Detection of Streptococcus pneumoniae strain cocolonization in the nasopharynx

Abstract

Colonization with more than one distinct strain of the same species, also termed cocolonization, is a prerequisite for horizontal gene transfer between pneumococcal strains that may lead to change of the capsular serotype. Capsule switch has become an important issue since the introduction of conjugated pneumococcal polysaccharide vaccines. There is, however, a lack of techniques to detect multiple colonization by S. pneumoniae strains directly in nasopharyngeal samples. Two hundred eighty-seven nasopharyngeal swabs collected during the prevaccine era within a nationwide surveillance program were analyzed by a novel technique for the detection of cocolonization, based on PCR amplification of a noncoding region adjacent to the pneumolysin gene (plyNCR) and restriction fragment length polymorphism (RFLP) analysis. The numbers of strains and their relative abundance in cocolonized samples were determined by terminal RFLP. The pneumococcal carriage rate found by PCR was 51.6%, compared to 40.0% found by culture. Cocolonization was present in 9.5% (10/105) of samples, most (9/10) of which contained two strains in a ratio of between 1:1 and 17:1. Five of the 10 cocolonized samples showed combinations of vaccine types only (n = 2) or combinations of nonvaccine types only (n = 3). Carriers of multiple pneumococcal strains had received recent antibiotic treatment more often than those colonized with a single strain (33% versus 9%, P = 0.025). This new technique allows for the rapid and economical study of pneumococcal cocolonization in nasopharyngeal swabs. It will be valuable for the surveillance of S. pneumoniae epidemiology under vaccine selection pressure.

FIG. 1.

Sensitivity of plyNCR PCR for the detection of colonization and cocolonization by S. pneumoniae strains. (A) plyNCR PCR was performed with serial dilutions of chromosomal DNA. The limit of detection was 0.001 ng. Bands at 50 and 10,380 bp in each gel lane represent internal size markers used by the bioanalyzer for sizing. (B) plyNCR RFLP was performed on an in vitro mixture of two pneumococcal strains. The detection limit for pneumococcal cocolonization is reached at a strain ratio of 1:100, indicated by the fading intensity of the band at 580 bp (arrow). Ladder peaks are shown and sizes are indicated on the left (bp = base pairs).

FIG. 2.

Detection of cocolonization by plyNCR PCR-RFLP is based on the comparison of the plyNCR PCR product size and the sum of fragment sizes obtained after restriction enzyme digestion of the PCR product. plyNCR stands for the undigested PCR product. (A) Example of a swab (sample 122 8799) without cocolonization: the sum of fragment sizes after enzyme digestion of the plyNCR PCR product is smaller than the size of the PCR product. (B) Example of a swab (sample 122 6470) with cocolonization: the sum of fragment sizes after enzyme digestion of the plyNCR PCR product exceeds the size of the PCR product. (C) Molecular types of cocolonizing strains can be inferred from plyNCR RFLP patterns by subtraction analysis if the following conditions are met: subtraction of the plyNCR RFLP pattern of a strain known to be in the mixture (due to culture results) allows determination of the pattern of the remaining strain(s) if cocolonization is indicated by only one restriction enzyme.

FIG. 3.

Pneumococcal cocolonization with two strains is detected by T-RFLP. The electropherogram shows results for two terminal fragments with labeled reverse primer (FAM [6-carboxyfluorescein], blue) and two fragments with labeled forward primer (VIC, green) which represent the two pneumococcal strains. The high resolution of T-RFLP is shown in the inset enlargements of the two VIC peaks. The two fragments are well separated, although their sizes differ by only 2 bp. The relative amounts of the two strains in the sample are indicated by the relative heights of the two peaks for a given primer. The estimation is best when the peaks lie close to each other, like the VIC peaks (strain ratio, 1.5:1) The red peaks represent the LIZ 1200 size standard, and the primer peak (unincorporated primers) is visible on the left. The y axis shows fluorescence units, and the x axis shows data collection points.