A Novel pH-Regulated, Unusual 603 bp Overlapping Protein Coding Gene pop Is Encoded Antisense to ompA in Escherichia coli O157:H7 (EHEC)

Abstract

Antisense transcription is well known in bacteria. However, translation of antisense RNAs is typically not considered, as the implied overlapping coding at a DNA locus is assumed to be highly improbable. Therefore, such overlapping genes are systematically excluded in prokaryotic genome annotation. Here we report an exceptional 603 bp long open reading frame completely embedded in antisense to the gene of the outer membrane protein ompA. An active σ 70 promoter, transcription start site (TSS), Shine-Dalgarno motif and rho-independent terminator were experimentally validated, providing evidence that this open reading frame has all the structural features of a functional gene. Furthermore, ribosomal profiling revealed translation of the mRNA, the protein was detected in Western blots and a pH-dependent phenotype conferred by the protein was shown in competitive overexpression growth experiments of a translationally arrested mutant versus wild type. We designate this novel gene pop (pH-regulated overlapping protein-coding gene), thus adding another example to the growing list of overlapping, protein coding genes in bacteria.

Evidence for Numerous Embedded Antisense Overlapping Genes in DiverseE. coliStrains

“…Error bars indicate standard deviations. Of note, OGC57 and OGC59, designated pop and asa , respectively, were studied in detail elsewhere (Vanderhaeghen et al, 2018; Zehentner et al, 2020).…”

“…Using this approach, almost-identical RNAs are expressed, which should not disturb an ncRNA’s function (Bobrovskyy and Vanderpool, 2013), while production of the protein ceases in the mutant. We have tested whether the plasmid copy number changes between a plasmid with the wild-type gene or the mutated gene while growing in competition, showing that this is not the case (Zehentner et al, 2020). The phenotype for the candidate genes was tested in conditions effective in the pooled experiment.…”

“…Ribosome profiling data were taken from the following studies: Data for E. coli strain Sakai (grown in LB) were taken from Hücker et al (2017). Data for E. coli K-12 (grown in LB) were taken from Woolstenhulme et al (2015), for E. coli strain BL21(DE3) grown in LB (abbreviated as BL21) from Meydan et al (2019), for E. coli strain LF82, grown in Schaedler broth as described in Zehentner et al (2020). For the latter, data was deposited in the NCBI Sequence Read Archive (accession number PRJNA654277).…”

“…Error bars indicate standard deviations. Of note, OGC57 and OGC59, designated pop and asa, respectively, were studied in detail elsewhere (Vanderhaeghen et al, 2018;Zehentner et al, 2020). In each strain, ORFs are shown which either have at least 1 read per million mapped reads (RPM) and 30% coverage (green), at least 1 RPM and 60% coverage (blue) or at least 1 RPM, 60% coverage, and also homologous to an ORF expressed above these thresholds in EDL933 (orange).…”

“…The potential widespread functionality of such overlapping ORFs has recently been proposed as a hypothesis to investigate, in light of evidence across diverse bacteria (Ardern et al, 2020). However, only very few cases of such gene pairs, in which their ORFs overlap extensively, have been characterized phenotypically in E. coli (Fellner et al, 2014;Fellner et al, 2015;Haycocks and Grainger, 2016;Hücker et al, 2018a;Hücker et al, 2018b;Vanderhaeghen et al, 2018;Zehentner et al, 2020) in strong contrast to more than 128.000 non-overlapping genes known in the E. coli pan-genome (Park et al, 2019).…”