Abstract
To study the evolution and virulence of pneumococcal populations, we used multilocus sequence typing to identify the major clones among 212 carriage and invasive isolates expressing capsular serogroup 6 from 39 countries. The global population consisted of 8 major complexes and 6 minor complexes of related clones and 32 clones of diverse origin. Surprisingly, serotype 6A clones evolved by mutation nearly as often as by recombination, whereas serotype 6B clones evolved almost exclusively by recombination (P = 0.0029). This is the first report of population genetic differences among serotypes of this species. The largest clonal complex was associated with invasive disease (P = 0.019) and included a common ancestor for five previously identified drug-resistant clones. The putative ancestors of the major clonal complexes were represented by a greater proportion of carriage isolates than were their descendents (P = 0.001), and the ancestors tended to be less virulent than their descendents in a mouse model of infection. These data suggested that virulent serogroup 6 clones have evolved multiple times from less-virulent ancestral clones.
Streptococcus pneumoniae causes more than 1 million deaths worldwide each year due to pneumonia and meningitis in young children alone (14, 15). More frequently, however, the pneumococcus colonizes the nasopharynx asymptomatically or causes less-serious diseases, such as otitis media and sinusitis. Although several bacterial products involved in virulence have been identified (19), the precise factors that predispose a clone for invasive disease versus carriage are unknown. Use of a population genetics approach to identify lineages of clones that are more or less virulent has been hampered by the tremendous diversity of phenotypes and genotypes presented by the species. The capsular polysaccharide, for example, a critical virulence factor and the traditional marker for identifying different strains, is expressed in more than 90 serologically distinct forms.
Genetic exchange between different pneumococcal clones has undoubtedly influenced the species' diversity. The cellular machinery for genetic transformation is present in most natural isolates (22). However, only recently have the molecular typing tools (8) and statistical methods (10) been developed to permit a quantitative estimation of the impact of recombination versus that of mutation on population structure. For the pneumococcus, it was estimated that individual housekeeping genes are ∼10 times more likely to evolve by recombination than by mutation, and individual nucleotides are ∼50 times more likely to evolve by recombination than by mutation (10). In the range of population structures that pathogenic bacteria present, the pneumococcus is among those more influenced by recombination (9).
These species-wide estimates of recombination and mutation are likely to be incomplete for at least two reasons. First, the samples analyzed were composed almost exclusively of invasive and drug-resistant isolates, yet most of the pneumococcal population exists in the carriage state and most natural isolates are still susceptible to antibiotics. The isolates causing invasive disease may represent only a subset of the total genotypic diversity in a local population (25). Second, the epidemiological differences that exist among different capsular serotypes can be so extensive that some have questioned whether S. pneumoniae should continue to be classified as a single etiological agent (28). As the risk of disease caused by different serotypes varies by age and geography, there is reason to suspect that serotypes may also differ in population structure.
Here, we study a global population of capsular serogroup 6 pneumococci, which includes the serotypes 6A and 6B. This serogroup consistently ranks in the top three as a cause of invasive pneumococcal disease worldwide (14, 28, 31). Multidrug-resistant clones have emerged in this serogroup, making it a particularly relevant group to study (2, 27, 32, 33). Some data suggest that in young children from the United States and Europe, the 6B serotype cause more invasive disease than does serotype 6A (14, 24). This potential epidemiological difference between the two serotypes is intriguing given that antibodies raised against one serotype are cross-reactive for the other (24). Our data show a striking difference in the population structures of serotypes 6A and 6B, a result that was not explained by the inherent transformability of the clones. Moreover, our data reveal novel patterns of evolution for antibiotic resistance and virulence in the global population.
MATERIALS AND METHODS
Bacterial isolates.
We analyzed a total of 212 pneumococcal isolates of serogroup 6 including serotypes 6A (n = 75) and 6B (n = 137). Carriage isolates were from the nasopharynx (n = 80) and pharynx (n = 5). Invasive isolates were from blood and cerebrospinal fluid (n = 82), pleural fluid (n = 3), and bronchial aspirate (n = 1). The remaining isolates were from undetermined sources or from cases of less serious disease (n = 41). The ages of the patients ranged from children ⩽5 years old (n = 149), children 6 to 17 years old (n = 8), adults 18 or older (n = 6), or unknown (n = 49). The dates of isolation were from 1989 to 1999 (n = 149), 1982 to 1988 (n = 8), or unknown (n = 55). Isolates were from 39 countries including Argentina (n = 2), Australia (n = 8), Botswana (n = 2), Brazil (n = 5), Bulgaria (n = 4), Canada (n = 16), Central African Republic (n = 2), Chile (n = 8), China (n = 6), Colombia (n = 5), Ethiopia (n = 3), Finland (n = 1), Gambia (n = 7), Germany (n = 1), Greece (n = 8), Greenland (n = 6), Hungary (n = 1), Iceland (n = 11), India (n = 3), Indonesia (n = 2), Israel (n = 15), Kuwait (n = 4), Lesotho (n = 1), Mexico (n = 9), New Guinea (n = 1), New Zealand (n = 3), Pakistan (n = 1), Philippines (n = 10), Poland (n = 3), Romania (n = 2), Slovokia (n = 1), South Africa (n = 12), South Korea (n = 5), Spain (n = 2), Sweden (n = 3), Thailand (n = 2), United States (Alabama [n = 3], Alaska [n = 19], Maryland [n = 1], and Tennessee[n = 7]), Uruguay (n = 5), and Zambia (n = 2). The 32 isolates from Alaska, China, and Tennessee were selected to represent diverse genetic backgrounds based on IS1167 and boxA genotyping procedures (25). The nine isolates from Alabama, Finland, Germany, Hungary, Maryland, and Spain, plus one isolate from South Africa, were laboratory or reference strains. The remaining 170 isolates were essentially random samples of serogroup 6 pneumococci. For comparison with the serogroup 6 isolates and to increase the diversity of the data set, we included an additional 35 isolates from 11 different serogroups (serogroups 4, 5, 7, 9, 11, 14, 15, 18, 19, 22, and 23). Seven isolates were from the pneumococcal genome diversity project and were chosen specifically to span diversity within the species (http://genome.microbio.uab.edu/strep/info).
Isolates were grown on blood agar plates or in Todd-Hewitt broth with 0.5% yeast extract (THY) overnight at 37°C in a candle jar. Optochin disks (Becton Dickinson) were used to differentiate pneumococci from other alpha-hemolytic streptococci. All isolates were capsule serotyped by the slide agglutination assay with commercial (Statens Seruminstitut) antisera. Cultures were stored at −70°C in a solution of THY plus 10% glycerol.
Multilocus sequence typing.
Seven housekeeping loci were used in a multilocus sequence typing (MLST) system by Enright and Spratt (8). The loci were aroE, gdh, gki, recP, spi, xpt, and ddl. We chose the recP locus in common with Enright and Spratt (8) for this study and elected to use six new loci expected to be selectively neutral and unlinked (>50 kb) from each other on the basis of their position in the completely sequenced TIGR4 strain (34). The choice was made in order to test whether seven loci can provide a reliable tool for studying evolutionary history in a species, such as the pneumococcus, where recombination is thought to be an important process for generating diversity. The amount of recombination that occurs in a species should be a factor in determining the number of loci that need to be examined in order to ensure that a “representative” sample of the chromosome is obtained. For a truly clonal species, a single locus could accurately reflect the evolutionary history of a group of isolates. With increasing amounts of recombination, a larger sample of loci is required because loci of differing evolutionary histories will become scattered around the chromosome. Moreover, as DNA sequencing becomes less expensive with time, the baseline of data provided by the six new loci of this study can be used for an expanded version of the Enright and Spratt (8) MLST system to address questions that require additional sequence data.
The six new loci and predicted protein products were the following: arcC (carbamate kinase), ldh (lactate dehydrogenase), scrK (fructokinase), hexA (DNA mismatch repair), aroB (dehydroquinate synthase), and dnaA (DNA polymerase initiator). Primers were designed to amplify ∼450-bp fragments and were used for both PCR and DNA sequencing on both strands. The primer sequences are available at www.dnaseq.uab.edu/popgen/index.html. The recP primers were as reported by Enright and Spratt (8).
Chromosomal DNA was isolated as described previously (4). PCRs consisted of 10 to 30 ng of chromosomal DNA, 1 μl of each primer from a 50-pmol stock, 2 μl of MgCl2, 5 μl of Q-solution (Qiagen), 12.5 μl of Master Mix (Qiagen), and 4 μl of sterile water (Sigma). Thermal cycling conditions for PCR were as follows: initial denaturation at 95°C for 3 min, 30 cycles of denaturation at 95°C for 1 min, annealing at 55°C for 1 min, extension at 72°C for 2 min, and a final extension at 72°C for 10 min. PCR products were purified with Qiaquick 96-well kits (Qiagen). Sequencing reactions using BigDye (PE Biosystems) or DYEnamic ET (Amersham Pharmacia) dye-terminators were performed according to the manufacturer's instructions. Sequencing products were purified with DyeEx 96-well kits (Qiagen). DNA sequences were obtained using ABI 310 or 377 automated sequencers (PE Biosystems). The Sequencher 4.1 program (Genecodes, Inc.) was used to edit the resulting electropherograms.
Unique DNA sequences defined alleles. Each allele represented by one or two isolates was rechecked and, if necessary, resequenced for quality control. The recP alleles were assigned using the scheme of Enright and Spratt (8), and all new recP alleles found in this study were submitted to their database. A unique combination of alleles at each of the seven loci defined a sequence type (ST) or clone.
Spontaneous transformation.
A few colonies from overnight growth on a blood agar plate were transferred to 3 ml of fresh competence medium (CM) containing THY, 0.01% CaCl2, and 0.2% bovine serum albumin (Sigma). Cultures were grown at 37°C to an optical density at 600 nm of ∼0.5. Each culture was used to seed two additional cultures by transferring 5 μl of the culture into 5 ml of prewarmed CM. Chromosomal DNA from the streptomycin-resistant strain, DP1617, was added to one culture at a concentration of ∼1 μg/ml, while no DNA was added to the second culture. Both cultures were grown at 37°C to an optical density at 600 nm of 0.6 to 0.8, serially diluted with lactated Ringer's solution in 96-well plates, and plated on blood agar or blood agar supplemented with streptomycin (Sigma) at a concentration of 100 μg/ml. The number of streptomycin-resistant colonies from the DNA-negative cultures was subtracted from the number from the DNA-positive cultures to correct for resistance arising as a result of spontaneous mutation. Transformation efficiency was then calculated as [(CFU/ml transformants)/(CFU/ml total)] × 100. Strains were tested 2 to 3 times each, and the results were averaged.
Mouse model of virulence.
Female 3- to 5-week-old BALB/cByJ mice (Jackson Laboratory) were housed for 2 weeks prior to use. Pneumococci were diluted in lactated Ringer's solution, and ∼7 × 106 CFU was injected intraperitoneally (i.p.) in a volume of 200 μl. Mice were monitored for 21 days. The virulence of each strain was tested using five mice. Virulent strains were defined as those causing a median number of days to death of less than 10 (1).
Data analysis.
The BURST program (by E. Feil and M. Chan [http://www.mlst.net]) was used to infer evolutionary relationships among clones. The relative contributions of recombination and mutation to the evolution of clones was calculated as in the work of Feil et al. (10). Contingency tables (2 × 2) were analyzed with Fisher's exact test. Comparison of transformation frequencies was done with the Mann-Whitney U test.
Nucleotide sequence accession numbers.
The DNA sequences reported in this study have been deposited in GenBank with accession numbers AF436690 to AF436776. The alleles for arcC, ldh, scrK, hexA, aroB, and dnaA along with a list of STs can also be retrieved at http://www.dnaseq.uab.edu/popgen/index.html.
RESULTS
Diversity of alleles.
An MLST database for pneumococci (8) contained results for ∼600 isolates at the beginning of our study. The database (8) contained 27 alleles at the recP locus, and our 212 serogroup 6 isolates yielded 9 new recP alleles, an increase of 25%. For recP and the six new loci that we sequenced, we identified a total of 112 alleles among the 212 serogroup 6 isolates. The mean number of alleles per locus was 16 with a range of 9 at dnaA to 26 at scrK. Of the total of 112 alleles, 47% (53 of 112) were shared by serotypes 6A and 6B, and 37% (41 of 112) were represented by single isolates. Only 16% (18 of 112) of the alleles can be considered to be unique to a particular serotype. These results suggested that the two serotypes were not sexually isolated populations. Even though strains of serotype 6A and 6B might not share the same niche due to the potential for immune exclusion, the data suggest that allelic transfer has occurred between the two serotypes, whether mediated through strains of other serotypes or by other means.
Diversity of clones.
Based on the observed number of alleles, we could theoretically distinguish ∼1.4 × 108 different STs or clones. The 212 isolates were classified into 131 distinct STs (Table 1; alleles and STs are listed at http://www.dnaseq.uab.edu/popgen/index.html). Ninety-nine of these STs were represented by a single isolate, and 32 STs were represented by 2 to 17 isolates. Of the 32 STs represented by multiple isolates, 30 STs expressed either serotype 6A or 6B. Serotype was therefore a relatively stable characteristic of a clone. Additionally, of the 32 recurring STs, 16 STs were from multiple countries (double-underlined in Fig. 1). The most abundant and geographically dispersed clone was the multidrug-resistant Spain6B-2 clone (18, 32), identified as ST19 in this study and isolated from 11 countries. From the product of the frequencies of the most common allele at each locus, the ST most likely to occur by chance was ST18 (P = 0.001), which was represented by one isolate from Uruguay. Interestingly, ST18 differed from ST19 at only a single locus and was therefore closely related. These data may indicate that a fitness advantage is associated with the combination of alleles found in STs 18 and 19.
TABLE 1.
List of allelic profiles for the 131 STs of serogroup 6
| ST | Alleles at the indicated locus
|
||||||
|---|---|---|---|---|---|---|---|
| arcC | ldh | scrK | hexA | recP | aroB | dnaA | |
| 1 | 6 | 4 | 1 | 1 | 6 | 1 | 1 |
| 2 | 3 | 4 | 1 | 1 | 6 | 1 | 1 |
| 3 | 3 | 4 | 1 | 1 | 6 | 13 | 1 |
| 4 | 6 | 4 | 1 | 1 | 18 | 1 | 1 |
| 5 | 5 | 4 | 1 | 1 | 6 | 1 | 1 |
| 6 | 5 | 1 | 1 | 1 | 6 | 1 | 1 |
| 7 | 1 | 1 | 1 | 1 | 6 | 1 | 1 |
| 8 | 1 | 1 | 1 | 1 | 8 | 1 | 1 |
| 9 | 2 | 1 | 6 | 3 | 1 | 1 | 2 |
| 10 | 2 | 1 | 6 | 3 | 2 | 1 | 2 |
| 11 | 2 | 9 | 6 | 3 | 2 | 1 | 2 |
| 12 | 3 | 9 | 6 | 3 | 2 | 1 | 2 |
| 13 | 3 | 12 | 6 | 3 | 2 | 1 | 2 |
| 14 | 3 | 1 | 6 | 3 | 2 | 1 | 2 |
| 15 | 3 | 1 | 1 | 3 | 2 | 1 | 2 |
| 16 | 3 | 1 | 6 | 3 | 2 | 2 | 2 |
| 17 | 3 | 1 | 6 | 1 | 2 | 1 | 2 |
| 18 | 3 | 1 | 1 | 1 | 2 | 1 | 2 |
| 19 | 3 | 1 | 1 | 1 | 2 | 4 | 2 |
| 20 | 3 | 1 | 1 | 6 | 2 | 4 | 2 |
| 21 | 3 | 1 | 6 | 3 | 2 | 1 | 1 |
| 22 | 3 | 1 | 6 | 3 | 2 | 1 | 11 |
| 23 | 3 | 1 | 6 | 3 | 16 | 1 | 2 |
| 24 | 2 | 1 | 6 | 3 | 2 | 1 | 5 |
| 25 | 2 | 10 | 6 | 3 | 2 | 1 | 2 |
| 26 | 20 | 9 | 6 | 3 | 2 | 1 | 2 |
| 27 | 3 | 3 | 6 | 3 | 2 | 1 | 2 |
| 28 | 1 | 3 | 6 | 3 | 2 | 1 | 2 |
| 29 | 7 | 3 | 6 | 3 | 2 | 1 | 2 |
| 30 | 3 | 3 | 6 | 3 | 5 | 4 | 2 |
| 31 | 26 | 3 | 6 | 3 | 5 | 4 | 2 |
| 32 | 1 | 3 | 6 | 3 | 5 | 4 | 2 |
| 33 | 9 | 1 | 11 | 3 | 2 | 1 | 2 |
| 34 | 9 | 1 | 4 | 3 | 2 | 1 | 2 |
| 35 | 14 | 1 | 11 | 3 | 2 | 1 | 2 |
| 36 | 5 | 9 | 6 | 3 | 37 | 1 | 2 |
| 37 | 1 | 9 | 20 | 3 | 2 | 1 | 2 |
| 38 | 3 | 9 | 20 | 12 | 2 | 1 | 2 |
| 39 | 5 | 1 | 31 | 3 | 2 | 1 | 2 |
| 40 | 2 | 1 | 4 | 2 | 1 | 9 | 2 |
| 41 | 1 | 1 | 4 | 2 | 1 | 9 | 2 |
| 42 | 2 | 1 | 29 | 2 | 1 | 9 | 2 |
| 43 | 2 | 8 | 4 | 2 | 1 | 9 | 2 |
| 44 | 2 | 1 | 4 | 1 | 1 | 9 | 2 |
| 45 | 2 | 1 | 4 | 1 | 1 | 2 | 2 |
| 46 | 2 | 3 | 4 | 1 | 1 | 9 | 2 |
| 47 | 7 | 1 | 4 | 6 | 1 | 9 | 2 |
| 48 | 9 | 1 | 1 | 1 | 5 | 1 | 1 |
| 49 | 9 | 1 | 1 | 1 | 5 | 4 | 1 |
| 50 | 2 | 2 | 1 | 1 | 6 | 1 | 1 |
| 51 | 3 | 4 | 19 | 1 | 6 | 1 | 1 |
| 52 | 3 | 4 | 22 | 1 | 6 | 1 | 1 |
| 53 | 6 | 1 | 1 | 1 | 6 | 1 | 1 |
| 54 | 6 | 3 | 1 | 1 | 6 | 1 | 1 |
| 55 | 6 | 3 | 1 | 6 | 6 | 1 | 1 |
| 56 | 6 | 3 | 1 | 1 | 29 | 1 | 1 |
| 57 | 6 | 3 | 1 | 1 | 18 | 1 | 1 |
| 58 | 1 | 1 | 10 | 4 | 10 | 3 | 3 |
| 59 | 1 | 1 | 10 | 4 | 10 | 4 | 3 |
| 60 | 6 | 1 | 10 | 4 | 10 | 4 | 3 |
| 61 | 1 | 10 | 21 | 1 | 2 | 4 | 3 |
| 63 | 13 | 1 | 23 | 1 | 4 | 3 | 3 |
| 64 | 13 | 1 | 23 | 10 | 4 | 3 | 3 |
| 65 | 13 | 1 | 11 | 1 | 32 | 3 | 3 |
| 66 | 13 | 1 | 17 | 3 | 28 | 3 | 3 |
| 67 | 4 | 1 | 17 | 3 | 28 | 3 | 3 |
| 68 | 4 | 3 | 17 | 1 | 1 | 3 | 4 |
| 69 | 1 | 3 | 17 | 1 | 1 | 3 | 4 |
| 70 | 4 | 3 | 4 | 1 | 1 | 3 | 4 |
| 71 | 11 | 3 | 17 | 1 | 1 | 3 | 4 |
| 72 | 3 | 3 | 17 | 1 | 16 | 1 | 4 |
| 73 | 5 | 1 | 16 | 3 | 10 | 2 | 3 |
| 74 | 5 | 1 | 16 | 3 | 10 | 1 | 3 |
| 75 | 24 | 1 | 16 | 3 | 10 | 2 | 3 |
| 76 | 3 | 3 | 10 | 3 | 4 | 3 | 1 |
| 77 | 3 | 3 | 27 | 3 | 4 | 3 | 1 |
| 78 | 4 | 3 | 10 | 6 | 1 | 3 | 3 |
| 79 | 4 | 7 | 10 | 6 | 1 | 3 | 3 |
| 80 | 11 | 3 | 10 | 6 | 1 | 3 | 3 |
| 81 | 9 | 3 | 10 | 6 | 4 | 3 | 3 |
| 82 | 7 | 3 | 10 | 6 | 13 | 3 | 3 |
| 83 | 5 | 3 | 10 | 3 | 4 | 3 | 3 |
| 84 | 7 | 3 | 1 | 6 | 4 | 1 | 3 |
| 85 | 7 | 3 | 10 | 6 | 33 | 3 | 3 |
| 86 | 2 | 1 | 25 | 1 | 30 | 3 | 6 |
| 87 | 2 | 1 | 26 | 1 | 30 | 3 | 6 |
| 88 | 3 | 1 | 25 | 1 | 30 | 3 | 6 |
| 89 | 1 | 1 | 16 | 6 | 18 | 3 | 7 |
| 90 | 1 | 1 | 16 | 1 | 18 | 3 | 7 |
| 91 | 12 | 9 | 11 | 3 | 15 | 11 | 1 |
| 92 | 12 | 9 | 1 | 9 | 5 | 11 | 1 |
| 93 | 1 | 1 | 14 | 8 | 19 | 3 | 3 |
| 94 | 1 | 1 | 14 | 8 | 19 | 4 | 3 |
| 95 | 4 | 1 | 8 | 6 | 5 | 1 | 6 |
| 96 | 4 | 1 | 21 | 6 | 5 | 1 | 6 |
| 97 | 1 | 9 | 24 | 1 | 1 | 9 | 1 |
| 98 | 2 | 9 | 24 | 1 | 5 | 9 | 1 |
| 99 | 7 | 1 | 34 | 6 | 18 | 3 | 3 |
| 101 | 16 | 3 | 32 | 1 | 5 | 1 | 2 |
| 102 | 2 | 1 | 17 | 4 | 1 | 1 | 2 |
| 103 | 1 | 1 | 24 | 2 | 38 | 16 | 3 |
| 104 | 14 | 11 | 1 | 6 | 13 | 2 | 1 |
| 105 | 1 | 1 | 10 | 9 | 10 | 10 | 4 |
| 106 | 15 | 3 | 22 | 1 | 4 | 4 | 3 |
| 107 | 2 | 1 | 22 | 1 | 12 | 1 | 2 |
| 109 | 1 | 4 | 16 | 1 | 5 | 1 | 3 |
| 110 | 2 | 3 | 11 | 1 | 2 | 3 | 7 |
| 111 | 5 | 3 | 22 | 6 | 4 | 12 | 3 |
| 113 | 16 | 1 | 1 | 1 | 16 | 1 | 3 |
| 114 | 2 | 9 | 6 | 11 | 2 | 3 | 3 |
| 115 | 6 | 1 | 1 | 3 | 1 | 1 | 1 |
| 116 | 9 | 1 | 14 | 1 | 5 | 1 | 3 |
| 117 | 23 | 9 | 10 | 1 | 1 | 10 | 2 |
| 118 | 4 | 1 | 1 | 1 | 16 | 10 | 1 |
| 119 | 20 | 9 | 1 | 4 | 50X | 15 | 3 |
| 120 | 22 | 1 | 17 | 1 | 40 | 4 | 3 |
| 122 | 21 | 1 | 1 | 1 | 39 | 10 | 3 |
| 123 | 6 | 1 | 1 | 1 | 6 | 3 | 2 |
| 124 | 3 | 3 | 1 | 1 | 6 | 3 | 10 |
| 125 | 9 | 10 | 12 | 6 | 4 | 3 | 3 |
| 126 | 4 | 3 | 1 | 3 | 8 | 1 | 3 |
| 128 | 3 | 1 | 1 | 1 | 12 | 3 | 5 |
| 129 | 1 | 1 | 4 | 6 | 5 | 3 | 3 |
| 130 | 13 | 1 | 30 | 1 | 8 | 1 | 2 |
| 131 | 25 | 3 | 12 | 1 | 12 | 1 | 2 |
| 132 | 1 | 3 | 4 | 3 | 8 | 1 | 4 |
| 133 | 9 | 1 | 33 | 1 | 5 | 1 | 3 |
| 135 | 2 | 3 | 6 | 3 | 5 | 4 | 2 |
| 136 | 9 | 9 | 17 | 4 | 5 | 1 | 3 |
| 138 | 11 | 1 | 18 | 4 | 1 | 3 | 3 |
| 142 | 17 | 3 | 25 | 3 | 25 | 3 | 2 |
FIG. 1.
Evolutionary relationships among clones of serogroup 6. CCs are defined as groups where each member's ST is identical at five of seven loci to at least one other ST in the group. The numbers inside the circles are the designation given to each unique ST. Sizes of the circles are proportional to the number of isolates in that ST. White circles represent serotype 6A, and black circles represent serotype 6B. Underlined numbers are STs from multiple countries. Outlined in grey are STs of drug-resistant reference clones. Length of branch indicate the number of allelic differences between STs. Short branches are SLVs, and long branches are DLVs. Dotted branches indicate ambiguity of relationship between the STs. This occurs when an ST defines a secondary ancestral ST but is not a SLV of another ancestral ST. Major CCs are named for the ST of the presumed ancestral ST of that group. Minor CCs are not named since the ancestors are uncertain. Singletons differ from other STs at three or more loci. A scale is given that indicates the number of isolates in an ST by the diameter of the circle.
A population of clonal complexes.
Attempts to reconstruct the evolutionary history of clones within a freely recombining population can be problematic. With data from such a population, algorithms that depict evolutionary history as a bifurcating tree can lead to poorly resolved trees, and algorithms that permit a reticulate evolutionary history can lead to overly complex networks. Both tree and network-based algorithms using allelic or nucleotide data generally yielded statistical support for small clusters of four or fewer STs (data not shown). We therefore used the BURST algorithm (6), which attempts to represent the relationships among closely related clones while ignoring the relationships among the more distantly related clones. Clonal complexes (CCs) were defined as groups where each member ST was identical at five of seven loci to at least one other ST in the group. Ancestral or founding STs for each CC were hypothesized to be those with the most single locus variants (SLVs), or in the case of a tie, double locus variants (DLVs), of any other ST in the group (5, 10).
We identified 8 major CCs within which an ancestral ST could be hypothesized, 6 minor CCs where the ancestor could not be confidently hypothesized, and 32 individual STs that were apparently unrelated to other STs in the data set (Fig. 1). In five of the eight major CCs, the hypothesized ancestral ST was also the most frequently occurring ST. The BURST analysis identified three ambiguities, all within CC14, where an ST defined a secondary ancestral ST but did not represent a SLV of another ancestral ST. A link was inferred between STs 27 and 30 because ST27 was the only SLV or DLV of ST30. A link could be inferred between ST18 and either ST 15 or 17, since both were SLVs of ST18. ST33 was a DLV of STs 10, 14, 15, and 39, so no link was inferred.
Of note, CCs 14, 40, and 1 matched exactly the three major lineages (called A, B, and C, respectively, in Fig. 3A of reference 26) identified from a population of invasive serotype 6B isolates in Alaska. This suggested that the CCs identified in this study were representative of the major serogroup 6 lineages worldwide.
As with individual clones, CCs were predominately of a single serotype and were often isolated from multiple countries. In two cases, those of CC40 and CC68, there was question as to the ancestral serotype of the groups (see Fig. 1). CC40 had an ancestral ST of mixed serotype. CC68 had an ancestral ST that expressed serotype 6B and three descendant STs that all expressed serotype 6A. Based on a frequency criterion of ancestry (5, 10), we suggest that the ancestor of CC40 expressed serotype 6A and that the ancestor of CC68 expressed serotype 6B. Most CCs have spread to multiple continents, but CC86 was isolated only from South Africa, and the lineage beginning with ST27 in CC14 was isolated only from South America.
Mechanisms of evolution.
To estimate the contributions of recombination and mutation to the divergence of clones, we used a method advocated by Guttman and Dykhuizen (12) and detailed by Feil et al. (10). This method examines the nucleotide differences between a hypothesized ancestor and those clones differing from the ancestor at one of seven loci (SLVs). Alleles that differ from the ancestor at multiple nucleotide sites are likely the result of recombination; it is unlikely that one housekeeping locus will experience multiple mutations while six other housekeeping loci experience no mutations. Alleles that differ from the ancestor at a single nucleotide site can be the result of recombination or mutation. To distinguish between these possibilities, the presence of the allele elsewhere in the data set is determined. The data set represents the sampled portion of the pneumococcal gene pool and includes alleles from the 212 isolates of serogroup 6 plus the 35 isolates of 11 other serogroups. Alleles differing from the ancestral allele at a single nucleotide site but shared with unrelated clones (i.e., a clone in another CC) are likely to be the result of recombination; it is more parsimonious to assume a single recombination event has imported the allele into the clone rather than to assume that multiple mutation events in unrelated clones have led to the same allele. Alleles differing from the ancestral allele at a single site but unique to a clone are considered to be the result of point mutation.
Our data set included a total of 55 pairs in which an ancestral ST was associated with an SLV (i.e., ancestor-SLV pairs). Remarkably, 5 of 14 pairs with serotype 6A ancestors were mutation events, but only 1 of 41 pairs with serotype 6B ancestors were mutation events (Table 2; P = 0.0029). This finding is particularly notable given that we analyzed nearly twice as many serotype 6B isolates as 6A isolates, which would provide more opportunities to sample mutation events from serotype 6B. The recombination-to-mutation ratio estimates the probability that an individual allele would change as a result of recombination versus mutation. The ratio for serotype 6A was 1.8 (9/5), and that for serotype 6B was 40 (40/1). Thus, serotype 6A clones evolved by mutation nearly as often as by recombination, whereas serotype 6B clones evolved almost exclusively by recombination. The ratio for all of serogroup 6 was 8.2 (49/6), which was similar to the 9.9 estimated previously for the species as a whole (10).
TABLE 2.
Evolution of alleles and nucleotides by recombination and mutation events
| Ancestral serotype→SLV serotype | No. of alleles that differ by 1 nucleotidea
|
No. of alleles that differ by >1 nucleotidea
|
||
|---|---|---|---|---|
| Uniqueb | Sharedc | Uniquec | Sharedc | |
| 6A→6A | 3 (3) | 2 (2) | 1 (2) | 3 (7) |
| 6A→6B | 2 (2) | 1 (1) | 0 | 2 (5) |
| 6B→6B | 1 (1) | 8 (8) | 3 (26) | 25 (81) |
| 6B→6A | 0 (0) | 0 (0) | 0 | 4 (13) |
Numbers in parentheses indicate number of nucleotide changes.
Mutation.
Recombination.
The per-site recombination-to-mutation ratio estimates the probability that an individual nucleotide site would change by recombination versus mutation. It is calculated by counting the number of nucleotide differences introduced by a recombination event versus a mutation event (Table 2). The per-site ratio for serotype 6A was 3.4 (17/5), and for serotype 6B it was 128 (128/1). The ratio for all of serogroup 6 was 24.2 (145/6), which was less than the ∼50 estimated previously for the species as a whole (10).
For nine of 55 ancestor-SLV pairs, the serotypes of the ancestral and descendant STs differed (6A→6B and 6B→6A in Table 2). In these nine cases, we inferred the serotype of the ancestor based on a frequency criterion (5, 10). It could be that these nine descendents evolved from an unsampled ancestor of like serotype. During our calculations, if we allowed the serotype of the ancestor to be the same as the descendant, even if no such ancestor was sampled, there remained a significant difference (P = 0.03) between the number of recombination and mutation events of the serotypes. The statistical difference also remained (P = 0.02) if we removed these nine pairs from the analysis.
The recombination-to-mutation ratio is sensitive to the diversity of strains included in the database. A unique allele that occurs in a clone that differs by a single nucleotide from another allele could have arisen as a mutation event or it could be an artifact of an incomplete database. A comparison of the proportion of unique versus shared alleles from our database of 247 isolates with the Enright and Spratt (8) database of 575 isolates makes it clear that our database is sufficiently diverse. Feil et al. (10) reported that of those alleles differing by a single nucleotide, 50% were unique to a given clone and 50% were shared by unrelated clones. In this study of serogroup 6, 35% (6 of 17) of alleles differing by single nucleotides were unique to a given clone and 65% (11 of 17) were shared by unrelated clones (Table 2). Thus, our database, which included a large number of carriage isolates, yielded a 15% greater recovery of shared alleles than the database of predominately invasive isolates typed by the Enright and Spratt (8) MLST system.
These findings raised the question of whether serotype 6A clones were transformation deficient or whether serotype 6B clones were hypertransformable. To investigate these possibilities, we performed spontaneous transformation experiments with 1 isolate each from 51 different STs spread among the 8 major CCs. There was no association between serotype and the number of spontaneously transformable STs; 10 of 19 STs of serotype 6A were spontaneously transformable, and 20 of 32 STs of serotype 6B were spontaneously transformable. Moreover, there was no difference in the transformation efficiencies of the two serotypes (data not shown). Either our assay did not represent the transformation process in vivo, or other processes besides transformation were involved in shaping the recombination-to-mutation ratio in this serogroup.
Relationships between drug-resistant clones.
To examine the origins of drug-resistant clones, we analyzed reference clones from Alaska (ST13) (27), Finland (ST16) (29), Germany (ST89) (23), Hungary (ST50) (23), Maryland (ST22) (11), South Africa (ST39) (30), and Spain (ST19) (32). Surprisingly, five of these seven clones were members of CC14 (outlined in grey in Fig. 1). Consider that the average number of different alleles for these five clones was three and that the number of different alleles we used to establish membership in a CC was two. Thus, our population genetic framework revealed relationships between these clones that would not have been suspected if the alleles of the clones were simply compared.
Evolving more-virulent clones.
The proportions of carriage isolates and invasive isolates differed between the 8 ancestors of the major CCs and their 30 SLVs (Table 3), (P = 0.001). The ancestral STs were more frequently carriage isolates, and their immediate-descendant STs were more frequently invasive isolates. This pattern was also reflected in a comparison of the 8 ancestors of the major CCs and all 78 other STs in the major CCs (Table 3) (P = 0.0003). These data indicated that the ancestors of the CCs in serogroup 6 pneumococci were better adapted for carriage and evolved into more-virulent clones rather than vice versa.
TABLE 3.
Distribution of isolates from carriage and invasive disease
| Group and comparison(s) | No. of carriage isolatesa | No. of invasive isolatesa | P valueb |
|---|---|---|---|
| Ancestral STs of major CCs | 20 (19) | 5 (5) | |
| vs. SLVs of ancestors | 16 (9) | 26 (11) | 0.001 (0.029) |
| vs. All STs other than ancestors | 38 (19) | 60 (23) | 0.0003 (0.0097) |
| Serotype 6A | 41 | 16 | |
| vs. Serotype 6B | 44 | 70 | <0.0001 |
| CC14 | 20 | 36 | |
| vs. 7 other major CCs | 38 | 28 | 0.019 |
Numbers in parentheses exclude data from clonal complex 14.
Based on Fisher's exact test.
The proportion of carriage and invasive isolates also differed by serotype and by CC. Members of serotype 6A were more frequently carriage isolates, whereas serotype 6B isolates were more frequently invasive isolates (Table 3) (P < 0.0001). CC14, which was noted above as including several drug-resistant clones, was represented by a higher proportion of invasive isolates than the other seven major CCs (Table 3) (P = 0.019). Since CC14 was associated with invasive disease and it was the most diverse CC, it could possibly account for the association between the virulence and evolutionary position of clones. After removing the data for CC14, however, there remained a statistically significant difference in the proportion of carriage and invasive isolates among the ancestors of the major CCs and either their immediate SLVs (Table 3) (P = 0.0285) or all remaining STs in the major CCs (Table 3) (P = 0.0097).
It is known that different strains of serogroup 6 can differ in virulence in mouse models of infection (Table 1 of reference 1), but the relationship between genetic background and mouse virulence has not been explored previously. The virulence of 15 STs representing 6 of the 8 major CCs was tested using i.p. infection of BALB/cByJ mice and monitoring mouse survival over 21 days (Table 4). An attempt was made to inoculate all mice with ∼7 × 106 CFU. Although there was variation in the actual inoculation dose, it showed no association with the survival time of the infected mice. Four of the STs were not virulent (median number of days to death, >21), whereas 11 STs each killed five of five mice in a median of 1 to 6 days (Table 4).
TABLE 4.
Virulence of strains in a BALB/cByJ i.p. infection model
| CC and ST(s)a | Strain | Serotype | No. of CFU injected | Median no. of days to deathb |
|---|---|---|---|---|
| CC1 | ||||
| ST1* | AGL6 | 6B | 10.75 × 106 | 3 |
| ST48 | AGC29 | 6B | 10.95 × 106 | 1 |
| ST52 | 703 | 6B | 6.32 × 106 | 2 |
| CC14 | ||||
| ST10 | ANZI2 | 6B | 7.5 × 106 | 4 |
| ST14* | V53 | 6B | 7.35 × 106 | >21 |
| ST19 | SP194 | 6B | 9.87 × 106 | >21 |
| ST30 | CL234 | 6B | 11.54 × 106 | 6 |
| ST33 | ASA5 | 6B | 9.68 × 106 | 5 |
| CC40 | ||||
| ST40* | ASW10 | 6A | 8.34 × 106 | >21 |
| ST44 | 2492 | 6B | 7.01 × 106 | 3 |
| CC59 | ||||
| ST59* | BG7322A | 6A | 7.42 × 106 | 3 |
| ST60 | AICC5 | 6A | 8.74 × 106 | 2 |
| CC68 | ||||
| ST68* | ABU441 | 6B | 5.25 × 106 | >21 |
| ST69 | AKU38 | 6A | 8.11 × 106 | 1 |
| CC78 | ||||
| ST78* | AISC6 | 6A | 7.86 × 106 | 6 |
*, ancestral clone. Each CC is named for the ST of its ancestor.
Results are based on five mice per strain.
There was no association between virulence in mice and whether the STs were isolated from carriage or invasive disease in humans. This probably reflects differences in the host environment of mice and humans and confirms a previous observation that the virulence of serotypes 6A and 6B in mice appears to show no relationship to whether the isolates were collected from carriage or invasive disease in humans (1). Nonetheless, in each of the CCs where two or more STs were examined, the ancestral ST was less virulent than the other STs (Table 4). This trend was most obvious in CCs 14, 40, and 68. The mouse virulence data was therefore consistent with the epidemiological data in this regard and indicated that the descendant STs have become more virulent than their ancestral STs.
DISCUSSION
These studies have demonstrated that patterns, mechanisms, and various consequences of evolution can be resolved within a pathogenic species of bacteria despite evidence of genetic recombination within the species. A necessary first step was the identification of clones by MLST and the inference of their evolutionary relationships using a novel parsimony analysis, called BURST analysis. Eight major complexes of clonally related bacteria were identified along with their putative ancestors. Like phylogenetic trees and networks, clonal complexes should be viewed as hypotheses of relationship, subject to testing and refinement by the sequencing of additional isolates and loci.
Natural populations of bacteria can exhibit different genetic structures that are largely influenced by the role of recombination versus mutation during the evolution of clones (9). Since recombination is unlinked to reproduction in bacteria, there is the potential for large differences in the rates of recombination within and between species. S. pneumoniae is naturally competent for genetic transformation, and population genetic evidence for recombination between clones has been presented (8, 13, 17, 20). It was previously suggested that pneumococcal clones evolved predominately by recombination (10). Using new loci and isolates from nasopharyngeal carriage as well as from invasive disease, we found the species estimates of recombination to be representative of serogroup 6 (although the per-site estimates for the species were greater than for serogroup 6), but the estimates differed significantly between serotypes 6A and 6B. Both recombination and mutation were found to have important roles in the evolution of serotype 6A clones, whereas recombination was the dominant mechanism of evolution for serotype 6B clones.
What could account for the different population structures of these serotypes? We found no differences in their spontaneous transformability. However, serotype 6A isolates were more frequently isolated from the carriage state, and serotype 6B isolates were more frequently isolated from invasive disease in our sample. This finding is in agreement with data suggesting that isolates of serotype 6B cause serious invasive disease in children more frequently than do those of serotype 6A (14, 24). At least 70% of our sample was from children ⩽5 years old. Thus, we cannot discount the possibility that a causal relationship exists between the epidemiology of these serotypes and their genetic structure.
The observation that five of seven drug-resistant clones belonged to the same clonal complex (CC14) could indicate that certain genetic backgrounds have a propensity to develop drug resistance. A role for the fib loci for penicillin-resistant S. pneumoniae (36) and the homologous fem loci for methicillin-resistant S. aureus (3) illustrates that genetic background can influence β-lactam resistance. If other forms of drug resistance can also be influenced by auxiliary loci, then an understanding of genetic background could be important to the control of drug resistance in natural populations. Alternatively, it could be that since clones of CC14 are more often invasive than clones of other CCs (Table 3), they are more often exposed to antibiotics during treatment, and therefore, selection has led to increased drug resistance. A thorough study of drug resistance would be needed to distinguish between these hypotheses.
The origin of the multidrug-resistant Spain6B-2 clone (ST19) deserves special attention as it was the most prevalent serogroup 6 clone worldwide and is of clinical concern (7). It has been suggested that a penicillin-susceptible but erythromycin- and clindamycin-resistant serotype 6B clone from Greece is the ancestor of the Spain6B-2 clone (33). The evidence was the similarity of the two clones based on restriction fragment end labeling and the assumption that resistant clones evolve from sensitive ones but not vice versa. An alternative hypothesis was that some members of the Spain6B-2 clone lost resistance to penicillin, which resulted in the Greek clone (33). Recently it was noted that the Spain6B-2 clone has indeed lost resistance to multiple antibiotics since its introduction into Iceland in the late 1980s (35). In this study, we observed one serotype 6B isolate from Greece that was penicillin and levofloxacin susceptible but erythromycin, clindamycin, and tetracycline resistant, and this isolate was identical to the Spain6B-2 clone at seven of seven loci. We obtained susceptibility data for these five antibiotics for 8 of the 17 isolates from the Spain6B-2 clone, and four different susceptibility patterns were observed (data not shown). Thus, variable susceptibilities in this clone may be common. The branching patterns of CC14 suggested that the Spain6B-2 clone evolved from ST18 (Fig. 1), which was tetracycline resistant and isolated only once from Uruguay.
The factors that dictate the relative virulence of pneumococcal strains are still rather poorly understood. While capsule is a critical virulence factor, several recent studies using signature-tagged mutagenesis to identify mutants affecting virulence indicate that a large number of gene loci can modulate virulence (16, 21). This fact, in combination with the extensive use of recombination in evolution as shown in this study and by others (10), makes it difficult to know the exact loci responsible for differences in virulence between ancestral STs and descendant STs. This could be explored in future studies involving this group of strains. Although capsular serotype was different in two or more of the ancestor/descendant pairs tested in the mouse virulence model, the more virulent member of the pair was 6A in one case (CC68) and 6B in another (CC40). Furthermore, increased virulence occurred in some descendant lines of CC14, when both ancestors and descendents were of the same serotype. The one exception was ST19 in CC14, of which the multidrug-resistant clone Spain6B-2 is a member. This ST was the most frequently and widely isolated clone in our sample. It is tempting to speculate that this descendant line may have acquired drug resistance rather than factors increasing virulence and that its frequent occurrence in our sample was an effect of antibiotic usage, but this hypothesis cannot be tested from this study design.
S. pneumoniae normally behaves as a commensal bacterium being carried in an asymptomatic fashion, but it is capable of causing severe disease and death when the bacteria reach the blood or other normally sterile tissue sites. We previously highlighted the importance of including both carriage and invasive isolates in studies of pneumococcal population biology, by showing differences in the genotypic diversity of these isolates in a cross-sectional sample of serotypes from a limited geographic region (25). Here, using epidemiological and mouse virulence data on global isolates of serogroup 6, we report a trend towards the evolution of more-virulent clones from less-virulent ancestors. Thus, it is clear that carriage isolates can have a prominent role in the genetic structure of natural pneumococcal populations. The long-term treatment and prevention of pneumococcal disease might therefore require the carriage population as well as those clones currently causing disease to be targeted for eradication.
Acknowledgments
We thank Metin Punar for his assistance in checking the data and Ed Feil for reviewing the manuscript. This study would not have been possible without the generous donation of isolates from the following individuals: Peter Applebaum, Gabriela Ech niz Aviles, Cristina Brandileone, Teresa Camou, Rose Capeding, Elizabeth Castañeda, Ron Dagan, Kathryn Edwards, Regine Hakenbeck, Ingrid Hethman, Preben Homøe, Maria Hortal, Molly Johny, Jim Kellner, Kyung Hyo Kim, Keith Klugman, Karl Kristinsson, Göran Kronvall, Myron Levine, Marguerite Lovgren, Diana Martin, Denise Murphy, Alan Parkinson, Mabel Regueira, Alicia Rossi, George Syrogiannopoulos, Maria Claudia Vela, and Malai Vorachit.
These studies were supported in part by NIH grants (AI-21548, AI-40645, and HL-58418) and a WHO grant (V23/181/47). The DNA Sequencing Core Facilities were supported by a grant from NIH to the Center for Aids Research (AI-27767), by a grant from the Tennessee Valley Authority to the Department of Microbiology at the University of Alabama at Birmingham, and by the Howard Hughes Foundation to the University of Alabama at Birmingham Medical School.
REFERENCES
- 1.Briles, D. E., M. J. Crain, B. M. Gray, C. Forman, and J. Yother. 1992. A strong association between capsular type and mouse virulence among human isolates of Streptococcus pneumoniae. Infect. Immun. 60:111-116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Carlisle, J. B., M. Gratten, and A. J. Leach. 2001. Molecular epidemiology of multiple drug resistant type 6B Streptococcus pneumoniae in the Northern Territory and Queensland, Australia. Epidemiol. Infect. 126:25-29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chambers, H. F. 1999. Penicillin-binding protein-mediated resistance in pneumococci and staphylococci. J. Infect. Dis. 179(Suppl. 2):S353-S359. [DOI] [PubMed] [Google Scholar]
- 4.Crain, M. J., J. S. Turner, D. A. Robinson, T. J. Coffey, A. Brooks-Walter, L. M. McDaniel, and D. E. Briles. 1996. Evidence for the simultaneous expression of two PspAs by a clone of capsular serotype 6B Streptococcus pneumoniae. Microb. Pathog. 21:265-275. [DOI] [PubMed] [Google Scholar]
- 5.Crandall, K. A., and A. R. Templeton. 1993. Empirical tests of some predictions from coalescent theory with applications to intraspecific phylogeny reconstruction. Genetics 134:959-969. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Day, N. P., C. E. Moore, M. C. Enright, A. R. Berendt, J. M. Smith, M. F. Murphy, S. J. Peacock, B. G. Spratt, and E. J. Feil. 2001. A link between virulence and ecological abundance in natural populations of Staphylococcus aureus. Science 292:114-116. [DOI] [PubMed] [Google Scholar]
- 7.Enright, M. C., A. Fenoll, D. Griffiths, and B. G. Spratt. 1999. The three major Spanish clones of penicillin-resistant Streptococcus pneumoniae are the most common clones recovered in recent cases of meningitis in Spain. J. Clin. Microbiol. 37:3210-3216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Enright, M. C., and B. G. Spratt. 1998. A multilocus sequence typing scheme for Streptococcus pneumoniae: identification of clones associated with serious invasive disease. Microbiology 144:3049-3060. [DOI] [PubMed] [Google Scholar]
- 9.Feil, E. J., E. C. Holmes, D. E. Bessen, M. S. Chan, N. P. Day, M. C. Enright, R. Goldstein, D. W. Hood, A. Kalia, C. E. Moore, J. Zhou, and B. G. Spratt. 2001. Recombination within natural populations of pathogenic bacteria: short-term empirical estimates and long-term phylogenetic consequences. Proc. Natl. Acad. Sci. USA 98:182-187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Feil, E. J., J. M. Smith, M. C. Enright, and B. G. Spratt. 2000. Estimating recombinational parameters in Streptococcus pneumoniae from multilocus sequence typing data. Genetics 154:1439-1450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gherardi, G., C. G. Whitney, R. R. Facklam, and B. Beall. 2000. Major related sets of antibiotic-resistant pneumococci in the United States as determined by pulsed-field gel electrophoresis and pbp1a-pbp2b-pbp2x-dhf restriction profiles. J. Infect. Dis. 181:216-229. [DOI] [PubMed] [Google Scholar]
- 12.Guttman, D. S., and D. E. Dykhuizen. 1994. Clonal divergence in Escherichia coli as a result of recombination, not mutation. Science 266:1380-1383. [DOI] [PubMed] [Google Scholar]
- 13.Hall, L. M. C., R. A. Whiley, B. Duke, R. C. George, and A. Efstratiou. 1996. Genetic relatedness within and between serotypes of Streptococcus pneumoniae from the United Kingdom: analysis of multilocus enzyme electrophoresis, pulsed-field gel electrophoresis, and antimicrobial resistance patterns. J. Clin. Microbiol. 34:853-859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hausdorff, W. P., J. Bryant, P. R. Paradiso, and G. R. Siber. 2000. Which pneumococcal serogroups cause the most invasive disease: implications for conjugate vaccine formulation and use, part I. Clin. Infect. Dis. 30:100-121. [DOI] [PubMed] [Google Scholar]
- 15.Klein, D. L. 2000. Pneumococcal disease and the role of conjugate vaccines, p. 467-477. In A. Tomasz (ed.), Streptococcus pneumoniae: molecular biology and mechanisms of disease. Mary Ann Liebert, Inc., Larchmont, N.Y.
- 16.Lau, G. W., S. Haataja, M. Lonetto, S. E. Kensit, A. Marra, A. P. Bryant, D. McDevitt, D. A. Morrison, and D. W. Holden. 2001. A functional genomic analysis of type 3 Streptococcus pneumoniae virulence. Mol. Microbiol. 40:555-571. [DOI] [PubMed] [Google Scholar]
- 17.Lomholt, H. 1995. Evidence of recombination and an antigenically diverse immunoglobulin A1 protease among strains of Streptococcus pneumoniae. Infect. Immun. 63:4238-4243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.McGee, L., L. McDougal, J. Zhou, B. G. Spratt, F. C. Tenover, R. George, R. Hakenbeck, W. Hryniewicz, J. C. Lefevre, A. Tomasz, and K. P. Klugman. 2001. Nomenclature of major antimicrobial-resistant clones of Streptococcus pneumoniae defined by the pneumococcal molecular epidemiology network. J. Clin. Microbiol. 39:2565-2571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Mitchell, T. J. 2000. Virulence factors and the pathogenesis of disease caused by Streptococcus pneumoniae. Res. Microbiol. 151:413-419. [DOI] [PubMed] [Google Scholar]
- 20.Müller-Graf, C. D., A. M. Whatmore, S. J. King, K. Trzcinski, A. P. Pickerill, N. Doherty, J. Paul, D. Griffiths, D. Crook, and C. G. Dowson. 1999. Population biology of Streptococcus pneumoniae isolated from oropharyngeal carriage and invasive disease. Microbiology 145:3283-3293. [DOI] [PubMed] [Google Scholar]
- 21.Polissi, A., A. Pontiggia, G. Feger, M. Altieri, H. Mottl, L. Ferrari, and D. Simon. 1998. Large-scale identification of virulence genes from Streptococcus pneumoniae. Infect. Immun. 66:5620-5629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Ramirez, M., D. A. Morrison, and A. Tomasz. 1997. Ubiquitous distribution of the competence related genes comA and comC among isolates of Streptococcus pneumoniae. Microb. Drug Resist. 3:39-52. [DOI] [PubMed] [Google Scholar]
- 23.Reichmann, P., E. Varon, E. Gunther, R. R. Reinert, R. Luttiken, A. Marton, P. Geslin, J. Wagner, and R. Hakenbeck. 1995. Penicillin-resistant Streptococcus pneumoniae in Germany: genetic relationship to clones from other European countries. J. Med. Microbiol. 43:377-385. [DOI] [PubMed] [Google Scholar]
- 24.Robbins, J. B., R. Austrian, C. J. Lee, S. C. Rastogi, G. Schiffman, J. Henrichsen, P. H. Makela, C. V. Broome, R. R. Facklam, R. H. Tiesjema, et al. 1983. Considerations for formulating the second-generation pneumococcal capsular polysaccharide vaccine with emphasis on the cross-reactive types within groups. J. Infect. Dis. 148:1136-1159. [DOI] [PubMed] [Google Scholar]
- 25.Robinson, D. A., K. M. Edwards, K. B. Waites, D. E. Briles, M. J. Crain, and S. K. Hollingshead. 2001. Clones of Streptococcus pneumoniae isolated from nasopharyngeal carriage and invasive disease in young children in central Tennessee. J. Infect. Dis. 183:1501-1507. [DOI] [PubMed] [Google Scholar]
- 26.Robinson, D. A., S. K. Hollingshead, J. M. Musser, A. J. Parkinson, D. E. Briles, and M. J. Crain. 1998. The IS1167 insertion sequence is a phylogenetically informative marker among isolates of serotype 6B Streptococcus pneumoniae. J. Mol. Evol. 47:222-229. [DOI] [PubMed] [Google Scholar]
- 27.Rudolph, K. M., M. J. Crain, A. J. Parkinson, and M. C. Roberts. 1999. Characterization of a multidrug-resistant clone of invasive Streptococcus pneumoniae serotype 6B in Alaska by using pulsed-field gel electrophoresis and PspA serotyping. J. Infect. Dis. 180:1577-1583. [DOI] [PubMed] [Google Scholar]
- 28.Scott, J. A., A. J. Hall, R. Dagan, J. M. Dixon, S. J. Eykyn, A. Fenoll, M. Hortal, L. P. Jette, J. H. Jorgensen, F. Lamothe, C. Latorre, J. T. Macfarlane, D. M. Shlaes, L. E. Smart, and A. Taunay. 1996. Serogroup-specific epidemiology of Streptococcus pneumoniae: associations with age, sex, and geography in 7,000 episodes of invasive disease. Clin. Infect. Dis. 22:973-981. [DOI] [PubMed] [Google Scholar]
- 29.Sibold, C., J. Wang, J. Henrichsen, and R. Hakenbeck. 1992. Genetic relationships of penicillin-susceptible and -resistant Streptococcus pneumoniae strains isolated on different continents. Infect. Immun. 60:4119-4126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Smith, A. M., and K. P. Klugman. 1997. Three predominant clones identified within penicillin-resistant South African isolates of Streptococcus pneumoniae. Microb. Drug Resist. 3:385-389. [DOI] [PubMed] [Google Scholar]
- 31.Sniadack, D. H., B. Schwartz, H. Lipman, J. Bogaerts, J. C. Butler, R. Dagan, G. Echaniz-Aviles, N. Lloyd-Evans, A. Fenoll, N. I. Girgis, J. Henrichsen, K. Klugman, D. Lehmann, A. K. Takala, J. Vandepitte, S. M. Gove, and R. F. Breiman. 1995. Potential interventions for the prevention of childhood pneumonia: geographic and temporal differences in serotype and serogroup distribution of sterile site pneumococcal isolates from children—implications for vaccine strategies. Pediatr. Infect. Dis. 14:503-510. [PubMed] [Google Scholar]
- 32.Soares, S., K. G. Kristinsson, J. M. Musser, and A. Tomasz. 1993. Evidence for the introduction of a multiresistant clone of serotype 6B Streptococcus pneumoniae from Spain to Iceland in the late 1980s. J. Infect. Dis. 168:158-163. [DOI] [PubMed] [Google Scholar]
- 33.Syrogiannopoulos, G. A., D. Bogaert, I. N. Grivea, N. G. Beratis, R. R. De Groot, and P. W. Hermans. 2001. Molecular epidemiology of penicillin-susceptible, multidrug-resistant serotype 6B pneumococci isolated from children in Greece. J. Clin. Microbiol. 39:581-585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Tettelin, H., K. E. Nelson, I. T. Paulsen, J. A. Eisen, T. D. Read, S. Peterson, J. Heidelberg, R. T. DeBoy, D. H. Haft, R. J. Dodson, A. S. Durkin, M. Gwinn, J. F. Kolonay, W. C. Nelson, J. D. Peterson, L. A. Umayam, O. White, S. L. Salzberg, M. R. Lewis, D. Radune, E. Holtzapple, H. Khouri, A. M. Wolf, T. R. Utterback, C. L. Hansen, L. A. McDonald, T. V. Feldblyum, S. Angiuoli, T. Dickinson, E. K. Hickey, I. E. Holt, B. J. Loftus, F. Yang, H. O. Smith, J. C. Venter, B. A. Dougherty, D. A. Morrison, S. K. Hollingshead, and C. M. Fraser. 2001. Complete genome sequence of a virulent isolate of Streptococcus pneumoniae. Science 293:498-506. [DOI] [PubMed] [Google Scholar]
- 35.Vilhelmsson, S. E., A. Tomasz, and K. G. Kristinsson. 2000. Molecular evolution in a multidrug-resistant lineage of Streptococcus pneumoniae: emergence of strains belonging to the serotype 6B Icelandic clone that lost antibiotic resistance traits. J. Clin. Microbiol. 38:1375-1381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Weber, B., K. Ehlert, A. Diehl, P. Reichmann, H. Labischinski, and R. Hakenbeck. 2000. The fib locus in Streptococcus pneumoniae is required for peptidoglycan crosslinking and PBP-mediated beta-lactam resistance. FEMS Microbiol. Lett. 188:81-85. [DOI] [PubMed] [Google Scholar]