The Coxiella burnetii cooperative database was set-up in 2014 by agregating
1-MLVA (Multiple Locus VNTR Analysis) genotyping data deduced from published whole genome sequence data by in silico analysis
2-data published in 2006 by Arricau-Bouvery et al. ("C. burnetii 2007 Orsay" database)
3-data, provided by Kinga Sulyok, Miklós Gyuranecz and col., Institute for Veterinary Medical Research, Budapest, Hungary ("C. burnetii 2014" Hungary database)
4-data, published and unpublished, produced since 2007 by Jeroen Tilburg et al. ("C. burnetii 2014 Nijmegen" database)
5-additional published data compiled from the litterature ("C. burnetii published others" database)
Until 2014, some of these data were presented in independant databases, because coding convention were not yet harmonised. Data has been adjusted before import in order to fit with the coding conventions recalled below. Adjustment is based on the Nine Mile code as indicated in the publication. Data from publications missing this essential information cannot be confidently aggregated.
Additional contributions are wellcome, contact any of the current participants for advice on how to proceed to associate a database to the cooperation.
VNTRs for genotyping Coxiella burnetii were initially evaluated by Svraka et al. and by Arricau-Bouvery et al. in 2006. The currently most popular assay MLVA6Nijmegen takes advantage of the six highly variable loci ms23, ms24, ms27, ms28, ms33, ms34, which were found to be the most valuable in the outbreak context in the Netherlands and correspond to VNTR panel 2 proposed by Arricau-Bouvery et al. with the exception of ms31. The use of additional loci would probably have added value when investigating a wider genetic diversity.
List of panel 1 loci + panel 2 locus ms31 by Arricau-Bouvery et al. ( bold typed and italicised : changes with respect to initial report ; changes may be due to design of new more robust primers, change in convention, or initial clerical errors).
ms01_16bp_198bp_4U GGCTCATTCAATTTTAGCTTCG AACGTGGGGAAGTTTGTTATTT
ms03_12bp_229bp_7U TTGTCGATAAATCGGGAAACTT CACTGGGAAAAGGAGAAAAAGA
ms07_126bp_1112bp_8U CTCTTAGCCATCGCTTACCACT AACGAAAATTGGTTTGCATTTT
ms12_126bp_1074bp_8U GAAAATTGGTTTGCGCTCTG CCTTCTCCCAAGAAGTTTAGCC
ms20_18bp_402bp_15U CTGAAACCAGTCTTCCCTCAAC CTTTATCTTGGCCTCGCCCTTC
ms21_12bp_210bp_6U AGCATCTGCCTTCTCAAGTTTC TGGGAGGTAGAAGAAAAGATGG
ms22_11bp_246bp_6U GGGGTTTGAACATAGCAATACC CAATATCTCTTTCTCCCGCATT
ms26_9bp_127bp_4U AGAATCAAACCTGCAAAACCTT TTGATTATTTTGACTTCGCTGGT
ms30_9bp_306bp_12U ATTTCCTCGACATCAACGTCTT AGTCGATTTGGAAACGGATAAA
ms31_7bp_285bp_5U GGGCATCTAATCGAGATAATGG TTTGAGAAAATTTTGGGTGCTT
ms36_9bp_477bp_4U GAAACCAGTCTTCCCTCAACAG ATAACCGTCATCGTCACCTTCT
The primers currently recommanded for typing the six loci in panel 2 minus ms31 (=MLVA6Nijmegen) are :
ms23_7bp_133bp_9U CGCMTAGCGACACAACCAC GACGGGCTAAATTACACCTGCT
ms24_7bp_261bp_27U TGGAGGGACTCCGATTAAAA GCCACACAACTCTGTTTTCAG
ms27_6bp_89bp_4U TCTTTATTTCAGGCCGGAGT GAACGACTCATTGAACACACG
ms28_6bp_111bp_6U AGCAAAGAAATGTGAGGATCG GCCAAAGGGATATTTTTGTCCTTC
ms33_7bp_104bp_9U TCGCGTAGCGACACAACC GTAGCCCGTATGACGCGAAC
ms34_6bp_101bp_5U TTCTTCGGTGAGTTGCTGTG GCAATGACTATCAGCGACTCGAA
The locus name "name_Xbp_Ybp_ZU" as indicated here contains the allele calling convention. X is the size of the repeat unit, Y is the PCR product size in the Nine Mile RSA493 NC_002971.3 reference genome as predicted by in silico analysis. Z is the assigned repeat copy number (this is a convention).
The MLVA6Nijmegen code for Nine Mile is 9-27-4-6-9-5.
Important remarks, including coding conventions :
1-ms23 : coding convention in Arricau-Bouvery et al. was 8U for Nine Mile in silico RSA493. It has been changed to 9U for harmonization with subsequent publications.
2-ms30 : initially considered as a 18 bp repeat unit locus by Arricau-Bouvery et al., coded 6U in Nine Mile in silico analysis, recoded 9bp and 12U because of 297 bp allele in RSA331 sequenced genome (accession number CP000890 or NC_010117.1) as compared to the 306 bp size in Nine Mile.
3-ms33 : an alternative convention by which ms33 is coded 4U instead of 9U has been proposed. However it is incorrect, as for instance strain RSA331 NC_010117.1 with a predicted 69 bp amplification product would then need to be coded -1 whereas the allele contains 4 repeat units.
4-ms20 and ms36 are adjacent. When using the primers published for ms36 by Arricau-Bouvery et al., both loci are amplified. Consequently the moderately polymorphic ms20 locus is redundant.
5-Table 3 in Arricau-Bouvery et al. contained three clerical errors regarding the indicated size of PCR products for ms30, ms31 and ms36 in Nine Mile. The correct values are as listed above.
MLVA6Nijmegen genotype numbering is proposed in the Coxiella burnetii database. About 100 different profiles are identified as of February 2014. Nine Mile 9-27-4-6-9-5 is MLVA6 genotype 1. However, although convenient, genotype numbering has some drawbacks. Firstly numbering does not reflect genotype proximity. Secondly, isolates which could not be fully amplified cannot be given an MLVA6 type number.
