Submitted to: BMC Microbiology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/17/2005
Publication Date: 6/23/2005
Citation: Bricker, B.J., Ewalt, D.R. 2005. Evaluation of the hoof-print assay for typing brucella abortus strains isolated from cattle in the united states: results with four performance criteria. BMC Microbiology. 5:37. Interpretive Summary: Brucellosis, an economically important disease in cattle, has almost been eliminated from the US. Though rare, the disease continues to be found in cattle herds, including four new incidents in Wyoming in 2003-2004. As a result, Wyoming cattle ranchers must now comply with new restrictions and regulations regarding the marketing and transporting of all cattle within the state. Sporadic cases are expected to be discovered within the US for several more years. An important aspect of disease eradication involves determining where the infection originated, to prevent additional transmission to other herds. Complicating this procedure is the continued presence of the disease in certain wildlife species, mainly elk and bison, in the greater Yellowstone area and possibly in private wildlife herds. Recently, a new technique was developed that creates DNA fingerprints of disease-causing strains. This manuscript describes our study to determine how well this method can discriminate among unrelated disease strains. The investigation found that with this technique, 98 unique DNA fingerprints were produced from 103 strains. The test group was composed of 97 independent field strains and 6 common laboratory strains. This unprecedented ability to specifically identify each disease strain will revolutionize how the trace-back of brucellosis outbreaks is performed, and will significantly improve the success rate of finding the disease source.
Technical Abstract: BACKGROUND: The epidemiology of infectious disease outbreaks benefits from the availability of highly discriminating typing methods. As new approaches for epidemiological typing are developed, it can be very difficult to decide which tests work best. The European Study Group for Epidemiological Markers established guidelines in 1994 for evaluating the performance of typing systems based of a number of criteria. Recently, a new method for typing Brucella abortus strains, called HOOF-Prints, has been described. The method measures hypervariability at eight tandem repeat loci. This paper describes a study with 97 independent field isolates and 6 common laboratory strains of B. abortus to test the HOOF-Print Assay by the criteria set out by the ESGEM. RESULTS: The HOOF-Print Assay was tested for typeability, reproducibility, power of discrimination and concordance to biotyping. Both typeability and reproducibility of the assay were excellent. Allele profiles and frequencies varied widely among the eight loci, ranging from 1 allele for Locus-8, to13 alleles each for Locus-1 and Locus-7. The distribution of alleles also varied, ranging from narrow (Locus-6: two to four repeat units) to broad (Locus-3: one to sixteen repeat units). As a result, the Hunter-Gaston discrimination index (HGDI) varied considerably by allele, ranging from 0 to 0.9, where absolute discrimination of all strains has a value of 1.0. None of the individual loci achieved the recommended HGDI of ' 0.95. However, when the alleles from all eight loci were combined, the DI of the composite fingerprints was 0.9992 [98 unique DNA fingerprints from 103 total strains evaluated], well above the recommended value. When the HGDI values were evaluated based on geographic regions, no obvious clustering of alleles or fingerprints occurred. By comparison, the HGDI value for biovar typing was 0.61 in a population biased with disproportionate numbers of the uncommon biovars. Based on the historical proportions of biotypes in the USA, the HGDI is closer to 0.23. Overall, the HOOF-Print results were concordant with the biovar typing results, except for one pair of isolates exhibiting identical fingerprints but characteristics of different biovars. CONCLUSION: HOOF-Prints genotyping met or exceeded all of the recommended levels for the performance criteria tested. These results support that multilocus VNTR analysis by HOOF-Print genotyping will provide a significant improvement in Brucella strain typing for epidemiological studies.