Genome-Wide Survey of Genes Under Positive Selection in Avian Pathogenic Escherichia coli Strains

Abstract

The ability to obtain bacterial genomes from the same host has allowed for comparative studies that help in the understanding of the molecular evolution of specific pathotypes. Avian pathogenic Escherichia coli (APEC) is a group of extraintestinal strains responsible for causing colibacillosis in birds. APEC is also suggested to possess a role as a zoonotic agent. Despite its importance, APEC pathogenesis still has several cryptic pathogenic processes that need to be better understood. In this work, a genome-wide survey of eight APEC strains for genes with evidence of recombination revealed that ∼14% of the homologous groups evaluated present signs of recombination. Enrichment analyses revealed that nine Gene Ontology (GO) terms were significantly more represented in recombinant genes. Among these GO terms, several were noted to be ATP-related categories. The search for positive selection in these APEC genomes revealed 32 groups of homologous genes with evidence of positive selection. Among these groups, we found several related to cell metabolism, as well as several uncharacterized genes, beyond the well-known virulence factors ompC, lamB, waaW, waaL, and fliC. A GO term enrichment test showed a prevalence of terms related to bacterial cell contact with the external environment (e.g., viral entry into host cell, detection of virus, pore complex, bacterial-type flagellum filament C, and porin activity). Finally, the genes with evidence of positive selection were retrieved from genomes of non-APEC strains and tested as were done for APEC strains. The result revealed that none of the groups of genes presented evidence of positive selection, confirming that the analysis was effective in inferring positive selection for APEC and not for E. coli in general, which means that the study of the genes with evidence of positive selection identified in this study can contribute for the better understanding of APEC pathogenesis processes.

Introduction

Comparative genomic analysis is an important tool for understanding how genomes evolve and adapt to new niches. The huge availability of genomic data nowadays allows researchers to use comparative genomic tools to survey how adaptation to distinct ecological niches changes the genomic composition of organisms. In the case of a pathogenic organism, positive selection and homologous recombination are determining factors for molecular adaptation. Both factors contribute to genetic variability among organisms, which in turn determines the population diversity and adaptation of pathogenic microorganisms to different groups of hosts (Wu et al., 1999; Chen et al., 2006; Lefébure and Stanhope, 2007, 2009; Petersen et al., 2007; Orsi et al., 2008; Aguileta et al., 2010; Xu et al., 2011).

Positive selection can be detected by the analysis of molecular data using sophisticated statistical models based on the frequency of synonymous (d_S) and nonsynonymous (d_N) substitutions in homologous codons. The ratio of nonsynonymous substitutions/synonymous substitutions (ω = d_N/d_S) is commonly calculated to indicate what type of selection pressure is acting on a particular protein site. Thus, values of ω close to one suggests the occurrence of neutral evolution, ω <1 indicates purifying selection, and ω >1 indicates positive selection (Yang and Nielsen, 2002). When a nonsynonymous mutation occurs and results in a phenotypic change with a decrease in fitness for individuals carrying the new allele, the phenotype will be under purifying selection pressure, and the alleles will quickly be removed from populations (Ng and Henikoff, 2001). Nevertheless, some positions in certain genes can present a significantly higher than expected frequency of nonsynonymous mutations, indicating selection for the emergence of new alleles in place of the old gene copy. In pathogenic bacterial strains, enhanced fitness is frequently related to virulence- and niche-associated factors because these factors ensure bacterial survival and should be fixed in the population. An analysis of the substitution rates of nonsynonymous and synonymous mutations can reveal the action of positive selection on specific genes, with many of these genes being frequently involved in the parasite-host relationship (Yang and Bielawski, 2000; Vandamme, 2003; Aguileta et al., 2009, 2010).

Escherichia coli typically colonizes the gastrointestinal tract of birds and mammals in a commensal relationship. However, some strains have acquired specific virulence factors that enable adaptation to new niches, allowing these strains to cause different diseases (Russo and Johnson, 2000; Kaper et al., 2004). These features are very diverse and are associated with a many characteristics, such as adhesion (Verma et al., 2016), invasiveness (Pilatti et al., 2016), and improved fitness (de Paiva et al., 2015). Avian pathogenic E. coli (APEC) causes colibacillosis in birds, a disease that may be present as a localized or systemic infection, resulting in loss of production, decrease in egg production, carcasses condemnations, increased spending on treatment cost, and mortality, all factors that are responsible for severe losses in the multibillion dollar poultry production industry worldwide (Barnes et al., 2003). A subset of APEC presents significant similarities with E. coli isolated from cases of bacteremia and urinary tract infections in humans (Maluta et al., 2014), supporting a possible role of APEC as a zoonotic agent, probably as a foodborne pathogen.

Considering the importance of this pathotype, a comparative genomic study may help highlight possible adaptation processes that enable these strains to recognize and successfully survive in their host. Indeed, the comparison of genomic data from different strains within the same species allows an investigation of their evolutionary process, as well as phenotype modeling, because the similarities are sufficiently close to identify specific changes, and the differences are sufficiently minor to enable the detection of a large number of homologous genes (Petersen et al., 2007).

The study of genes under positive selection has previously been performed with E. coli strains. Using a branch-site model, 29 positively selected genes were identified in a uropathogenic E. coli (UPEC) strain (Chen et al., 2006), and 23 genes with evidence of positive selection were detected in E. coli and Shigella flexneri strains (Chen et al., 2006; Petersen et al., 2007). However, despite the huge importance of APECs for bird health, up to now there is not a comprehensive study searching and contextualizing the set of genes evolving with signs of recombination or positive selection in this pathotype.

Therefore, the goal of the present study is to perform a genome-wide search in APEC genomes present in the GenBank database to identify genes with evidence of recombination or evolving under positive selection. Knowledge of these genes can contribute to better understanding of APEC pathogenicity. Furthermore, these genes represent good candidates for future studies that would seek to analyze individual gene contributions to biological processes related to APEC pathogenesis.

Materials and Methods

Data acquisition and homology inference

All APEC genome and plasmid sequences were obtained from the GenBank database. The genome set analyzed included eight APEC genome sequences, comprising two complete genomes and six draft assemblies (Table 1).

Table 1.

Characteristics of the Avian Pathogenic Escherichia coli Strains Used in Positive Selection Analyses

Genome and plasmids	Serotype	Disease	Genome size (Mb)	GenBank Accession No.	Reference
SEPT362	OR:H10	Septicemia	5.28	AOGL00000000.1	Rojas et al. (2013)
S17	O113:H4	Septicemia	4.59	AOGN00000000.1	Rojas et al. (2013)
O08	O38:H10	Omphalitis	5.08	AOGM00000000.1	Rojas et al. (2013)
SCI-07	ONT:H31	Swollen head syndrome	4.97	AJFG00000000.1	Rojas et al. (2012)
APEC O1	O1:K1:H7	Colibacillosis	5.08	CP000468.1	Johnson et al. (2007)
pAPEC-O1-ColBM	—	—	0.17	DQ381420.1	Johnson et al. (2006a)
pAPEC-O1-R	—	—	0.24	DQ517526.1	Johnson et al. (2006b)
APECO78	O78	Colibacillosis	4.80	CP004009.1	Mangiamele et al. (2013)
Chi7122	O78:H9	Colibacillosis	4.77	CAJR00000000.1	Dziva et al. (2013)
IMT2125	O78:H9	Airsacculitis	4.75	CAJT00000000.1	Dziva et al. (2013)

The groups of homologous genes were determined by OrthoMCL software (Li et al., 2003) using default parameters. We filtered out any sequences with no start codon and/or no stop codon (based on the NCBI bacterial codon table 11), sequences with ambiguous nucleotides and nonmultiples of three.

To verify whether the genes inferred to be under positive selection in APEC genomes are also under positive selection in other E. coli genomes that occupy other ecological niches, we selected E. coli strains belonging to two groups: one of them composed of only nonpathogenic strains from different origins, while the other was composed of human pathogenic strains (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/fpd) to perform positive selection test. A scheme demonstrates how the tests were done (Fig. 1).

FIG. 1.

Experimental approach to detect genes under positive selection.

Data filtering and sequence alignment

We used a preliminary version of POTION to perform all the analysis from now on, except where explicitly stated otherwise (Hongo et al., 2015). It is well known that APEC represents a very diverse pathotype, with considerable variation in genome length and gene content (Table 1) (Maturana et al., 2011; Dziva et al., 2013). Therefore, analyzing only the 1:1 orthologs present in all eight genomes would result in the loss of valuable information. Furthermore, six of the eight genomes studied are draft versions that contain plasmid genes together with their assembled contigs. For these reasons, we chose to analyze the 1:1 orthologs present in at least four genomes, which is also a reasonable number from which to infer positive selection.

We also filtered out sequences from genomes with evidence of paralogy in any lineage, that is, duplicated sequences, leaving only putative 1:1 orthologs in the final set of sequences within each group. Finally, all groups with fewer than four sequences were excluded from further analyses. The groups remaining after all filtering steps are referred to hereafter as valid groups of homologs.

All the valid groups of homolog genes from the previous step were translated using the NCBI bacterial codon table 11, and the protein sequences were aligned using PRANK with default parameters (Löytynoja and Goldman, 2010). POTION uses an internal subroutine to generate protein-guided codon alignment files, which were used in downstream analyses.

Detection of homologous recombination and enrichment analysis

Homologous recombination was inferred from the codon alignment files for each valid group of homologs with PhiPack software, using all three recombination tests (Phi, NSS, and MaxChi²) (Bruen et al., 2006). All groups of homologous genes with a q-value smaller than 0.05, for at least two of the three tests, were considered as recombinants and were removed from the search for positive selection.

The groups with evidence of recombination were surveyed for possible enrichment in the context of Gene Ontology (GO) terms, using Blast2GO (Conesa et al., 2005). First, we annotated the longest sequence of each valid group of homologous using Blast2GO with default parameters. Afterward, the groups with evidence of recombination were surveyed for enrichment of specific GO terms (Fisher's Exact test, false discovery rate [FDR], and Benjamini–Hochberg correction). We used a FDR of 0.05 to control the inflation of Type 1 errors due to the multiple hypothesis testing scenarios.

Detection of genes under positive selection

The individual, multiple alignment files of protein sequences from all valid groups of homologs without evidence of recombination were used to reconstruct phylogenetic trees using proml from the PHYLIP package with 1000 bootstraps (Retief, 2000). The resulting phylogenetic trees and the multiple alignment files of codons previously obtained were utilized as input to the codeml from PAML 4.3 to detect possible groups of homologous genes with evidence of positive selection (Yang, 2007). We evaluated two popular nested codon evolution models (M1a × M2 and M7 × M8) to search for specific sites with evidence of positive selection in each alignment of a valid group. The search for positive selection in codeml was performed by comparing the log-likelihood values of codon models that do not allow sites with positive selection (M1a and M7) with the values of the more general coupled models that also do allow for site classes with the occurrence of positive selection (M2 and M8, respectively). The p-values were calculated as 2Δℓ (twice the difference in likelihood of the two nested models evaluated) based on the χ ² distribution with two degrees of freedom. We used an FDR value of 0.1 to control the occurrence of false positive cases of positive selection due to the multiple testing scenarios.

Positive selection test was also performed on non-APEC strains. For that, the same genes detected in positive selection test performed on APEC strains were retrieved from the non-APEC genomes and grouped according to the homology. The same procedures for positive selection detection as described for APEC strains were performed separated for pathogenic and nonpathogenic strains.

Results

We obtained a total of eight APEC genomes and two plasmid sequences annotated from GenBank (Table 1) and identified 5572 homologous groups after OrthoMCL analysis. The genes likely to inflate false-positive selection detection, such as gene fragments, ambiguous coding sequences (CDS) or possible pseudogenes, and assembly errors, were then removed using the filtering procedures (“Materials and Methods” section). We also filtered out any genes with evidence of genomic duplication because orthologous and paralogous genes are likely to evolve under distinct selective pressures (Koonin, 2005). After the cleaning procedures, we obtained 4143 putative 1:1 orthologous groups in at least four APEC genomes. These groups, referred to as valid groups of homologs, were then subject to recombination and positive selection tests.

Recombination analysis determined that ∼14% (605) of the valid homologous groups showed significant evidence of homologous recombination. We performed an enrichment analysis searching for GO categories overrepresented in the list of genes with evidence of recombination (test set) compared with the total set of valid homologous groups (background set), to eventually gain a deeper insight about which gene categories may be target of recombination process in APECs. We found nine GO terms from the molecular function and biological process categories enriched in these genes. Enriched GO categories included ATP-dependent DNA helicase activity, ATP binding, DNA repair, tryptophan biosynthetic process, cofactor binding, metal ion transmembrane transporter activity, protein kinase activity, GTP binding, and ligase activity forming carbon-nitrogen bonds (Table 2).

Table 2.

Enrichment Analysis Results of the Genes

GO-ID	Term	Category	FDR	p
Genes with evidence of homologous recombination
GO:0004003	ATP-dependent DNA helicase activity	F	1.91E−06	9.29E−09
GO:0005524	ATP binding	F	3.33E−06	1.71E−08
GO:0006281	DNA repair	P	6.16E−03	5.31E−05
GO:0000162	Tryptophan biosynthetic process	P	8.39E−03	9.82E−05
GO:0048037	Cofactor binding	F	1.70E−02	2.16E−04
GO:0046873	Metal ion transmembrane transporter activity	F	3.60E−02	4.90E−04
GO:0004672	Protein kinase activity	F	3.60E−02	5.05E−04
GO:0005525	GTP binding	F	4.58E−02	6.80E−04
GO:0016879	Ligase activity, forming carbon-nitrogen bonds	F	4.58E−02	6.92E−04
Genes under positive selection
GO:0046718	Viral entry into host cell	P	2.87E−01	7.73E−04
GO:0009597	Detection of virus	P	2.87E−01	7.73E−04
GO:0046930	Pore complex	C	8.18E−01	3.75E−03
GO:0009420	Bacterial-type flagellum filament	C	1.00E+00	1.25E−02
GO:0008919	Lipopolysaccharide glucosyltransferase I activity	F	1.00E+00	1.25E−02
GO:0017150	tRNA dihydrouridine synthase activity	F	1.00E+00	1.87E−02
GO:0008918	Lipopolysaccharide 3-alpha-galactosyltransferase activity	F	1.00E+00	1.87E−02
GO:0004555	Alpha, alpha-trehalase activity	F	1.00E+00	1.87E−02
GO:0015288	Porin activity	F	1.00E+00	1.87E−02
GO:0003849	3-Deoxy-7-phosphoheptulonate synthase activity	F	1.00E+00	2.48E−02
GO:0009360	DNA polymerase III complex	C	1.00E+00	2.48E−02
GO:0046961	Proton-transporting ATPase activity, rotational mechanism	F	1.00E+00	3.10E−02
GO:0015768	Maltose transport	P	1.00E+00	3.10E−02
GO:0042777	Plasma membrane ATP synthesis coupled proton transport	P	1.00E+00	3.10E−02
GO:0045261	Proton-transporting ATP synthase complex, catalytic core F(1)	C	1.00E+00	3.10E−02
GO:0046933	Proton-transporting ATP synthase activity, rotational mechanism	F	1.00E+00	4.31E−02
GO:0005991	Trehalose metabolic process	P	1.00E+00	4.31E−02
GO:0006265	DNA topological change	P	1.00E+00	4.31E−02
GO:0003916	DNA topoisomerase activity	F	1.00E+00	4.31E−02

C, cellular component; F, molecular function; P, biological process.

FDR, false discovery rate.

The 3538 valid groups of homologs with no evidence of recombination were tested for positive selection using the likelihood ratio test (LRT) for comparing two sets of popular nested codon evolution models implemented in codeml (M1a × M2 and M7 × M8). After multiple test correction considering a q-value smaller than 0.1, 69 genes with significant evidence of positive selection in at least one test remained. A total of 32 genes were detected by both models and were considered to be under positive selection for the APEC strains analyzed (Table 3). We performed enrichment analysis of these genes in the same way that was described for genes with homologous recombination, but in this case, the genes under positive selection were counterpoised against all valid group of homologous. In this analysis, 19 different GO classes were enriched (Table 2). Among these classes, the first three with smaller p-values and with two genes (viral entry into the host cell and detection of virus and pore complex) could be associated with bacterial defense. Furthermore, genes well known to be related with bacterial host interaction/contact were also detected (fliC, waaW, waaL, ompC). This preliminary analysis was able to show that several of the genes under positive selection can be viewed as important examples of the host-parasite defense/attack system, illustrating the premise that genes under positive selection are frequently related to virulence- and niche-associated factors.

Table 3.

Genes Under Positive Selection

Name or locus tag	Description	D ^a (M1a, M2)	D ^a (M7, M8)	p-value (M1a, M2)	p-value (M7, M8)	q-value (M1a, M2)	q-value (M7, M8)
ynaB	Side tail fiber protein homolog from lambdoid prophage Rac	345.28045	358.74151	0.00000	0.00000	0.00000	0.00000
fliC	Flagellin	119.53389	119.63426	0.00000	0.00000	0.00000	0.00000
ygeH	Uncharacterized protein YgeH	107.58482	108.53150	0.00000	0.00000	0.00000	0.00000
aroH	Phospho-2-dehydro-3-deoxyheptonate aldolase	91.85176	97.38464	0.00000	0.00000	0.00000	0.00000
ynjI	Inner membrane protein YnjI	89.85381	90.01318	0.00000	0.00000	0.00000	0.00000
hyfD	Hydrogenase-4 component D	68.43640	66.71463	0.00000	0.00000	0.00000	0.00000
atpG	ATP synthase gamma chain	68.23229	68.14872	0.00000	0.00000	0.00000	0.00000
A364_15858	Uncharacterized protein	61.28512	63.00258	0.00000	0.00000	0.00000	0.00000
ompC	Outer membrane protein C	57.27197	50.55480	0.00000	0.00000	0.00000	0.00000
yjgL	Uncharacterized protein YjgL	56.93260	58.29760	0.00000	0.00000	0.00000	0.00000
yibA	Protein YibA	52.89469	52.47698	0.00000	0.00000	0.00000	0.00000
yoaG	Protein YoaG	49.75668	49.80356	0.00000	0.00000	0.00000	0.00000
ybjE	Lysine exporter LysO	47.57903	51.42612	0.00000	0.00000	0.00000	0.00000
ykgF	Uncharacterized electron transport protein YkgF	40.14336	41.08301	0.00000	0.00000	0.00000	0.00000
lamb	Maltoporin	39.70085	40.65415	0.00000	0.00000	0.00000	0.00000
ymfN	Putative terminase large subunit	39.07911	40.66824	0.00000	0.00000	0.00000	0.00000
phoH	Protein PhoH	33.66622	33.97075	0.00000	0.00000	0.00001	0.00001
waaW	UDP-galactose:(Galactosyl) LPS alpha 1,2-galactosyltransferase	32.62220	34.34270	0.00000	0.00000	0.00001	0.00000
flmC	Stable plasmid inheritance protein	29.05988	29.06001	0.00000	0.00000	0.00006	0.00005
waaL	Lipid A-core:surface polymer ligase WaaL	28.60949	41.92750	0.00000	0.00000	0.00007	0.00000
lacY	MFS transporter	26.83302	26.84275	0.00000	0.00000	0.00016	0.00016
dusA	tRNA-dihydrouridine(20/20a) synthase	26.29841	26.51280	0.00000	0.00000	0.00019	0.00017
A364_13802	Uncharacterized protein	20.33342	20.42138	0.00004	0.00004	0.00367	0.00324
rarD	Protein RarD	18.82603	18.83449	0.00008	0.00008	0.00720	0.00643
yrdD	Putative DNA topoisomerase	18.85201	19.01847	0.00008	0.00007	0.00720	0.00607
pspA	Phage shock protein	16.09634	20.67627	0.00032	0.00003	0.02715	0.00297
hemp	Uncharacterized protein	15.77697	15.79701	0.00038	0.00037	0.03071	0.02661
yehZ	Glycine betaine-binding protein YehZ	15.24351	16.47216	0.00049	0.00027	0.03872	0.02025
yicS	Uncharacterized protein YicS	15.17010	16.30599	0.00051	0.00029	0.03883	0.02129
ydjO	Uncharacterized protein YdjO	13.88418	14.29050	0.00097	0.00079	0.07147	0.05167
treA	Periplasmic trehalase	13.39038	14.70744	0.00124	0.00064	0.08862	0.04318
dnaN	DNA polymerase III subunit beta	13.31820	19.95040	0.00128	0.00005	0.08910	0.00395

The p-value and q-value have a cutoff value smaller than 0.05 and 0.1, respectively.

D represents the results of the likelihood ratio test for the set of models selected.

To confirm whether the positive selection results obtained for APEC strains were valid for E. coli in general, we separated two groups of E. coli strains, nonpathogenic and human pathogenic strains (Supplementary Table S1), retrieved from their genome the gene homologues to those 32 detected on APEC positive selection analysis and performed the same tests described above for APEC. No genes found to be under positive selection in APEC genomes were under positive selection in non-APEC genomes.

Discussion

Positive selection and recombination are important evolutionary forces involved in the phenotype-modeling process. The changes resulting from these two processes are responsible for genetic variations that may increase intraspecies genetic diversity and provide variability for the evolutionary process to act on (Xu et al., 2011). Recombination is an important force in bacterial pathogen adaptation, for several reasons. This phenomenon occurs at higher rates than mutation and, consequently, is a major force to introduce variability in a genome.

In this study, the enrichment analysis of the groups with evidence of recombination indicated that most genes were placed in ATP binding, ATP-dependent DNA helicase activity, GTP binding, DNA repair, and cofactor binding classes. These enriched classes include proteins with diverse biological functions, such as transport activity, sensory regulators, cell division, metabolism, chaperones, and resistance to antimicrobials. The high number of GO terms related to ATP-interacting proteins can lead us to think that this could be the result of a recombination bias for this type of protein because of the ATP needs of RecA (recombinase A) (Goodman, 2014).

The comparative, evolutionary genomic study of different strains through the detection of genes that are under positive selection allows for the identification of factors that may provide an adaptive advantage to these strains. In general, it has been shown that a large proportion of genes under positive selection have a close relationship with the organism lifestyle, in the specific case of pathogenic strains, with parasitism-related genes (Chen et al., 2006; Petersen et al., 2007; Xu et al., 2011). Thus, greater knowledge of genes under positive selection may help further understand the pathogenicity processes and, consequently, aid the development of vaccines and drugs (Polley et al., 2003).

As a pathogen of major importance, several pathotypes of E. coli have been previously surveyed for positive selection. In a search for positive selection in specific lineages of phylogenetic tree (branch-site tests), 29 genes under positive selection were identified in the UTI89 strain. Several of the detected genes were involved in cell surface structure, nutrient acquisition, DNA metabolism, and urinary tract infection processes. Another comparative genome study identified 23 genes with evidence of positive selection in E. coli and S. flexneri strains (Petersen et al., 2007). Most of the gene products were predicted to be located on the cell surface and involved in interactions with the host, phage, or other bacteria, suggesting a coevolutionary context (Chen et al., 2006; Petersen et al., 2007).

When we compare the genes that were identified to be under positive selection with genes identified in the E. coli-related studies described above, we note that the ompC product is commonly identified in all studies (Chen et al., 2006; Petersen et al., 2007). OmpC is a trimeric porin with a B-barrel domain localized on the outer membrane that allows permeability for the transport of ions and hydrophilic solutes across membranes (Nikaido, 1994; Schirmer, 1998). OmpC is one of the most abundant proteins expressed in E. coli and has been demonstrated to be expressed at significantly higher levels during urinary tract infections. High expression of OmpC in vivo may be an advantage in environments, where small molecular toxins are plentiful because the pore formed by OmpC also serves as an exit path for toxins (Nikaido, 1994). Another important feature of this protein is that the pore formed by OmpC also functions as a receptor for various phages (Yu et al., 2000).

The beta barrel porin LamB, found to be under positive selection in this study and in the study conducted by Petersen et al. (2007), is a maltose outer membrane porin, the main function of which is the diffusion of maltose (Wang et al., 1997). The gene name refers to the Lambda phage because this porin also operates as a receptor for this phage (Randall-Hazelbauer and Schwartz, 1973). The four sites found to be under positive selection were residues 408, 409, 410, and 411 of the 446-amino acid LamB protein, corresponding to the loop 9 region involved in phage binding (Newton et al., 1996). These results are similar to those obtained by Petersen et al. (2007), suggesting that LamB positive selection may be the result of selection to avert phage binding.

Two other genes under positive selection found in our study, waaL and waaW, are related to the biosynthesis of the outer membrane lipopolysaccharide (LPS). The gene waaW product is the enzyme UDP-galactose, (galactosyl) LPS alpha 1,2-galactosyltransferase, that acts during LPS biosynthesis. LPS consists of three parts as follows: lipid A, the core oligosaccharide (OS) and O-antigen (Raetz and Whitfield, 2002; Valvano, 2011). The gene waaL product is required to ligate the terminal sugar of lipid A to OS (Valvano, 2011); the absence of WaaL results in rough bacteria with no O-antigen polysaccharide (Mulford and Osborn, 1983; McGrath and Osborn, 1991). WaaL is a promising target for antivirulence factors due to O-antigen host–pathogen interactions and its localization in the periplasmic space (Ruan et al., 2012).

Pathogenic E. coli strains can be identified by O:H serotyping, which corresponds to the cell surface LPS O-antigen and the flagellar antigen H (Orskov et al., 1977). The fliC gene, observed to be under positive selection in our study, encodes the structural subunits of flagellin. The comparison of several homologous flagellin sequences revealed that the N- and C-terminal regions are well conserved among species. In contrast, the central region is rather variable and is responsible for the antigenic property of the filament, closely interacting with the host defense system (Wieler et al., 1997). Our positive selection analysis revealed that the sites with higher nonsynonymous substitution rates are in the central region of the protein sequence. As fliC positive selection can be the result of the intimate contact between the host and bacterial cells during infection, selection may function to avoid bacterial recognition by the host immune system.

The phage-shock-protein A (PspA) was first discovered during an E. coli filamentous phage infection due to its high concentrations (Brissette and Russel, 1990). The Psp system includes proteins PspA, B, C, D, E, F, and G and responds to several factors responsible for damage to membrane integrity, such as filamentous phage infection, medium downshifts, extreme temperatures, the mislocation of some envelope proteins, osmolarity, ethanol concentration, and the presence of proton ionophores, such as carbonylcyanide m-chlorophenylhydrazone (CCCP) (Model et al., 1997). This regulon is required to maintain the proton-motive force that protects the inner membrane, and this stress adaptation is important for bacterial growth under stress conditions and for pathogenicity (Darwin, 2005). The sensory proteins PspB and PspC positively regulate the expression of PspA through protein–protein interactions under stress conditions resulting in the release of the inhibitory complex PspA-PspF and in an increased expression of other Psp proteins (Jovanovic et al., 2006). PspA, which is under positive selection, can be considered an example of how selection acts on factors that enable bacterial survival under stress conditions.

After reviewing all genomic information currently available, we found that several genes have still not been characterized. In this positive selection study, some genes corresponding to uncharacterized protein functions were detected (yjgL, ymfN, ydiE, yicS, ydjO, A364_13802, and A364_15858). The identification of unknown genes with occurrence of positive selection may help in further understanding of the processes of APEC pathogenicity. Even though several of these gene activities remain cryptic, they constitute promising targets for study and, if their role in bacterial pathogenesis were demonstrated, they may constitute interesting targets for the development of attenuated vaccine strains.

Conclusions

In conclusion, the genes found to be under positive selection in APEC were not under positive selection in non-APEC E. coli. This result may indicate that, in our restricted analysis, the genes that are under positive selection in APEC appear to be specific of this pathotype and may represent an APEC response to environmental pressure.

Footnotes

Acknowledgments

The authors thank the Laboratório Multiusuário de Bioinformática of Embrapa Informática Agropecuária for the computational resources. This project was supported by grant numbers 2010/51421-8 (FAPESP), 2012/04931-6 (FAPESP), PNPD0146080 (CAPES), and 485279/2011-8 (CNPq).

Disclosure Statement

No competing financial interests exist.

References

Aguileta

, Lengelle

, Marthey

, Chiapello

, Rodolphe

, Gendrault

, Yockteng

, Vercken

, Devier

, Fontaine

, Wincker

, Dossat

, Cruaud

, Couloux

, Giraud

. Finding candidate genes under positive selection in Non-model species: Examples of genes involved in host specialization in pathogens. Mol Ecol, 2010; 19:292–306.

Aguileta

, Refrégier

, Yockteng

, Fournier

. Giraud T. Rapidly evolving genes in pathogens: Methods for detecting positive selection and examples among fungi, bacteria, viruses and protists. Infect Genet Evol, 2009; 9:656–670.

Barnes

, Vaillancourt

, Gross

. Colibacillosis. In: Saif

(ed.). Diseases of Poultry. Ames: Iowa State University Press, 2003, pp. 631–656.

Brissette

, Russel

. Secretion and membrane integration of a filamentous phage-encoded morphogenetic protein. J Mol Biol, 1990; 211:565–580.

Bruen

, Philippe

, Bryant

. A simple and robust statistical test for detecting the presence of recombination. Genetics, 2006; 172:2665–2681.

Chen

, Hung

C-S

, Xu

, Reigstad

, Magrini

, Sabo

, Blasiar

, Bieri

, Meyer

, Ozersky

, Armstrong

, Fulton

, Latreille

, Spieth

, Hooton

, Mardis

, Hultgren

, Gordon

. Identification of genes subject to positive selection in uropathogenic strains of Escherichia coli: A comparative genomics approach. Proc Natl Acad Sci U S A, 2006; 103:5977–5982.

Conesa

, Götz

, García-Gómez

, Terol

, Talón

, Robles

. Blast2GO: A universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics, 2005; 21:3674–3676.

Darwin

. The phage-shock-protein response. Mol Microbiol, 2005; 57:621–628.

de Paiva

, Leite

, da Silva

, Rojas

, de Pace

, Conceição

, Sperandio

, da Silveira

. Influence of the major nitrite transporter NirC on the virulence of a Swollen Head Syndrome avian pathogenic E. coli (APEC) strain. Vet Microbiol, 2015; 175:123–131.

10.

Dziva

, Hauser

, Connor

, van Diemen

, Prescott

, Langridge

, Eckert

, Chaudhuri

, Ewers

, Mellata

, Mukhopadhyay

, Curtiss

, Dougan

, Wieler

, Thomson

, Pickard

, Stevens

. Sequencing and functional annotation of avian pathogenic Escherichia coli serogroup O78 strains reveal the evolution of E. coli lineages pathogenic for poultry via distinct mechanisms. Infect Immun, 2013; 81:838–849.

11.

Goodman

. The discovery of error-prone DNA polymerase V and its unique regulation by RecA and ATP. J Biol Chem, 2014; 289:26772–26782.

12.

Hongo

, de Castro

, Cintra

, Zerlotini

, Lobo

. POTION: An end-to-end pipeline for positive Darwinian selection detection in genome-scale data through phylogenetic comparison of protein-coding genes. BMC Genomics, 2015; 16:567.

13.

Johnson

, Johnson

, Nolan

. Complete DNA sequence of a ColBM plasmid from avian pathogenic Escherichia coli suggests that it evolved from closely related ColV virulence plasmids. J Bacteriol, 2006a;188:5975–83.

14.

Johnson

, Kariyawasam

, Wannemuehler

, Mangiamele

, Johnson

, Doetkott

, Skyberg

, Lynne

, Johnson

, Nolan

. The genome sequence of avian pathogenic Escherichia coli strain O1:K1:H7 shares strong similarities with human extraintestinal pathogenic E. coli genomes. J Bacteriol, 2007; 189:3228–36.

15.

Johnson

, Wannemeuhler

, Scaccianoce

, Johnson

, Nolan

. Complete DNA sequence, comparative genomics, and prevalence of an IncHI2 plasmid occurring among extraintestinal pathogenic Escherichia coli isolates. Antimicrob Agents Chemother, 2006b;50:3929–33.

16.

Jovanovic

, Lloyd

, Stumpf

, Mayhew

, Buck

. Induction and function of the phage shock protein extracytoplasmic stress response in Escherichia coli . J Biol Chem, 2006; 281:21147–21161.

17.

Kaper

, Nataro

, Mobley

HLT

. Pathogenic Escherichia coli . Nat Rev Microbiol, 2004; 2:123–140.

18.

Koonin

. Orthologs, paralogs, and evolutionary genomics. Annu Rev Genet, 2005; 39:309–338.

19.

Lefébure

, Stanhope

. Evolution of the core and pan-genome of Streptococcus: Positive selection, recombination, and genome composition. Genome Biol, 2007; 8:R71.

20.

Lefébure

, Stanhope

. Pervasive, genome-wide positive selection leading to functional divergence in the bacterial genus Campylobacter. Genome Res, 2009; 19:1224–1232.

21.

, Stoeckert

Jr. , Roos

. OrthoMCL: Identification of ortholog groups for eukaryotic genomes. Genome Res, 2003; 13:2178–2189.

22.

Löytynoja

, Goldman

. webPRANK: A phylogeny-aware multiple sequence aligner with interactive alignment browser. BMC Bioinformatics, 2010; 11:579.

23.

Maluta

, Logue

, Tiba Casas

, Meng

, Lopes Guastalli

, Galvao Rojas

, Montelli

, Sadatsune

, Ramos

MdC

, Nolan

, da Silveira

. Overlapped sequence types (STs) and serogroups of avian pathogenic (APEC) and human extra-intestinal pathogenic (ExPEC) Escherichia coli isolated in Brazil. PLoS One, 2014; 9:e105016.

24.

Mangiamele

, Nicholson

, Wannemuehler

, Seemann

, Logue

, Li

, Tivendale

, Nolan

. Complete Genome Sequence of the Avian Pathogenic Escherichia coli Strain APEC O78. Genome Announc, 2013; 1:e0002613.

25.

Maturana

, de Pace

, Carlos

, Mistretta Pires

, Amabile de Campos

, Nakazato

, Guedes Stheling

, Logue

, Nolan

, Dias da Silveira

. Subpathotypes of avian pathogenic Escherichia coli (APEC) exist as defined by their syndromes and virulence traits. Open Microbiol J, 2011; 5:55–64.

26.

McGrath

, Osborn

. Localization of the terminal steps of O-antigen synthesis in Salmonella Typhimurium. J Bacteriol, 1991; 173:649–654.

27.

Model

, Jovanovic

, Dworkin

. The Escherichia coli phage-shock-protein (psp) operon. Mol Microbiol, 1997; 24:255–261.

28.

Mulford

, Osborn

. An intermediate step in translocation of lipopolysaccharide to the outer membrane of Salmonella Typhimurium. Proc Natl Acad Sci U S A, 1983; 80:1159–1163.

29.

Newton

, Klebba

, Michel

, Hofnung

, Charbit

. Topology of the membrane protein LamB by epitope tagging and a comparison with the X-ray model. J Bacteriol, 1996; 178:3447–3456.

30.

, Henikoff

. Predicting deleterious amino acid substitutions. Genome Res, 2001; 11:863–874.

31.

Nikaido

. Porins and specific diffusion channels in bacterial outer membranes. J Biol Chem, 1994; 269:3905–3908.

32.

Orsi

, Maron

, Nightingale

, Jerome

, Tabor

, Wiedmann

. Lineage specific recombination and positive selection in coding and intragenic regions contributed to evolution of the main Listeria monocytogenes virulence gene cluster. Infect Genet Evol, 2008; 8:566–576.

33.

Orskov

, Orskov

, Jann

. Serology, chemistry, and genetics of O and K antigens of Escherichia coli . Bacteriol Rev, 1977; 41:667–710.

34.

Petersen

, Bollback

, Dimmic

, Hubisz

, Nielsen

. Genes under positive selection in Escherichia coli . Genome Res, 2007; 17:1336–1343.

35.

Pilatti

, Boldrin de Paiva

, Rojas

, Leite

, Conceição

, Nakazato

, Dias da Silveira

. The virulence factor ychO has a pleiotropic action in an Avian Pathogenic Escherichia coli (APEC) strain. BMC Microbiol, 2016; 16:35.

36.

Polley

, Chokejindachai

, Conway

. Allele frequency-based analyses robustly map sequence sites under balancing selection in a malaria vaccine candidate antigen. Genetics, 2003; 165:555–561.

37.

Raetz

, Whitfield

. Lipopolysaccharide endotoxins. Annu Rev Biochem, 2002; 71:635–700.

38.

Randall-Hazelbauer

, Schwartz

. Isolation of the bacteriophage lambda receptor from Escherichia coli . J Bacteriol, 1973; 116:1436–1446.

39.

Retief

. Phylogenetic analysis using PHYLIP. Methods Mol Biol, 2000; 132:243–258.

40.

Rojas

, Maluta

, Parizzi

, Koenigkan

, Yang

, Yu

, Pereira

, Dias da Silveira

. Genome Sequences of Avian Pathogenic Escherichia coli Strains Isolated from Brazilian Commercial Poultry. Genome Announc, 2013; 1:e0011013.

41.

Rojas

, Parizzi

, Tiba

, Chen

, Pereira

, Sangal

, Yang

, Yu

, Dias da Silveira

. Draft genome of a Brazilian avian-pathogenic Escherichia coli strain and in silico characterization of virulence-related genes. J Bacteriol, 2012; 194:3023.

42.

Ruan

, Loyola

, Marolda

, Perez-Donoso

, Valvano

. The WaaL O-antigen lipopolysaccharide ligase has features in common with metal ion-independent inverting glycosyltransferases. Glycobiology, 2012; 22:288–299.

43.

Russo

, Johnson

. Proposal for a new inclusive designation for extraintestinal pathogenic isolates of Escherichia coli: ExPEC. J Infect Dis, 2000; 181:1753–1754.

44.

Schirmer

. General and specific porins from bacterial outer membranes. J Struct Biol, 1998; 121:101–109.

45.

Valvano

. Common themes in glycoconjugate assembly using the biogenesis of O-antigen lipopolysaccharide as a model system. Biochemistry (Mosc), 2011; 76:729–735.

46.

Vandamme

. Basic concepts of molecular evolution. In Salemi

, Vandamme

(eds.): The Phylogenetic Handbook: A Pratical Approach to DNA and Protein Phylogeny. Cambridge: Cambridge University Press, 2003, pp 1–21.

47.

Verma

, Rojas

, Maluta

, Leite

, da Silva

, Nakazato

, Dias da Silveira

. Fimbria-encoding gene yadC has a pleiotropic effect on several biological characteristics and plays a role in avian pathogenic Escherichia coli pathogenicity. Infect Immun, 2016; 84:187–193.

48.

Wang

, Dutzler

, Rizkallah

, Rosenbusch

, Schirmer

. Channel specificity: Structural basis for sugar discrimination and differential flux rates in maltoporin. J Mol Biol, 1997; 272:56–63.

49.

Wieler

, McDaniel

, Whittam

, Kaper

. Insertion site of the locus of enterocyte effacement in enteropathogenic and enterohemorrhagic Escherichia coli differs in relation to the clonal phylogeny of the strains. FEMS Microbiol Lett, 1997; 156:49–53.

50.

, Chiang

, Shiue

, Wang

, Sheen

, Huang

, Syu

. Recombination of hepatitis D virus RNA sequences and its implications. Mol Biol Evol, 1999; 16:1622–1632.

51.

, Chen

, Zhou

. Genome-wide evidence for positive selection and recombination in Actinobacillus pleuropneumoniae . BMC Evol Biol, 2011; 11:203.

52.

Yang

. PAML 4: Phylogenetic analysis by maximum likelihood. Mol Biol Evol, 2007; 24:1586–1591.

53.

Yang

, Bielawski

. Statistical methods for detecting molecular adaptation. Trends Ecol Evol, 2000; 15:496–503.

54.

Yang

, Nielsen

. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol Biol Evol, 2002; 19:908–917.

55.

, Ko

, Chen

, Chang

, Syu

. Characterization of the distal tail fiber locus and determination of the receptor for phage AR1, which specifically infects Escherichia coli O157:H7. J Bacteriol, 2000; 182:5962–5968.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB