Periodic Reporting for period 1 - EBM (The evolvability of bacterial multicellularity)
Berichtszeitraum: 2018-09-01 bis 2020-08-31
The evolution of multicellularity transformed life on earth. There are numerous independent origins of multicellularity across the tree of life, many of which resulted in a bewildering diversity of organismal forms and function. Yet, how multicellularity evolves and what determines their evolvability, i.e. their capacity to undergo evolutionary change, remains largely unknown. Here, we investigated the evolvability of one of the simplest multicellular phenotypes: bacterial colony formation. Bacterial colonies are widespread in nature and perform many functions relevant to humans, either because they are harmful (e.g. pathogenicity) or because they have beneficial side-effects (e.g. biocontrol agents in agriculture). Despite their simplicity, bacterial colonies can show a large diversity. This is arguably most apparent in the Bacilli, a group of highly potent colony formers that grow in the soil. There is a good understanding of the gene regulatory network underlying colony formation in the Bacilli, but it is unknown how this network evolves and thereby shapes the evolvability of colony formation. Our aim was to understand the evolvability of colony formation in the Bacilli. We focused on three primary research objectives. In the first research objectives, we examined how the gene regulatory network underlying colony formation evolved across the phylogenetic tree of the Bacillales. In the second research objective, we determined how mutations in the gene regulatory network affect colony-level properties. In the third and final research objective, we determined the functional benefits of these mutations. Why are certain colony-level properties favored or disfavored by selection? Through these research objectives, we strongly advanced our understanding of colony formation in the Bacilli, which may ultimately allow us to control those colonies to the benefit of human health and ecosystem functioning.
Here, we provide a brief overview of the research performed in each research objective and the main results obtained thus far.
(1) Research objective 1: Phylogenetic analysis
Before starting the phylogenetic analysis, we first performed an in-depth characterization of the gene regulatory network in Bacillus subtilis. This showed a strong modular organization, where sets of transcriptional regulators regulate the same downstream genes. Using a machine learning approach, called auto-associative artificial neural networks, we subsequently analyzed hundreds of previously acquire transcriptomic profiles. This revealed that the identified regulatory modules are sequentially expressed over the life cycle of Bacillus subtilis. In other words, the modularity in gene regulation reflects the modularity in life cycle biology with three more-or-less discrete life stages: motility, colony formation and sporulation. We examined how the regulatory modules were conserved across the Bacillales, by comparing more than 380 Bacillales genomes to that of Bacillus subtilis. This revealed a striking mosaic conservation patterns, where regulatory modules were lost and conserved as discrete units. Interestingly, genes underlying colony formation showed an overall strong conservation across the Bacillales, suggesting that the gene regulatory network underlying colony formation is essential for survival in a wide range of ecological conditions.
(2) Research objective 2: Mutants and colony-level properties
We performed a large-scale evolution experiment to determine how genetic mutations, which spontaneously arise in the lab, affect colony-level properties. To account for a wide diversity of genetic backgrounds, we selected eight Bacillales strains/species and exposed them to a simple selection regime where colonies from each strain/species were grown on an agar plate for one week, after which cells from the outermost colony edge where selected and transferred to a fresh agar plate. The same growth cycle was performed for over 10 consecutive weeks. Each week a small volume of cell suspension was stored in the -80C freezer for further analyses.
Interestingly, the strains showed distinct evolutionary response: whereas some strains underwent rapid evolutionary change in colony size, colonies from other strains remained practically unaltered for the duration of the experiment. This suggests that strains differ strongly in their capacity to respond to the imposed selection regime. For further analysis, we sequenced all strains that showed a strong phenotypic response. We sequenced both the ancestor and evolved clones, as well as a few clones from intermediate stages of the evolution experiment. Interestingly, most mutations enhancing colony growth occurred in the global transcription regulators identified under research objective 1, resulting in large expression changes in the downstream regulatory modules.
(3) Research objective 3: Functional benefit of colony-level properties
In order to characterize the mutants in more depth, we performed common garden experiments, where, for each mutant, we determined the colony growth dynamics, their cell state composition and gene expression profiles. This analysis revealed that mutants showed dramatic changes in the cell state composition with accompanying gene expression changes. First, many colonies showed highly reduced or complete loss of sporulation. Sporulation was likely selected against because it delays the outgrow of cells after transfer to a fresh agar plate. Second, some colonies showed increase in fraction of filamentous cells. Filamentous cells have previously been shown to enhance colony spreading. Third, some colonies showed altered rates of polysaccharide production, which in combination with filament formation might have further boosted colony expansion.
(1) Research objective 1: Phylogenetic analysis
Before starting the phylogenetic analysis, we first performed an in-depth characterization of the gene regulatory network in Bacillus subtilis. This showed a strong modular organization, where sets of transcriptional regulators regulate the same downstream genes. Using a machine learning approach, called auto-associative artificial neural networks, we subsequently analyzed hundreds of previously acquire transcriptomic profiles. This revealed that the identified regulatory modules are sequentially expressed over the life cycle of Bacillus subtilis. In other words, the modularity in gene regulation reflects the modularity in life cycle biology with three more-or-less discrete life stages: motility, colony formation and sporulation. We examined how the regulatory modules were conserved across the Bacillales, by comparing more than 380 Bacillales genomes to that of Bacillus subtilis. This revealed a striking mosaic conservation patterns, where regulatory modules were lost and conserved as discrete units. Interestingly, genes underlying colony formation showed an overall strong conservation across the Bacillales, suggesting that the gene regulatory network underlying colony formation is essential for survival in a wide range of ecological conditions.
(2) Research objective 2: Mutants and colony-level properties
We performed a large-scale evolution experiment to determine how genetic mutations, which spontaneously arise in the lab, affect colony-level properties. To account for a wide diversity of genetic backgrounds, we selected eight Bacillales strains/species and exposed them to a simple selection regime where colonies from each strain/species were grown on an agar plate for one week, after which cells from the outermost colony edge where selected and transferred to a fresh agar plate. The same growth cycle was performed for over 10 consecutive weeks. Each week a small volume of cell suspension was stored in the -80C freezer for further analyses.
Interestingly, the strains showed distinct evolutionary response: whereas some strains underwent rapid evolutionary change in colony size, colonies from other strains remained practically unaltered for the duration of the experiment. This suggests that strains differ strongly in their capacity to respond to the imposed selection regime. For further analysis, we sequenced all strains that showed a strong phenotypic response. We sequenced both the ancestor and evolved clones, as well as a few clones from intermediate stages of the evolution experiment. Interestingly, most mutations enhancing colony growth occurred in the global transcription regulators identified under research objective 1, resulting in large expression changes in the downstream regulatory modules.
(3) Research objective 3: Functional benefit of colony-level properties
In order to characterize the mutants in more depth, we performed common garden experiments, where, for each mutant, we determined the colony growth dynamics, their cell state composition and gene expression profiles. This analysis revealed that mutants showed dramatic changes in the cell state composition with accompanying gene expression changes. First, many colonies showed highly reduced or complete loss of sporulation. Sporulation was likely selected against because it delays the outgrow of cells after transfer to a fresh agar plate. Second, some colonies showed increase in fraction of filamentous cells. Filamentous cells have previously been shown to enhance colony spreading. Third, some colonies showed altered rates of polysaccharide production, which in combination with filament formation might have further boosted colony expansion.
The project made a number important advances, including both biological advances and technical advances:
(1) Biological advances:
(i) We show that the life cycle of Bacillus subtilis is characterized by more-or-less discrete life stages, each of which expresses a unique set of regulatory modules.
(ii) We showed that regulatory modules show parallel evolutionary changes over both the short time span of our evolution experiment and the long time span of species divergence in the Bacillales.
(iii) We showed that species show dramatic differences in their responsiveness to the same selective regime in the lab. The role of the genetic background has been highlighted in previous evolution experiments before, but we substantiate this evidence for multicellular phenotypes as well.
(2) Technical advances:
(i) We for the first time analyzed hundreds of bacterial transcriptomic profiles using a machine learning approach called auto-associative artificial neural network. Although this type of pseudotime analysis has successfully been applied to eukaryotes before, it was never employed to analyze prokaryotic life cycles.
(ii) We developed a high-throughput screening approach, using flow cytometry, to determine the cell state composition in Bacilli colonies. A simple staining procedure allows one to recognize multiple cell states and screen millions of cells in a short amount of time. This high-throughput screen can be applied across species and will form an important tool for future research, e.g. characterization of mutant libraries.
(1) Biological advances:
(i) We show that the life cycle of Bacillus subtilis is characterized by more-or-less discrete life stages, each of which expresses a unique set of regulatory modules.
(ii) We showed that regulatory modules show parallel evolutionary changes over both the short time span of our evolution experiment and the long time span of species divergence in the Bacillales.
(iii) We showed that species show dramatic differences in their responsiveness to the same selective regime in the lab. The role of the genetic background has been highlighted in previous evolution experiments before, but we substantiate this evidence for multicellular phenotypes as well.
(2) Technical advances:
(i) We for the first time analyzed hundreds of bacterial transcriptomic profiles using a machine learning approach called auto-associative artificial neural network. Although this type of pseudotime analysis has successfully been applied to eukaryotes before, it was never employed to analyze prokaryotic life cycles.
(ii) We developed a high-throughput screening approach, using flow cytometry, to determine the cell state composition in Bacilli colonies. A simple staining procedure allows one to recognize multiple cell states and screen millions of cells in a short amount of time. This high-throughput screen can be applied across species and will form an important tool for future research, e.g. characterization of mutant libraries.