Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

7.1 Introduction

Whole-genome sequencing has become a powerful and informative approach for determining the genetic basis of known bacterial properties, predicting new properties, and enabling post-genomic tools such as transcriptomics and proteomics. It also shapes the choice of representative bacterial strains and species to serve as model organisms for experimental work. However, genome sequencing and annotation are most useful in the context of comparative genomic and evolutionary analysis, which allows the determination of phylogenetic relationships between extant organisms, provides insights into the evolution of different biological systems, and sheds light on processes accounting for organismal diversity. Classification of organisms based solely on their phenotypic characteristics is no longer recognized as useful for understanding the evolutionary relationships between organisms. Therefore, modern evolutionary studies are focused on DNA- and protein-based phylogenetics, generally using universally distributed genes such as ribosomal rRNA or informational protein-coding genes. However, the evolutionary relationships between species derived from different genes are rarely consistent with each other, due to incomplete incorporation of knowledge about complex processes of sequence evolution and limited evolutionary information available within a single gene. The availability of whole-genome sequences provides an opportunity to address these limitations and obtain a more objective and comprehensive understanding of evolutionary relationships among microbes and their genomes.

Recent and ongoing comparative genomics and ultrastructural (see Chaps. 2, 3 and 11) studies have generated fundamentally important and exciting insights into the evolution of biological systems within organisms of the PVC superphylum (a group named for three of its component phyla: Planctomycetes, Verrucomicrobia, and Chlamydiae). While some bacteria within this group (e.g., many of the Chlamydiae) have the simple cell structure common among bacteria, all the characterized planctomycetes, several verrucomicrobia, and the only cultured species in phylum Lentisphaerae have a common cell plan that features an additional intracellular membrane and is unique among bacteria (Fieseler et al. 2004; Fuerst 2005; Lee et al. 2009; Lindsay et al. 2001). It should also be noted that some PVC species currently considered to lack this plan (e.g., members of Chlamydiae) have not been examined by state-of-the-art cryotechniques and may yet be shown to possess it. Planctomycetes exhibit variations upon the common PVC plan, featuring additional membrane-enclosed compartments with functions and biological consequences that are either known (e.g., anammoxosome, anaerobic ammonia-oxidizing compartment, in anammox bacteria) or undetermined (e.g., double-layered membrane envelope surrounding the condensed genomic DNA of Gemmata obscuriglobus) (Fuerst 2005; Fuerst and Webb 1991). The availability of genome sequences from members of the PVC superphylum allowed comparative/structural genomic analysis, followed by experimental immunogold localization demonstrating the presence of proteins structurally resembling eukaryotic membrane coat proteins (Santarella-Mellwig et al. 2010). While sterols are generally considered to be eukaryote-specific molecules, the G. obscuriglobus genome sequence established a foundation for the discovery of sterols and sterol biosynthesis genes in this organism (Pearson et al. 2003). The almost complete genome of Candidatus Kuenenia stuttgartiensis assembled from a community genome has permitted deduction of the biochemical pathway of anaerobic ammonium oxidation, a technologically important biochemical reaction (Strous et al. 2006). Identification of canonical pathogenicity determinants (encoding a putative type III secretion system) in the genome of Verrucomicrobium spinosum led to work suggesting that this species, not previously known to interact with eukaryotes, is capable of pathogenic or symbiotic relationships with invertebrates (Sait et al. 2011).

Accumulation of PVC genomes has also spurred genome-scale evolutionary analyses of individual genes to better understand the mechanisms of protein sequence evolution driving diversification of PVC organisms and underlying the emergence of unusual ecology, cell biology, and physiology within this group. The large phylogenetic distances separating members of the PVC superphylum challenge classical methods of evolutionary analysis and have stimulated development of novel sequence analysis approaches. Study of indel (insertion/deletion) evolution within PVC organisms has shown that indels evolve under differential selective pressure in different regions of protein-coding genes and on different lineages of the PVC superphylum (Kamneva et al. 2010). As well as providing insight into the emergence of new biological function, this study established a new framework for molecular evolutionary studies. The framework is generally applicable to any group of proteins, especially if available sequence data are sparse and separated by large evolutionary distances.

This chapter provides an overview of recent insights into the PVC superphylum that have emerged from whole-genome sequence comparisons and their implications for molecular, cellular, and evolutionary biology. The first section will review current knowledge of bacterial genome structure and the evolutionary processes contributing to this structure, followed by a summary of current progress in PVC genome sequencing, and several examples of genome-scale analysis of PVC organisms. Lastly, this chapter will introduce the PVC Genome Database—a resource for comparative genomics and evolutionary analysis of the PVC superphylum.

7.2 Structure and Evolution of Bacterial Genomes in Light of Comparative Genomics

In this section, we provide an overview of bacterial genome structure, its relevance for comparative genomics, and evolutionary processes shaping bacterial genomes. The main features of bacterial genome structure arise from fundamental molecular mechanisms facilitating either the flow of encoded genetic information from DNA to functional RNAs and proteins or vertical transfer of this information during reproduction. Bacterial genomes comprise single or multiple, circular or linear chromosomes, sometimes with the addition of extrachromosomal elements (plasmids). In addition to plasmids, bacterial genomes contain other mobile elements such as insertion elements, integrons, and different kinds of transposons. Bacterial genome size does not vary as much as that of eukaryotes, ranging from around 0.5 megabase pairs (Mbp) in intracellular pathogens and symbionts to 10 Mbp in developmentally complex free-living bacteria. A “typical” genome size for free-living organisms or those without obligate host associations is 4–5 Mbp.

Fig. 7.1
figure 00071

General organization of bacterial operon and flow of genetic information in bacteria. RBS—ribosome binding site (green rectangles); −10, −35—TA-rich elements of a core promoter located at approximately −10 and −35 position relative to translation start site (blue rectangles)

Full size image

Bacterial genomes contain stretches of DNA (usually uninterrupted by introns) encoding proteins and different types of RNAs. It is common in bacteria for protein-encoding open reading frames (ORFs) involved in a set of related processes to be organized into operons, which provide a simple general mechanism of coordinating gene expression (Fig. 7.1). Genes in an operon are regulated by a single set of regulatory elements (including promoter sequences) and are transcribed as a polycistronic RNA. The core of a single promoter generally consists of −10 and −35 (relative to the start site of transcription) AT-rich sequence segments, which are recognized by different sigma factors, accessory proteins to the RNA polymerase holoenzyme. This provides an additional mechanism of co-regulation of genes involved in producing certain physiological responses. Genetic information is further transmitted into proteins, in the case of protein-coding genes. Translation in bacteria is initiated when the 30S ribosomal subunit binds to the ribosome binding site (RBS) preceding an ORF on the mRNA. Relative to eukaryotic genomes, bacterial genomes harbor very little noncoding DNA (generally made up of regulatory elements), explaining why genome size and gene number are strongly correlated. All the features of bacterial genomes described above are often conserved through evolution and therefore relevant to comparative and evolutionary genomics, influencing genome annotation and genome-based functional predictions.

A number of evolutionary processes shape bacterial genomes and affect coding and noncoding DNA sequences, allowing for different phenotypes to arise and creating an additional level of complexity for evolutionary analysis. The main events of genome content evolution are gene duplications and losses, and horizontal gene transfer (Mira et al. 2001; Ochman et al. 2000). Active mobile elements in bacteria are a major force for this kind of genome evolution, as they facilitate rapid genome rearrangements and acquisition of new genes. High numbers of mobile elements are known to be associated with high genome plasticity. Duplication and horizontal transfer of genes provide raw material for evolution and bring brand-new biological functions into bacterial genomes, potentially increasing an organism’s fitness. However, there is an associated metabolic cost of propagating extra DNA. In the case of horizontal gene transfer, newly acquired genes might also be incompatible with the preexisting genetic background of the host and result in a fitness reduction. Mutation events (nucleotide substitutions, small insertions, and deletions) also shape the coding and noncoding regions of bacterial genomes, contributing to the evolution of novel biological functions. Within protein-coding genes, the need to maintain necessary biochemical functions often exerts considerable pressure of stabilizing selection on mutation events. Thus, the interplay between natural selection and rates of different evolutionary events allows bacterial genomes to diversify, while recombination among compatible lineages maintains the genetic integrity of the bacterial population.

7.3 History and Current Status of Genome Sequencing of PVC Organisms

The PVC superphylum includes bacteria that are interesting for diverse reasons (evolutionary, ecological, cell biological, biochemical, medical), and this has influenced the selection and timing of genome sequencing projects (Table 7.1). Organisms of phylum Chlamydiae are important to human health and possess genomes that are among the smallest known for cellular organisms. Therefore, genome sequencing projects for chlamydiae started rather early, with the genome of Chlamydia trachomatis D/UW-3/CX published in 1998 (Stephens et al. 1998) followed by that of Chlamydophila pneumoniae CWL029 in 1999 (Kalman et al. 1999). Further genome sequencing efforts established phylum Chlamydiae as a system for studying evolution of intracellular pathogens and symbionts in a comparative genomics framework, with reports on sequencing projects for Protochlamydia amoebophila , Parachlamydia acanthamoebae, and Waddlia chondrophila in 2004, 2009, and 2010 (Bertelli et al. 2010; Greub et al. 2009; Horn et al. 2004).

Table 7.1 PVC genome sequencing projects
Full size table

In contrast to the chlamydiae, the primary motivations for genome sequencing of planctomycetes and verrucomicrobia have been their unusual cell biology or biochemistry as well as their distinct evolutionary position on the bacterial phylogenetic tree. All characterized planctomycetes and verrucomicrobia have a common cell organization featuring an intracellular membrane, unique among the Bacteria (Fuerst 2005; Lee et al. 2009). Additionally some planctomycete bacteria exhibit variations on this common plan, featuring additional membrane-enclosed compartments of known or unknown function (Fuerst 2005). These cell biology properties helped to establish first planctomycetes, and then verrucomicrobia, as model systems for studying evolution of biological complexity and to launch several independent genome sequencing projects. The genome sequence of Pirellula st. 1 was the first published planctomycete genome (Glöckner et al. 2003). Discovery of the anammox reaction and marine “anammox” planctomycetes capable of this reaction established the biotechnological importance of planctomycete bacteria for the development of sustainable wastewater treatment systems and led to sequencing of the Candidatus K. stuttgartiensis genome (Strous et al. 2006). These initial projects laid the foundation for a new wave of experimental work in planctomycetes and verrucomicrobia, where genetic, biochemical, and functional studies have arisen from genomic information. In turn, these experimental studies have prompted new rounds of genome sequencing from additional strains, such that genomes currently available or in progress span the phylogenetic, metabolic, and lifestyle diversity of the PVC superphylum (Table 7.1).

7.4 Phylogenetic Position of PVC Organisms

The relative phylogenetic positions of organisms belonging to phyla Planctomycetes, Verrucomicrobia, Lentisphaerae, and Chlamydiae within the tree of life have been debated (Embley et al. 1994; Griffiths and Gupta 2007; Hedlund et al. 1997; Jenkins and Fuerst 2001; Van de Peer et al. 1994; Pilhofer et al. 2008; Roenner et al. 1991; Schloss and Handelsman 2004; Stackebrandt et al. 1984; Wagner and Horn 2006; Ward et al. 2000, 2006). Initially, planctomycetes were considered to be a deep-branching bacterial lineage (Roenner et al. 1991; Stackebrandt et al. 1984). This hypothesis was later rejected based on analysis of larger data sets with more sophisticated methods (Embley et al. 1994; Van de Peer et al. 1994). Some early studies suggested chlamydiae to be the closest relative of planctomycetes; later it was shown that verrucomicrobia are the closest living relatives of chlamydiae (Griffiths and Gupta 2007). More recently, it has been proposed that the four established phyla (Planctomycetes, Verrucomicrobia, Chlamydia, and Lentisphaerae) and two candidate phyla (OP3 and Poribacteria) (the phylogenetic position of which has not been clearly established due to limited sequence availability for the representative species) form a coherent group of organisms named the PVC superphylum (Pilhofer et al. 2008; Schloss and Handelsman 2004; Wagner and Horn 2006). Since publication of these reports, the availability of genome sequences from additional PVC taxa has allowed us to evaluate the robustness and internal structure of the superphylum by including a larger number of phylogenetic markers and as many divergent PVC organisms as possible. Our large-scale phylogenetic study (Kamneva et al. 2012) used multiple protein families and a maximum-likelihood approach on concatenated phylogenetic markers to reveal the evolutionary history of 26 members of the PVC superphylum (Fig. 7.2). We observed species relationships that are largely consistent with most recently published 16S rRNA-, 23S rRNA-, and protein-based phylogenies (Hou et al. 2008; Pilhofer et al. 2008; Wagner and Horn 2006). Species within the four distinct phyla formed four well-supported monophyletic groups. Planctomycetes occupied a separate position from the rest of the superphylum, and Kuenenia stuttgartiensis appeared to be the most ancestral lineage among planctomycetes, as was observed in previous studies (Wagner and Horn 2006). Within the rest of the superphylum, Lentisphaerae formed a cluster with Chlamydiae, which contradicts previously published phylogenies where Lentisphaerae species were more closely related to phylum Verrucomicrobia (Hou et al. 2008). We also detected a hypothetical sister clade to the PVC superphylum containing phyla Spirochaetes , Bacteroidetes,and Chlorobi (Fig. 7.2). This relationship was also recovered in previous studies conducted using a different set of species and phylogenetic markers (Hou et al. 2008). This section is modified from reference (Kamneva et al. 2012), with permission of Oxford University Press, Society for Molecular biology and Evolution.

Fig. 7.2
figure 00072

Evolutionary relationships within the PVC superphylum. Species-tree topology was recovered for the entire set of 99 bacterial species from various bacterial phyla (including 26 PVC organisms), as a consensus tree averaging over gene trees of 41 phylogenetic markers. Divergence times were estimated using a concatenated alignment of all 41 phylogenetic markers (EngD, PrfA, SecY, MraW, NusG, ObgE, RRF, DNA primase, family 22 peptidase, 29 ribosomal proteins, and 3 tRNA synthases). Non-PVC clades were collapsed for clarity, and corresponding lineages are shown in grey. Names of PVC phyla are shown on the right. Phylum names are shown on the right

Full size image

7.5 General Features of PVC Genomes

The major features of the sequenced PVC genomes are summarized in Table 7.2. The number of predicted protein-coding genes ranges from 940 in Chlamydia muridarum to 7,989 in the planctomycete Gemmata obscuriglobus. Gene number is highly correlated with lifestyle and suggests substantial variation in gene loss and gain rates among different evolutionary lineages. The genomes differ in total number of rRNA genes (16S, 5S, and 23S rRNA genes), which ranges from 3 to 4 in some Planctomycetes and Verrucomicrobia species to 12–15 in others. The underlying rRNA copy number variation somewhat correlates with the number of tRNA genes (Table 7.2). The number of mobile elements (approximated using the number of transposase genes) in a given genome was variable, ranging from none in the genomes of C. muridarum and C. pneumoniae to 3–4 % of all the protein-coding genes in W. chondrophila, G. obscuriglobus, and L. araneosa. This suggests different levels of genome plasticity in these organisms. Several chlamydial organisms harbor plasmids. Homologs of many plasmid-encoded genes are found on the main chromosomes of plasmid-free chlamydial strains, suggesting the presence of the plasmid in the ancestor of extant Chlamydiae (Collingro et al. 2011).

Table 7.2 Genome information summary
Full size table

7.6 Analysis of Genome Properties

We investigated the genomic content of 24 PVC superphylum members through application of Genome Properties, a system that detects key biological properties encoded in prokaryotic genomes through the use of standardized computational methods and controlled vocabularies (Haft et al. 2005; Selengut et al. 2007). The output of this process consisted of more than 600 individual properties. Rendering and analysis of this large number of properties is beyond the scope of this chapter and will constitute a feature of the future PVC Genomics Database (see Sect. 7.11 below). The potential usefulness of the Genome Properties approach is illustrated here through comparative analysis of the distribution of 119 representative Genome Properties (Fig. 7.3). Visualization of the Genome Properties output through dual dendrogram heatmaps showed that, by and large, clustering of genomes according to shared genome properties (upper dendrogram) reflects established phylum-level relationships (Fig. 7.3). Minor discrepancies cannot be meaningfully interpreted due to the fact that the selection of properties for this analysis was somewhat arbitrary.

Fig. 7.3
figure 00073

Genome Properties analysis of PVC organisms. Colors on the heatmap represent the probability of any given genome possessing a certain property. The rows and columns of the heatmap are arranged according to hierarchical clustering of Euclidian distances for rows and columns, respectively. Names of the properties are shown on the right; species names are at the bottom of the figure. Three major clusters of the properties are denoted by purple, grey, and orange bars on the left of the heatmap. The counts of properties for major functional categories are shown for every cluster of the properties (Met = metabolism; DNA = DNA handling; Cell Surf = cell surface component; Transp = transport; Self El = selfish genetic elements; Vir = virulence; Reg Sys = regulatory systems; Cell Gr = cell growth, organization, and division). Organism names are abbreviated as follows: S. negevensis (Simkania negevensis Z), P. acanthamoebae (Parachlamydia acanthamoebae str. Hall’s coccus), P. amoebophila (Candidatus Protochlamydia amoebophila UWE25), W. chondrophila (Waddlia chondrophila WSU 86-1044), C. muridarum (Chlamydia muridarum Nigg), C. pneumoniae (Chlamydophila pneumoniae AR39), A. muciniphila (Akkermansia muciniphila ATCC BAA−835), V. spinosum (Verrucomicrobium spinosum DSM 4136), C. flavus (Chthoniobacter flavus Ellin428), M. infernorum (Methylacidiphilum infernorum V4), P. parvula (Pedosphaera parvula Ellin514), C. akajimensis (Coraliomargarita akajimensis DSM 45221), O. bacterium (Opitutaceae bacterium TAV2), O. terrae (Opitutus terrae PB90−1), V. bacterium (Verrucomicrobiae bacterium DG1235), L. araneosa (Lentisphaera araneosa HTCC2155), V. vadensis (Victivallis vadensis ATCC BAA−548), B. marina (Blastopirellula marina DSM 3645), P. staleyi (Pirellula staleyi DSM 6068), R. baltica (Rhodopirellula baltica SH 1), P. brasiliensis (Planctomyces brasiliensis DSM 5305), P. maris (Planctomyces maris DSM 8797), P. limnophilus (Planctomyces limnophilus DSM 3776), I. pallida (Isosphaera pallida ATCC 43644), G. obscuriglobus (Gemmata obscuriglobus UQM 2246), K. stuttgartiensis (Candidatus Kuenenia stuttgartiensis)

Full size image

Other information that can be extracted from the heatmap analysis includes the prevalence of individual properties, with more universally distributed properties arrayed at the top of the heatmap, and less frequently occurring properties at the bottom. As might be expected, the majority of widely distributed properties are “housekeeping” functions such as the processing of the informational molecules DNA and RNA. For example, RNA polymerase (transcription) and ribosomal units (translation) are universally present, as are the GroEL/GroES and DnaK-DnaJ-GrpE chaperone systems. Other widely distributed properties include the metabolic pathways such as the TCA and pentose phosphate cycles, as well as glycogen storage. DNA repair pathways are also represented in this group of “universal” or “almost universal” properties, which form the top cluster in the left-hand dendrogram. At the bottom of this cluster are five properties which show less universal distribution, including peptidoglycan (murein) biosynthesis. As expected, the planctomycetes lack (all or part of) peptidoglycan biosynthesis, but it is interesting to note that the verrucomicrobial termite symbiont Opitutaceae bacterium TAV2 also appears to lack peptidoglycan (based on failure to detect the required components: d-alanine–d-alanine ligase, murABCEFGI, and peptidoglycan biosynthetic transglycosylases). While the presence/absence of peptidoglycan in the cell wall of Opitutaceae bacterium TAV2 is not established, peptidoglycan-less verrucomicrobia have been previously reported (Yoon et al. 2010), but genomes are not available for these Cerasicoccus species and thus could not be included in our Genome Properties analysis.

Two additional clusters of properties can be observed. The upper part of the middle cluster on the figure displays a strong segregation of phylum Chlamydiae from the other phyla, based primarily on their known lack of many amino acid synthesis pathways. Other properties separating the Chlamydiae from other PVC phyla include the lack of phosphate ABC transporters, Tat (Sec-independent) protein export, and the synthesis of inosine-5-phosphate from ribose-5-phosphate. In the lower part of the middle cluster, properties that differ in distribution within the Chlamydiae can also be seen. These include the presence in Waddlia chondrophila (uniquely among the Chlamydiae examined) of proline biosynthesis from glutamate and de novo uridine-5-phosphate biosynthesis, as previously reported by Bertelli et al. (2010). As also found in P. amoebophila, W. chondrophila possesses a glycine cleavage system, class II (B12-dependent) ribonucleotide reductase, and F1/F0 ATPase (Bertelli et al. 2010). Other variable properties within the Chlamydiae separate the amoebal symbionts from the mammalian chlamydiae C. pneumoniae and C. muridarum.

The last property cluster (lower part of diagram) contains properties that occur infrequently, and in some cases, in only one genome of the PVC representatives analyzed. Unique occurrences include an ectoine ABC transporter in B. marina. Ectoine is a compatible solute, suggesting a mechanism for salt homeostasis in B. marina, which is quite halotolerant (Schlesner and Stackebrandt 1986). Genes required for the production of a biosynthetically related compatible solute, 5-hydroxyectoine, have been previously reported in B. marina (Bursy et al. 2007). Glutathione, another compound that protects against cellular stresses (Masip et al. 2006), appears to be uniquely synthesized in L. araneosa. A third stress compound, polygammaglutamate (which also has a role for pathogenesis), is predicted only for R. baltica, and the genomic potential for this synthesis has been previously reported (Candela et al. 2010). These examples, and others which will be revealed by full comparative analysis of Genome Properties for the PVC superphylum, provide interesting starting points and testable hypotheses for future experimental work.

7.7 Genes Encoded in PVC Genomes

Robust identification of a gene family set (genes derived from the same ancestral gene) is a necessary first step for reliable evolutionary/comparative genomic analysis of any group of organisms. By using the computational procedure described and implemented previously within OrthoMCL software (Li et al. 2003), an initial set of protein families was constructed. Many clusters obtained included paralogous genes and xenologs that evolved via duplication and HGT events at different stages of evolution. On the other hand, many gene families contained just one member (singletons). This can be attributed to the fact that construction of gene families for distantly related taxa, such as different bacteria of the PVC superphylum, leads to smaller groups and a greater fraction of one-member clusters. Altogether, we identified 17,608 gene families that included two or more sequences from PVC genomes and the outgroup genome, leaving 63,679 singletons with no detected close homologs within other PVC genomes.

The conserved core of genes, present in all 25 PVC species under consideration and 74 outgroup genomes, consisted of 44 gene families. Analysis of functional distribution of these core gene families showed that the majority encode components of the information-processing systems (translation, transcription, and replication) and some enzymes from core biosynthetic pathways. Furthermore, 13, 35, 64, and 2 gene families were exclusively detected within the PVC phyla Planctomycetes, Lentisphaerae, Chlamydiae, or Verrucomicrobia, respectively. One of these genomic markers for planctomycete bacteria encodes proteins containing a cytochrome C assembly protein (PF01578) domain. In all planctomycete species, these genes were located next to the glutamyl-tRNA reductase gene and might be co-regulated with this housekeeping gene. A second planctomycete genomic marker was a gene family encoding proteins containing the domain found at the N-terminus of the chaperone SurA (PF13624). These genes were located in close proximity to those encoding DNA topoisomerase I.

Another interesting genetic module not associated with a particular phylum, but rather conserved across a number of Planctomycetes, Lentisphaerae, and Verrucomicrobia species, features the domain of unknown function DUF1501. A large number of genetic clusters containing PSCyt1/PSCyt2/PSD1 and DUF1501 (Planctomycetes-specific cytochromes and Planctomycetes-specific domain of unknown function) containing proteins of varying domain composition and structure are preferentially encoded in the genomes of I. pallida, G. obscuriglobus, Planctomyces and Pirellula species, P. parvula, V. spinosum, C. flavus, C. akajimensis, and L. araneosa. The most complex gene clusters included four genes: (1) DUF1501, sometimes with twin-arginine signal peptide, (2) protein with weak support for one or several PPC domains normally found in secreted bacterial peptidases (Yeats et al. 2003) and conserved regions without characterized signatures, (3) PSCyt1/Big_2/PSCyt2/PSD1 protein, and (4) PSCyt1/WD40 protein. The three former proteins also contain predicted type I signal peptides. Domains Big_2 and WD40 are known to be involved in protein–protein interaction (Kelly et al. 1999; Xu and Min 2011) and probably are responsible for protein complex assembly or substrate recognition. Twin-arginine signal peptide is often found in proteins transported through the membrane in the folded state because of prosthetic groups acquired in the cytoplasm. Planctomycetes-specific cytochrome domains contain a highly conserved CxxCH motif responsible for heme binding within other cytochrome domains. All these suggest that these proteins form a complex either outside the cell or within the periplasm and carry out undetermined enzymatic reactions.

7.8 Influence of Indel Substitutions on Evolution of Protein-Coding Genes in PVC Genomes

Indel substitutions represent a common type of sequence variation contributing to the evolution of both coding and regulatory/noncoding sequences (Brandstrom and Ellegren 2007; Britten et al. 2003; Britten 2002; Chan et al. 2007, 2010; Osterberg et al. 2002; Podlaha et al. 2005; Podlaha and Zhang 2003; Schully and Hellberg 2006). As with amino acid replacement substitutions, interplay between natural selection and other factors results in the differential fixation of indels. In case studies of individual genes, including the Catsper1 calcium ion channel genes in mammals (Podlaha et al. 2005; Podlaha and Zhang 2003) and the Acp26Aa gene in Drosophila species (Schully and Hellberg 2006), it has been found that positive diversifying selection acts upon indels. In Kamneva et al. (2010), we have expanded on this general concept in a genome-wide study of indel substitutions. To investigate the patterns of selective constraints on indel substitutions in a genome-wide manner, we estimated secondary structure-specific insertion and deletion rates for every lineage of every gene family in the data set using gapped ancestral sequence reconstruction (Edwards and Shields 2004) and then compared the observed distribution of insertion/deletion rates with the expected distributions obtained using simulations under neutral conditions. We applied this approach to a data set of 17 genomes from members of the PVC superphylum to evaluate how insertions and deletions have affected the evolution of this group (Kamneva et al. 2010). This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

7.8.1 Rates of Indel Substitutions in Proteins from PVC Genomes

It has been previously shown that different types of secondary structure have different susceptibility to insertions and deletions (Benner and Gerloff 1991). Loops or coils accommodate indels more easily than alpha-helices or beta-strands. To evaluate secondary structure-specific patterns of indel substitutions, alignments in every gene family were split based on predicted secondary structure. Branch lengths of gene trees were reevaluated using generated alignment partitions, and insertion/deletion rates were recalculated for every branch of every gene phylogeny for each type of secondary structure (loops, alpha-helices, and beta-strands). As expected, most gene tree lineages show no insertion or deletion events. However, in full-length proteins and in loops, there is a more pronounced local maximum of density at about five insertions/deletions per unit of sequence divergence per unit of alignment length (Fig. 7.4). This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

Fig. 7.4
figure 00074

Expected and observed insertion/deletion rate distributions derived from gene families encoded in PVC genomes. Adapted from Kamneva et al. (2010). Rate distributions are shown for every type of secondary structure (coils, alpha-helices, and beta-strands) and for full-length alignments. For every type of event (insertions and deletions), distributions are depicted with a histogram (x-axes: event rates, number of events from 0 to 50 per unit of evolutionary distance, per unit of alignment length; y-axes: density) and a boxplot of the entire data set (x-axes: class of the data; y-axes: event rates, number of events per unit of evolutionary distance, per unit of alignment length). In both cases, blue and red colors denote observed (Obs) and expected (Ex) distributions, respectively (Adapted from Kamneva et al. 2010)

Full size image

7.8.2 Indel Size Distribution

This study represents the first analysis of indel substitutions in the genomes of distantly related organisms, providing insights into the general characteristics of insertions and deletions in the set of divergent protein sequences as well as into their patterns of selective constraints. We identified 37,365 insertion and 53,557 deletion events along the branches of the gene trees in full-length alignments. Observing larger number of deletions than insertions is consistent with what has been shown in other studies of protein-coding sequences from nematodes (Wang et al. 2009) and in a rat/mouse comparison (Taylor et al. 2004). It seems that the presence of small genomes from chlamydial species might have influenced our results for insertion/deletion frequency; it has been shown in eukaryotes that DNA loss is one of the underlying mechanisms of genome shrinkage (Petrov 2002). However, we examined the evolution of individual genes, whereas processes associated with dramatic genome size changes in pathogenic bacteria occur on a larger scale with loss of whole genes or large parts of genomes containing several open reading frames (Gregory 2004; Mira et al. 2001; Moran and Mira 2001; Nilsson et al. 2005).

The longest insertion identified in our data set was 217 amino acids, whereas the longest deletion was 190 amino acids. The most common insertion or deletion event was a one amino acid-long substitution, independent of the type of secondary structure under consideration. The mean length value of observed insertions/deletions was 3.77/3.22 amino acids for full-length proteins. Observed insertions generally tended to be longer than deletions in all the types of structural elements. This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

7.8.3 Detecting Strength of Natural Selection on Indels

In order to be able to differentiate between varying strengths of selective pressure on indel substitutions, respective null distributions were generated for every observed distribution using randomization. Our results showed that specific branches of many gene trees possess significantly higher number of insertions/deletions than would be expected by chance. For many partitions, the maximum observed event rate is several orders of magnitude higher than the maximum rate in randomized data. An insertion/deletion rate value significantly higher than that expected by chance on a branch of the gene phylogeny is consistent with positive Darwinian selection on insertion/deletion substitutions on that particular branch. The magnitude of the indel influence on the overall evolutionary trend might be estimated as a percentage of the branches where it was possible to detect positive selection on insertions or deletions (Table 7.3). Insertions and deletions on up to 12 % of all the branches in the data set evolved under positive selection. This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

Table 7.3 Number (#) and percentage of branches showing evidence for positive selection on insertions (ins) and deletions (del) in different length groups and secondary structural units
Full size table

7.8.4 Indels in Proteins of Different Biological Functions

Selection is observed at the level of the individual gene/protein but actually occurs in the context of broader cellular biology. We used KEGG metabolic pathways (Kanehisa et al. 2010) to classify gene families in the data set and systematically identify molecular pathways affected by indel processes. We linked every gene family with information from the KEGG Molecular Pathway Database using a Blast search against the database. We were able to map all full-length gene families onto 106 groups of cellular pathways. However, the total number of pathways obtained varied depending on the specific types of secondary structure in which indels occurred. We employed a binomial test to identify pathways with positive selection on insertions/deletions of different length in varying secondary structural elements consistently overrepresented among gene families. Different types of transporters (ABC transporters, pore ion channels) as well as several pathways related to general metabolism (cysteine and methionine, thiamine, selenoamino acid, phenylalanine, sphingolipid metabolism, base excision repair, glycosaminoglycan degradation, terpenoid backbone biosynthesis, ribosome, bacterial secretion systems, and protein export) were consistently overrepresented among gene families with positive selection on insertions/deletions of different length. Noticeably, ABC-type transporters and ion-coupled transporters show elevated rates of deletions and insertions in coils. This may suggest a general pattern of evolution for these types of proteins. Insertions (deletions) in coiled regions might change the structural composition of the protein by introducing (eliminating) structural elements, in the case of long indels containing alpha-helices or beta-strands. In the case of indels that do not affect structural composition of the protein, they may alter flexibility of the existing protein fold in terms of positioning of structural elements relative to each other or to binding partners. In some cases, this might also change the thermodynamic stability of proper protein folding (Meenan et al. 2010; Viguera and Serrano 1997). As described below, we examined the structural and functional consequences of indel events in example gene families that exhibited evidence for positive selection on indel substitutions. This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

7.8.5 Insertions in Ammonium Transporter Proteins in Planctomycetes and Verrucomicrobia

One of the ion-coupled transporters with an unexpectedly high number of insertions in loop regions is the ammonium transporter from planctomycete and verrucomicrobia species (Fig. 7.5). Several branches of the gene phylogeny for this protein family exhibit elevated levels of insertions in coils. Additionally, mapping of insertions on the tertiary structure of E. coli AmtB showed clustering of otherwise conserved insertions in periplasmic loops. There are no known binding partners that would interact with the periplasmic domain of AmtB. However, a previous study of the E. coli protein allowed identification of several mutations in the periplasmic domain of the pore entrance that significantly increased ammonium uptake (Javelle et al. 2008). W148A is particularly interesting as it is located in the periplasmic coil between the fourth and fifth transmembrane helices, adjacent to a small periplasmic helical element. In the proteins of the planctomycete and verrucomicrobia clade, the periplasmic helix contained several small indels. Furthermore, part of the loop adjacent to the fifth transmembrane helix contained an additional protein segment conserved among members of the family. Ammonium has been found to induce surface attachment and biofilm formation in R. baltica (Frank et al. 2011); therefore, the observed evolutionary changes in AmtB might have led to emergence of new regulatory interactions with other proteins within the periplasmic domain. An alternative explanation might be that changes in AmtB structure created a more efficient ammonium transporter, which would be a beneficial trait for organisms living in generally low-nutrient conditions, as many planctomycetes and verrucomicrobia do. This section is modified from reference (Kamneva et al. 2010), with permission of Oxford University Press, Society for Molecular biology and Evolution.

Fig. 7.5
figure 00075

Ammonium transporter protein family and representatives in the PVC organisms. Gene family with insertions in coils under positive selection. Adapted from (Kamneva et al. 2010). (a) Phylogenetic tree of the ammonium transporter protein family (ion-coupled transporter, according to KEGG), individual sequences are designated by GenBank accession numbers and species name. Branches with significantly high level of insertions in coils are shown in red. Two additional names correspond to PDB accession numbers for the homologous sequence with determined tertiary structure and to the corresponding secondary structural elements identified based on tertiary structure. (b) Corresponding multiple sequence alignment of the members of the protein family and E. coli AmtB sequence from PDB (the parts of alignment corresponding to transmembrane helices have been trimmed). Bright and light shadings correspond to 50 % identical or similar (based on PAM250) residues in the sequences of the protein family. Red palette—periplasmic parts; yellow palette—trimmed transmembrane parts; green palette—cytoplasmic parts. Last line represents types of secondary structure elements, based on tertiary structure of E. coli AmtB 1XQE:A (Zheng et al. 2004). Residue W148 at the beginning of periplasmic coil between transmembrane helices four and five is marked with asterisk. (c) Structural model 1XQE:A of E. coli AmtB ammonium transporter was used to show periplasmic side coiled and α-helical regions with an unexpectedly high number of insertions (marked in red) (Adapted from Kamneva et al. 2010)

Full size image

7.9 Genome Content Evolution in PVC

It is commonly acknowledged that genome dynamics (gene family acquisition, expansion, and contraction) contribute significantly to the general evolution of bacterial species as well as to the emergence of particular ecological and physiological properties of microorganisms. Genome content can vary a great deal, even between closely related bacterial species, in terms of the presence and size of particular gene families. This is mostly attributed to horizontal gene transfer and rapid loss of non-beneficial genes, due to large population sizes and high pressure of natural selection in bacteria.

Reconstruction of gene family evolution is fundamental for understanding the evolution of living organisms; a number of methods for studying genome content evolution have been developed over the years. These methods can be broadly divided into two major classes: gene-tree species-tree reconciliation-based and phyletic pattern-based methods grouped by the type of input information and parsimony-based and likelihood-based methods grouped by the utilized statistical framework. The majority of existing approaches take into account at least two major types of evolutionary events contributing to gene family evolution, i.e., gene duplication and loss. Some of them consider horizontal gene transfer as well, which makes such methods especially useful for characterizing evolution of bacterial species. This section is modified from reference (Kamneva et al. 2012), with permission of Oxford University Press, Society for Molecular biology and Evolution.

7.9.1 Gene Family Dynamics in PVC Genomes

In order to evaluate rates and patterns of gene family gain, loss, expansion, and contraction, we performed analysis of all the gene families inferred as described in Sect. 7.6 above, using a parsimonious gene-tree species-tree reconciliation procedure implemented in the AnGST program (David and Alm 2010). The algorithm identifies evolutionary events (HGT, gene duplication, and loss) necessary to explain discrepancy between gene and species phylogeny. We performed gene-tree species-tree reconciliation for every gene family which allowed us to explicitly infer the evolutionary history of every gene family in the PVC superphylum and to evaluate genome size for every ancestral genome on the PVC species tree (Kamneva et al. 2012). The summarized results of this analysis are depicted in Fig. 7.6.

Fig. 7.6
figure 00076

Genome content evolution in the PVC superphylum. Adapted from Kamneva et al. (2012). Events of genome content evolution were mapped onto lineages of the species tree; only the PVC clade is shown here. Numbers at every node, either ancestral or extant, represent genome size and number of genes in multigene families (for instance, the I. pallida genome contains 3,722 genes, out of which 604 genes are predicted to be members of multigene families). Numbers above and below every lineage represent the number of birth/duplication/transfer and loss events, respectively, predicted to occur on the branch. Numbers shown in red or blue correspond to accelerated or decelerated rates of events on the branch (for instance, on the lineage leading to V. spinosum, 1,761 births, 646 duplications, 1,292 transfers, and 265 loss events occurred. This observed event count implies elevated gene birth and duplication rates on this lineage and low gene loss rate). Species names are abbreviated as in Fig. 7.3 (Adapted from Kamneva et al. 2012)

Full size image

Our analysis suggested that the common ancestor of all PVC organisms had a genome containing 3,106 genes, of which 786 were predicted to be in multigene families. Thus, the origin of the four PVC phyla involved extensive loss of ancestral genes on some lineages and acquisition of novel genes through various mechanisms on other lineages. After the Planctomycetes split from the rest of the superphylum, a number of planctomycete lineages underwent acquisition of gene families and expansion of existing families. This process resulted in the largest genomes in the group, seen in the extant species G. obscuriglobus and R. baltica, where many new families appeared. The ancestor of Verrucomicrobia, Chlamydiae, and Lentisphaerae is predicted to have possessed a relatively small genome containing 2,557 genes. This ancestral gene set was shaped primarily by gene loss and gene family contraction on the lineage leading to the common ancestor of Lentisphaerae and Chlamydiae and by gene gain and gene family expansion on the lineage leading to the ancestor of Verrucomicrobia. The ancestral genome of Chlamydiae and Lentisphaerae was further minimized by gene loss and gene family contraction on the lineage leading to the ancestor of all Chlamydiae and even further on the lineage leading to obligate intracellular pathogens belonging to the genera Chlamydia and Chlamydophila. Conversely, gene gain and gene duplication contributed significantly to the evolution of genomes on many lineages of phyla Lentisphaerae and Verrucomicrobia, with the exception of A. muciniphila and M. infernorum lineages. This section is modified from reference (Kamneva et al. 2012), with permission of Oxford University Press, Society for Molecular biology and Evolution.

7.9.2 Horizontal Gene Transfer Among PVC Organisms and from Members of Other Bacterial Groups

Horizontal (lateral) gene transfer (HGT) is a major source of diversity in bacterial genomes (Jain et al. 2003). The frequency of HGT between different organisms depends on a number of factors such as genome size and similarity in GC content, as well as ecological factors such as carbon source utilization and oxygen tolerance (probably pointing to similarity in ecological habitat) (Jain et al. 2003; Smillie et al. 2011). Other factors which intuitively should affect frequency of HGT between organisms include differences in codon usage and divergence of regulatory motifs between donor and recipient organisms. Modes of interaction in an ecosystem must also be critical to HGT. A number of transfer events between different organisms were detected in various gene families within our data set, asserted on the basis of gene-tree species-tree reconciliation data (Table 7.4). Several genes were predicted to be acquired from the Candidatus Solibacter usitatus Ellin6076 lineage on different branches of Planctomycetes- and Verrucomicrobia-specific clades as well as from Deltaproteobacteria (D. vulgaris Miyazaki F and S. cellulosum So ce 56 lineages) on various superphylum lineages but excluding the chlamydial clade. We also detected a large number of genes transferred laterally to A. muciniphila from the B. fragilis NCTC 9343 lineage. These findings suggest previous ecological contexts shared between the recipient PVC lineages and donor lineages outside the superphylum. This section is modified from reference (Kamneva et al. 2012), with permission of Oxford University Press, Society for Molecular biology and Evolution.

Table 7.4 Non-PVC organisms (extant or ancestral) frequently acting as donors in lateral transfer events
Full size table

7.10 Large Outer Membrane Autotransporter Barrel Domain Protein Family in Verrucomicrobia

A number of protein-sorting systems acting within bacterial cells are known. The PEP-CTERM/EpsH system has been recently proposed to facilitate outer membrane/cell wall directed trafficking of proteins in environmental microorganisms (Haft et al. 2006). The system includes two main components, the first being EpsH (exopolysaccharide locus protein H). It is predicted to act as a signal peptidase upon proteins containing a conserved carboxy-terminal PEP motif, followed by a stretch of hydrophobic residues and a short segment of positively charged amino acids (PEP-CTERM, TIGR02595) (Haft et al. 2006). V. spinosum has the largest known number of proteins bearing the PEP-CTERM motif. One of the proteins containing a PEP-CTERM domain is a divergent multimember family of large outer membrane autotransporter barrel domain proteins. These large proteins, in addition to PEP-CTERM, are predicted to contain an autotransporter-associated beta-strand repeat domain (PF12951) and type I signal sequences, along with a putative lipid attachment site. However, we lack exact functional predictions (and experimental validation) for these protein domains.

Sixteen autotransporter barrel domain proteins have been detected within V. spinosum, and two within C. flavus. Analysis of the genomic neighborhood revealed the presence of four hypothetical genes in proximity to the autotransporter genes. These genes are organized in one or two operons and are present only within these genomic clusters. The structure of one representative region is shown in Fig. 7.7. While these four genes lack explicit functional assignment, they possess conserved domains TIGR02596, TIGR02600, TIGR02599, and TIGR02598, respectively. The highly organized structure of genomic regions containing these genes, along with PEP-CTERM bearing autotransporter-encoding genes, suggests functional relatedness of these proteins and EpsH. While the exact molecular mechanisms and functional importance of this association are yet to be uncovered, it is an exciting example of a gene family related to protein sorting, considering the distinctive cell plan of PVC organisms.

Fig. 7.7
figure 00077

Schematic of one representative genomic region containing large outer membrane autotransporter barrel domain protein gene, and four co-located hypothetical genes arranged in an operon. Locus names are shown for all the loci (Vspi), domain structure shown schematically for Vspi_22965, names of domains detected in four genes in the operon are shown under the map, number of genomic clusters found in V. spinosum and C. flavus genomes is indicated

Full size image

7.11 PVC Genomics Database

In order to facilitate comparative genomic analysis of distantly related PVC organisms, we are currently constructing the PVC Genome Database (anticipated location: http://www.pvcgenomics.org) (Kamneva et al. unpublished). The database will include all Planctomycetes, Verrucomicrobia, and Lentisphaerae species with publicly available genome sequences and representative species from phylum Chlamydiae. The database will contain information on organisms (including taxonomy, isolation site, culture collection accessibility, cellular structure), their genomes (status of genome project, genome structure, number of features in a genome), genes and gene products (including protein annotation, predicted PFAM protein domains, hydrophobicity patterns, signal peptides, subcellular localization, and functional sites), orthologous protein families (including high-quality orthologous groups, alignments, gene phylogeny, and history of gene families in terms of duplication, loss, and transfer events), predicted operons, orthologous groups of operons, and putative upstream genomic regions for every operon in the group (including predicted conserved DNA motifs, representing transcription factor binding sites). The database will be integrated with BLAST (Altschul et al. 1990) and Psi-Square programs (Glazko et al. 2006). BLAST will be used to search for sequences similar to a query sequence and sequences in the PVC Genome Database. The Psi-Square program will be used to search for features with specific phyletic patterns. The PVC Genome Database will aim to provide a high-quality genomics and evolutionary biology resource for the PVC research community and will be freely available for use.

7.12 Concluding Remarks

The field of PVC superphylum genomics and bioinformatics is currently at an exciting stage. The PVC research community, together with sequencing center partners, has generated genome data and analysis from organisms representing the breadth and diversity of the superphylum. These data are invaluable for understanding the genomic basis for various intriguing properties of PVC superphylum members, predicting previously unknown properties, and formulating hypotheses for experimental testing. They also allow us to begin to unravel the evolutionary history of these fascinating organisms and their genomes. Lastly, the data challenge us to develop new tools for optimally extracting biologically significant information from the genomes. We hope that this brief survey of PVC comparative genomics and bioinformatics provides a useful resource for the PVC research community and stimulates both deeper bioinformatic analysis and new experimental studies.