Francisella tularensis subspecies holarctica FTNF002-00 strain was originally obtained from the first known clinical case of bacteremic F. tularensis pneumonia in Southern Europe isolated from an immunocompetent individual. The FTNF002-00 complete genome contains the RD23 deletion and represents a type strain for a clonal population from the first epidemic tularemia outbreak in Spain between 1997–1998. Here, we present the complete sequence analysis of the FTNF002-00 genome. The complete genome sequence of FTNF002-00 revealed several large as well as small genomic differences with respect to two other published complete genome sequences of F. tularensis subsp. holarctica strains, LVS and OSU18. The FTNF002-00 genome shares >99.9% sequence similarity with LVS and OSU18, and is also ∼5 MB smaller by comparison. The overall organization of the FTNF002-00 genome is remarkably identical to those of LVS and OSU18, except for a single 3.9 kb inversion in FTNF002-00. Twelve regions of difference ranging from 0.1–1.5 kb and forty-two small insertions and deletions were identified in a comparative analysis of FTNF002-00, LVS, and OSU18 genomes. Two small deletions appear to inactivate two genes in FTNF002-00 causing them to become pseudogenes; the intact genes encode a protein of unknown function and a drug:H+ antiporter. In addition, we identified ninety-nine proteins in FTNF002-00 containing amino acid mutations compared to LVS and OSU18. Several non-conserved amino acid replacements were identified, one of which occurs in the virulence-associated intracellular growth locus subunit D protein. Many of these changes in FTNF002-00 are likely the consequence of direct selection that increases the fitness of this subsp. holarctica clone within its endemic population. Our complete genome sequence analyses lay the foundation for experimental testing of these possibilities.
Citation: Barabote RD, Xie G, Brettin TS, Hinrichs SH, Fey PD, Jay JJ, et al. (2009) Complete Genome Sequence of Francisella tularensis Subspecies holarctica FTNF002-00. PLoS ONE4(9): e7041. https://doi.org/10.1371/journal.pone.0007041
Editor: Niyaz Ahmed, University of Hyderabad, India
Received: May 15, 2009; Accepted: July 29, 2009; Published: September 16, 2009
This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Funding: The genome sequencing was supported by funds from the Intelligence Technology Innovation Center. The sequencing of the genome was performed at the JGI production genomics facility in Walnut Creek, CA. Genome finishing was performed by the JGI at Los Alamos National Laboratory, Los Alamos, NM. Annotation was performed at BioHealthBase, a Bioinformatics Resource Center supported by the NIH/NIAID. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
F. tularensis, the etiologic agent of tularemia, is a Gram-negative, facultative intracellular zoonotic pathogen originally isolated from ground squirrels in 1911 during a plague investigation in Tulare County, CA . It is known to infect numerous animal species including humans , . The organism has been weaponized and is currently listed as a Category A Select Agent posing a significant threat to humans especially due to its extremely low infectious dose when acquired through the inhalation route .
F. tularensis is comprised of three subspecies: subsp. tularensis (also Type A or F.t. tularensis hereafter), subsp. holarctica (also Type B or F.t. holarctica), and subsp. mediaasiatica. F. novicida, officially considered a separate species, is often considered a fourth F. tularensis subspecies and will be treated as such in this publication . Epidemiology suggests only the tularensis and holarctica subspecies are clinically significant in humans , and between them it is believed that Type A causes more serious disease , .
Comparative genome hybridization (CGH) studies have demonstrated small differences between F.t. tularensis and F.t. holarctica populations based on genome contents. Eight regions of difference (RD) were found to distinguish all F.t. tularensis from F.t. holarctica strains tested from a geographically diverse collection , . These RD include the subsp. tularensis-specific RDtularensis6, a segment of the Francisella Pathogenicity Island (FPI) encompassing the pdpD gene shown to be involved in virulence differences between F. tularensis subspecies , and which is truncated and non-functional in F.t. holarctica. In addition, other comparative genome studies have shown extensive genome rearrangements including translocations and inversions between F.t. tularensis and F.t. holarctica , . These changes were highly conserved in strains from multiple geographic locations, particularly in F.t. holarctica populations, indicating a likely early origin during subspecies divergence . The rearrangement events likely caused changes in expression patterns of genes adjacent to the rearrangement sites, possibly explaining some of the virulence differences between the two subspecies.
Beyond differences in virulence, the F. tularensis subspecies also differ in geographic distribution, with F.t. tularensis largely limited to North America whereas F.t. holarctica has been documented throughout the N. Hemisphere . Within the tularensis subspecies, genetic diversity is well documented, with distinct genetic groups A.I and A.II found in central and western parts of the U.S., respectively , , . Genetic analyses of F.t. holarctica strains from globally derived populations showed that strains from different continents could similarly be found in nearly all of the major F.t. holarctica clusters . Collectively, these studies suggest F.t. holarctica may be relatively newer and genetically more homogeneous than F.t. tularensis.
With respect to F.t. holarctica, phylogeographic genome variation has been detected among global F.t. holarctica populations ,  including parts of Europe as demonstrated by a new RD, termed RD23. RD23 was found in all DNA samples from a subset of strains from Spain and France, which by phylogenetic analysis comprise a highly clonal population likely to have recently emerged on the Iberian Peninsula . Most of the strains analyzed were from an epidemic outbreak primarily in rabbits and humans in Spain between 1997–1998, and which was the first reported tularemia outbreak in that country , . The outbreak involved over 500 human cases of tularemia , , , ; as expected given the geographic location, all samples tested were identified as F.t. holarctica , , , .
In the present work, we report on the first complete genome sequence (CGS) of F.t. holarctica strains carrying RD23. The sequence, designated FTNF002-00, is from a strain isolated from a previously published clinical case in France involving an immunocompetent 56-year old male with bacteremic F.t. holarctica pneumonia . FTNF002-00 revealed several genomic differences with respect to other published F.t. holarctica sequences as will be described here. These differences suggest that the F.t. holarctica subspecies may be more diverse than suggested to date, and also include polymorphisms which may contribute to the overall fitness of FTNF002-00 and clonal success of strains containing its genotype within its defined population niche.
General Genome Features
An overview of the FTNF002-00 genome as well as the complete genomes of six other previously sequenced F. tularensis strains is presented in Table 1. FTNF002-00 possesses the smallest genome of the fully sequenced F. tularensis strains to date. Its genome is smaller by ∼5 MB as compared to the other two sequenced subsp. holarctica strains (Table 1). Incidentally, the same amount of difference (∼5.5 MB) exists between the genome sizes of the sequenced A.II strain and the two A.I strains (Table 1), while the sequenced strain of the novicida subspecies, which are non-pathogenic in humans, has a genome that is larger by about 11.5 MB than the largest genome among the pathogenic F. tularensis strains. The % G+C content of the different strains is approximately 32%, with slight variation between the subspecies. All seven strains listed in Table 1 encode identical numbers of tRNAs, rRNAs and other structural and regulatory RNAs. The numbers of pseudogenes as well as predicted proteins in the seven fully sequenced F. tularensis strains do not correlate with their genome size and vary considerably, likely due to the fact that many of these genomes have not been manually curated; it is noteworthy, however, that the larger non-virulent subsp. novicida genome contains substantially fewer predicted pseudogenes than the F.t. tularensis and F.t. holarctica genomes by comparison.
The FTNF002-00 genome shares high sequence similarity (>99.9% on average) with the other two subspecies holarctica genomes, LVS and OSU18. The FTNF002-00 and the OSU18 genomes start at the same nucleotide bases, however the first nucleotide in the LVS genome maps to nucleotide position number 275 in both FTNF002-00 and OSU18. The organization and synteny of the FTNF002-00 genome is almost identical to that of the other subspecies holarctica genomes (data not shown) except for a single ∼3.8 kb inversion in the FTNF002-00 genome, with respect to both LVS and the original assembly of OSU18 (Fig. 1). Interestingly, OSU18 has subsequently been re-assembled and found to contain two unique inverted regions of ∼5 kb and ∼18 kb, respectively . The single inversion in FTNF002-000 was probably caused by homologous recombination between two 874 bp identical inverted repeats that mark the two edges of the inverted genomic fragment. The inverted repeats overlap two ISFtu1 transposase ORFs (FTA_0277 and FTA_0281) at the two ends of the inverted fragment that encodes a single 331 aa long hypothetical protein (FTA_0279) that is as yet unique to F. tularensis. The protein bears weak sequence similarity to a short region within the pfam04488 glycosyltransferase sugar-binding region containing DXD motif. Multiple homologs of this protein occur in the sequenced genomes of the subsp. tularensis and holarctica strains. The FTNF002-00, LVS, and OSU18 genomes encode two, three, and three homologs, respectively. The genomes of two strains of subsp. tularensis subtype A.I, FSC198 and SCHU-S4, encode three homologs each, while the single completely sequenced genome of subtype A.II of the subsp. tularensis strain WY96-3418 encodes four copies of the hypothetical protein. It is interesting to note that the sequenced genome of subsp. novicida strain U112 does not contain a homolog of this hypothetical protein suggesting that the hypothetical protein may play a role in the pathogenic strains of F. tularensis.
Arrows indicate the direction of gene transcription. Genes and intergenic regions are not to scale. Homologous genes in the three genomes have been assigned the same color. Solid arrows indicate putative functional genes and hatched or grey arrows represent pseudogenes. The horizontal box on the FTNF002-00 chromosomal fragment indicates the 3,868 bp fragment (positioned at 247478–251345) that is inverted with respect to the other two genomes. A vertical box over the three genomes indicates the location of a 962 bp region that is different in OSU18 (see Table 2). The two equal sized horizontal boxes on the OSU18 genomic fragment indicate tandem duplications of a 1933 bp region (see Table 2).
Regions of Difference (RD) within subspecies holarctica
Each of the three F.t. holarctica genomes contain genomic regions that are absent in one or both of the other two subsp. holartica genomes (Fig. 2A). The FTNF002-00 genome revealed three regions that are different compared to LVS and OSU18 (Table 2). The first region, an approximately 1.5 kb deletion, termed the RD23 region, has been previously noted in FTNF002-00 . The RD23 sequence encodes either hypothetical genes (in LVS) or pseudogenes (in OSU18), and may simply represent a non-functional region as it shows varying degrees of degradation in the different F.t. holarctica strains. However, it is worth noting that the RD23 deletion occurs in close proximity upstream of a regulator that is predicted to belong to the xenobiotic response element family of DNA-binding transcriptional regulators. Possibly, the RD23 deletion alters the expression of this transcriptional regulator in FTNF002-00.
(A) Total numbers of regions of difference (RDs), (B) number of small insertions and deletions (indels), and (C) number of single nucleotide polymorphisms (SNPs) that are either shared or unique between the three strains. Further details of these differences are included in Tables 2, 3, and 4.
A second small genomic deletion of 276 bp occurs within a pseudogene (FTA_1387) in FTNF002-00. The corresponding genes in OSU18 (FTH_1284) as well as in all three sequenced subspecies tularensis genomes are also pseudogenes, suggesting that the protein is dispensable for pathogenicity. The third region reveals that a 362 bp fragment, at 1819665–1820026 in the FTNF002-00 chromosome, replaces a second genomic region of about 1.6 kb found in LVS and OSU18. The two edges of the fragment partially overlap transposase sequences on either side. This fragment along with the flanking transposase genes (a total of 2.9 kb region) occurs in 3 copies in both LVS and OSU18, and only occurs in two copies in FTNF002-00. Interestingly, one of the copies in FTNF002-00 occurs in the inverted genomic region (see above), suggesting that the genomic inversion and the deletion events may have been linked evolutionarily.
Next, FTNF002-00 and LVS share three genomic regions that are different in OSU18 (Fig. 2A, Table 2), one of which involves a ∼2 kb tandem duplication in OSU18. Of the remaining two, one region (993 bp) is immediately downstream of the inverted fragment in FTNF002-00 (see Fig. 1) and is predicted to encode a small hypothetical protein (FTA_0282) and overlaps a 628 aa chaperone protein, HtpG (FTA_0283). This ∼1 kb sequence occurs in single copy in FTNF002-00, LVS, all three subsp. tularensis genomes as well as in subsp. novicida U112. However, the corresponding region in OSU18 is replaced by a completely different sequence 962 bp long that carries partial sequence of ISFtu1 and encodes two divergently-oriented pseudogenes (Fig. 1). The other, a 187 bp fragment shared by FTNF002-00 and LVS, but deleted in OSU18, occurs in the middle of an ORF predicted to encode a 671 aa competence-related protein. Homologs of this protein are predicted in LVS, subsp. tularensis A.II strain WY96-3418, as well as in subsp. novicida strain U112; however, genomes of the two subsp. tularensis A.I strains (FSC198 and SCHU-S4) contain pseudogenes carrying a frameshift mutation.
Further, there are a total of six genomic fragments in FTNF002-00 that are shared with OSU18, but absent in LVS (Table 2), two of which include tandemly duplicated regions in LVS. The remaining four include: (a) an approximately 1.5 kb region (at 407597–409077); (b) a ∼0.5 kb fragment (at 361986–362511); (c) a ∼0.2 kb fragment (at 484148–484319); and (d) a ∼0.2 kb fragment (at 1041146–1041356). Fragment (c) above is also lacking in the subtype A.I subsp. tularensis genomes.
Lastly, as mentioned above, both LVS and OSU18 each possess tandem duplications in certain regions that are found only in one copy in the other two genomes (Table 2). The first region involves a 1933 bp fragment in OSU18, carrying a copy of ISFtu1, that is not present in the corresponding regions in the LVS and FTNF002-00 genomes. This region occurs in two tandem copies in OSU18 and only the second copy is shared in OSU18, LVS and FTNF002-00. Two other regions occur as tandem duplications in LVS but in single copy in the other F. tularensis genomes. One of these regions in LVS involves tandem duplication of a ∼3.9 kb long fragment (at 1392385–1396298 in the LVS chromosome). Only one copy of the 3.9 kb region occurs in the other six sequenced F. tularensis genomes. The second region involves tandem duplication of a 468 bp fragment in LVS (at 1169486–1169953 in the LVS chromosome) and overlaps the 5′ end of a gene encoding a hypothetical protein.
Despite these regions of difference, there are no genomic regions where all three subsp. holarctica strains are different.
Small Insertions and Deletions
In order to identify the unique variations in FTNF002-00 as compared to those of LVS and OSU18, we performed a detailed analysis of small stretches of insertions and deletions. Forty-two short sequences ranging from 10–100 nucleotides were identified that were either inserted or deleted in one or more of the three genomes compared (Fig. 2B, Table 3). Of these, fifteen regions are different in FTNF002-00, while 18 are different in OSU18, and 9 in LVS. Eleven of the fifteen short sequences are absent in the corresponding position in the FTNF002-00 genome, while four short sequences are lacking in the other two holarctica genomes. The insertions/deletions are either mostly in intergenic regions (five) or in genes that are either pseudogenes (two) or ORFs encoding hypothetical proteins (two) in the holarctica genomes. However, interestingly, two deletions appear to inactivate two distinct genes in FTNF002-00 causing them to become pseudogenes specifically in FTNF002-00. The first one is a frame-shifting deletion of 16 nucleotides within the FTA_1097 pseudogene, the corresponding functional gene in OSU18 encodes a 188 amino acid long protein (FTH_1016) of unknown function that is conserved in all the fully sequenced subsp. holarctica as well as subsp. tularensis strains. The second deletion of 20 bp also results in a frame-shifted pseudogene (FTA_1713) whose functional counterpart in OSU18 (FTH_1570) encodes a transporter of the major facilitator superfamily (MFS). This transporter is predicted to be a drug:H+ antiporter involved in the efflux of drugs and antimicrobial compounds.
The OSU18 genome contains fourteen small deletions and four small insertions compared to the FTNF002-00 and LVS genomes (Table 3). Almost half of the insertions/deletions are either in intergenic regions or in pseudogenes. Interesting in-frame insertions and deletions are also observed in two OSU18 proteins. One is a 30 bp deletion in the pgsA gene that is predicted to encode a phosphatidylglycerophosphate synthase (FTH_0277), which is shorter by 10 amino acids than the homologs in LVS and FTNF002-00. The PgsA protein in other organisms is known to be important in the biogenesis of membrane lipids. The second sequence, an in-frame insertion of 21 bp in the ATP-dependent DNA helicase gene (FTH_0671), adds a copy of a 7 aa repeat sequence, ESSPKEQ, in the 483 aa long protein in OSU18. It is worthwhile to note that homologs of this protein in the other pathogenic strains of Francisella are shorter (476 aa), while subsp. novicida GA99-3548 encodes a longer (490 aa) protein. In addition, two more sequence alterations in OSU18 inactivate two separate genes. A 25 bp insertion in the UDP-N-acetylmuramate-L-alanine ligase gene (FTH_0166) of OSU18 appears to cause frameshift mutations that are predicted to inactivate the gene. The ligase is known to be important in the biosynthesis of peptidoglycan in the bacterial cell wall in other bacteria. Another 55 bp deletion causes a pseudogene (FTH_1316) that is supposed to encode a 97 aa protein of unknown function. This small protein is encoded in FTNF002-00 (FTA_1431) and is conserved in other Francisella genomes. The gene encoding this hypothetical protein is only 23 nucleotides downstream of the gene that encodes the DNA replication and repair protein RecF protein (FTA_1430) in FTNF002-00 as well as in LVS. The short intergenic distance suggests that the two genes are part of a single operon and that they are involved in a common metabolic function - DNA replication and repair.
Single Nucleotide Polymorphisms
Comparative alignments between regions shared between all three holarctica genomes revealed 1364 polymorphic sites. Surprisingly, only a single site was found where each of three genomes contained a different nucleotide. This site occurs at position 731044 in FTNF002-00 and forms the third base of the codon encoding Alanine-175 in the glucose-inhibited division protein A (FTA_0780). The FTNF002-00, LVS, and OSU18 genomes contain T, G, and A, respectively, at the corresponding positions. The mutation causes no change in amino acid. Of the remaining 1363 polymorphic sites, six occur within the regions of difference discussed above, 504 sites contained a different nucleotide in FTNF002-00 as compared to LVS and OSU18, while LVS alone carried a different nucleotide at 412 sites, and OSU18 alone showed differences at 447 sites (Fig. 2C).
Among the 504 polymorphic sites in FTNF002-00, 172 sites are either intergenic (85 sites) or are within a gene (87 sites) but cause no change in amino acid in the encoded protein. However, 103 polymorphic sites in FTNF002-00 cause amino acid mutations in 99 proteins (Table 4). Approximately 20% of these mutations change the amino acid to an isoleucine in FTNF002-00 and the most frequent change is from valine in LVS and OSU18 to isoleucine in FTNF002-00. Based on the BLOSUM62 matrix used for scoring BLAST search alignments, the V→I change is a conserved amino acid change. Using the BLOSUM62 matrix, the most non-conservative amino acid replacements in FTNF002-00 were identified to be R→C, P→L, R→I, G→C, I→N, and D→Y mutations. Interestingly, the above non-conserved amino acid changes predominantly occurred in proteins that function in the essential cellular functions such as replication, transcription, translation, ATP and cell wall synthesis. Notably, using this approach, the intracellular growth locus, subunit D protein was also identified as one of the proteins carrying a non-conserved C→F replacement in FTNF002-00 compared to the other two subsp. holarctica genomes.
As previously noted, the FTNF002-00 genome shares >99.9% sequence similarity with the LVS and OSU18, and the organization of the three genomes appears to be largely similar with a few exceptions. FTNF002-00 is the first fully sequenced genome of F.t. holarctica carrying RD23. The strain was isolated from a French patient with F.t. holarctica bacteremia. Strains carrying RD23 are believed to be highly clonal and recently emerged in France and the Iberian Peninsula, and are virulent like FTNF002-00 as demonstrated by their association with the numerous human cases from the 1997–1998 Spanish outbreak. With respect to relative virulence of the subsp. holarctica strains compared in this analysis, LVS is attenuated, whereas both OSU18 and FTNF002-00 represent virulent strains, the former isolated from a dead beaver, and the latter from a human with F.t. holarctica bacteremia as mentioned above. Due to the uniqueness of FTNF002-00 based on the presence of the RD23 deletion and its observed pathogenic success in its geographic niche, it is important to thoroughly examine its genomic differences with respect to other sequenced F.t. holarctica genomes to understand the genomic traits that define this pathogen. Our analyses showed that the FPI-associated pdpD gene in FTNF002-00 is truncated and present as a pseudogene as it is in the other F.t. holarctica genomes, and no obvious differences in known F. tularensis virulence genes were found in the FTNF002-00 genome (data not shown). However, we did identify interesting features that include several potentially new loci that may possibly contribute to its geographically defined fitness and pathogenic success.
The smaller genome size of FTNF002-00 is interesting. Genome reduction is a hallmark of intracellular pathogens. Presumably, progressive reduction or reorganization of the genome, such that can inactivate non-essential genes resulting in net increase in the number of pseudogenes, provides optimal genotype for enhanced replication and growth of the pathogen, evasion of host immune response, and increased expression of the pathogen's virulence arsenal. This has been shown by comparative genome analyses among strains within the Mycobacterium tuberculosis complex and Yersinia species , . For example, the more virulent etiologic agent of plague, Y. pestis, is believed to have evolved from the less virulent enteric pathogen, Y. pseudotuberculosis; this process appears to have involved a massive net inactivation/loss of genes and substantial gain of pseudogenes as shown within the Y. pestis genome as compared with Y. pseudotuberculosis . Although, experimental evidences lack behind to successfully elucidate the relationship between genomic deletions/reorganization and the virulence potential of F. tularensis strains, the observed difference in the genome size of the non-pathogenic subsp. novicida and the pathogenic F.t. tularensis and F.t. holarctica strains is intriguing. Furthermore, the smaller genome of the more potent A.I pathogen than the A.II type  is suggestive of a linkage between the two properties. In this regard, when compared to the other two subsp. holarctica genomes, the RD23 deletion as well as the overall size reduction of FTNF002-00 as accounted for by the combination of RDs and insertions/deletions (Tables 2–3) observed within its genome may be a result of natural selection towards enhancing its survival, replication, and disease potential; but this would require experimental evaluation.
The RD23 deletion in FTNF002-00 may alter expression of the xenobiotic response element family transcriptional regulator, given the proximity of the two. While the function of this family of regulators remains poorly understood in bacteria, studies suggest their involvement in stress response. Enhanced response to stress could conceivably be a selective trait, especially in intracellular pathogens. Reversed orientation of one of the two noted pfam04488 homologs in FTNF002-00 caused by the 3.8 kb inversion may also play a role in its virulence as compared to other F.t. holarctica strains, perhaps due to altered expression or regulation of that gene or surrounding region. For example, downstream of the inverted region is the 1 kb RD in common with LVS, but not OSU18, predicted to encode the 628 aa HtpG chaperone protein, which itself may play a role in acclimation or fitness of FTNF002-00, even though it is likely inconsequential in LVS due the strain's attenuated state. HtpG has been shown to have a role in virulence in Porphyromonas gingivalis , and more recently has been demonstrated to be essential for survival and growth in vivo by F. t. subsp. novicida .
Further, the predicted longer pgsA gene in FTNF002-00 as compared with its 10 aa shorter homolog in OSU18 may have an important role in its pathogenic potential, especially considering the protein's role in other organisms involving biogenesis of membrane lipids, which may alter host-pathogen interactions. Likewise intriguing are potential differences in the function of some of the other genes, to include the hypothetical protein associated with the operon encoding the RecF protein and subsequent involvement in DNA replication and repair, as well as the 99 proteins in FTNF002-00 affected by amino acid changes due to SNPs. Many of these polymorphisms, especially the ones noted in genes involved in virulence and essential cellular functions, may contribute to FTNF002-00's overall fitness and pathogenic competence.
While our analyses of FTNF002-00 show numerous polymorphisms with respect to the other F.t. holarctica sequences compared and support the hypothesis that the subspecies is genetically more diverse than previously described based on a whole-genomic scale, another point of interest involves the evolutionary relationship between FTNF002-00, OSU18, and LVS. Vogler et al. recently performed detailed SNP analysis and multi-locus variable numbered tandem repeat analysis (MLVA) of a set of 496 globally diverse F. tularensis isolates . Their findings correlate with the marked diversity observed from our comparative CGS analysis approach, and together these data provide possible clues about the evolutionary relationship. For example, the hypothesis that F.t. holarctica originated in N. America and subsequently spread to Eurasia is supported based on the presence of both N. American and Scandinavian strains within the subclade defined by SNPs derived from the OSU18 CGS . The analysis tends to suggest that subclades defined by SNPs from FTNF002-00 and LVS diverged from subclades of Scandinavian origin, and corroborates previous findings  indicating that the FTNF002-00 subclade emerged very recently.
Using our comparative CGS analysis method, we can hypothesize the evolutionary events leading to the divergence of the FTNF002-00, OSU18, and LVS strains by comparing the polymorphisms between the genomes (Fig. 3). Our CGS analyses reveal likely genomic events that may have led to the emergence of the LVS and FTNF002 subclades. In the three genomes compared, the number of genomic changes observed in each of the three genomes is substantially the same (Fig. 2). This leads us to propose a model for the divergence of LVS and FTNF002 from the more ancestral OSU18 (Fig. 3). As is evident, large and small deletions, insertions, duplications, as well as SNPs have shaped the emergence of the two subclades. Availability of more genomes within each subclade in the future may provide more resolution to this evolutionary model.
Whereas SNP and analysis by other methods such as MLVA provide very robust phylogeographic genetic differentiation, comparative CGS analysis is essential to provide detailed genetic comparisons at the level of potential gene expression, such that may define pathogenicity or fitness, as presented by our analyses. Vogler et al. observed the presence of several MLVA genotypes for most of the SNP-defined subclades , and this suggests that there may be substantial genetic diversity at the CGS scale. Such potential genomic diversity may be essential to understanding the fitness of relative genotypes within the subclades, and this should be further evaluated by sequencing and comparing more representative strains. If a consensus of genetic features thus far unique to FTNF002-00 can be established among additional strains within the subclade, then it may serve as a basis for studying the functional aspects of those features as pertains to clonal success including pathogenicity of the FTNF002-00 subclade.
Materials and Methods
Strain Cultivation and Nucleic Acid extraction
F. tularensis strain FTNF002-00 was propagated on chocolate agar at 37°C in 5% CO2. Glycerol fermentation characteristics were determined using Biolog® (Biolog, Inc., Hayward, CA). In addition to biochemical characteristics, FTNF002-00 was tested for subspecies-specific genome alterations by PCR using previously published primers specific for RDtularensis1 . High quality DNA was extracted using a large-scale bacterial genomic DNA preparation method  with omission of the CsCl step. These operations strictly complied with Biosafety in Microbiological and Biomedical Laboratories (BMBL), 4th Edition , which also encompasses 42-Code of Federal Regulations (CFR) Part 73 to govern the possession, use, and transfer of select agents. Site-specific standard operating procedures (SOPs) regulated such operations to include entry and exit, waste disposal, personal protective equipment (PPE) gowning and disrobing, occupational-health surveillance, and post-extraction DNA non-viablility certification.
The random shotgun method was used in sequencing the genome of F. tularensis FTNF002-00. Medium (8 kb) and small (3 kb) insert random sequencing libraries in pMCL200 and pUC-hinc cloning vectors (respectively) were subjected to paired-end Sanger sequencing to an average depth of 19× coverage. After the shotgun stage, reads were assembled with parallel phrap (High Performance Software, LLC). Possible mis-assemblies were corrected with Dupfinisher (unpublished, C. Han) or transposon bomb of bridging clones (EZ-Tn5 <P6Kyori/KAN-2> Tnp Transposome kit, Epicentre Biotechnologies). Gaps between contigs were closed by editing, custom primer walks or PCR amplification. The completed genome sequence of F. tularensis FTNF002-00 was finished to a standard of less than 1 error in 100,000 bases.
Gene prediction and Annotation
Genome annotation was performed at BioHealthBase (www.biohealthbase.org) using TIGR's gene prediction algorithm, Glimmer3 , automated annotation pipeline, AutoAnnotate , followed by manual curation using Manatee . The genome sequence was reformatted so that dnaA was the first gene and sequence coordinates were reassigned before submitting the sequence for gene predictions. Glimmer 3.02 was used to predict protein coding genes using default parameters. The genes coding for tRNAs were identified using tRNAscanSE  (version 1.23, parameters: -B). A sequence similarity search using Exonerate  (version 1.0.0, parameters: -m NER) was used to identify 16S and 23S ribosomal RNA genes. A sequence similarity search against RFam  (release 7.0, rfam_scan.pl version 0.1, parameters: –bt 0.1 –nobig), a comprehensive database of non-coding RNA (ncRNA) families was used to identify genes coding for other non-coding RNAs, (such as 5S ribosomal RNAs).
For assignment of protein function, the first step was a sequence similarity search BLAST  for each of the predicted protein against a non-redundant amino acid database (nraa) made up of all proteins available from GenBank, PIR and SWISS-PROT using the BLAST-Extend-Repraze (BER) algorithm. All of the proteins from the genome sequence were also searched against the Pfam  (release 20.0) HMMs (hmmpfam HMMER version 2.3.2, parameters: -E 0.1 –cut_ga) and TIGRFAMs  (release 6.0) built from highly curated multiple alignments of proteins thought to share the same function or to be members of the same family. Each HMM has an associated cutoff score above which hits are known to be significant. Additional data used to support annotations included prediction of trans-membrane helices using TMHMM , prediction of signal peptide with signalP , lipoprotein motif and Clusters of Orthologous Groups (COG) . These data used by AutoAnnotate to make functional predictions for proteins were then made available in the Manatee interface for manual evaluation of the predicted function. Manual curation involved correcting the start sites for the predicted ORFs, reviewing overlapping ORFs, frameshifts and premature stops, analysis of predicted pseudogenes, transposons and transport proteins.
The complete genome sequence of FTNF002-00 has been deposited in Genbank under the accession number CP000803 and the Refseq accession is NC_009749.
Data acquisition and Analysis
Genome sequence files of completely sequenced genomes of seven Fransicella tularensis strains listed in Table 1 were obtained from the NCBI ftp site (ftp://ftp.ncbi.nih.gov/genomes/Bacteria/) in September 2007. The executable BLAST (Altschul et al., 1997) programs were obtained from the National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/BLAST/download.shtml). The complete genome and predicted proteome of FTNF002-00 were analyzed against those of the other two F. tularensis subsp. holarctica genomes, LVS and OSU18. Whole genome and whole proteome analyses were performed locally using BLAST programs. Results of whole genome DNA BLAST search (filtering was set to off) were used to carefully align the whole genomes of FTNF002-00 with either LVS or OSU18 genomes. Genomic regions of differences and polymorphic sites were identified from the aligned genomes using short perl codes.
We are grateful to Douglas Landry for his expertise in production of the figures. The views expressed in this article are those of the authors and do not reflect the official policy or position of the U.S. Air Force, U.S. Army, Department of Defense, or U.S. Government.
Conceived and designed the experiments: RDB TSB SHH PDF CL MPD. Performed the experiments: RDB GX JLE SDG JMN RHS LWZ MPD. Analyzed the data: RDB JJJ SDG JMN RHS LWZ MPD. Contributed reagents/materials/analysis tools: RDB JJJ JLE SDG JMN RHS LWZ MPD. Wrote the paper: RDB MPD.
- 1. McCoy GW, Chapin CW (1912) Further observations on a plague-like disease of rodents with a preliminary note on the causative agent, Bacterium tularense. J Infect Dis 10: 61–72.GW McCoyCW Chapin1912Further observations on a plague-like disease of rodents with a preliminary note on the causative agent, Bacterium tularense.J Infect Dis106172
- 2. Hopla CE, Hopla AK (1994) Tularemia;. In: Beran GW, editor. CRC Press. pp. 113–126.CE HoplaAK Hopla1994Tularemia;GW BeranCRC Press113126
- 3. Johansson A, Farlow J, Larsson P, Dukerich M, Chambers E, et al. (2004) Worldwide genetic relationships among Francisella tularensis isolates determined by multiple-locus variable-number tandem repeat analysis. J Bacteriol 186: 5808–5818.A. JohanssonJ. FarlowP. LarssonM. DukerichE. Chambers2004Worldwide genetic relationships among Francisella tularensis isolates determined by multiple-locus variable-number tandem repeat analysis.J Bacteriol18658085818
- 4. Evans ME, Friedlander AM (1997) Chapter 24: Tularemia. In: Zajtchuk R, Belamy R, editors. Text Book of Military Medicine Part I: Medical Aspects of Chemical and Biological Warfare. Washington, D.C.: Office of the Surgeon General. 10 p.ME EvansAM Friedlander1997Chapter 24: Tularemia.R. ZajtchukR. BelamyText Book of Military Medicine Part I: Medical Aspects of Chemical and Biological WarfareWashington, D.C.Office of the Surgeon General10
- 5. Keim P, Johansson A, Wagner DM (2007) Molecular epidemiology, evolution, and ecology of Francisella. Ann N Y Acad Sci 1105: 30–66.P. KeimA. JohanssonDM Wagner2007Molecular epidemiology, evolution, and ecology of Francisella.Ann N Y Acad Sci11053066
- 6. Ellis J, Oyston PC, Green M, Titball RW (2002) Tularemia. Clin Microbiol Rev 15: 631–646.J. EllisPC OystonM. GreenRW Titball2002Tularemia.Clin Microbiol Rev15631646
- 7. Svensson K, Larsson P, Johansson D, Bystrom M, Forsman M, et al. (2005) Evolution of subspecies of Francisella tularensis. J Bacteriol 187: 3903–3908.K. SvenssonP. LarssonD. JohanssonM. BystromM. Forsman2005Evolution of subspecies of Francisella tularensis.J Bacteriol18739033908
- 8. Staples JE, Kubota KA, Chalcraft LG, Mead PS, Petersen JM (2006) Epidemiologic and molecular analysis of human tularemia, United States, 1964-2004. Emerg Infect Dis 12: 1113–1118.JE StaplesKA KubotaLG ChalcraftPS MeadJM Petersen2006Epidemiologic and molecular analysis of human tularemia, United States, 1964-2004.Emerg Infect Dis1211131118
- 9. Broekhuijsen M, Larsson P, Johansson A, Bystrom M, Eriksson U, et al. (2003) Genome-wide DNA microarray analysis of Francisella tularensis strains demonstrates extensive genetic conservation within the species but identifies regions that are unique to the highly virulent F. tularensis subsp. tularensis. J Clin Microbiol 41: 2924–2931.M. BroekhuijsenP. LarssonA. JohanssonM. BystromU. Eriksson2003Genome-wide DNA microarray analysis of Francisella tularensis strains demonstrates extensive genetic conservation within the species but identifies regions that are unique to the highly virulent F. tularensis subsp. tularensis.J Clin Microbiol4129242931
- 10. Samrakandi MM, Zhang C, Zhang M, Nietfeldt J, Kim J, et al. (2004) Genome diversity among regional populations of Francisella tularensis subspecies tularensis and Francisella tularensis subspecies holarctica isolated from the US. FEMS Microbiol Lett 237: 9–17.MM SamrakandiC. ZhangM. ZhangJ. NietfeldtJ. Kim2004Genome diversity among regional populations of Francisella tularensis subspecies tularensis and Francisella tularensis subspecies holarctica isolated from the US.FEMS Microbiol Lett237917
- 11. Nano FE, Zhang N, Cowley SC, Klose KE, Cheung KK, et al. (2004) A Francisella tularensis pathogenicity island required for intramacrophage growth. J Bacteriol 186: 6430–6436.FE NanoN. ZhangSC CowleyKE KloseKK Cheung2004A Francisella tularensis pathogenicity island required for intramacrophage growth.J Bacteriol18664306436
- 12. Dempsey MP, Nietfeldt J, Ravel J, Hinrichs S, Crawford R, et al. (2006) Paired-end sequence mapping detects extensive genomic rearrangement and translocation during divergence of Francisella tularensis subsp. tularensis and Francisella tularensis subsp. holarctica populations. J Bacteriol 188: 5904–5914.MP DempseyJ. NietfeldtJ. RavelS. HinrichsR. Crawford2006Paired-end sequence mapping detects extensive genomic rearrangement and translocation during divergence of Francisella tularensis subsp. tularensis and Francisella tularensis subsp. holarctica populations.J Bacteriol18859045914
- 13. Petrosino JF, Xiang Q, Karpathy SE, Jiang H, Yerrapragada S, et al. (2006) Chromosome rearrangement and diversification of Francisella tularensis revealed by the type B (OSU18) genome sequence. J Bacteriol 188: 6977–6985.JF PetrosinoQ. XiangSE KarpathyH. JiangS. Yerrapragada2006Chromosome rearrangement and diversification of Francisella tularensis revealed by the type B (OSU18) genome sequence.J Bacteriol18869776985
- 14. Petersen JM, Schriefer ME (2005) Tularemia: emergence/re-emergence. Vet Res 36: 455–467.JM PetersenME Schriefer2005Tularemia: emergence/re-emergence.Vet Res36455467
- 15. Farlow J, Wagner DM, Dukerich M, Stanley M, Chu M, et al. (2005) Francisella tularensis in the United States. Emerg Infect Dis 11: 1835–1841.J. FarlowDM WagnerM. DukerichM. StanleyM. Chu2005Francisella tularensis in the United States.Emerg Infect Dis1118351841
- 16. Dempsey MP, Dobson M, Zhang C, Zhang M, Lion C, et al. (2007) Genomic deletion marking an emerging subclone of Francisella tularensis subsp. holarctica in France and the Iberian Peninsula. Appl Environ Microbiol 73: 7465–7470.MP DempseyM. DobsonC. ZhangM. ZhangC. Lion2007Genomic deletion marking an emerging subclone of Francisella tularensis subsp. holarctica in France and the Iberian Peninsula.Appl Environ Microbiol7374657470
- 17. Vogler AJ, Birdsell D, Price LB, Bowers JR, Beckstrom-Sternberg SM, et al. (2009) Phylogeography of Francisella tularensis: global expansion of a highly fit clone. J Bacteriol 191: 2474–2484.AJ VoglerD. BirdsellLB PriceJR BowersSM Beckstrom-Sternberg2009Phylogeography of Francisella tularensis: global expansion of a highly fit clone.J Bacteriol19124742484
- 18. de la Puente-Redondo VA, del Blanco NG, Gutierrez-Martin CB, Garcia-Pena FJ, Rodriguez Ferri EF (2000) Comparison of different PCR approaches for typing of Francisella tularensis strains. J Clin Microbiol 38: 1016–1022.VA de la Puente-RedondoNG del BlancoCB Gutierrez-MartinFJ Garcia-PenaEF Rodriguez Ferri2000Comparison of different PCR approaches for typing of Francisella tularensis strains.J Clin Microbiol3810161022
- 19. Bachiller Luque P, Perez Castrillon JL, Martin Luquero M, Mena Martin FJ, de la Lama Lopez-Areal J, et al. (1998) [Preliminary report of an epidemic tularemia outbreak in Valladolid]. Rev Clin Esp 198: 789–793.P. Bachiller LuqueJL Perez CastrillonM. Martin LuqueroFJ Mena MartinJ. de la Lama Lopez-Areal1998[Preliminary report of an epidemic tularemia outbreak in Valladolid].Rev Clin Esp198789793
- 20. Bellido-Casado J, Perez-Castrillon JL, Bachiller-Luque P, Martin-Luquero M, Mena-Martin FJ, et al. (2000) Report on five cases of tularaemic pneumonia in a tularaemia outbreak in Spain. Eur J Clin Microbiol Infect Dis 19: 218–220.J. Bellido-CasadoJL Perez-CastrillonP. Bachiller-LuqueM. Martin-LuqueroFJ Mena-Martin2000Report on five cases of tularaemic pneumonia in a tularaemia outbreak in Spain.Eur J Clin Microbiol Infect Dis19218220
- 21. García Peňa FJ, Sáurez Mayoral P, Cogolluda Cogolluda C, Arriola Garrote C, Anadón Navarro E (1998) An outbreak of tularemia in Castilla-Leon. First isolation of Francisella tularensis in Spain. Medicina Veterinaria 15: 418–423.FJ García PeňaP. Sáurez MayoralC. Cogolluda CogolludaC. Arriola GarroteE. Anadón Navarro1998An outbreak of tularemia in Castilla-Leon. First isolation of Francisella tularensis in Spain.Medicina Veterinaria15418423
- 22. Perez-Castrillon JL, Bachiller-Luque P, Martin-Luquero M, Mena-Martin FJ, Herreros V (2001) Tularemia epidemic in northwestern Spain: clinical description and therapeutic response. Clin Infect Dis 33: 573–576.JL Perez-CastrillonP. Bachiller-LuqueM. Martin-LuqueroFJ Mena-MartinV. Herreros2001Tularemia epidemic in northwestern Spain: clinical description and therapeutic response.Clin Infect Dis33573576
- 23. Garcia Del Blanco N, Dobson ME, Vela AI, De La Puente VA, Gutierrez CB, et al. (2002) Genotyping of Francisella tularensis strains by pulsed-field gel electrophoresis, amplified fragment length polymorphism fingerprinting, and 16S rRNA gene sequencing. J Clin Microbiol 40: 2964–2972.N. Garcia Del BlancoME DobsonAI VelaVA De La PuenteCB Gutierrez2002Genotyping of Francisella tularensis strains by pulsed-field gel electrophoresis, amplified fragment length polymorphism fingerprinting, and 16S rRNA gene sequencing.J Clin Microbiol4029642972
- 24. Garcia del Blanco N, Gutierrez CB, de la Puente VA, Rodriguez Ferri EF (2004) Biochemical characterisation of Francisella tularensis strains isolated in Spain. Vet Rec 154: 55–56.N. Garcia del BlancoCB GutierrezVA de la PuenteEF Rodriguez Ferri2004Biochemical characterisation of Francisella tularensis strains isolated in Spain.Vet Rec1545556
- 25. Haristoy X, Lozniewski A, Tram C, Simeon D, Bevanger L, et al. (2003) Francisella tularensis bacteremia. J Clin Microbiol 41: 2774–2776.X. HaristoyA. LozniewskiC. TramD. SimeonL. Bevanger2003Francisella tularensis bacteremia.J Clin Microbiol4127742776
- 26. Puiu D, Salzberg SL (2008) Re-assembly of the genome of Francisella tularensis subsp. holarctica OSU18. PLoS ONE 3: e3427.D. PuiuSL Salzberg2008Re-assembly of the genome of Francisella tularensis subsp. holarctica OSU18.PLoS ONE3e3427
- 27. Brosch R, Pym AS, Gordon SV, Cole ST (2001) The evolution of mycobacterial pathogenicity: clues from comparative genomics. Trends Microbiol 9: 452–458.R. BroschAS PymSV GordonST Cole2001The evolution of mycobacterial pathogenicity: clues from comparative genomics.Trends Microbiol9452458
- 28. Chain PS, Carniel E, Larimer FW, Lamerdin J, Stoutland PO, et al. (2004) Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis. Proc Natl Acad Sci U S A 101: 13826–13831.PS ChainE. CarnielFW LarimerJ. LamerdinPO Stoutland2004Insights into the evolution of Yersinia pestis through whole-genome comparison with Yersinia pseudotuberculosis.Proc Natl Acad Sci U S A1011382613831
- 29. Lopatin DE, Combs A, Sweier DG, Fenno JC, Dhamija S (2000) Characterization of heat-inducible expression and cloning of HtpG (Hsp90 homologue) of Porphyromonas gingivalis. Infect Immun 68: 1980–1987.DE LopatinA. CombsDG SweierJC FennoS. Dhamija2000Characterization of heat-inducible expression and cloning of HtpG (Hsp90 homologue) of Porphyromonas gingivalis.Infect Immun6819801987
- 30. Weiss DS, Brotcke A, Henry T, Margolis JJ, Chan K, et al. (2007) In vivo negative selection screen identifies genes required for Francisella virulence. Proc Natl Acad Sci U S A 104: 6037–6042.DS WeissA. BrotckeT. HenryJJ MargolisK. Chan2007In vivo negative selection screen identifies genes required for Francisella virulence.Proc Natl Acad Sci U S A10460376042
- 31. Wilson K (1997) Preparation of Genomic DNA from Bacteria. In: Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, et al., editors. Current Protocols in Molecular Biology: John Wiley & Sons. pp. 2.4.1–2.4.5.K. Wilson1997Preparation of Genomic DNA from Bacteria.FM AusubelR. BrentRE KingstonDD MooreJG SeidmanCurrent Protocols in Molecular Biology: John Wiley & Sons22.214.171.124.5
- 32. USDHHS (1999) Biosafety in Microbiological and Biomedical Laboratories (BMBL) 4th Edition. 4th ed. U.S. Government Printing Office. USDHHS1999Biosafety in Microbiological and Biomedical Laboratories (BMBL) 4th Edition. 4th ed.U.S. Government Printing Office
- 33. Delcher AL, Harmon D, Kasif S, White O, Salzberg SL (1999) Improved microbial gene identification with GLIMMER. Nucleic Acids Res 27: 4636–4641.AL DelcherD. HarmonS. KasifO. WhiteSL Salzberg1999Improved microbial gene identification with GLIMMER.Nucleic Acids Res2746364641
- 34. TIGR (2006) Manatee Home Page. TIGR2006Manatee Home Page.<http://manatee.sourceforge.net/>. <http://manatee.sourceforge.net/>.
- 35. Lowe TM, Eddy SR (1997) tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res 25: 955–964.TM LoweSR Eddy1997tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence.Nucleic Acids Res25955964
- 36. Slater GS, Birney E (2005) Automated generation of heuristics for biological sequence comparison. BMC Bioinformatics 6: 31.GS SlaterE. Birney2005Automated generation of heuristics for biological sequence comparison.BMC Bioinformatics631
- 37. Griffiths-Jones S, Moxon S, Marshall M, Khanna A, Eddy SR, et al. (2005) Rfam: annotating non-coding RNAs in complete genomes. Nucleic Acids Res 33: D121–124.S. Griffiths-JonesS. MoxonM. MarshallA. KhannaSR Eddy2005Rfam: annotating non-coding RNAs in complete genomes.Nucleic Acids Res33D121124
- 38. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402.SF AltschulTL MaddenAA SchafferJ. ZhangZ. Zhang1997Gapped BLAST and PSI-BLAST: a new generation of protein database search programs.Nucleic Acids Res2533893402
- 39. Bateman A, Birney E, Durbin R, Eddy SR, Howe KL, et al. (2000) The Pfam protein families database. Nucleic Acids Res 28: 263–266.A. BatemanE. BirneyR. DurbinSR EddyKL Howe2000The Pfam protein families database.Nucleic Acids Res28263266
- 40. Haft DH, Loftus BJ, Richardson DL, Yang F, Eisen JA, et al. (2001) TIGRFAMs: a protein family resource for the functional identification of proteins. Nucleic Acids Res 29: 41–43.DH HaftBJ LoftusDL RichardsonF. YangJA Eisen2001TIGRFAMs: a protein family resource for the functional identification of proteins.Nucleic Acids Res294143
- 41. Krogh A, Larsson B, von Heijne G, Sonnhammer EL (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305: 567–580.A. KroghB. LarssonG. von HeijneEL Sonnhammer2001Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes.J Mol Biol305567580
- 42. Bendtsen JD, Nielsen H, von Heijne G, Brunak S (2004) Improved prediction of signal peptides: SignalP 3.0. J Mol Biol 340: 783–795.JD BendtsenH. NielsenG. von HeijneS. Brunak2004Improved prediction of signal peptides: SignalP 3.0.J Mol Biol340783795
- 43. Tatusov RL, Fedorova ND, Jackson JD, Jacobs AR, Kiryutin B, et al. (2003) The COG database: an updated version includes eukaryotes. BMC Bioinformatics 4: 41.RL TatusovND FedorovaJD JacksonAR JacobsB. Kiryutin2003The COG database: an updated version includes eukaryotes.BMC Bioinformatics441