| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Universidade de Lisboa, Faculdade de Ciências, Centro de Genética e Biologia Molecular and Instituto de Ciência Aplicada e Tecnologia, Edifício ICAT, Campus FCUL, Campo Grande. 1749-016 Lisboa, Portugal 2CEVDI, Instituto Nacional de Saúde Dr Ricardo Jorge, 2965 Águas de Moura, Portugal
Correspondence Líbia Zé-Zé lmze-ze{at}fc.ul.pt
Received November 10, 2004
Accepted June 2, 2005
Mediterranean spotted fever (MSF) is a tick-borne rickettsiosis caused by ‘Rickettsia conorii complex’ strains. In Portugal, R. conorii and Israeli tick typhus (ITT) are the aetiological agents of this disease. A novel 65 bp tandem repeat was identified by the analysis of the R. conorii Malish 7 whole genome sequence with an appropriate algorithm for searching for repeated sequences. The variable number tandem repeat (VNTR) was named VNTR Rc-65 and this locus was amplified by PCR and sequenced in order to characterize the repeat diversity within different rickettsial strains including Portuguese strains isolated from clinical and vector samples. The VNTR Rc-65 has seven alleles within the rickettsial strains studied and a diversity index value of 0.71, meaning that this locus has a great discriminatory capacity and therefore can be used for identification of closely related strains. PCR amplification of the Rc-65 locus can be used to differentiate between the Portuguese R. conorii Malish-like and Israeli tick typhus strains, enabling a more accurate and rapid identification of these rickettsial isolates.
The GenBank/EMBL/DDBJ accession numbers for the VNTR Rc-65 locus sequences of the 24 strains tested are AY820021–AY820045.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Mediterranean spotted fever (MSF), also known as Boutonneuse fever, is an acute, febrile tick-transmitted rickettsiosis caused by strains of ‘Rickettsia conorii complex'. MSF is caused by two strains: R. conorii Malish, which is endemic in the Mediterranean area, and Israeli tick typhus (ITT). The latter strain was believed to be restricted to Israel, but was isolated from a patient in Portugal for the first time in 1997 (Bacellar et al., 1999). In Portugal, MSF is endemic and is an obligatory reported disease, with an incidence rate of 9.8 per 105 inhabitants (de Sousa et al., 2003). The number of reported cases and the fatality rate of the disease has been increasing (Bacellar et al., 2003). Moreover, severe forms of the disease have been described in several patients.
To date, the diagnosis of a rickettsial illness has been often carried out by microimmunofluorescence assay and identification is accomplished by ompA PCR-RFLP gene analysis (Regnery et al., 1991), and sequencing of the citrate synthase (gltA) (Roux et al., 1997) and ompA genes (Fournier et al., 1998). Although ompA gene analysis is suited for accurate identification of rickettsiae from spotted fever group, it cannot be used to identify several other rickettsial species. Likewise, the gltA gene sequence has little nucleotide variation to distinguish among all rickettsial species (Roux et al., 1997). Therefore, an accurate, discriminatory, reproducible and faster typing test should lead to an improvement in the epidemiological surveillance and diagnosis of rickettsiosis.
Variable number tandem repeats (VNTRs) consist of multimeric repeats that can often vary in copy number, showing inter-individual length polymorphisms that can be detected by a PCR assay since the regions flanking the repeats are generally well-conserved targets for primer design (van Belkum et al., 1998). Polymorphisms at a tandem repeat locus can also occur as a result of nucleotide sequence changes between individual repeat units. VNTR sequences display high variability and diversity, and hence they have great discriminatory capacity. The variability observed in VNTR loci in repeat numbers and sequence degeneracy is thought to be primarily caused by slipped-strand mispairing (SSM) increased by inadequate DNA mismatch repair or, alternatively, explained by DNA recombination between multiple loci of similar repeat motifs (van Belkum et al., 1998).
The availability of whole genome sequences and appropriate algorithms for searching for repetitive sequences has led to the application of the VNTR typing approach, namely the development of multilocus VNTR analysis (MLVA) typing systems for several pathogens, such as Bacillus anthracis (Keim et al., 2000), Yersinia pestis (Klevytska et al., 2001), Francisella tularensis (Farlow et al., 2001), Borrelia burgdorferi (Farlow et al., 2002), Legionella pneumophila (Pourcel et al., 2003) and Mycobacterium tuberculosis (Spurgiesz et al., 2003). The MLVA approach has been proven to present high discriminatory power, reproducibility and portability, making it a strong candidate for the increased development of reference databases that allow online strain identification services (Supply et al., 2001; Onteniente et al., 2003).
In this study, we identified a 65 bp tandem repeat by screening for tandem repeat regions in the R. conorii Malish 7 genome sequence data. PCR amplification and sequencing of this locus was performed to characterize the repeat diversity within different rickettsial strains including Portuguese clinical and vector isolates.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
Culture conditions and DNA extraction.
Rickettsial isolation from human total blood and from tick haemolymph was achieved by shell-vial technique (Marrero & Raoult, 1989), followed by 7 days of incubation in VERO cells (E-6 clone) in Eagle's minimal essential medium (MEM; GibcoBRL) supplemented with 1 % glutamine and 10 % fetal bovine serum at 32 °C for 5–6 days. Infected cell cultures were monitored for the presence of rickettsiae. Harvesting was done when the degree of infection was optimal, which was determined by Gimenez staining (Gimenez, 1964). Infected cells were mechanically suspended in the medium. The suspension was frozen and unfrozen three times, then centrifuged at 1000 r.p.m. for 30 min. Supernatants were recovered and rickettsiae were harvested by centrifugation at 8000 r.p.m. for 30 min and resuspended in ultrapure water. DNA was extracted using Qiagen columns (DNAeasy Protocol for Animal Tissue Kit) according to the manufacturer's instructions.
Tandem repeat identification.
The complete genome sequence of R. conorii Malish 7 (GenBank accession no. AE006914) was screened for tandem repeat loci using the software Tandem Repeat Finder (TRF, version 3.21), developed by Benson (1999). The TRF software was downloaded from the web page of the Department of Biomathematical Sciences at Mount Sinai School of Medicine (http://tandem.bu.edu/trf/trf.html). A tandem repeat locus with a repeat unit of 65 bp in length, 98 % similiarity between adjacent copies overall and five copies of the repeat unit was selected for further analysis. The locus was named Rc-65 and is located in position 1 168 767 to 1 169 087 of the R. conorii Malish 7 genome. BLASTN (Altschul et al., 1997) searches within R. prowazekii Madrid E genome (GenBank accession no. AJ235269), Rickettsia sibirica 246 (GenBank accession no. AABW00000000) and Rickettsia rickettsii (GenBank accession no. AADJ00000000) were performed to determine the presence of related repeat sequences in these genomes.
PCR amplification and sequencing.
PCR amplification of the VNTR Rc-65 locus was accomplished using specific primers Rc-65-Fw (5'-TTGAGAAGGTTTATATCCCATAG-3') and Rc-65-Rv (5'-TAC TACCGCATATCCAATTAAAAA-3') located 31 nucleotides upstream and 78 nucleotides downstream of the Rc-65 tandem repeat in the R. conorii Malish 7 genome. PCR was performed in a 50 µl reaction mixture containing 1 pmol of each primer, 200 µM (each) dATP, dGTP, dCTP and dTTP (Invitrogen), 1.75 U Taq polymerase (Invitrogen), 2 mM MgCl2, 0.5x BSA, 1x Taq buffer and 50–100 ng genomic DNA. Amplification was carried out in a DNA thermocycler (TGradient; Biometra) under the following conditions: 4 min of initial denaturation at 96 °C, then 35 cycles of 94 °C for 1 min, 45 °C for 1 min and 72 °C for 1 min. The amplification was completed by holding for 5 min at 72 °C to allow complete extension of the PCR products. PCR products were visualized by ethidium bromide staining after electrophoresis in a 1 % agarose gel and their sizes were estimated by comparison with a molecular mass standard (1 kb plus DNA ladder; GibcoBRL). The PCR products were purified using Jet Quick-PCR Purification Kit (Genomed) as described by the manufacturer. The purified PCR products were sequenced in an automated DNA capillary sequencer CEQ 2000-XL (Beckman Coulter) by a dye-labelled dideoxy-termination method (DTCS, Dye Terminator Cycle Sequencer start kit, Beckman Coulter). All sequences were determined by the consensus of the forward and the reverse sequence analysis. Low quality sequences were repeated using different conditions for the PCR sequencing reaction, namely denaturation temperature and DNA concentration. GenBank accession numbers for the nucleotide sequences are listed in Table 2.
|
|
(allele frequency)2 (Weir, 1990). The complete VNTR Rc-65 region was used to assess the genetic relationships among the isolates using the UPGMA clustering method of the software package PAUP4a (D. Swofford, Sinauer Associates). UPGMA analysis was performed with the simple matching coefficient to estimate genetic distances. |
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This locus proved to be polymorphic by PCR amplification of several rickettsial strains (Fig. 1), and hence can be referred to as a VNTR locus. However, five species failed to yield VNTR Rc-65 PCR amplification, Rickettsia typhi, Rickettsia bellii, Rickettsia canadensis, Rickettsia australis and Rickettsia felis. Among all the rickettsiae tested, six different amplicon lengths were identified, corresponding to six distinct alleles. Taking into account the analysis of R. conorii Malish 7, a total of seven alleles are considered in the present study. Based on the number of alleles and the frequency of each allele, a diversity index value of 0.71 was determined, which provides a measure of the VNTR Rc-65 discriminatory power.
The Rc-65 loci of 25 rickettsiae, which consist of distinct alleles, as determined by agarose gel electrophoresis analysis (Fig. 1), were sequenced to assess the nucleotide consensus sequences and the number of repeats within each allele (Table 2). To consider the sequence variability among individual units and the number of arrays, the entire VNTR region was compared by sequence alignment, and cluster analysis was performed using the UPGMA method (Fig. 2).
|
In the majority of the rickettsiae species tested, the repeat-unit length was 65 bp, as predicted by TRF software (Benson, 1999). Rickettsia rhipicephali, Rickettsia massiliae and Rickettsia sp. Bar29 revealed a tandem repeat unit of 64 bp (Fig. 2), which corresponds to the minimal amplicon length detected, 180 bp. Two distinct alleles were detected within ITT Portuguese isolates; three isolates revealed an additional repeat unit (Fig. 1, Table 2).
Eight tandem repeats within the Rc-65 locus were identified in all of the R. conorii Portuguese isolates, giving them a different array size to the R. conorii Malish 7 genome (GenBank accession no. AE006914), which consisted of five tandem repeat units of 65 bp (Table 2). Moreover, the consensus pattern of R. conorii Malish-like Portuguese isolates determined by the TRF software (Benson, 1999), which applies the majority rule, consisted of 63 bp repeat units (Fig. 2). However, within the eight tandem repeats there were four that had 65 bp, which reflects the high level of nucleotide sequence polymorphism between individual repeats (Table 2, Fig. 2) detected in these strains.
The nucleotide sequence structure of the Rc-65 locus is illustrated by dot plot analysis (Fig. 3). The centre diagonal line in each panel represents the sequence identity with itself, while parallel diagonals indicate directly repeating sequences. The difference in the number of repeats between the R. conorii Malish-like and ITT strains is easily observed in these plots. The nucleotide structure of R. conorii PoTiR12, which has eight tandem repeats, is represented by one central diagonal and seven parallel lines (Fig. 3c), while the ITT strains show one central diagonal and two or three parallel lines (Fig. 3a, b). It is also possible to observe the nucleotide sequence degeneracy between the repeats, by the lack of continuity in the parallel lines.
|
We also tested whether the VNTR Rc-65 locus could be used to identify Rickettsia species directly from vector and clinical samples. An expected amplicon size was achieved by PCR amplification from tick lysates previously screened for rickettsiae by ompA gene amplification (Regnery et al., 1991). The Rc-65 locus was also specifically amplified in ticks coinfected with Borrelia species (data not shown). Considering the skin biopsies or blood samples, which were previously known to be positive for rickettsiae, the VNTR Rc-65 locus was not successfully amplified by PCR. Even so in some cases a faint band was observed (data not shown). This could probably be due to the low number of bacteria present in these kinds of samples. Indeed, the amplification of other genes, such as the ompA or gltA genes, in these samples was also difficult, and was only achieved by nested PCR. In the future, a nested-PCR approach using an additional primer pair will be attempted in these clinical samples.
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This VNTR has a diversity index value of 0.71, which means that this locus is highly informative, and hence possesses great discriminatory power for identification of genetically similar strains. Indeed, the main finding of this study is the accurate discrimination of ‘R. conorii complex’ strains, R. conorii Malish and Israeli tick typhus, solely by PCR amplification of this novel polymorphic VNTR locus. This is particularly useful in the MSF diagnosis of Portuguese patients, since recent data from the National Institute of Health Dr Ricardo Jorge have pointed out that half of MSF cases occurring in Portugal are caused by ITT strains, revealing a prevalence of infection similar to R. conorii Malish-like strains (Bacellar et al., 2003), which were previously thought to be the most common agent of MSF disease in Portugal.
Indeed, the VNTR genotypes of the Portuguese isolates are in agreement with the serotyping and sequencing data previously obtained, which validates the usefulness of this approach as a diagnostic tool to distinguish between R. conorii Malish-like and ITT isolates. Regarding ITT isolates two distinct alleles were detected (Table 2), which points out the capacity of this methodology to reveal some heterogeneity within different ITT isolates.
The usefulness of the Rc-65 VNTR region for typing R. conorii strains was also observed in a recent study developed by Fournier et al. (2004), which was published after we finished our analysis. That study, based on the analysis of R. conorii and R. prowazekii genomes, compares the capacity of three different types of sequences (variable coding genes, genes degraded in R. conorii but intact in R. prowazekii, and conserved and variable intergenic spacers) for R. conorii strain typing. Although some differences can be pointed out in the consensus analysis determined (mainly due to the types of approaches employed) the dksA–xerC intergenic spacer is also considered one of the most polymorphic regions to be used in genotyping rickettsiae.
The VNTR typing results showed that this assay can be used to distinguish between different rickettsial species, mainly those belonging to the R. rickettsii group, which is comprised of many different species including the ‘R. conorii complex’ strains (Sekeyova et al., 2001). As predicted by a BLASTN search within the R. prowazekii genome (Andersson et al., 1998), the Rc-65 locus was not amplified in this species nor in R. typhi. Interestingly, the other species that did not yield amplification results were those previously described as being more distant phylogenetically (Sekeyova et al., 2001; Roux et al., 1997; Fournier et al., 1998; Vitorino et al., 2003). Moreover, it has been suggested that R. canadensis and R. bellii, as well as an unidentified bacterium from Adalia bipunctata, were the first representatives of the genus Rickettsia to diverge, and hence they have major genetic differences (Stothard et al., 1994). In view of this fact, the unsuccessful PCR amplification of the Rc-65 locus in these rickettsiae species could mean that they may have different nucleotide sequences in the primer regions. Therefore, a new set of primers designed from an alignment consensus sequence of the dksA and xerC genes might enable the VNTR Rc-65 analysis for these species.
In the dendrogram obtained from the UPGMA clustering analysis of the VNTR region some rickettsiae groups were observed, mainly the R. massiliae group (Fig. 2, indicated by ‘B') and the R. rickettsii group (Fig. 2, indicated by ‘A'). R. rickettsii revealed mismatches between both adjacent copy repeats, and so the presented nucleotide indetermination (according to IUB code) in the VNTR consensus sequence means that different nucleotides are present in the two repeats in the identified position. Thus, this VNTR consensus sequence represents distinctive nucleotide differences when compared with other rickettsiae from the R. rickettsii group (Fig. 2). Rickettsia akari shows the highest variability within the consensus sequence.
Concerning the R. conorii strains, there is 3 % sequence variation between tandem repeat arrays. R. conorii Malish 7 displays sequence variation of 2 % (Table 2). This variation is due to mismatches and indels (insertions/deletions) (1 or 2 nucleotides) between adjacent copies of the sequence. Heterogeneity among repeats is thought to be indicative of more frequent recombination (van Belkum et al., 1998), and thus R. conorii Malish-like isolates may be more prone to suffer recombination events than the other rickettsiae species included in this study. Within the Portuguese R. conorii Malish-like isolates there is the same indel, which accounts for a 63 bp consensus array.
In fact, Fournier et al. (2004) described a repeated sequence array ranging in size from 63 bp to 102 bp within the dksA–xerC intergenic spacer of R. conorii strains. In that study there are also some strains possessing eight tandem repeat units, including a Portuguese isolate, which is in agreement with our results. Nevertheless, there are two other Portuguese isolates included in the study that revealed different VNTR lengths. Hence, these strains must be tested by this approach, and a wide range of local isolates should be screened to find out if there is a different VNTR Rc-65 locus in the R. conorii Malish-like Portuguese strains.
Regarding the ITT strains, no sequence variation among individual units is detected, though there is some variation in the number of repeats within three different isolates, which could reflect some heterogeneity among the ITT Portuguese isolates. However, no correlation could be made between this VNTR profile and the geographic source, the isolation year or even the patient's clinical manifestations. Nevertheless, further analysis of other potential VNTR loci as well as a multilocus sequencing study based in several genes should be developed to elucidate the heterogeneity of these isolates.
As pointed out by Fournier et al. (2004), the Rc-65 locus has a low G+C content, in contrast to other intergenic regions scattered across the R. conorii Malish 7 genome (Ogata et al., 2001b), and since the rickettsial genome is AT-rich and is thought to be in a reductive evolution process (Andersson et al., 1998; Ogata et al., 2001b), this locus may represent remnants of a rickettsial gene on its way to decay. From this perspective it would be reasonable to consider that VNTR polymorphisms are most likely to occur by deletion of repeat units, either by DNA recombination between repeats or by SSM. Indeed, it seems that a higher percentage of adenosine and thymine may increase the possibility of SSM and therefore enhance deletion events (van Belkum et al., 1998).
Nevertheless, this tandem repeat region could be somehow important to the rickettsiae, given the fact that sequence repeats have been described as important elements in bacterial adaptation and pathogenesis, particularly when located within surface-protein coding genes or within potential virulence genes, respectively (van Belkum et al., 1998). Even in the case of extragenic VNTRs, they may contribute to a high mutation rate of flanking genes, allowing the bacterium to react quickly to deleterious environmental conditions (Moxon et al., 1994). Since the 5' end of the VNTR Rc-65 contingency gene locus encodes the dnaK suppressor protein homologue, whose function is related to stress conditions, it would be interesting to study whether array repeat number variation plays any role in the regulation of gene expression, as for many pathogenic bacteria that have VNTR-mediated phase variation of virulent factor expression (Peak et al., 1996).
This VNTR analysis has shown to be a valuable approach for rickettsiae typing, in agreement with the results of Fournier et al. (2004) obtained only for R. conorii strains, with excellent potential to be further optimized and used in a MLVA approach for diagnostics and application to large-scale epidemiological studies. Future work will involve the improvement of rickettsiae detection in clinical samples (biopsy and blood) and the analysis of several other tandem repeat loci, as well as automated gel analysis, through the use of fluorescent-labelled primers, to accomplish a high-throughput system that reduces time and determination errors in results. The combination of several VNTR loci should improve the discriminatory capacity of the VNTR typing system, providing greater resolution and probably resulting in the development of a novel and accurate diagnostic PCR-based approach for rickettsiae detection and identification.
This new assay constitutes an improvement in the accuracy and swiftness of rickettsiae strain identification, since the result can be analysed directly by gel agarose electrophoresis because there is no need for further time expenses in restriction analysis or sequencing, unlike ompA or gltA gene typing assays.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
