Contrary to what’s typically described in biology textbooks, bacteria and archaea can have transcription proceed in opposite directions on the genome. This occurs thanks to bidirectional promoters—DNA sequences where RNA polymerases can hop on and travel one way or the other to produce mRNA transcripts. Such promoters aren’t rare occurrences: 19 percent of all transcription start sites (TSSs) in Escherichia coli are associated with a bidirectional promotor, according to a study published May 6 in Nature Microbiology

“We were really surprised,” says study coauthor Emily Warman, a molecular microbiology postdoc at the University of Birmingham in the UK. While previous research had described bidirectional promoters in eukaryotes, as well as in a few bacteria and archaea species, the new study establishes divergent transcription—the reading of genes in both directions—as a widespread feature conserved across all three domains of life. 

Bidirectional promoters across biology

In eukaryotic cells, DNA winds around histone proteins and is packaged into chromatin. Stretches of DNA that aren’t tightly coiled are accessible to RNA polymerase and other proteins needed for transcription. In some cases, these regions contain two TSSs, one on each strand of DNA, oriented in opposite directions; these TSSs can be separated by hundreds or thousands of base pairs. Scientists have identified these types of bidirectional promoters in a variety of eukaryotic cells, from yeast to mouse macrophages. 

Bacteria don’t have histones. But some do have histone-like nucleoid structuring (H-NS) protein, which binds to DNA and aids in folding bacterial chromosomes. In a 2014 study published in Genes & Development, researchers found that in E. coli, H-NS suppresses promoters picked up via horizontal gene transfer, which is the transmission of genetic material between organisms outside of reproduction. Intriguingly, they noted that many of the H-NS–suppressed promoters were for noncoding RNAs and located in the middle of other genes. 

One of Warman’s first jobs when she was a PhD student in David Grainger’s lab at Birmingham was to characterize the activity of those promoters by inserting them in front of a reporter gene in a plasmid and measuring the resulting gene expression. “A lot of the information we had about these regions didn’t tell us which direction transcription would run in,” she says. “To cover all our bases, I was taking all these regions and I was putting them in both ways.” Surprisingly, many of the promoter fragments produced activity in either orientation, meaning that the same segment of DNA could drive transcription in both directions.  

To explore whether bidirectional promoters were common throughout the E. coli genome, the team analyzed previously obtained datasets that mapped TSSs. They found 5,292 divergent TSSs, which were between 7 and 25 base pairs apart but on different strands of DNA. These TSS pairs accounted for 19 percent of all TSSs in E. coli. The most common distance between the sites was 18 base pairs—much closer than the distances observed in eukaryotic cells. This close spacing positions promoter elements, DNA sequences that are critical for RNA polymerase recruitment, opposite each other on the two DNA strands. Thus, the authors propose, RNA polymerase can attach the same section of DNA in two different orientations and proceed to initiate transcription in either direction. 

Bidirectional promoters include two closely positioned transcriptional start sites (TSSs), one on each strand of DNA. These sites are upstream of gene coding regions. Researchers found that in E. coli, such TSSs are commonly separated by 18 base pairs, a distance that places promoter elements that are required for recruitment of RNA polymerase opposite each other on the two DNA strands.
The Scientist staff

They went on to examine TSSs from more bacteria species and found that divergent pairs were abundant. In proteobacteria and actinobacteria, the TSS pairs were typically 18 or 19 base pairs apart. The team also looked at previously published TSS maps for two archaeal species and uncovered many divergent TSS pairs. 

“Bidirectional transcription is a feature of eukaryotic transcription as well, but importantly, this paper shows that the mechanism in bacteria is different than the mechanism in eukaryotes,” Seth Darst, a biophysicist at the Rockefeller University who was not involved in the study, tells The Scientist by email.  

In a 2018 study published in BMC Genomics, researchers reported a similar finding in Pseudomonas aeruginosa, a pathogen that causes infections in humans. They found 105 TSS pairs on opposite strands, exactly 18 base pairs apart. 

“We just looked at Pseudomonas, and we found them and thought it was unusual,” says Peter Unrau, a biochemist at Simon Fraser University in British Columbia and a coauthor of the 2018 study. “They’re apparently everywhere in the bacteria and the archaea—so that’s actually really cool.” 

Bidirectional gene regulation

The authors propose that bidirectional promoters could allow for coordinated regulation of genes running in opposite directions. For example, transcription factors that bind a promoter region could modulate the expression of two adjacent genes simultaneously. These molecular details of this and other potential forms of RNA-dependent regulation are still open questions, Unrau says. 

In a 2019 study published in Nature Microbiology, Shixin Liu, a biophysicist at the Rockefeller University, and colleagues made a complementary discovery about transcription in E. coli: some convergent genes running toward each other share a bidirectional transcription terminator.  

Bacteria have relatively compact genomes, says Liu. “These [bidirectional elements] seem to be a way to encode more sophisticated regulatory functions in their small genomes, so that one promoter can control two divergent genes, or one terminator can simultaneously control two convergent genes.”  

The prevalence of bidirectional promoters could be noteworthy for biotechnology applications, where scientists aim to use efficient promoters to generate gene products, Warman says. “I think it’s just something that anyone who’s interested in gene expression needs to be aware of.”