Research Article: Large Direct Repeats Flank Genomic Rearrangements between a New Clinical Isolate of Francisella tularensis subsp. tularensis A1 and Schu S4

Date Published: February 3, 2010

Publisher: Public Library of Science

Author(s): Ufuk Nalbantoglu, Khalid Sayood, Michael P. Dempsey, Peter C. Iwen, Stephen C. Francesconi, Ravi D. Barabote, Gary Xie, Thomas S. Brettin, Steven H. Hinrichs, Paul D. Fey, Niyaz Ahmed.

Abstract: Francisella tularensis subspecies tularensis consists of two separate populations A1 and A2. This report describes the complete genome sequence of NE061598, an F. tularensis subspecies tularensis A1 isolated in 1998 from a human with clinical disease in Nebraska, United States of America. The genome sequence was compared to Schu S4, an F. tularensis subspecies tularensis A1a strain originally isolated in Ohio in 1941. It was determined that there were 25 nucleotide polymorphisms (22 SNPs and 3 indels) between Schu S4 and NE061598; two of these polymorphisms were in potential virulence loci. Pulsed-field gel electrophoresis analysis demonstrated that NE061598 was an A1a genotype. Other differences included repeat sequences (n = 11 separate loci), four of which were contained in coding sequences, and an inversion and rearrangement probably mediated by insertion sequences and the previously identified direct repeats I, II, and III. Five new variable-number tandem repeats were identified; three of these five were unique in NE061598 compared to Schu S4. Importantly, there was no gene loss or gain identified between NE061598 and Schu S4. Interpretation of these data suggests there is significant sequence conservation and chromosomal synteny within the A1 population. Further studies are needed to determine the biological properties driving the selective pressure that maintains the chromosomal structure of this monomorphic pathogen.

Partial Text: Francisella tularensis is a highly pathogenic gram-negative cocco-bacillus that is the causative agent of tularemia, commonly referred to as “rabbit fever.” The large majority of disease is ulceroglandular in nature and can be traced to contact with an infected host (e.g. rabbit or cat) or vector (e.g. tick or mosquito); however more serious forms of disease such as pneumonic tularemia can be life-threatening, and therefore F. tularensis is considered a potential biowarfare agent. There are three recognized subspecies of F. tularensis including tularensis (commonly referred to as type A), holarctica (commonly referred to as type B), and mediasiatica as well as a closely related species F. novicida. These subspecies are associated with important geographic differences in their distribution with F. tularensis holarctica found throughout the northern temperate regions of both hemispheres whereas subspecies tularensis is found primarily in North America. In addition, the population of F. tularensis subspecies tularensis consists of two major, geographically isolated clades, A1 and A2 [1], [2]. The A2 population has been isolated in the western United States whereas the A1 population is found east of the Rocky Mountains, primarily in the Ozark mountain regions of Missouri, Oklahoma and Arkansas. The genomes of two F. tularensis subspecies tularensis A1 isolates (Schu S4 and FSC198) have recently been sequenced; FSC198 was isolated from Slovakia in 1986 whereas Schu S4, an often-utilized virulent laboratory strain, is a clinical isolate obtained from Ohio in 1941 [3], [4]. In addition, a draft sequence of a separate F. tularensis subsp. tularensis A.I isolate, FSC033, was also recently published [5]. FSC033 was isolated from a squirrel in Georgia, USA. Genomic comparisons between FSC198 and Schu S4 revealed remarkable sequence conservation; only 8 SNP and three variable number tandem repeat (VNTR) differences were noted [3]. Chaudhri et al. [3] have suggested that the close similarity between FSC198 and Schu S4 indicated that the FSC198 strain may have derived from Schu S4. Preliminary analysis between a recent human clinical isolate of F. tularensis subsp. tularensis obtained in 1998 in Nebraska and Schu S4 revealed distinguishing characteristics [6]. This presented an opportunity to further examine the genomic diversity within the A1 population, and therefore, the complete sequence of a F. tularensis subspecies tularensis A1 isolate NE061598 was determined. The genomes of the four A1 isolates that have been fully or partially sequenced (SchuS4, FSC198, NE061598 and FSC033) were compared in light of their temporal and spatial separation. This analysis demonstrated that the F. tularensis subsp. tularensis A1 population, as represented by these isolates, is highly clonal and displays a high degree of DNA sequence conservation and chromosomal synteny. The primary chromosomal differences between NE061598 and Schu S4/FSC198/FSC033 were due to rearrangements occurring between large direct repeats and insertion sequences.

Due to the remarkable sequence conservation between Schu S4 and FSC198 [3], speculation was made that these two isolates may have the same origin. Therefore, we proposed to sequence a separate virulent isolate of F. tularensis subsp. tularensis A1 and compare it with Schu S4 to evaluate the issue of sequence divergence over time. NE061598 was isolated in Nebraska in 1998 from the blood of a patient with ulceroglandular tularemia, Schu S4 was derived in 1941 and FSC198 was isolated in 1986. The availability of a recent clinically virulent isolate of F. tularensis subsp. tularensis A.I isolate obtained in the mid-western portion of the United States provided the opportunity for an in-depth sequence comparison with other A.I. isolates. Because of the significant temporal separation (45 years) between Schu S4 and NE061598, the sequence conservation between these two isolates was unexpected. Even though VNTR analysis yielded 11 distinct polymorphisms (see Table 2), analysis of the entire genome only yielded 25 additional SNPs/indels. The most significant difference detected was an inversion associated with LCB 2 and rearrangements associated with LCBs 4 and 5 (see Figures 1 and 2); both events were predictably mediated through IS element recombination (LCB 2) or rearrangement mediated by large duplicated regions (LCBs 4 and 5). Significantly, there was no net gain (or loss) of genes within the NE061598 genome in relationship to Schu S4. These data may suggest that the minimal differences observed in pulsed-field RFLP patterns of the F. tularensis subsp. tularensis A1 population may be due to IS- or direct repeat-mediated rearrangements and is not due to the acquisition of new genes [1], [2], [6]. Furthermore, these data support the notion that this highly monomorphic pathogen [14] may have undergone a recent population bottleneck which may be related to its specific host preference (e.g. lagomorphs, humans) and vectors (e.g. ticks). The further elucidation of the natural reservoir, hosts, and vectors of F. tularensis may lead to novel hypotheses of the selective pressure of this A1 population.