Musser et al. Reply to "Emergence of the Same Successful Clade among Distinct Populations of emm89 Streptococcus pyogenes in Multiple Geographic Regions".

REPLY

The gist of the letter by Friaes et al. (1) seems to be that an emerged clone of emm89 Streptococcus pyogenes documented in three peer-reviewed publications (2–4) to be present in the United Kingdom, United States, Iceland, and Finland also may be present in Portugal. This conclusion was reached by analysis of publicly available full-genome sequence data for >900 emm89 strains and comparisons performed with very limited genetic information (data from multilocus sequence type characterization, PCR analysis of the hasABC region, and DNA sequencing of the nga gene and its promoter region) for 95 Portuguese emm89 strains. These Portuguese strains appear to be a convenience sample, rather than population based, an investigative approach that has led to erroneous conclusions, as described in previous studies (4). Unfortunately, the validity of the speculative claim by Friaes et al. (1) cannot be appropriately evaluated for at least three reasons. First, full-genome sequencing of the strains recovered from Portuguese patients was not conducted, which means that genomic relationships among the Portuguese and previously studied strains (2–4) can only be inferred, not precisely determined. Second, in the absence of full-genome data and high-quality closed Portuguese reference genomes, the precise reason for the PCR data presented in the table cannot be fully discerned. Third, the sequence data for the nga promoter region and nga gene for the Portuguese strains were not publicly accessible as of 23 October 2015. Although the presence of the emergent emm89 clone in Portugal may very well be the case, full-genome analyses are required. Moreover, in the absence of a detailed description of the analysis methods used (bioinformatics and otherwise), it is simply not possible to rigorously evaluate their claim. The authors should strive to publish their data in a peer-reviewed journal format. Documenting the presence of this emm89 clone in Portugal would be an important observation if proven true.

The authors claim (1) that deposited genome data of 907 strains met quality standards required in their analysis for both the has and nga region loci. This implies that some deposited data (2, 3, 5) did not meet their quality standards, but the standards used are unfortunately neither described nor validated. Similarly, the claim is made (1) that “multilocus sequence typing alleles (MLST) could be confidently determined for 886 of these strains.” It is not clear if “these” refers to 886 of the 907 strains or 886 of the deposited strains, nor is it clear how “confidently” was defined or validated in this context.

The authors claim (1) that “virulence studies using animal models indicate an increased virulence of the emm89-new strains,” and they cite two publications (3, 5). The goal of the mBio paper by Zhu et al. (5) was to use a panel of isogenic mutant strains to rigorously probe the contribution of variation in the level of expression of secreted toxins SPN (S. pyogenes NADase) and SLO (streptolysin O) and of the hyaluronic acid capsule to virulence in an animal infection model. As we showed (5), the animal infection data suggest that increased production of SPN and SLO is a more important contributor to virulence than hyaluronic acid capsule production in the model used. In Zhu et al. (5), we conclude our manuscript as follows: “The sum of the data provide additional evidence to support the idea we recently put forward that upregulation of SPN and SLO production is a key trigger for epidemic disease caused by S. pyogenes.” In Zhu et al. (3), on the basis of extensive biochemical, pathogenesis, and ex vivo studies of a panel of precisely defined isogenic mutant strains, we state the following: “These data emphasize the critical role that SNPs in the promoter spacer region upstream of the nga, ifs, and slo operon play in enhancing pathogen fitness and dissemination.” We also state (3), “Our data suggest that enhanced production of SPN and SLO creates a pathogen with increased fitness in the upper respiratory tract, resulting in ability to disseminate in a very short time frame and cause abundant human infections in susceptible persons.” We specifically investigated the upper respiratory phenotype in the most relevant animal model available because of the strong possibility that enhanced survival in the upper respiratory tract plays an important role in enhanced person-to-person spread and an increased number of human infections. A linkage between enhanced human upper respiratory tract abundance and increased transmission was reported 70 years ago (6). It is clear that proving that a new strain of a human-specific pathogen is more virulent than an old strain is a very difficult if not impossible task. And, of course, it is reasonable to think—as we and others do—that virulence may include the relative ability of an organism to survive in a host and spread between hosts. In summary, we stand by our assessment that in the aggregate, the peer-reviewed and published data (2–5) provide evidence for upregulation of SPN and SLO production being a key trigger for some epidemics caused by S. pyogenes. It will continue to be important to investigate this idea using many different lines of study and in many different human populations and to add the resulting data to the peer-reviewed literature.

In the present era of rapid and inexpensive bacterial genome sequencing, we encourage analyses that include study of comprehensive population-based rather than convenience samples. Full-genome analyses have rapidly become the gold standard for investigating dissemination of bacterial strains in natural populations (2–4, 7, 8). In addition, comparative studies using isogenic mutant strains and relevant animal infection models and related testing, as commonly done for decades in bacterial pathogenesis research, are critical for assessing virulence and other phenotypic differences among strains. Further, we advocate for depositing all primary sequence data to publicly accessible databases and publishing interpretations thereof in the literature via the conventional peer-review process, rather than alternative means.