| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Infectious, Parasitic and Immune-Mediated Diseases, Istituto Superiore di Sanità, Rome, Italy;
2 Molecular Biology Section, Army Medical and Veterinary Research Center, Rome, Italy;
3 Istituto Zooprofilattico Sperimentale della Puglia e della Basilicata-Anthrax Reference Institute of Italy, Foggia, Italy;
4 Istituto Zooprofilattico Sperimentale delle Regioni Lazio e Toscana, Rome, Italy;
5 Cattedra di Allergologia e Immunologia Clinica, II Facoltà di Medicina, Università di Roma "La Sapienza", Rome, Italy;
6 Direzione Generale della Sanità Militare, Rome, Italy;
aaddress correspondence to this author at: Department Infectious, Parasitic and Immuno-Mediated Diseases, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy; fax 39-06-49387112, e-mail alecara{at}iss.it
Abstract
Background: Molecular genotyping of Bacillus anthracis, the etiologic agent of anthrax, is important for differentiating and identifying strains from different geographic areas and for tracing strains deliberately released in a bioterrorism attack. We previously described a multiple-locus variable-number tandem repeat (VNTR) analysis (MLVA) based on 25 marker loci. Although the method has great differentiating power and reproducibility, faster genotyping at low cost may be requested to accurately identify B. anthracis strains in the field.
Methods: We used the High Resolution Melter-1 (Idaho Technology) and a saturating dye of double-stranded DNA (LCGreen I) to identify alleles via PCR and melting-curve analysis of the amplicons. We applied high-resolution melting analysis (HRMA) to a collection of 19 B. anthracis strains.
Results: HRMA produced reproducible results for 6 of the 25 B. anthracis loci tested. These easily interpretable and distinguishable melting curve results were consistent with MLVA results obtained for the same alleles. The feasibility of this method was demonstrated in testing of different allelic variants for the 6 selected loci.
Conclusions: The described HRMA application for screening B. anthracis VNTR loci is fast and widely accessible and may prove particularly useful under field conditions.
Bacillus anthracis, the etiologic agent of anthrax, is a spore-forming gram-positive bacterial species that, because of its highly pathogenic nature, environmental resistance, and relatively easy dissemination, is included among the major pathogens considered likely bioterrorism agents. Although genetically homogeneous (1), this bacterium has been genotyped (2), and molecular genotyping of B. anthracis played an important role in differentiating and identifying the strains used in the 2001 bioterrorism attack in the US (1)(2). Recent efforts to evaluate the diversity of B. anthracis isolates from different geographic areas have improved our knowledge of the phylogeny and geographic distribution of different genetic variants (gene clusters) of this species (3)(4)(5)(6)(7). Studies have also tried to trace strains deliberately released into the environment back to their origins (8). The method universally adopted for genotyping B. anthracis strains is multiple-locus variable-number tandem repeat (VNTR) analysis (MLVA). This technique involves amplifying multiple chromosomal loci carrying VNTRs and sizing the fragments to classify allelic variants on the basis of length polymorphisms (8)(9). The level of intraspecific polymorphism for tandemly repeated sequences varies among loci, and this variation needs to be evaluated experimentally with collections of representative strains.
Previous reports proposed that the MLVA method be used with 6 chromosomal and 2 plasmid marker loci (MLVA8) and that amplicon size be analyzed by agarose gel electrophoresis (9)(10)(11). This method has recently been updated for the analysis of 25 marker loci (MLVA25) by measuring PCR fragment size with an automated capillary DNA sequencer (12). The latter method allows a more sensitive differentiation of closely related strains, but accurate analyses require expensive platforms and well-organized laboratory facilities. In this study, we have examined the feasibility of a high-resolution melting analysis (HRMA) for rapidly and easily recognizing and differentiating VNTR allelic variants. This technique uses the High Resolution Melter-1 (HR-1; Idaho Technology), which has recently been adopted for genotyping and screening mutations in several different research and clinical applications (13)(14)(15)(16)(17). This method consists of a rapid (2 min after the PCR), closed-tube assay that detects sequence variation within specific genetic loci via melting curve analysis of the amplicons with a saturating dye of double-stranded DNA (LCGreen I). We applied HRMA initially to a B. anthracis collection at the Istituto Superiore di Sanità, Rome, Italy. This collection consists of 12 strains—10 previously analyzed with MLVA25 and 2 new isolates of a B. anthracis strain involved in a recent case of human anthrax in Italy (18). One of these latter isolates was from necropsy materials from a sheep killed by anthrax, and one was isolated from the shepherd’s cutaneous lesion. Eight of the 10 previously characterized strains were from animals (ST1761 and ST2844 strains belong to B. anthracis clusters B2 and D, respectively), and 2 were vaccine strains (Pasteur and Sterne strains) (Table 1
). We extracted total genomic DNA from the 12 strains and amplified the 25 VNTR loci as previously described (12) with the LCGreen I dye to fluorescently label double-stranded DNA in a LightCycler 2.0 instrument (Roche Diagnostics).
|
Our study showed that DNA quality and final concentration were critical variables for obtaining reproducible melting curves for identical loci. We tested the reproducibility of this assay with triplicate samples of 0.5–500 ng of column-purified chromosomal DNA and PCR products as potential templates for PCR and HRMA. We also tested DNA obtained by boiling bacterial suspensions for 20 min (19); we compared the results with these DNA templates with those obtained for column-purified chromosomal DNA. We obtained optimal reproducibility with 20–40 ng of column-purified chromosomal DNA per PCR reaction tube and further improved reproducibility by reducing the amount of LCGreen I dye used per reaction tube to half that recommended by the manufacturer (13)(14)(15)(16)(17). Reducing the dye amount probably limits the availability of the fluorescent dye, thereby producing a constant number of labeled molecules; use of half the recommended LCGreen I produces lower fluorescent signals in the PCR (see Fig. 1A
in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol53/issue7). This modification produces better melting curve reproducibility in the HRMA for different amplicons and different amounts of DNA templates in each PCR reaction tube [see (16) and Figs. 1 and 2 in the online Data Supplement].
|
Under these conditions, HRMA gave reproducible intraexperimental melting curves for the same alleles, which permitted the differentiation of alleles with different numbers of repeat units (ru) (Fig. 1
). Bovo et al. (20) have described multiple-band artifacts that are produced by improper reannealing of repetitive microsatellite sequences when too many PCR cycles are used. We did not observe this effect for the B. anthracis loci under our experimental conditions (see above and Fig. 3 in the online Data Supplement). These observations suggest that HRMA can be used as an initial and low-cost method for screening B. anthracis VNTR allelic variants. We also tested the interexperimental reproducibility of HRMA by comparing the melting curves obtained for the same allelic variant in different experimental runs. Despite our application of the standardized conditions we have described, interexperimental reproducibility needs to be improved. Melting curve shape was reproducible, but the curves did not always perfectly overlap. Nonetheless, our use of triplicate measurements of each reference allelic variant can improve interexperimental comparisons of allele melting curves.
We applied this experimental approach to all 25 B. anthracis VNTR loci and eventually selected 6 loci that yielded easily differentiated melting curves and a high capacity for distinguishing different strains.
We then included 7 additional strains (12) with different allelic variants for these 6 loci in the collection of tested strains (Table 1
) to verify the feasibility of HRMA to detect all the different variants previously described for these loci (Fig. 1
).
For the large amplicon of the BamS31 locus, which MLVA25 scored as having a high differentiating power, we added a 5'-CGC CGC CCG CCG CCC GCC-3' clamp at the 5' ends of the previously described forward and reverse primers to improve melting curve resolution for alleles carrying 15, 55, 64, 65, 83, or 84 ru (panel BamS31 in Fig. 1
) (12)(17). It is noteworthy that HRMA was able to correctly identify all of the other allelic variants (differing by just 1 ru to 69 ru) for the rest of the loci, with the exception of the BamS23 10.5 ru variant, which we could not differentiate from the 10 ru allele. For example, the allelic variants for the VrrA, VrrB1, and VrrB2 loci that differed by 1 ru yielded nicely differentiated melting curves (panels VrrA, VrrB1, and VrrB2 in Fig. 1
). The assay is also reproducible, because duplicate amplifications and HR-1 analyses produced identical curves for each allelic variant (Fig. 1
).
We applied optimized HRMA to a cluster analysis of the 12 strains. A genotype cluster analysis of the 6 loci evaluated with HRMA distinguished 7 genotypes, whereas a cluster analysis based on the MLVA results distinguished 10 genotypes (see Fig. 4 in the online Data Supplement). The HRMA method was still able to identify the major branches, however. HRMA was unable to separate the very close relationships between the LT5, LT2, LT4, and Pasteur strains (0.1 linkage distance) identified with the MLVA25 method (12).
We obtained identical allelic variants for the sheep and shepherd (farmer) isolates, demonstrating that the patient’s cutaneous lesion was caused by the B. anthracis strain in the animal. This conclusion was tentatively suggested during the epidemiologic investigation of this anthrax case (18), and we have confirmed that inference. Based on the HRMA and the MLVA25 analyses, we have assigned these 2 strains to B. anthracis cluster B3.
Our application of the HRMA method to B. anthracis genotyping opens the possibility of a rapid screening method for B. anthracis VNTR loci and for implementing this diagnostic procedure in the field or in the first-event laboratory. HRMA may be helpful for evaluating B. anthracis allelic variants if a large number of samples have to be rapidly analyzed and compared, as may be required in the event of a bioterrorism attack.
Acknowledgments
Grant/funding support: This work is supported by the Progetto Antrace- ISS-Ministero della Salute, within the framework of the "Italy-US Collaboration Program".
Financial Disclosures: None declared.
Acknowledgments: We are grateful to Claudia Lucarelli for her skillful assistance.
References
uzewski S, Rakin A, Jagielski M, Zasada A, Jakubczak A, et al. Intriguing diversity of Bacillus anthracis in eastern Poland the molecular echoes of the past outbreaks. FEMS Microbiol Lett 2004;239:235-240.[CrossRef][ISI][Medline]
[Order article via Infotrieve]The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  J.-H. Lin, C.-P. Tseng, Y.-J. Chen, C.-Y. Lin, S.-S. Chang, H.-S. Wu, and J.-C. Cheng Rapid Differentiation of Influenza A Virus Subtypes and Genetic Screening for Virus Variants by High-Resolution Melting Analysis J. Clin. Microbiol., March 1, 2008; 46(3): 1090 - 1097. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
