Moving toward the Inclusion of Epigenomics in Bacterial Genome Evolution: Perspectives and Challenges.

The universality of DNA methylation as an epigenetic regulatory mechanism belongs to all biological kingdoms. However, while eukaryotic systems have been the primary focus of DNA methylation studies, the molecular mechanisms in prokaryotes are less known. Nevertheless, DNA methylation in prokaryotes plays a pivotal role in many cellular processes such as defense systems against exogenous DNA, cell cycle dynamics, and gene expression, including virulence. Thanks to single-molecule DNA sequencing technologies, genome-wide identification of methylated DNA is becoming feasible on a large scale, providing the possibility to investigate more deeply the presence, variability, and roles of DNA methylation. Here, we present an overview of the multifaceted roles of DNA methylation in prokaryotes and suggest research directions and tools which can enable us to better understand the contribution of DNA methylation to prokaryotic genome evolution and adaptation. In particular, we emphasize the need to understand the presence and role of transgenerational inheritance, as well as the impact of epigenomic signatures on adaptation and genome evolution. Research directions and the importance of novel computational tools are underlined.

Ruthenium(II) Polypyridyl Complexes and Metronidazole Derivatives: A Powerful Combination in the Design of Photoresponsive Antibacterial Agents Effective under Hypoxic Conditions.

Ruthenium(II) polypyridyl complexes (RPCs) are gaining momentum in photoactivated chemotherapy (PACT), thanks to the possibility of overcoming the classical reliance on molecular oxygen of photodynamic therapy while preserving the selective drug activation by using light. However, notwithstanding the intriguing perspectives, the translation of such an approach in the development of new antimicrobials has been only barely considered. Herein, MTZH-1 and MTZH-2, two novel analogues of metronidazole (MTZ), a mainstay drug in the treatment of anaerobic bacterial infections, were designed and inserted in the strained ruthenium complexes [Ru(tpy)(dmp)(MTZ-1)]PF6 (Ru2) and [Ru(tpy)(dmp)(MTZ-2)]PF6 (Ru3) (tpy = terpyridine, dmp = 2,9-dimethyl-1,10-phenanthroline) (Chart 1). Analogously to the parental compound [Ru(tpy)(dmp)(5NIM)]PF6 (Ru1) (5-nitroimidazolate), the Ru(II)-imidazolate coordination of MTZ derivatives resulted in promising Ru(II) photocages, capable to easily unleash the bioactive ligands upon light irradiation and increase the antibacterial activity against Bacillus subtilis, which was chosen as a model of Gram-positive bacteria. The photoreleased 5-nitroimidazole-based ligands led to remarkable phototoxicities under hypoxic conditions (<1% O2), with the lead compound Ru3 that exhibited the highest potency across the series, being comparable to the one of the clinical drug MTZ. Besides, the chemical architectures of MTZ derivatives made their interaction with NimAunfavorable, being NimA a model of reductases responsible for bacterial resistance against 5-nitroimidazole-based antibiotics, thus hinting at their possible use to combat antimicrobial resistance. This work may therefore provide fundamental knowledge in the design of novel photoresponsive tools to be used in the fight against infectious diseases. For the first time, the effectiveness of the "photorelease antimicrobial therapy" under therapeutically relevant hypoxic conditions was demonstrated.

A Meta-Analysis Approach to Defining the Culturable Core of Plant Endophytic Bacterial Communities.

Endophytic bacteria are key members of the plant microbiome, which phylogenetic diversity has been widely described through next-generation sequencing technologies in the last decades. On the other side, a synopsis of culturable plant endophytic bacteria is still lacking in the literature. However, culturability is necessary for biotechnology innovations related to sustainable agriculture, such as biofertilizer and biostimulant agents' development. In this review, 148 scientific papers were analyzed to establish a large data set of cultured endophytic bacteria, reported at the genus level, inhabiting different compartments of wild and farmed plants, sampled around the world from different soil types and isolated using various growth media. To the best of our knowledge, this work provides the first overview of the current repertoire of cultured plant endophytic bacteria. Results indicate the presence of a recurrent set of culturable bacterial genera regardless of factors known to influence the plant bacterial community composition and the growth media used for the bacterial isolation. Moreover, a wide variety of bacterial genera that are currently rarely isolated from the plant endosphere was identified, demonstrating that culturomics can catch previously uncultured bacteria from the plant microbiome, widening the panorama of strains exploitable to support plant holobiont health and production.

Nitroimidazole-Based Ruthenium(II) Complexes: Playing with Structural Parameters to Design Photostable and Light-Responsive Antibacterial Agents.

5-Nitroimidazole (5NIMH), chosen as a molecular model of nitroimidazole derivatives, which represent a broad-spectrum class of antimicrobials, was incorporated into the ruthenium complexes [Ru(tpy)(phen)(5NIM)]PF6 (1) and [Ru(tpy)(dmp)(5NIM)]PF6 (2) (tpy = terpyridine, phen = phenanthroline, dmp = 2,9-dimethyl-1,10-phenanthroline). Besides the uncommon metal coordination of 5-nitroimidazole in its imidazolate form (5NIM), the different architectures of the spectator ligands (phen and dmp) were exploited to tune the "mode of action" of the resulting complexes, passing from a photostable compound where the redox properties of 5NIMH are preserved (1) to one suitable for the nitroimidazole phototriggered release (2) and whose antibacterial activity against B. subtilis, chosen as cellular model, is effectively improved upon light exposure. This study may provide a fundamental knowledge on the use of Ru(II)-polypyridyl complexes to incorporate and/or photorelease biologically relevant nitroimidazole derivatives in the design of a novel class of antimicrobials.

Scent of a Symbiont: The Personalized Genetic Relationships of Rhizobium-Plant Interaction.

Many molecular signals are exchanged between rhizobia and host legume plants, some of which are crucial for symbiosis to take place, while others are modifiers of the interaction, which have great importance in the competition with the soil microbiota and in the genotype-specific perception of host plants. Here, we review recent findings on strain-specific and host genotype-specific interactions between rhizobia and legumes, discussing the molecular actors (genes, gene products and metabolites) which play a role in the establishment of symbiosis, and highlighting the need for research including the other components of the soil (micro)biota, which could be crucial in developing rational-based strategies for bioinoculants and synthetic communities' assemblage.

Crossing Bacterial Genomic Features and Methylation Patterns with MeStudio: An Epigenomic Analysis Tool.

DNA methylation is one of the most observed epigenetic modifications. It is present in eukaryotes and prokaryotes and is related to several biological phenomena, including gene flow and adaptation to environmental conditions. The widespread use of third-generation sequencing technologies allows direct and easy detection of genome-wide methylation profiles, offering increasing opportunities to understand and exploit the epigenomic landscape of individuals and populations. Here, we present a pipeline named MeStudio, with the aim of analyzing and combining genome-wide methylation profiles with genomic features. Outputs report the presence of DNA methylation in coding sequences (CDSs) and noncoding sequences, including both intergenic sequences and sequences upstream of the CDS. We apply this novel tool, showing the usage and performance of MeStudio, on a set of single-molecule real-time sequencing outputs from strains of the bacterial species Sinorhizobium meliloti.

Differential Response of Wheat Rhizosphere Bacterial Community to Plant Variety and Fertilization.

The taxonomic assemblage and functions of the plant bacterial community are strongly influenced by soil and host plant genotype. Crop breeding, especially after the massive use of nitrogen fertilizers which led to varieties responding better to nitrogen fertilization, has implicitly modified the ability of the plant root to recruit an effective bacterial community. Among the priorities for harnessing the plant bacterial community, plant genotype-by-microbiome interactions are stirring attention. Here, we analyzed the effect of plant variety and fertilization on the rhizosphere bacterial community. In particular, we clarified the presence in the bacterial community of a varietal effect of N and P fertilization treatment. 16S rRNA gene amplicon sequence analysis of rhizospheric soil, collected from four wheat varieties grown under four N-P fertilization regimes, and quantification of functional bacterial genes involved in the nitrogen cycle (nifH; amoA; nirK and nosZ) were performed. Results showed that variety played the most important role and that treatments did not affect either bacterial community diversity or bacterial phyla abundance. Variety-specific response of rhizosphere bacterial community was detected, both in relation to taxa (Nitrospira) and metabolic functions. In particular, the changes related to amino acid and aerobic metabolism and abundance of genes involved in the nitrogen cycle (amoA and nosZ), suggested that plant variety may lead to functional changes in the cycling of the plant-assimilable nitrogen.

Robustness encoded across essential and accessory replicons of the ecologically versatile bacterium Sinorhizobium meliloti.

Bacterial genome evolution is characterized by gains, losses, and rearrangements of functional genetic segments. The extent to which large-scale genomic alterations influence genotype-phenotype relationships has not been investigated in a high-throughput manner. In the symbiotic soil bacterium Sinorhizobium meliloti, the genome is composed of a chromosome and two large extrachromosomal replicons (pSymA and pSymB, which together constitute 45% of the genome). Massively parallel transposon insertion sequencing (Tn-seq) was employed to evaluate the contributions of chromosomal genes to growth fitness in both the presence and absence of these extrachromosomal replicons. Ten percent of chromosomal genes from diverse functional categories are shown to genetically interact with pSymA and pSymB. These results demonstrate the pervasive robustness provided by the extrachromosomal replicons, which is further supported by constraint-based metabolic modeling. A comprehensive picture of core S. meliloti metabolism was generated through a Tn-seq-guided in silico metabolic network reconstruction, producing a core network encompassing 726 genes. This integrated approach facilitated functional assignments for previously uncharacterized genes, while also revealing that Tn-seq alone missed over a quarter of wild-type metabolism. This work highlights the many functional dependencies and epistatic relationships that may arise between bacterial replicons and across a genome, while also demonstrating how Tn-seq and metabolic modeling can be used together to yield insights not obtainable by either method alone.

Lung and Gut Microbiota Changes Associated with Pseudomonas aeruginosa Infection in Mouse Models of Cystic Fibrosis.

Cystic fibrosis (CF) disease leads to altered lung and gut microbiomes compared to healthy subjects. The magnitude of this dysbiosis is influenced by organ-specific microenvironmental conditions at different stages of the disease. However, how this gut-lung dysbiosis is influenced by Pseudomonas aeruginosa chronic infection is unclear. To test the relationship between CFTR dysfunction and gut-lung microbiome under chronic infection, we established a model of P. aeruginosa infection in wild-type (WT) and gut-corrected CF mice. Using 16S ribosomal RNA gene, we compared lung, stool, and gut microbiota of C57Bl/6 Cftr tm1UNCTgN(FABPCFTR) or WT mice at the naïve state or infected with P. aeruginosa. P. aeruginosa infection influences murine health significantly changing body weight both in CF and WT mice. Both stool and gut microbiota revealed significantly higher values of alpha diversity in WT mice than in CF mice, while lung microbiota showed similar values. Infection with P. aeruginosa did not changed the diversity of the stool and gut microbiota, while a drop of diversity of the lung microbiota was observed compared to non-infected mice. However, the taxonomic composition of gut microbiota was shown to be influenced by P. aeruginosa infection in CF mice but not in WT mice. This finding indicates that P. aeruginosa chronic infection has a major impact on microbiota diversity and composition in the lung. In the gut, CFTR genotype and P. aeruginosa infection affected the overall diversity and taxonomic microbiota composition, respectively. Overall, our results suggest a cross-talk between lung and gut microbiota in relation to P. aeruginosa chronic infection and CFTR mutation.

Comparative genomics and pangenome-oriented studies reveal high homogeneity of the agronomically relevant enterobacterial plant pathogen Dickeya solani.

BACKGROUND: Dickeya solani is an important plant pathogenic bacterium causing severe losses in European potato production. This species draws a lot of attention due to its remarkable virulence, great devastating potential and easier spread in contrast to other Dickeya spp. In view of a high need for extensive studies on economically important soft rot Pectobacteriaceae, we performed a comparative genomics analysis on D. solani strains to search for genetic foundations that would explain the differences in the observed virulence levels within the D. solani population. RESULTS: High quality assemblies of 8 de novo sequenced D. solani genomes have been obtained. Whole-sequence comparison, ANIb, ANIm, Tetra and pangenome-oriented analyses performed on these genomes and the sequences of 14 additional strains revealed an exceptionally high level of homogeneity among the studied genetic material of D. solani strains. With the use of 22 genomes, the pangenome of D. solani, comprising 84.7% core, 7.2% accessory and 8.1% unique genes, has been almost completely determined, suggesting the presence of a nearly closed pangenome structure. Attribution of the genes included in the D. solani pangenome fractions to functional COG categories showed that higher percentages of accessory and unique pangenome parts in contrast to the core section are encountered in phage/mobile elements- and transcription- associated groups with the genome of RNS 05.1.2A strain having the most significant impact. Also, the first D. solani large-scale genome-wide phylogeny computed on concatenated core gene alignments is herein reported. CONCLUSIONS: The almost closed status of D. solani pangenome achieved in this work points to the fact that the unique gene pool of this species should no longer expand. Such a feature is characteristic of taxa whose representatives either occupy isolated ecological niches or lack efficient mechanisms for gene exchange and recombination, which seems rational concerning a strictly pathogenic species with clonal population structure. Finally, no obvious correlations between the geographical origin of D. solani strains and their phylogeny were found, which might reflect the specificity of the international seed potato market.

Chromids Aid Genome Expansion and Functional Diversification in the Family Burkholderiaceae.

Multipartite genomes, containing at least two large replicons, are found in diverse bacteria; however, the advantage of this genome structure remains incompletely understood. Here, we perform comparative genomics of hundreds of finished ß-proteobacterial genomes to gain insights into the role and emergence of multipartite genomes. Almost all essential secondary replicons (chromids) of the ß-proteobacteria are found in the family Burkholderiaceae. These replicons arose from just two plasmid acquisition events, and they were likely stabilized early in their evolution by the presence of core genes. On average, Burkholderiaceae genera with multipartite genomes had a larger total genome size, but smaller chromosome, than genera without secondary replicons. Pangenome-level functional enrichment analyses suggested that interreplicon functional biases are partially driven by the enrichment of secondary replicons in the accessory pangenome fraction. Nevertheless, the small overlap in orthologous groups present in each replicon's pangenome indicated a clear functional separation of the replicons. Chromids appeared biased to environmental adaptation, as the functional categories enriched on chromids were also overrepresented on the chromosomes of the environmental genera (Paraburkholderia and Cupriavidus) compared with the pathogenic genera (Burkholderia and Ralstonia). Using ancestral state reconstruction, it was predicted that the rate of accumulation of modern-day genes by chromids was more rapid than the rate of gene accumulation by the chromosomes. Overall, the data are consistent with a model where the primary advantage of secondary replicons is in facilitating increased rates of gene acquisition through horizontal gene transfer, consequently resulting in replicons enriched in genes associated with adaptation to novel environments.

Tissue specificity and differential effects on in vitro plant growth of single bacterial endophytes isolated from the roots, leaves and rhizospheric soil of Echinacea purpurea.

BACKGROUND: Echinacea-endophyte interaction might affect plant secondary metabolites content and influence bacterial colonization specificity and plant growth, but the underlying mechanisms need deepening. An in vitro model, in which E. purpurea axenic plants as host species and E. angustifolia and Nicotiana tabacum as non-host species inoculated with single endophytes isolated from stem/leaf, root and rhizospheric soil, were used to investigate bacterial colonization. RESULTS: Colonization analysis showed that bacteria tended to reach tissues from which they were originally isolated (tissue-specificity) in host plants but not in non-host ones (species-specificity). Primary root elongation inhibition as well as the promotion of the growth of E. purpurea and E. angustifolia plants were observed and related to endophyte-produced indole-3-Acetic Acid. Bacteria-secreted substances affected plant physiology probably interacting with plant regulators. Plant metabolites played an important role in controlling the endophyte growth. CONCLUSIONS: The proposed in vitro infection model could be, generally used to identify novel bioactive compounds and/or to select specific endophytes contributing to the host metabolism properties.

Multidisciplinary approaches for studying rhizobium-legume symbioses.

The rhizobium-legume symbiosis is a major source of fixed nitrogen (ammonia) in the biosphere. The potential for this process to increase agricultural yield while reducing the reliance on nitrogen-based fertilizers has generated interest in understanding and manipulating this process. For decades, rhizobium research has benefited from the use of leading techniques from a very broad set of fields, including population genetics, molecular genetics, genomics, and systems biology. In this review, we summarize many of the research strategies that have been employed in the study of rhizobia and the unique knowledge gained from these diverse tools, with a focus on genome- and systems-level approaches. We then describe ongoing synthetic biology approaches aimed at improving existing symbioses or engineering completely new symbiotic interactions. The review concludes with our perspective of the future directions and challenges of the field, with an emphasis on how the application of a multidisciplinary approach and the development of new methods will be necessary to ensure successful biotechnological manipulation of the symbiosis.

Cell Cycle Control by the Master Regulator CtrA in Sinorhizobium meliloti.

In all domains of life, proper regulation of the cell cycle is critical to coordinate genome replication, segregation and cell division. In some groups of bacteria, e.g. Alphaproteobacteria, tight regulation of the cell cycle is also necessary for the morphological and functional differentiation of cells. Sinorhizobium meliloti is an alphaproteobacterium that forms an economically and ecologically important nitrogen-fixing symbiosis with specific legume hosts. During this symbiosis S. meliloti undergoes an elaborate cellular differentiation within host root cells. The differentiation of S. meliloti results in massive amplification of the genome, cell branching and/or elongation, and loss of reproductive capacity. In Caulobacter crescentus, cellular differentiation is tightly linked to the cell cycle via the activity of the master regulator CtrA, and recent research in S. meliloti suggests that CtrA might also be key to cellular differentiation during symbiosis. However, the regulatory circuit driving cell cycle progression in S. meliloti is not well characterized in both the free-living and symbiotic state. Here, we investigated the regulation and function of CtrA in S. meliloti. We demonstrated that depletion of CtrA cause cell elongation, branching and genome amplification, similar to that observed in nitrogen-fixing bacteroids. We also showed that the cell cycle regulated proteolytic degradation of CtrA is essential in S. meliloti, suggesting a possible mechanism of CtrA depletion in differentiated bacteroids. Using a combination of ChIP-Seq and gene expression microarray analysis we found that although S. meliloti CtrA regulates similar processes as C. crescentus CtrA, it does so through different target genes. For example, our data suggest that CtrA does not control the expression of the Fts complex to control the timing of cell division during the cell cycle, but instead it negatively regulates the septum-inhibiting Min system. Our findings provide valuable insight into how highly conserved genetic networks can evolve, possibly to fit the diverse lifestyles of different bacteria.

Omics approaches on fresh-cut lettuce reveal global molecular responses to sodium hypochlorite and peracetic acid treatment.

BACKGROUND: Lettuce is a leafy vegetable that is extensively commercialized as a ready-to-eat product because of its widespread use in human nutrition as salad. It is well known that washing treatments can severely affect the quality and shelf-life of ready-to-eat vegetables. The study presented here evaluated the effect of two washing procedures on fresh-cut lettuce during storage. RESULTS: An omics approach was applied to reveal global changes at molecular level induced by peracetic acid washing in comparison with sodium hypochlorite treatment. Microbiological analyses were also performed to quantify total bacterial abundance and composition. The study revealed wide metabolic alterations induced by the two sanitizers. In particular, transcriptomic and proteomic analyses pointed out a number of transcripts and proteins differentially accumulated in response to peracetic acid washing, mainly occurring on the first day of storage. In parallel, different microbiota composition and significant reduction in total bacterial load following washing were also observed. CONCLUSION: The results provide useful information for the fresh-cut industry to select an appropriate washing procedure preserving fresh-like attributes as much as possible during storage of the end product. Molecular evidence indicated peracetic acid to be a valid alternative to sodium hypochlorite as sanitizer solution. © 2017 Society of Chemical Industry.

Subfunctionalization influences the expansion of bacterial multidrug antibiotic resistance.

BACKGROUND: Antibiotic resistance is a major problem for human health. Multidrug resistance efflux pumps, especially those of the Resistance-Nodulation-Cell Division (RND) family, are major contributors to high-level antibiotic resistance in Gram-negative bacteria. Most bacterial genomes contain several copies of the different classes of multidrug resistance efflux pumps. Gene duplication and gain of function by the duplicate copies of multidrug resistance efflux pump genes plays a key role in the expansion and diversification of drug-resistance mechanisms. RESULTS: We used two members of the Burkholderia RND superfamily as models to understand how duplication events affect the antibiotic resistance of these strains. First, we analyzed the conservation and distribution of these two RND systems and their regulators across the Burkholderia genus. Through genetic manipulations, we identified both the exact substrate range of these transporters and their eventual interchangeability. We also performed a directed evolution experiment, combined with next generation sequencing, to evaluate the role of antibiotics in the activation of the expression of these systems. Together, our results indicate that the first step to diversify the functions of these pumps arises from changes in their regulation (subfunctionalization) instead of functional mutations. Further, these pumps could rewire their regulation to respond to antibiotics, thus maintaining high genomic plasticity. CONCLUSIONS: Studying the regulatory network that controls the expression of the RND pumps will help understand and eventually control the development and expansion of drug resistance.

A Different Microbiome Gene Repertoire in the Airways of Cystic Fibrosis Patients with Severe Lung Disease.

In recent years, next-generation sequencing (NGS) was employed to decipher the structure and composition of the microbiota of the airways in cystic fibrosis (CF) patients. However, little is still known about the overall gene functions harbored by the resident microbial populations and which specific genes are associated with various stages of CF lung disease. In the present study, we aimed to identify the microbial gene repertoire of CF microbiota in twelve patients with severe and normal/mild lung disease by performing sputum shotgun metagenome sequencing. The abundance of metabolic pathways encoded by microbes inhabiting CF airways was reconstructed from the metagenome. We identified a set of metabolic pathways differently distributed in patients with different pulmonary function; namely, pathways related to bacterial chemotaxis and flagellar assembly, as well as genes encoding efflux-mediated antibiotic resistance mechanisms and virulence-related genes. The results indicated that the microbiome of CF patients with low pulmonary function is enriched in virulence-related genes and in genes encoding efflux-mediated antibiotic resistance mechanisms. Overall, the microbiome of severely affected adults with CF seems to encode different mechanisms for the facilitation of microbial colonization and persistence in the lung, consistent with the characteristics of multidrug-resistant microbial communities that are commonly observed in patients with severe lung disease.

Antagonistic interactions between endophytic cultivable bacterial communities isolated from the medicinal plant Echinacea purpurea.

In this work we have studied the antagonistic interactions existing among cultivable bacteria isolated from three ecological niches (rhizospheric soil, roots and stem/leaves) of the traditional natural medicinal plant Echinacea purpurea. The three compartments harboured different taxonomic assemblages of strains, which were previously reported to display different antibiotic resistance patterns, suggesting the presence of differential selective pressure due to antagonistic molecules in the three compartments. Antagonistic interactions were assayed by the cross-streak method and interpreted using a network-based analysis. In particular 'within-niche inhibition' and 'cross-niche inhibition' were evaluated among isolates associated with each compartment as well as between isolates retrieved from the three different compartments respectively. Data obtained indicated that bacteria isolated from the stem/leaves compartment were much more sensitive to the antagonistic activity than bacteria from roots and rhizospheric soil. Moreover, both the taxonomical position and the ecological niche might influence the antagonistic ability/sensitivity of different strains. Antagonism could play a significant role in contributing to the differentiation and structuring of plant-associated bacterial communities.

Evolution of Intra-specific Regulatory Networks in a Multipartite Bacterial Genome.

Reconstruction of the regulatory network is an important step in understanding how organisms control the expression of gene products and therefore phenotypes. Recent studies have pointed out the importance of regulatory network plasticity in bacterial adaptation and evolution. The evolution of such networks within and outside the species boundary is however still obscure. Sinorhizobium meliloti is an ideal species for such study, having three large replicons, many genomes available and a significant knowledge of its transcription factors (TF). Each replicon has a specific functional and evolutionary mark; which might also emerge from the analysis of their regulatory signatures. Here we have studied the plasticity of the regulatory network within and outside the S. meliloti species, looking for the presence of 41 TFs binding motifs in 51 strains and 5 related rhizobial species. We have detected a preference of several TFs for one of the three replicons, and the function of regulated genes was found to be in accordance with the overall replicon functional signature: house-keeping functions for the chromosome, metabolism for the chromid, symbiosis for the megaplasmid. This therefore suggests a replicon-specific wiring of the regulatory network in the S. meliloti species. At the same time a significant part of the predicted regulatory network is shared between the chromosome and the chromid, thus adding an additional layer by which the chromid integrates itself in the core genome. Furthermore, the regulatory network distance was found to be correlated with both promoter regions and accessory genome evolution inside the species, indicating that both pangenome compartments are involved in the regulatory network evolution. We also observed that genes which are not included in the species regulatory network are more likely to belong to the accessory genome, indicating that regulatory interactions should also be considered to predict gene conservation in bacterial pangenomes.