Adherence to host cells is the first step in the progression from S. pneumoniae carriage to disease . While adherence to host cells by S. pneumoniae has been shown in a number of studies [10, 12, 17, 20–22, 29, 30], this is the first investigation to compare adherence across a large sample of S. pneumoniae isolates from pneumonia patients. An important conclusion of this study is that S. pneumoniae isolates' ability to adhere to lung cells vary by more than 1,000-fold. Additionally, S. pneumoniae with higher binding activity to host cells have a prolonged activation of the NFκB pathway.
Patients infected with high-binding isolates could be at increased risk compared to patients infected with the more common low-binding isolates. While high-binding S. pneumoniae strains are relatively uncommon among patient isolates, they could still be important clinically. The United States' Centers for Disease Control expects at least 100,000 cases of pneumococcal pneumonia every year in the United States alone . Thus, 2,000 of those patients would be expected to be infected with high-binding strains if the percentages reported here are representative of S. pneumoniae isolates overall. Although this investigation does not compare outcomes of patients infected with high- and low-binding S. pneumoniae, studies are in progress to evaluate this association. These studies require the evaluation of more patients since high-binding bacteria comprise such a small minority (2%) of patient isolates.
There was no correlation between isolate ability to bind to host cells and serotype, although other studies have shown a relationship between pathogenicity and serotype  and invasiveness and serotype . In this study, 18 serotypes were found among 298 patient isolates. While 3 of 6 high-binding isolates were serotype 6B, this capsular type was not statistically linked to increased adherence, since 6B is a common serotype in this sample (n = 39; 13%).
High-binding S. pneumoniae used a bacterial surface protein, CbpA, to bind to C3 on lung epithelial cells, while low-binding strains did not use CbpA or C3. This interaction between CbpA and C3 has been noted by others [21, 22]. CbpA is also capable of binding to other cellular receptors, such as the polymericimmuno globulin receptor (pIgR) on nasopharyngeal epithelial cells [30, 34] and PAFr [17, 29]. In this study blocking host PAFr did not reduce the adherence of high-binders to A549 cells.
Orihuela et al. observed that CbpA is required for nasopharyngeal colonization, progression from the nasopharynx to the lower respiratory tract, and the progression from bacteremia to meningitis in a mouse model but is not needed for S. pneumoniae survival in murine lungs or blood . In contrast, Paton and Berry showed that CbpA does not contribute to S. pneumoniae pathogenicity in mice infected via intraperitoneal injection . Future studies are needed to elucidate the importance of CbpA on virulence in humans infected with S. pneumoniae.
Others have shown that high-binding S. pneumoniae strains have a transparent colony morphology, while colonies of low-binding S. pneumoniae are likely to be opaque [12, 37] Further, it has been observed that transparent S. pneumoniae strains contain more CbpA than do opaque strains  and that individual S. pneumoniae isolates can fluctuate between highly-adherent transparent and less-adherent opaque phases [37, 38]. Phase variation was not noted when comparing high-binding S. pneumoniae isolates with low-binding S. pneumoniae isolates. All clinical strains used exhibited stable binding activity and colony morphology over at least 4 years and after numerous passages from the freezer to agar. Additionally, artificially-evolved high binders had the same colony morphology as did their low-binding progenitors. Thus, in this instance it does not seem that colony morphology was responsible for differences in binding ability of these clinical isolates. It is possible that certain colony morphotypes were selected out in the host or during the infectious process. Those studies would require additional patient isolates and extensive review of patient history. Interestingly, four of the high-binding S. pneumoniae isolates came from patients under the age of 5 years. This observation could imply that high-binding S. pneumoniae using CbpA are more prevalent in children than in adults. This interesting finding requires further study.
CbpA has been considered as a vaccine candidate. CbpA is immunogenic in rabbits and S. pneumoniae human volunteers [39, 40]. Recently, Zhang et al. found that children with higher serum levels of anti-CbpA antibodies had a reduced risk of S. pneumoniae carriage . Thus, CbpA-mediated binding to host cells could be an important contributor to S. pneumoniae pathogenicity and could have implications for the development of vaccines directed against CbpA.
In this study, high-binding S. pneumoniae were observed to strongly activate the NFκB pathway in lung epithelial cells. NFκB activation was independent of serotype and genetic background. Although NFκB activation often leads to secretion of many different cytokines, in this study only IL-8 was released by cells treated with either high- or low-binding S. pneumoniae. Epithelial cells surprisingly produced more IL-8 after 8 hours of incubation with low binding S. pneumoniae than with high binding isolates. This indicates differences in intracellular pathways within the epithelial cells after activation of different receptors by S. pneumoniae. Additionally, the interaction and recruitment of other cell types to the site of infection may be effected by this differential activation. A study by Jones et al. showed that IL-1 and TNF-α were rapidly released by murine lungs in response to S. pneumoniae infection . Thus, although the results from the current study indicate a difference in the initial interaction of S. pneumoniae with host epithelial cells alone, further investigation in the presence of other cell types is needed to understand the implications of this difference in epithelial cell activation.
NFκB activation and IL-8 secretion by pneumocytes recruit neutrophils to the lung [43, 44]. While neutrophils kill bacteria, they also damage lung tissue . It is not clear whether the increased IL-8 secretion in pneumocytes treated with low-binding S. pneumoniae is ultimately protective or destructive. It is also unclear why the increased NFκB activation seen in cells treated with high-binding S. pneumoniae did not result in increased IL-8 secretion.