A curated multi-locus sequence typing (MLST) database for Haemophilus parasuis

Authors: Register, Karen; Mullins, Michael; Bayles, Darrell; JOLLEY, KEITH - University Of Oxford

Submitted to: American Society for Microbiology General Meeting
Publication Type: Abstract Only
Publication Acceptance Date: 2/28/2011
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Background. Haemophilus parasuis is the etiologic agent of Glasser's disease and pneumonia in swine. Serotyping has traditionally been used for classification of strains but results are subjective and not highly reproducible and the required reagents are expensive to produce, not widely available, and not uniform among laboratories. Sequence-based methods, such as MLST, are objective and reproducible and results are uniform and easily shared among laboratories. An MLST scheme recently proposed for H. parasuis performed favorably when applied to a group of field isolates but the design is not optimal and the method has a number of shortcomings. Here we report several modifications that enhance and simplify the method and describe the development and launch of a publicly available MLST database based on the modified scheme. Methods. Partial gene sequences of 7 housekeeping genes from 35 genetically distinct strains of H. parasuis, including the 15 serovar reference strains, were determined by sequencing PCR amplicons obtained using the MLST primers originally reported. Gene-specific DNA sequence alignments were examined for variability and quality of match within the original primer-binding regions. Vector NTI Advance software was used for sequence editing and analysis and to assist in primer redesign. DNA sequences were trimmed, concatenated and used to construct neighbor-joining trees. Results. Ten of the 14 original primers were redesigned to eliminate mismatches and avoid regions of high sequence heterogeneity revealed by our analyses. To simplify the method for end users, primer redesign included standardization of melting temperature for all primer sets; a single set of cycling conditions under which robust and reproducible amplification of all targets was also identified. The number and % of variable positions within the dataset obtained using the modified method (240 of 3154 positions; 7.6%) was slightly less than for the original method (408 of 3806 positions; 10.7%) but the grouping of strains in major clusters was unaffected. Data from the modified MLST were used to populate a publicly available H. parasuis MLST database. An accompanying isolates database captures descriptive information. Conclusions. The modified MLST method described is more robust, reproducible and user-friendly than the original. Establishment of a publicly accessible MLST database provides a novel resource for epidemiologic investigation of H. parasuis outbreaks and for tracking strain evolution. Researchers worldwide may now access and view isolate and typing information and also contribute to the database through submission to the curator.