Identification of the Hypervirulent Lineages of Neisseria meningitidis, the ST-8 and ST-11 Complexes, by Using Monoclonal Antibodies Specific to NmeDI

Abstract

Most cases of serogroup C meningococcal disease are caused by the clonal lineages ST-8 and ST-11. The gene encoding the putative restriction-modification system NmeDI is specific to these lineages. We report here a monoclonal antibody directed against the NmeDI endonuclease as a tool for their rapid and spe-cific identification.

Neisseria meningitidis causes meningitis and septicemia in humans. Molecular epidemiology has revealed that most cases are caused by only a few clonal lineages, whereas the genetic diversity is higher in meningococci recovered from healthy carriers (3, 4). Two important hypervirulent meningococcal lineages are the sequence type 8 (ST-8) and ST-11 complexes, both of which are responsible for a significant proportion of serogroup C disease worldwide (1); a search of the Neisseria Multilocus Sequence Typing (MLST) homepage (http://www.neisseria.org/nm/typing/mlst/; March 2003) revealed that of a total of 325 invasive serogroup C strains reported to the MLST center, 257 (79%) belonged to the ST-8 and ST-11 complexes, respectively. The ST-8 and ST-11 complexes are related, since both of them express related PorA and PorB proteins (class 2), they share differentially distributed restriction-modification (R-M) systems (5, 7), and the founder STs share three alleles at seven loci analyzed in the MLST scheme.

We recently demonstrated by DNA-DNA hybridization that the putative R-M system NmeDI is specific to these two lineages (5, 7). The R-M system was inserted between the pheS and the pheT genes, and other meningococcal lineages harbored different genes at this location. The MLST reference strain collection (9), including 103 mostly pathogenic isolates, was recently used to demonstrate the specificity of NmeDI for the ST-8 and ST-11 complexes (5), which was 95% (there were four false positives and 83 true negatives; specificity was calculated by using the formula specificity = true negatives/[false positives + true negatives]). We reevaluated this specificity by using the Bavarian meningococcal carrier strain collection (822 isolates, 322 STs), for which a complete MLST data set was available (6). The collection comprised 10 isolates belonging to the ST-8 and ST-11 complexes, which were all positive for NmeDI upon DNA-DNA hybridization in accordance with the protocol published recently (5). Of the remaining 812 isolates, only 5 (<1%) were NmeDI positive without belonging to the ST-8 and ST-11 complexes (4 isolates of a total of 136 isolates belonging to the ST-44 complex; 1 isolate was ST-914). This finding indicates that after horizontal gene transfer, NmeDI is probably not genetically fixed in lineages other than the ST-8 and ST-11 complexes. Because of the low number of ST-8 and ST-11 complex strains, the positive predictive value (true positives/[true positives + false positives]) of DNA-DNA hybridization with NmeDI was only 67% for the Bavarian meningococcal carriage isolate collection. However, the negative predictive value (true negatives/[true negatives + false negatives]) was 100% and the specificity was 99.4%. These results prompted us to generate monoclonal antibodies (MAbs) against the NmeDI endonuclease in order to obtain an easy-to-use reagent for the detection of ST-8 and ST-11 complex meningococci. Mice were immunized with a recombinant peptide comprising amino acids 1 to 143 of NmeDI (National Center for Biotechnology Information accession no. CAB59898), and MAbs were raised in accordance with standard protocols (10). Affinity-purified MAb 3/9-2 specifically reacted with NmeDI in meningococcal lysates, as demonstrated by Western blot analysis and by dot immunoblotting. As expected, a 40-kDa protein was detected in Western blots (data not shown). The meningococcal protein was not detectable by enzyme-linked immunosorbent assay, although the recombinant peptide expressed in Escherichia coli was detectable in this assay (data not shown). Therefore, dot blot assays were used for further analyses. We suggest that the epitope in the correctly folded protein is not accessible to the antibody and that denaturing methods have to be used. Approximately 10⁹ CFU were suspended in 1,000 μl of a sample solution containing 5% β-mercaptoethanol and 2% sodium dodecyl sulfate (11). Ten microliters was dotted onto nitrocellulose membranes (Schleicher & Schüll, Dassel, Germany) and dried. Unspecific binding sites on the membranes were blocked by incubation with phosphate-buffered saline-5% skim milk powder-0.1% Tween 20 for 60 min. A 1:5,000 dilution in phosphate-buffered saline-1% skim milk powder-0.1% Tween 20 of the affinity-purified MAb 3/9-2 stock solution containing 2.26 mg of protein/ml was added, and the mixture was incubated for 60 min. The blots were developed as described previously (11). MAb 3/9-2 reacted with the 15 NmeDI-positive meningococcal isolates of the Bavarian isolate collection, suggesting constitutive expression. It did not react with 15 selected NmeDI-negative isolates from the same collection (Fig. 1). Therefore, the results obtained with MAb 3/9-2 correlated with the reactivity of the strains with an NmeDI probe. MAb 3/9-2 reacted with eight strains of the ST-8 complex and nine strains of the ST-11 complex, which were isolated between 1967 and 2000 in nine different countries (data not shown). This finding again underlines the association of NmeDI with these lineages. The MAb did not react with a collection of 26 genetically diverse strains of N. lactamica described recently (2; data not shown). It also did not react with a variety of phylogenetically diverse bacterial species (Table 1). However, some of the Staphylococcus aureus isolates used were positive. This reaction was due to nonspecific binding of MAb 3/9-2 by staphylococcal protein A, because the strains also bound mouse MAbs directed against the meningococcal serogroup B capsule (MAb 735) (8) and human complement factor C3 (MAb 755) (11), respectively (data not shown). In conclusion, we report on MAb 3/9-2, which specifically identified meningococci of the ST-8 and ST-11 complexes. We recommend the use of the antibody in reference laboratories for rapid assignment of isolates to the ST-8 and ST-11 complexes prior to further sequence typing. The antibody showed no cross-reactivity with other neisserial species and several other bacterial genera, with the exception of an unspecific binding to S. aureus. Use of the antibody for detection of NmeDI in clinical speci-mens would require elimination of the Fc portion of the antibody.

FIG. 1.

Dot blot showing the reactivity of meningococcal lysates with MAb 3/9-2. Representative meningococci of the Bavarian carrier isolate collection were used, including all strains harboring the gene for NmeDI. A1, positive-control strain 2120 (ST-11) (5); A2, negative-control strain MC58 (ST-74; donated by Richard E. Moxon, Oxford, England). Strains positive for the gene for NmeDI all reacted with MAb 3/9-2 (eight ST-11 complex strains, two ST-8 complex strains, four ST-44 complex strains, and one ST-914 strain), i.e., A3, B4 to -6; D2, -3, -5, and -6; E2 to -6; and F1 and -2. Genetically diverse strains lacking the gene for NmeDI did not react with MAb 3/9-2.

TABLE 1.

Reactivity of MAb 3/9-2 with diverse bacterial species other than N. meningitidis^a

Species	DSMZ	ATCC	LMG	NCTC	Other source	MAb 3/9-2
N. animalis				0212		No
N. canis			8383			No
N. caviae		14659				No
N. cinerea	4630					No
N. denitrificans		33394				No
N. elongata			5124			No
N. elongata subsp. glycolytica				11050		No
N. elongata subsp. nitroreducens		49377				No
N. elongata subsp. nitroreducens		49378				No
N. flavescens			5297			No
N. flavescens		13115				No
N. gonorrhoeae FA 1090					Mark Achtman, Berlin, Germany	No
N. gonorrhoeae		43069				No
N. iguanae		51483				No
N. lactamica	4691					No
N. macacae		33926				No
N. mucosa	4631					No
N. mucosa			5136			No
N. ovis			8381			No
N. perflava			5284			No
N. polysaccharea		43768				No
N. sicca			5290			No
N. subflava			5313			No
N. weaveri		51223				No
Acinetobacter lwoffii	2403					No
Citrobacter freundii		8090				No
Enterococcus faecalis		19433				No
Escherichia coli		25922				No
Eubacterium lentum		43055				No
Fusobacterium nucleatum						No
Haemophilus influenzae		49766				No
Mycobacterium fortuitum		6841				No
Prevotella nigrescens	13386					No
Pseudomonas aeruginosa		27853				No
S. aureus (MRSA, S32, ST-22)					Our laboratory	No
S. aureus (MRSA, S89, ST-22)					Our laboratory	Yesb
S. aureus (MRSA, S5, ST-225)					Our laboratory	No
S. aureus (MRSA, S25, ST-225)					Our laboratory	Yesb
S. aureus (MRSA, S47, ST-255)					Our laboratory	Yesb
S. aureus (MRSA, S3, ST-228)					Our laboratory	Yesb
S. aureus (MRSA, S16, ST-228)					Our laboratory	Yesb
S. aureus (MRSA, S33, ST-228)					Our laboratory	Yesb
S. aureus (MRSA, S2)					Our laboratory	No
S. aureus (MRSA, S23)					Our laboratory	No
S. aureus (MRSA, S40)					Our laboratory	No
S. aureus (MRSA, S6)					Our laboratory	No
S. aureus (MRSA, S42)					Our laboratory	No
S. aureus (MRSA, S61)					Our laboratory	Yesb
S. aureus (MRSA, S34)					Our laboratory	Yesb
S. aureus		3538				Yesb
Streptococcus gordonii		33399				No
Streptococcus mitis	12043					No

^aDSMZ, German strain collection Braunschweig; ATCC, American Type Culture Collection; LMG, Belgian Coordinated Collections of Microorganisms; NCTC, National Collection of Type Cultures; MRSA, methicillin-resistant S. aureus; S32 etc, numbering of our own strain collection. Binding of MAb 3/9-2 was assessed by using dot blot assays.

^bBinding of MAb 3/9-2 to S. aureus isolates was nonspecific.