Mechanisms That Shape Microbial Pangenomes.

Genome-scale network reconstructions are useful tools for understanding cellular metabolism, and comparisons of such reconstructions can provide insight into metabolic differences between organisms. Recent efforts toward comparing genome-scale models have focused primarily on aligning metabolic networks at the reaction level and then looking at differences and similarities in reaction and gene content. However, these reaction comparison approaches are time-consuming and do not identify the effect network differences have on the functional states of the network. We have developed a bilevel mixed-integer programming approach, CONGA, to identify functional differences between metabolic networks by comparing network reconstructions aligned at the gene level. We first identify orthologous genes across two reconstructions and then use CONGA to identify conditions under which differences in gene content give rise to differences in metabolic capabilities. By seeking genes whose deletion in one or both models disproportionately changes flux through a selected reaction (e.g., growth or by-product secretion) in one model over another, we are able to identify structural metabolic network differences enabling unique metabolic capabilities. Using CONGA, we explore functional differences between two metabolic reconstructions of Escherichia coli and identify a set of reactions responsible for chemical production differences between the two models. We also use this approach to aid in the development of a genome-scale model of Synechococcus sp. PCC 7002. Finally, we propose potential antimicrobial targets in Mycobacterium tuberculosis and Staphylococcus aureus based on differences in their metabolic capabilities. Through these examples, we demonstrate that a gene-centric approach to comparing metabolic networks allows for a rapid comparison of metabolic models at a functional level. Using CONGA, we can identify differences in reaction and gene content which give rise to different functional predictions. Because CONGA provides a general framework, it can be applied to find functional differences across models and biological systems beyond those presented here.

Introduction In the previous chapters, we had discussed methods for studying the genetic content of bacteria on a genomic scale. In the process, we highlighted characteristics of certain bacterial genomes, besides discussing properties common to many bacterial genomes. We also documented intra-specific variation in gene content between bacteria. Different strains of the same bacterial species might show considerable variation in gene content, as a result of mutations getting fixed in a particular context, or the dynamics of gene acquisition and loss. Many of these variations might arise from selection resulting from their niches and/or lifestyles. However, this does not reflect the fact that the same bacterium might be phenotypically distinct under different environmental or cellular contexts. Much of these differences might be attributed to gene expression changes, i.e., under a given condition, a bacterium expresses only a subset of its genes, with the remaining genes being silent. Or, the expression level of the same gene might be quantitatively different between two conditions, in contrast to the dramatic on–off distinction made by the previous statement. Thus, the genetic content of a bacterium can be interpreted by a ‘gene expression machinery’ in different ways at different times. The expression of genes, and the manner in which it is regulated, have been studied in recent years using genome-scale techniques. Many of these approaches use DNA microarrays or next-generation sequencing, the basics of which have been covered in the previous two chapters. In the present, core chapter of this book, we will discuss the application of these technologies to the study of gene expression and its control in bacteria. The discussion of techniques and data analysis will focus on gene expression measurements and investigation of genomic regions that bind to a protein of interest. We will also present particular exemplary research as case studies in the process.

Transposable elements (TEs) are ubiquitous components of eukaryotic genomes and can create variation in genome organization and content. Most maize genomes are composed of TEs. We developed an approach to define shared and variable TE insertions across genome assemblies and applied this method to four maize genomes (B73, W22, Mo17 and PH207) with uniform structural annotations of TEs. Among these genomes we identified approximately 400000 TEs that are polymorphic, encompassing 1.6Gb of variable TE sequence. These polymorphic TEs include a combination of recent transposition events as well as deletions of older TEs. There are examples of polymorphic TEs within each of the superfamilies of TEs and they are found distributed across the genome, including in regions of recent shared ancestry among individuals. There are many examples of polymorphic TEs within or near maize genes. In addition, there are 2380 gene annotations in the B73 genome that are located within variable TEs, providing evidence for the role of TEs in contributing to the substantial differences in annotated gene content among these genotypes. TEs are highly variable in our survey of four temperate maize genomes, highlighting the major contribution of TEs in driving variation in genome organization and gene content. OPEN RESEARCH BADGES: This article has earned an Open Data Badge for making publicly available the digitally-shareable data necessary to reproduce the reported results. The data is available at https://github.com/SNAnderson/maizeTE_variation; https://mcstitzer.github.io/maize_TEs.

Variation in protein gene and intron content among land plant mitogenomes

Two sequenced nodulation regions of lupin Bradyrhizobium sp. WM9 carried the majority of genes involved in the Nod factor production. The nod region I harbored: nolA, nodD, nodA, nodB, nodC, nodS, nodI, nodJ, nolO, nodZ, fixR, nifA, fixA, nodM, nolK and noeL. This gene arrangement resembled that found in the nodulation region of Bradyrhizobium japonicum USDA110, however strain WM9 harbored only one nodD gene copy, while the nodM, nolK and noeL genes had no counterparts in the 410 kb symbiotic region of strain USDA110. Region II harbored nolL and nodW, but lacked an nodV gene. Both regions carried ORFs that lacked similarity to the published USDA110 sequences, though they had homologues in symbiotic regions of Rhizobium etli, Sinorhizobium sp. NGR234 and Mesorhizobium loti. These differences in gene content, as well as a low average sequence identity (70%) of symbiotic genes with respect to B. japonicum USDA110 were in contrast with the phylogenetic relationship of USDA110 and WM9 revealed by the analysis of 16S rDNA and dnaK sequences. This most likely reflected an early divergence of symbiotic loci, and possible co-speciation with distinct legumes. During this process the loss of a noel gene and the acquisition of a nolL gene could be regarded as an adaptation towards these legumes that responded to Nod factors carrying 4-O-acetylfucose rather than 2-O-methylfucose. This explained various responses of lupins and serradella plants to infection by mutants in nodZ and nolL genes, knowing that serradella is a stringent legume while lupins are more promiscuous legumes.

Lactobacillus plantarum is a lactic acid bacterium that promotes animal intestinal health as a probiotic and is found in a wide variety of habitats. Here, we investigated the genomic features of different clusters of L. plantarum strains via pan-genomic analysis. We compared the genomes of 108 L. plantarum strains that were available from the NCBI GenBank database. These genomes were 2.9-3.7 Mbp in size and 44-45% in G+C content. A total of 8,847 orthologs were collected, and 1,709 genes were identified to be shared as core genes by all the strains analyzed. On the basis of SNPs from the core genes, 108 strains were clustered into five major groups (G1-G5) that are different from previous reports and are not clearly associated with habitats. Analysis of group-specific enriched or depleted genes revealed that G1 and G2 were rich in genes for carbohydrate utilization (L-arabinose, L-rhamnose, and fructooligosaccharides) and that G3, G4, and G5 possessed more genes for the restriction-modification system and MazEF toxin-antitoxin. These results indicate that there are critical differences in gene content and survival strategies among genetically clustered L. plantarum strains, regardless of habitats.

Komagataeibacter strains are important bacterial cellulose producers, yet closely related isolates can differ in cellulose yield, pellicle properties, and genetic stability during propagation. Such variability suggests that lineage structure and mobile genetic elements both contribute to strain-level genomic divergence. Here, complete genome comparisons were used to integrate vertical relatedness, gene-content structure, cellulose-associated signatures, and mobilome heterogeneity across 22 closed Komagataeibacter assemblies. A maximum likelihood phylogeny inferred from 642 single copy core genes provided the lineage scaffold. An anvi'o pangenome analysis defined a constant core gene cluster component across genomes and a noncore fraction that accounted for most of the genome differences in gene content. Targeted features linked to cellulose biosynthesis and local c-di-GMP-associated context were extracted from each genome. These features captured differences in bcs neighborhood composition and the presence of nearby GGDEF and EAL domain signals. The resulting feature matrix was projected by principal component analysis to summarize between-genome variation. Mobilome profiles were strongly strain dependent. Plasmid homology clustering identified 12 clusters comprising 36 plasmids from 13 genomes, including two dominant clusters of seven and six plasmids. Mash-based distance summaries further distinguished clusters consistent with conserved backbones from clusters consistent with heterogeneous, module-driven relationships. Prophage sequences, assessed as VIBRANT-predicted regions, were widespread but sparse per genome and dominated by medium length fragments. Insertion sequence burden ranged from 50 to 181 elements per genome, indicating substantial differences in transposition-associated sequence content. Pairwise association tests did not support robust cross module covariation beyond expected relationships among pangenome composition metrics at the current sampling depth. Overall, these results provide a complete genome reference framework linking lineage structure and mobilome heterogeneity, and they define reusable resources for comparative studies in bacterial cellulose biotechnology.

ABSTRACTSalmonella enterica serovar Typhimurium definitive type 2 (DT2) is host restricted to Columba livia (rock or feral pigeon) but is also closely related to S. Typhimurium isolates that circulate in livestock and cause a zoonosis characterized by gastroenteritis in humans. DT2 isolates formed a distinct phylogenetic cluster within S. Typhimurium based on whole-genome-sequence polymorphisms. Comparative genome analysis of DT2 94-213 and S. Typhimurium SL1344, DT104, and D23580 identified few differences in gene content with the exception of variations within prophages. However, DT2 94-213 harbored 22 pseudogenes that were intact in other closely related S. Typhimurium strains. We report a novel in silico approach to identify single amino acid substitutions in proteins that have a high probability of a functional impact. One polymorphism identified using this method, a single-residue deletion in the Tar protein, abrogated chemotaxis to aspartate in vitro. DT2 94-213 also exhibited an altered transcriptional profile in response to culture at 42°C compared to that of SL1344. Such differentially regulated genes included a number involved in flagellum biosynthesis and motility.

Having the ability to predict the protein-encoding gene content of an incomplete genome or metagenome-assembled genome is important for a variety of bioinformatic tasks. In this study, as a proof of concept, we built machine learning classifiers for predicting variable gene content in Escherichia coli genomes using only the nucleotide k-mers from a set of 100 conserved genes as features. Protein families were used to define orthologs, and a single classifier was built for predicting the presence or absence of each protein family occurring in 10%-90% of all E. coli genomes. The resulting set of 3,259 extreme gradient boosting classifiers had a per-genome average macro F1 score of 0.944 [0.943-0.945, 95% CI]. We show that the F1 scores are stable across multi-locus sequence types and that the trend can be recapitulated by sampling a smaller number of core genes or diverse input genomes. Surprisingly, the presence or absence of poorly annotated proteins, including "hypothetical proteins" was accurately predicted (F1 = 0.902 [0.898-0.906, 95% CI]). Models for proteins with horizontal gene transfer-related functions had slightly lower F1 scores but were still accurate (F1s = 0.895, 0.872, 0.824, and 0.841 for transposon, phage, plasmid, and antimicrobial resistance-related functions, respectively). Finally, using a holdout set of 419 diverse E. coli genomes that were isolated from freshwater environmental sources, we observed an average per-genome F1 score of 0.880 [0.876-0.883, 95% CI], demonstrating the extensibility of the models. Overall, this study provides a framework for predicting variable gene content using a limited amount of input sequence data. IMPORTANCE Having the ability to predict the protein-encoding gene content of a genome is important for assessing genome quality, binning genomes from shotgun metagenomic assemblies, and assessing risk due to the presence of antimicrobial resistance and other virulence genes. In this study, we built a set of binary classifiers for predicting the presence or absence of variable genes occurring in 10%-90% of all publicly available E. coli genomes. Overall, the results show that a large portion of the E. coli variable gene content can be predicted with high accuracy, including genes with functions relating to horizontal gene transfer. This study offers a strategy for predicting gene content using limited input sequence data.

Recent genomic studies have revealed extensive variation in natural populations of many pathogenic bacteria. However, the evolutionary processes which contribute to much of this variation remain unclear. A previous whole-genome DNA microarray study identified variation at a large chromosomal region (RD13) of Staphylococcus aureus which encodes a family of proteins with homology to staphylococcal and streptococcal superantigens, designated staphylococcal exotoxin-like (SET) proteins. In the present study, RD13 was found in all 63 S. aureus isolates of divergent clonal, geographic, and disease origins but contained a high level of variation in gene content in different strains. A central variable region which contained from 6 to 10 different set genes, depending on the strain, was identified, and DNA sequence analysis suggests that horizontal gene transfer and recombination have contributed to the diversification of RD13. Phylogenetic analysis based on the RD13 DNA sequence of 18 strains suggested that loss of various set genes has occurred independently several times, in separate lineages of pathogenic S. aureus, providing a model to explain the molecular variation of RD13 in extant strains. In spite of multiple episodes of set deletion, analysis of the ratio of silent substitutions in set genes to amino acid replacements in their products suggests that purifying selection (selective constraint) is acting to maintain SET function. Further, concurrent transcription in vitro of six of the seven set genes in strain COL was detected, indicating that the expression of set genes has been maintained in contemporary strains, and Western immunoblot analysis indicated that multiple SET proteins are expressed during the course of human infections. Overall, we have shown that the chromosomal region RD13 has diversified extensively through episodes of gene deletion and recombination. The coexpression of many set genes and the production of multiple SET proteins during human infection suggests an important role in host-pathogen interactions.

The Baculovirus 10-kDa Protein

Owing to their small size and paucity of phenotypic characters, progress in the evolutionary biology of microbes in general, and human pathogenic fungi in particular, has been linked to a series of advances in DNA sequencing over the past quarter century. Phylogenetics was the first area to benefit, with the achievement of a basic understanding of fungal phylogeny. Population genetics was the next advance, finding cryptic species everywhere, and recombination in species previously thought to be asexual. Comparative genomics saw the next advance, in which variation in gene content and changes in gene family size were found to be important sources of variation. Fungal population genomics is showing that gene flow among closely related populations and species provides yet another source of adaptive, genetic variation. Now, two means to associate genetic variation with phenotypic variation, "reverse ecology" for adaptive phenotypes, and genome-wide association of any phenotype, are letting evolutionary biology make a profound contribution to molecular developmental biology of pathogenic fungi.

GENE REGULATION

Date palms (Phoenix dactylifera) are the most significant perennial crop in arid regions of the Middle East and North Africa. Here, we present a comprehensive catalogue of approximately seven million single nucleotide polymorphisms in date palms based on whole genome re-sequencing of a collection of 62 cultivars. Population structure analysis indicates a major genetic divide between North Africa and the Middle East/South Asian date palms, with evidence of admixture in cultivars from Egypt and Sudan. Genome-wide scans for selection suggest at least 56 genomic regions associated with selective sweeps that may underlie geographic adaptation. We report candidate mutations for trait variation, including nonsense polymorphisms and presence/absence variation in gene content in pathways for key agronomic traits. We also identify a copia-like retrotransposon insertion polymorphism in the R2R3 myb-like orthologue of the oil palm virescens gene associated with fruit colour variation. This analysis documents patterns of post-domestication diversification and provides a genomic resource for this economically important perennial tree crop.

Natural killer (NK) cells have diverse roles in hominid immunity and reproduction. Modulating these functions are the interactions between major histocompatibility complex (MHC) class I molecules that are ligands for two NK cell surface receptor types. Diverse killer cell immunoglobulin-like receptors (KIR) bind specific motifs encoded within the polymorphic MHC class I cell surface glycoproteins, while, in more conserved interactions, CD94:NKG2A receptors recognize MHC-E with bound peptides derived from MHC class I leader sequences. The hominid lineage presents a choreographed co-evolution of KIR with their MHC class I ligands. MHC-A, -B, and -C are present in all great apes with species-specific haplotypic variation in gene content. The Bw4 epitope recognized by lineage II KIR is restricted to MHC-B but also present on some gorilla and human MHC-A. Common to great apes, but rare in humans, are MHC-B possessing a C1 epitope recognized by lineage III KIR. MHC-C arose from duplication of MHC-B and is fixed in all great apes except orangutan, where it exists on approximately 50% of haplotypes and all allotypes are C1-bearing. Recent study showed that gorillas possess yet another intermediate MHC organization compared to humans. Like orangutans, but unlike the Pan-Homo species, duplication of MHC-B occurred. However, MHC-C is fixed, and the MHC-C C2 epitope (absent in orangutans) emerges. The evolution of MHC-C drove expansion of its cognate lineage III KIR. Recently, position −21 of the MHC-B leader sequence has been shown to be critical in determining NK cell educational outcome. In humans, methionine (−21M) results in CD94:NKG2A-focused education whereas threonine (−21T) produces KIR-focused education. This is another dynamic position among hominids. Orangutans have exclusively −21M, consistent with their intermediate stage in lineage III KIR-focused evolution. Gorillas have both −21M and −21T, like humans, but they are unequally encoded by their duplicated B genes. Chimpanzees have near-fixed −21T, indicative of KIR-focused NK education. Harmonious with this observation, chimpanzee KIR exhibit strong binding and, compared to humans, smaller differences between binding levels of activating and inhibitory KIR. Consistent between these MHC-NK cell receptor systems over the course of hominid evolution is the evolution of polymorphism favoring the more novel and dynamic KIR system.