Neisseria meningitidis Subtype Nomenclature

We believe the novel subtyping nomenclature proposed for Neisseria meningitidis (7) is flawed and, if adopted, will create chaos for meningococcal epidemiology with serious implications for public health monitoring and vaccine development. The meningococcal subtype is determined by amino acid sequence variation in the cell-surface-exposed loops I and IV of the porin protein PorA (5, 10). The antigenically variable loops are encoded by variable regions 1 and 2 (VR1 and VR2) of the porA gene, respectively. Although this variation was first recognized by the use of murine monoclonal antibodies (MAbs) (1), it is more reliably deduced from the nucleotide sequence of the porA gene, considering the deficiencies inherent in the serological approach: the panel of MAb subtyping reagents is not comprehensive (4, 7, 9); it is becoming increasingly obsolete, as the antigenic composition of PorA continually evolves under the immune selection imposed by the host (6); murine MAbs have limited relevance to vaccine design, when epitopes recognized in the human immune response are required (3, 9); and the early use of antibiotic therapy in meningococcal disease has increased the dependence on PCR-based diagnoses and subtyping (2).

Sacchi et al. (7) propose a nomenclature that attempts to encompass both genotypic, or DNA-based, subtyping data and phenotypic subtyping data which is based on the reactions of PorA with a panel of MAbs. The proposal is unsound for a number of important reasons: (i) it is unnecessary, and potentially confusing, to attempt to convey different types of information in a single nomenclature; (ii) the proposed nomenclature ignores the fundamental relationship between genotype and phenotype, naming genetically related antigens disparately contingent upon the availability of MAb reagents; (iii) their assignment of VRs to families on the basis of our 80% amino acid identity cutoff using a denominator that makes no allowance for genetic insertions, duplications, and deletions results in genetically related PorA proteins having unrelated names; and (iv) such radical changes to the previously published names of meningococcal PorA epitopes will create unnecessary confusion in meningococcal epidemiology. Finally, the combined use of uppercase, lowercase, and underlined letters for the nomenclature is excessively complicated and difficult to understand, even for those familiar with meningococci.

We maintain that the meningococcal subtype nomenclature should be based primarily on the amino acid sequence deduced from the nucleotide sequence of porA but should accommodate existing names assigned originally from serological data (8). This scheme has several advantages: closely related VR sequences have similar names regardless of their reactivity with a specific MAb; the potential for a nontypeable result is eliminated; it can be readily expanded to include novel sequences; and, since porA is under immune selection in humans, it provides more reliable information for the design of candidate vaccines based on PorA than analyses based on murine MAbs. The genetic relationships between the porA genes of the meningococcus are best conveyed by using the conventional genetic designation porA followed by a number representing each unique allele. We have made a comprehensive collection of these data available on a website, http://mlst.zoo.ox.ac.uk/Meningococcus, that is regularly updated to accommodate novel sequences. New VR and allele numbers can be obtained on request.