Transcriptome analysis of Escherichia coli using high-density oligonucleotide probe arrays

Abstract

Microarrays traditionally have been used to analyze the expression behavior of large numbers of coding transcripts. Here we present a comprehensive approach for high-throughput transcript discovery in Escherichia coli focused mainly on intergenic regions which, together with analysis of coding transcripts, provides us with a more complete insight into the organism's transcriptome. Using a whole genome array, we detected expression for 4052 coding transcripts and identified 1102 additional transcripts in the intergenic regions of the E.coli genome. Further classification reveals 317 novel transcripts with unknown function. Our results show that, despite sophisticated approaches to genome annotation, many cellular transcripts remain unidentified. Through the experimental identification of all RNAs expressed under a specific condition, we gain a more thorough understanding of all cellular processes.

Figure 1

The pts operon in E.coli. The enlarged pictures of a microarray show the high expression level of the neighboring genes ptsH, ptsI and crr in the pts operon. The bottom picture (from RegulonDB; http://kinich.cifn.unam.mx:8850/db/regulondb_intro.frameset) shows the location of the genes within the genome.

Figure 2

Operon detection using oligonucleotide probe intensities. Individual oligonucleotide probe intensities (PM – MM) from three conditions are shown to validate the microarray-predicted hnr–galU operon. Intensities for individual probes interrogating hnr, the 200 bp intergenic region and galU are shown. This operon was independently confirmed using RT–PCR (Fig. 3).

Figure 3

RT–PCR for seven predicted operons. The DNA bands represent the PCR products for the following operons, including the expected size of the PCR product in parentheses (see also Supplementary Material, Table S1): lane 1, rpsR–rplI (0.9 kb); lane 2, ytfQ–ytfR (2.4 kb); lane 3, yaeR–mesJ (1.4 kb); lane 4, rplA–rplL (1.5 kb); lane 5, ptsH–ptsJ (1.7 kb); lane 6, purA–yjeB (1.8 kb); lane 7, hnr–galU (2 kb); lane St, 1 kb standard DNA ladder.

Figure 4

5′-UTR detection upstream of ompA. Individual oligonucleotide probe intensities (PM – MM) from three conditions are shown to validate the microarray-detected 5′-UTR upstream of ompA. Intensities for individual oligonucleotide probes interrogating ompA, the 356 bp intergenic region and sulA are shown. The arrows above the indicated genes show the direction of transcription.