Complete genome sequence of Actinobacillus equuli subspecies equuli ATCC 19392T

Actinobacillus equuli subsp. equuli is a member of the family Pasteurellaceae that is a common resident of the oral cavity and alimentary tract of healthy horses. At the same time, it can also cause a fatal septicemia in foals, commonly known as sleepy foal disease or joint ill disease. In addition, A. equuli subsp. equuli has recently been reported to act as a primary pathogen in breeding sows and piglets. To better understand how A. equuli subsp. equuli can cause disease, the genome of the type strain of A. equuli subsp. equuli, ATCC 19392T, was sequenced using the PacBio RSII sequencing system. Its genome is comprised of 2,431,533 bp and is predicted to encode 2,264 proteins and 82 RNAs.


Introduction
Actinobacillus equuli subsp. equuli, previously known as 'Bacillus viscosum-equi', or 'Shigella equirulis', is a common resident of the oral flora of healthy horses, as well as that of the alimentary and genital tracts [1,2]. It has also been reported to be present in other host species such as mice, seemingly without ill effect [3] and on rare occasions, has been transmitted through bite wounds to humans [4]. A. equuli subsp. equuli is the etiological agent of sleepy foal disease, an acute form of fatal septicemia in neonatal foals that may progress to a chronic form, joint ill disease, producing lesions in the kidneys, joints, and lungs [5][6][7][8]. Horses with A. equuli infection can present with arthritis, bronchitis, pneumonia, pleuritis, peritonitis, sepsis, endocarditis, pericarditis, nephritis, meningitis, metritis, and abortion [7,[9][10][11][12]. A. equuli subsp. equuli was previously proposed to act as a secondary pathogen in foals; however, a recent study by Layman and colleagues [13] has revealed that A. equuli subsp. equuli has the potential to act as a primary pathogen given favourable conditions. Recently, it has been reported to also be a primary pathogen in sows and piglets [14,15].
The hemolytic counterpart of this bacterium, A. equuli subsp. haemolyticus, is isolated more frequently from the respiratory tract rather than the oral cavity. It can also cause septicemia and sequelae such as arthritis and meningitis, respiratory tract infections, and mare reproductive loss syndrome [8,10,16].
The similar colonial morphology and biochemical markers and shared 16S rRNA sequences make differentiation of A. equuli from Actinobacillus suis difficult [8]. In addition, little is known about the virulence factors of A. equuli subsp. equuli. To be better able to identify and to improve our understanding of the mechanism of pathogenhost interactions [7], the genome of the type strain A. equuli subsp. equuli strain ATCC 19392 T was sequenced. This strain was isolated from foal blood and deposited in the American Type Culture Collection by the Equine Research Station (New Market, UK) in 1953 [17].

Classification and features
As a member of the genus Actinobacillus, A. equuli subsp. equuli belongs to the family Pasteurellaceae, class Gammaproteobacteria [18] (Table 1). Phylogenetic analysis using 16S rRNA sequences suggests that A. equuli subsp. equuli is most closely related to A. suis and A. hominis (Figure 1).

Figure 1
Phylogenetic tree based on 16S rRNA sequences of Actinobacillus sensu stricto species plus A. capsulatus and H. parasuis as outgroups.
A. equuli subsp. equuli is indicated in bold. The RDP aligner, which applies the Jukes-Cantor corrected distance model to align sequences, and the RDP Tree Builder, which implements the Weighbor algorithm [36] for tree construction were used. Tree building also involved a bootstrapping process in which the values to the left of the branches illustrate the frequency of occurrence of a branch in 100 replicates [37].

Genome sequencing information
Genome project history A. equuli subsp. equuli was selected for sequencing because of its importance to the horse industry as the etiologic agent of sleepy foal disease and joint ill disease [7]. Sequencing was done at the McGill University and Génome Québec Innovation Centre (Montréal, QC, Canada) using the PacBio RS II DNA Sequencing System, and assembled using PacBio RS II software and Celera Assembler. A. equuli subsp. equuli was annotated using the NCBI Prokaryotic Genome Annotation   Table 2 [38].

Growth conditions and genomic DNA preparation
A. equuli subsp. equuli was grown from a frozen (-70°C) seed stock on sheep blood agar plates overnight in an atmosphere of 5% CO 2 at 37°C. After subculture, wellisolated colonies were used for genomic DNA isolation. Cells were lysed using modified B1 (150 mM Tris · Cl, 50 mM EDTA, 0.5% Tween®-20, 0.5% Triton X-100, pH 8.0) and B2 (750 mM NaCl, 50 mM MOPS, 15% isopropanol, 0.15% Triton X-100, pH 7.0) buffers. DNA was then column purified using a QIAGEN Plasmid Midi Kit (Qiagen, Germany) following manufacturer's protocol for binding and elution. The resultant DNA preparation was characterized using a NanoDrop model ND1000 Spectrophotometer and was diluted to a concentration of~0.47 mg/μl.

Genome sequencing and assembly
Single Molecule, Real-Time DNA sequencing (Pacific Biosciences) [39] was done to obtain the genome sequence of the A. equuli subsp. equuli ATCC 19392 T . A total of 133,616 raw subreads were generated with an average length of 4,348 bp using two SMRT Cells in a PacBio RSII sequencer. The resultant subread length cutoff value, 29.42, was used in the Basic Local Alignment with Successive Refinement step [40] where short reads were used to correct for errors on long reads [39]. The corrected reads were assembled into contigs according to the Hierarchical Genome Assembly Process (HGAP) workflow using the Celera Assembler and refined using BLASR to align raw reads on contigs [39]. Final processing was conducted using Quiver, a variant calling algorithm, to generate high quality consensus sequences [39]. There were a total of 4,777 corrected reads with an average length of 7,804 bp and a final product of one contig.

Genome annotation
Genes were identified using the NCBI Prokaryotic Genome Annotation Pipeline. The prediction software, GeneMark, is integrated into the pipeline and performs unsupervised gene finding using heuristic Markov Models [41]. Additional gene prediction analysis and functional annotation was performed within the Integrated Microbial Genomes (IMG) platform [42] developed by the Joint Genome Institute [43] (Table 3).

Genome properties
The genome of A. equuli subsp. equuli is a single circular chromosome that is 2,431,533 bp in length with a G + C content of approximately 40.3%. It is predicted to contain 2,264 genes, of which 2,182 code for proteins and 82 for RNA; 11 pseudogenes are also present (Table 3 and Figure 3). Approximately 3/4 of the predicted genes can be assigned to one of 25 functional COG categories (Table 4). Of particular note with regard to virulence are several lipopolysaccharide genes predicted to encode biosynthetic enzymes for the O-antigen and lipid A components. Adhesins of different types were  observed including several autotransporters; a tight adherence locus; prepilins, and fimbriae; a filamentous hemagglutinin homolog was also detected. In addition, several putative iron acquisition systems are present including those for siderophores, hemoglobin and transferrin. A number of toxin and hemolysin genes were also identified including an aqxCABD operon, although compared to the aqxCABD of A. equuli subsp. haemolyticus there are many point mutations and sizable deletions at both ends of the aqxA gene. Other regions of particular interest include an integron and Mu-like phage, identified using PHAST [44].

Insights from the genome sequence
Given the marked similarities of A. equuli and A. suis there has been some debate as to whether these organisms should be a single species. In the current study we determined that the A. equuli subsp. equuli 16S genes are 99% identical to those of both A. suis H91-0380 and the A. suis type strain, ATCC 33415, consistent with membership in the same species. Further, as can be seen in the circular maps below, the genome of A. equuli subsp. equuli is very similar to that of A. suis again suggesting that A. equuli subsp. equuli and A. suis might be the same species (Figure 4). On the other hand, when genomes of A. suis H91-0380 and A. suis ATCC 33415 were compared with that of A. equuli subsp. equuli using the ANI calculator [45], the ANI value of both comparisons was 93.82%, which is lower than 95%, the recommended cutoff value for delineating species [46]. In-silico DNA-DNA hybridization, done using a Genome Blast Distance Phylogeny approach to generate genome based distance measures for phylogenetic inferences, also demonstrated differences between A. equuli and A. suis. The Genome-to-Genome Distance Calculator [47] revealed a distance of 0.0685 between A. suis H91-0380 and A. equuli subsp. equuli, with a DDH estimate of 51.40% +/-2.66. A DDH similarity below 70% is interpreted as two species being distinct; 79% is used to discriminate between subspecies [48]. The DDH estimate exceeding the 70% species threshold was determined from logistic regression to be 23.14%. In terms of subspecies relatedness, the probability of exceeding the 79% threshold was 4.82% between A. equuli subsp. equuli and A. suis H91-0380. The distance calculated between A. suis ATCC 33415 and A. equuli subsp. equuli and their DDH estimate was 0.0681 and 51.60% +/-2.66, respectively. The probability that DDH exceeded 70% and 79% for A. suis ATCC 33415 and A. equuli subsp. equuli were 23.66% and 4.94%, respectively.
Taken together, these analyses are consistent with the notion that A. suis and A. equuli subsp. equuli are related but distinct species, and care is needed to correctly identify them.

Conclusions
A. equuli subsp. equuli can induce fatal septicemia in foals resulting in significant economic losses in the equine industry; as well, A. equuli subsp. equuli has recently been reported to cause septicemia in swine of all ages. Our analysis of the A. equuli subsp. equuli genome indicates that A. suis and A. equuli subsp. equuli are closely related yet distinct species. At the present time little is known about how A. equuli subsp. equuli causes disease or the factors that control species and tissue tropism. More research including biological experiments is required to better understand the pathogenesis of A. equuli and it is hoped this reported genome sequence of A. equuli subsp. equuli ATCC 19392 T will provide vital information for such studies. In addition, pathway analysis and genome studies may help improve our understanding of host-pathogen interactions of A. equuli subsp. equuli and other Actinobacillus species and aid in the design of diagnostic tools and antimicrobial agents.