Hierarchical evolution of the bacterial sporulation network

Abstract

Genome sequencing of multiple species makes it possible to understand the main principles behind the evolution of developmental regulatory networks. It is especially interesting to analyze the evolution of well-defined model systems in which conservation patterns can be directly correlated with the functional roles of various network components. Endospore formation (sporulation), extensively studied in Bacillus subtilis, is driven by such a model bacterial network of cellular development and differentiation. In this review, we analyze the evolution of the sporulation network in multiple endospore-forming bacteria. Importantly, the network evolution is not random but primarily follows the hierarchical organization and functional logic of the sporulation process. Specifically, the sporulation sigma factors and the master regulator of sporulation, Spo0A, are conserved in all considered spore-formers. The sequential activation of these global regulators is also strongly conserved. The feed-forward loops, which are likely used to fine-tune waves of gene expression within regulatory modules, show an intermediate level of conservation. These loops are less conserved than the sigma factors but significantly more than the structural sporulation genes, which form the lowest level in the functional and evolutionary hierarchy of the sporulation network. Interestingly, in spore-forming bacteria, gene regulation is more conserved than gene presence for sporulation genes, while the opposite is true for non-sporulation genes. The observed patterns suggest that, by understanding the functional organization of a developmental network in a model organism, it is possible to understand the logic behind the evolution of this network in multiple related species.

Figure 1. Morphological stages of the B. subtilis life cycle

The temporal and compartment-specific activity of each sporulation sigma (σ factor is indicated. During vegetative growth, cells divide by binary fission to generate two identical daughter cells. Sporulation is initiated in response to starvation. In the predivisional sporulating cell, the chromosomes (red) are oriented with their origin-proximal region anchored at the cell poles. During asymmetric division, two membrane-bounded compartments are generated: a small forespore and a large mother cell. After asymmetric division, the remainder of the forespore chromosome (i.e. the origin-distal region) is pulled into the forespore by translocation. Engulfment of the forespore by the mother cell results in the release of the forespore as a free protoplast in the mother cell. The cortex (composed of modified peptidoglycan, gray) is synthesized between the two membranes surrounding the forespore. The coat (black) is a complex structure made of at least 70 distinct proteins that assemble around the forespore surface. Following mother cell lysis, the mature spore is released in the environment. B. subtilis cells can remain in a dormant spore state for an extended period of time, but spores will germinate in response to presence of small molecules (e.g. single amino acids, sugars or fragments of peptidoglycan) and resume vegetative growth.

Figure 2. Modular architecture of the sporulation regulatory network in B. subtilis

The temporal progression of sporulation is from the top to bottom. Each cellular compartment (pre-divisional cell, forespore and mother cell) is surrounded by dotted lines. Sigma factors are shown in purple, transcription factors in green, and regulons in yellow. Signaling proteins are shown in red, and the feeding tube in olive. Transcriptional regulation is indicated by black arrows, gene expression (protein synthesis) by blue arrows, and signaling pathways by red arrows. Coherent feed-forward loops are indicated by a plus sign, and incoherent feed-forward loops by a minus sign.

Figure 3. Phylogenetic tree of 24 representative endospore-forming species

The tree was calculated using PHYLIP program [97] based on 16S ribosomal RNA sequences from the Ribosomal Database Project [98]. B. subtilis belongs to the cluster (yellow) of 12 species of aerobic bacteria from the class Bacilli. The other cluster (cyan) includes 12 species from the anaerobic class Clostridia. The anaerobic Clostridia cluster can be subdivided further into two subclusters: one corresponding to the Clostridiaceae family, which includes C. acetobutylicum, the other including the Thermoanaerobacteraceae and Peptococcaceae families.

Figure 4. Conservation of gene presence and regulation in the sporulation regulatory network for the 24 representative spore-formers. (a) The percentage of conserved genes b) The percentage of conserved regulations, given the presence of the regulated gene

The conservations are shown in percentages compared to B. subtilis. Conservation of gene presence was established using bidirectional Blast hits, while conservation of gene regulation was established by searching for sigma factor binding sites using the PSSMs available in DBTBS [13]. The figure demonstrates that the observed conservation patterns follow the functional and structural hierarchy of the sporulation network: the sigma factors (red) are most conserved, followed by feed-forward loops (green) and inter-compartmental interactions (blue), followed by all downstream genes that are directly regulated by the sporulation network (yellow).

Figure 5. Conservation of gene presence (a) and regulation (b) in different functional categories of the sporulation network

The percentages are shown for gene regulated by the sporulation sigma factors (σ^F, σ^G, σ^E, σ^K, σ^H) in the 24 representative organisms. The following functional categories are considered: 1.) Sporulation (179 genes, brown): Genes encoding proteins specifically involved in sporulation, except for spore coat proteins, regulation, cell wall biosynthesis, and metabolism. 2.) Metabolism (45 genes, blue): Genes encoding metabolic enzymes. 3.) Coat (38 genes, yellow): Genes encoding spore coat proteins. 4.) Regulation (31 genes, orange): Genes encoding proteins with regulatory functions (e.g. transcription factors and proteins involved in signaling) 5.) Cell wall or spore cortex (14 genes, purple): Genes whose products are involved in cell wall biosynthesis or hydrolysis. Interestingly, while the conservation of genes presence is weak for coat proteins, conservation of its regulation is relatively strong.

Figure 6. The estimated conservation of regulatory interactions for genes regulated by sporulation (red) and non-sporulation (green) sigma factors

The conservation is shown in percentages compared to B. subtilis. The conservation results were corrected for the limited sensitivity of predicting sigma factor binding sites using PSSMs. (a) The estimated conservation of regulation, given the presence of regulated gene, for the 307 genes regulated by the sporulation sigma factors, and the 211 genes regulated by the non-sporulation sigma factors. (b) The conservation of gene presence versus the estimated conservation of regulation for the 24 representative organisms. The conservation is shown in red for regulations by the sporulation sigma factors (σ^F, σ^G, σ^E, σ^K, σ^H), and in green for regulations by the non-sporulation sigma factors (σ^B, σ^D, σ^W, σ^X). The initials of each organism name are shown, with B. subtilis appearing in the upper-right corner. The solid red and green lines represent a linear fit between the conservation of gene presence and the estimated conservation of gene regulation for sporulation and non-sporulation sigma factors, respectively. The dashed diagonal line (in blue) corresponds to the equal conservation of gene presence and gene regulation. Interestingly, for the sporulation sigma factors, gene presence evolves faster than gene regulation, while the reverse is true for the non-sporulation sigma factors.