Nucleoid-associated proteins shape chromatin structure and transcriptional regulation across the bacterial kingdom

Abstract

Genome architecture has proven to be critical in determining gene regulation across almost all domains of life. While many of the key components and mechanisms of eukaryotic genome organization have been described, the interplay between bacterial DNA organization and gene regulation is only now being fully appreciated. An increasing pool of evidence has demonstrated that the bacterial chromosome can reasonably be thought of as chromatin, and that bacterial chromosomes contain transcriptionally silent and transcriptionally active regions analogous to heterochromatin and euchromatin, respectively. The roles played by histones in eukaryotic systems appear to be shared across a range of nucleoid-associated proteins (NAPs) in bacteria, which function to compact, structure, and regulate large portions of bacterial chromosomes. The broad range of extant NAPs, and the extent to which they differ from species to species, has raised additional challenges in identifying and characterizing their roles in all but a handful of model bacteria. Here we review the regulatory roles played by NAPs in several well-studied bacteria and use the resulting state of knowledge to provide a working definition for NAPs, based on their function, binding pattern, and expression levels. We present a screening procedure which can be applied to any species for which transcriptomic data are available. Finally, we note that NAPs tend to play two major regulatory roles - xenogeneic silencers and developmental regulators - and that many unrecognized potential NAPs exist in each bacterial species examined.

Figure 1.

Nucleoid associated proteins mediate a variety of DNA conformations and are abundant in different stages of growth. Overview of highly abundant NAPs and their relative abundance in the different stages of growth in E. coli. Single subunit molecular weights (kD), oligomeric states, and binding preferences were obtained from [221], binding capabilities (filaments, bendings, etc.) were obtained from [28,221]. The H-NS binding depicted is an example of a bridged DNA filament formed between Hha, StpA, and H-NS

Figure 2.

Comparison of abundance of NAPs to a local regulator. Molecules per cell (log₂) of nucleoid associated proteins (NAPs; green) and the transcription factor LacI (gray) in E. coli grown in rich media (data from [221])

Figure 3.

Identification of known and potential E. coli NAPs through expression analysis. Left: Plots of the expression levels of all genes flagged as NAP candidates by our screening approach (see Text for details); all data are taken in glucose minimal media (raw sequencing data are provided in the github link from the main text). Genes are categorized by manual curation (based on information from Ecocyc [221]) into global regulators, local regulators, classical nucleoid-associated proteins (NAP), possible new nucleoid-associated proteins (Potential NAP), or DNA binding proteins with non-regulatory primary functions (Other). Red dashed lines indicate the 80th and 90th percentile thresholds used to filter for potential NAPs. Right: Correlation of RNA levels and protein production per cell cycle (data from [344]) for all E. coli proteins identified in our screen; all data are taken in glucose minimal media

Figure 4.

Identification of known and potential B. subtilis NAPs through expression analysis. Left: Plot of the expression levels of all genes flagged as NAP candidates by our screening approach (see Text for details); vegetative growth refers to the 8-hour time point in [345], all data were drawn from [345]. Red dashed lines indicate the 80th and 90th percentile thresholds used to filter for potential NAPs. Genes were categorized by manual curation (based on information from Subtiwiki [346]) into the categories described for Figure 3. Right: Expansion of inset from plot on the left

Figure 5.

Identification of putative NAPs from example Gram-negative species. Shown are all proteins in P. aeruginosa (left) and N. meningitidis that pass the NAP-screening procedures described in the text. The hits were then manually assigned to one of the functional categories shown in Figure 3 on the basis of their annotations and expression levels, with the unknown category reserved primarily for lower-expression cases with unclear functions. Red dashed lines indicate the 80th and 90th percentile thresholds used to filter for potential NAPs. All transcriptional data were processed from [347], using the “control” and “stationary phase” conditions for each organism