Proc. Natl. Acad. Sci. USA Vol. 95, pp. 3140–3145, March 1998 Microbiology Multilocus sequence typing: A portable approach to the identification of clones within populations of pathogenic microorganisms (molecular typingyNeisseria meningitidisyhousekeeping genesyWorld-Wide Webyhyper-virulent clones) MARTIN C. J. MAIDEN*, JANE A. BYGRAVES†,EDWARD FEIL‡,GIOVANNA MORELLI§,JOANNE E. RUSSELL†, RACHEL URWIN*, QING ZHANG‡,JIAJI ZHOU*, KERSTIN ZURTH§,DOMINIQUE A. CAUGANT¶,IAN M. FEAVERS†, i MARK ACHTMAN§ , AND BRIAN G. SPRATT*‡ *Wellcome Trust Centre for the Epidemiology of Infectious Disease, Department of Zoology, University of Oxford, Oxford OX1 3PS, United Kingdom; †Division of Bacteriology, National Institute for Biological Standards and Controls, Blanche Lane, South Mimms, Potters Bar EN6 3QG, United Kingdom; ‡School of Biological Sciences, University of Sussex, Brighton BN1 9QG, United Kingdom; §Max-Planck-Institut fu¨r Molekulare Genetik, Ihnestrasse 73, 14195 Berlin, Germany; and ¶World Health Organization Collaborating Centre for Reference and Research on Meningococci, National Institute of Public Health, P.O. Box 4404, Torshov, N-0403 Oslo, Norway Edited by John Maynard Smith, University of Sussex, Brighton, United Kingdom, and approved January 6, 1998 (received for review October 16, 1997) ABSTRACT Traditional and molecular typing schemes methods is the difficulty of comparing the results achieved by for the characterization of pathogenic microorganisms are different laboratories. poorly portable because they index variation that is difficult Molecular typing methods are used to address two very to compare among laboratories. To overcome these problems, different kinds of problem. First, are the isolates recovered we propose multilocus sequence typing (MLST), which ex- from a localized outbreak of disease the same or different ploits the unambiguous nature and electronic portability of strains (short term or local epidemiology)? Second, how are nucleotide sequence data for the characterization of micro- strains causing disease in one geographic area related to those organisms. To evaluate MLST, we determined the sequences isolated world-wide (long term or global epidemiology)? Dif- of '470-bp fragments from 11 housekeeping genes in a ferent methods may be appropriate for investigating local and reference set of 107 isolates of Neisseria meningitidis from global epidemiology, but in both cases they should be highly invasive disease and healthy carriers. For each locus, alleles discriminatory such that isolates assigned to the same molec- were assigned arbitrary numbers and dendrograms were ular type are likely to be descended from a recent common constructed from the pairwise differences in multilocus allelic ancestor, and isolates that share a more distant common profiles by cluster analysis. The strain associations obtained ancestor are not assigned to the same type. were consistent with clonal groupings previously determined High levels of discrimination can be achieved in two quite by multilocus enzyme electrophoresis. A subset of six gene different ways. In one approach, individual loci, or uncharac- fragments was chosen that retained the resolution and con- terized regions of the genome, that are highly variable within gruence achieved by using all 11 loci. Most isolates from the bacterial population are identified. For bacterial patho- hyper-virulent lineages of serogroups A, B, and C meningo- gens, several methods based on this approach are currently cocci were identical for all loci or differed from the majority popular, e.g., ribotyping, pulsed-field gel electrophoresis type at only a single locus. MLST using six loci therefore (PFGE), and PCR with repetitive element primers, or arbi- reliably identified the major meningococcal lineages associ- trary primers (1). In these methods, restriction enzymes (or ated with invasive disease. MLST can be applied to almost all PCR primers) are chosen that give maximal variation within bacterial species and other haploid organisms, including the population; consequently, the variation that is indexed is those that are difficult to cultivate. The overwhelming advan- evolving very rapidly, usually for unknown reasons. The sec- tage of MLST over other molecular typing methods is that ond approach, typified by multilocus enzyme electrophoresis sequence data are truly portable between laboratories, per- (MLEE), is to use variation that is accumulating very slowly in mitting one expanding global database per species to be placed the population and that is likely to be selectively neutral. on a World-Wide Web site, thus enabling exchange of molec- Although only a small number of alleles can be identified ular typing data for global epidemiology via the Internet. within the population by using this type of variation, high levels of discrimination are achieved by analyzing many loci. Methods that index rapidly evolving variation are useful for The ability to identify accurately the strains of infectious short term epidemiology but may be misleading for global agents that cause disease is central to epidemiological surveil- epidemiology. Several studies have shown that techniques such lance and public health decisions, but there are no wholly as PFGE resolve isolates that are indistinguishable by MLEE. satisfactory methods of achieving this goal (1). All of the For example, MLEE studies of populations of Salmonella numerous methods that are currently used suffer from signif- enterica have shown that isolates of serovar Typhi from typhoid icant drawbacks, including various combinations of inadequate fever belong to one of two closely related electrophoretic types discrimination, limited availability of reagents, poor reproduc- (ETs) (2). In contrast, isolates of serovar Typhi are relatively ibility within and between laboratories, and an inability to quantitate the genetic relationships between isolates. How- This paper was submitted directly (Track II) to the Proceedings office. ever, perhaps the most important limitation of current typing Abbreviations: ET, electrophoretic type; MLST, multilocus sequence typing; MLEE, multilocus enzyme electrophoresis; PFGE, pulsed- The publication costs of this article were defrayed in part by page charge field gel electrophoresis; ST, sequence type. Data deposition: The nucleotide sequences described in this paper payment. This article must therefore be hereby marked ‘‘advertisement’’ in have been deposited in the GenBank database (accession nos. accordance with 18 U.S.C. §1734 solely to indicate this fact. AF037753–AF037981). i © 1998 by The National Academy of Sciences 0027-8424y98y953140-6$2.00y0 To whom reprint requests should be addressed. e-mail: achtman@ PNAS is available online at http:yywww.pnas.org. mpimg-berlin-dahlem.mpg.de. 3140 Downloaded by guest on September 23, 2021 Microbiology: Maiden et al. Proc. Natl. Acad. Sci. USA 95 (1998) 3141 diverse according to PFGE (3). PFGE is therefore useful for strains by using PCR with the following primers: abcZ-P1, studying individual outbreaks of typhoid fever because, unlike 59-AATCGTTTATGTACCGCAGG-39 and abcZ-P2, 59-GT- MLEE, it identifies the microvariation that is needed to TGATTTCTGCCTGTTCGG-39; adk-P1, 59-ATGGCAGTT- distinguish between strains circulating within a geographic TGTGCAGTTGG-39 and adk-P2, 59-GATTTAAACAGCG- area. However, this technique is too discriminatory for long ATTGCCC-39; aroE-P1, 59-ACGCATTTGCGCCGACA- term epidemiology because it does not indicate that isolates TC-39 and aroE-P2, 59-ATCAGGGCTTTTTTCAGGTT-39; that cause typhoid fever are members of a single globally gdh-P1, 59-ATCAATACCGATGTGGCGCGT-39 and gdh- distributed clonal lineage of S. enterica. To use a common P2, 59-GGTTTTCATCTGCGTATAGAG-39; mtg-P1, 59-CG- metaphor, PFGE and other similar methods fail to see the GCATCTTTATCTTTTTCAA-39 and mtg-P2, 59-TCAGTC- forest for the trees. CGTAyGTCNCTTyCTCNGG-39; pdhC-P1, 59-GGTTTCC- The most appropriate of the current techniques for long AACGTATCGGCGAC-39 and pdhC-P2, 59-ATCGGCTTT- term epidemiology, and for the identification of lineages that GATGCCGTATTT-39; pgm-P1, 59-CTTCAAAGCCTACG- have an increased propensity to cause disease, is undoubtedly ACATCCG-39 and pgm-P2, 59-CGGATTGCTTTCGATGA- MLEE. This approach also has contributed most to our CGGC-39; pilA-P1, 59-AAGGGCTGAAAGACGGCAA-39 understanding of the global epidemiology and population and pilA-P2, 59-CAATCCAGCAGTCGGTCCACA-39; pip- structure of infectious agents. For many pathogens, MLEE P1, 59-CGGATACTTGCAGGTGTCTG-39 and pip-P2, 59-C- successfully has identified clusters of closely related strains TCAACCGCCTGAACCAACG-39; ppk-P1 59-GAACAAA- (clones or clonal complexes) that are particularly liable to ACCGCATCCTCTGC-39 and ppk-P2, 59-ATCGTTTTGCA- cause disease (1, 4). A major problem with MLEE, and all GGTCGGCTTC-39; and serC-P1, 59 CTGCCAGCCTAAA- other current typing methods, is that the results obtained in ATCGGGCGGGTTATTG-39 and serC-P2, 59-CAACATC- different laboratories are difficult to compare. We have there- GGGACATCAACCG-39. Sequencing of both strands of the fore chosen to adapt the proven concepts and methods of amplified fragments was achieved by using an Applied Bio- MLEE by identifying alleles directly from the nucleotide systems Prism 377 automated sequencer with dRhodamine- sequences of internal fragments of housekeeping genes rather labeled terminators (PE Applied Biosystems). The following than by comparing the electrophoretic mobilities of the en- primers were used for sequencing: abcZ-P1 and