, gefL, gip, glcG, glgS, glnBK, glpEGLTBK, glvBC, gntV, greA, grxA, gutM, hofB, hrsA, hyaE, hybFG, hycABFGHI, hydN, hypAC, kdgT, kdpC, hdtB, kefC, lacZYA, lar, leuO, malG, mbhA, mcrC, mhpB, mreD, nanA, nikBE, nirCD, nlp, nrdG, nrfBCFG, oraA, pheM, phnCDL, pinO, ppdABCD, priC, prmA, pshM, ptrB, racC, rfaZ, rhaD, rhsC, rnk, rpiR, rspB, sdaB, sieB, smpA, sms, sprT, srlABDR, tdcB, thrS, tnaB, ttdAB, ublH and uhpT. Genes of unknown function are not listed.

One hundred ten genes that were “silent” during transition of a culture from the exponential phase to the stationary phase in minimal medium are as follows: agaASW, ais, alkB, araFH, arsR, asr, bglG, ccmBCD, celA, cheWY, cpsG, cspBF, cvpA, cycW, dedA, dicBC, dsbE, dsrB, emrY, endA, evgS, fdnI, flgBCDGM, fliEFGHJLNOPQRT, gatR, glcD, gntKU, gusC, hdhA, hemK, hipB, hisMQ, hofFGH, holE, hrpA, hyaC, hybDE, hypD, kduI, lysR, malI, marBR, motA, napBCFH, narV, nuoAK, nupG, ogrK, pgsA, prfH, psrA, pspBC, pstA, pryL, rcsA, recT, relF, rhaR, rnb, sapBC, sdhD, sdiA, sfa, soxS, tdcCR, tdk, tpr, trkG, ubiX, uldAB, umuD, usg, and uxaB. Not listed are genes whose function is yet to be elucidated.