Pneumococcal surface protein C group 4 of Streptococcus pneumoniae is a significant factor to human invasive pneumococcal diseases

Streptococcus pneumoniae is a major causative pathogen of non-invasive and invasive disease worldwide. Although many virulence factors of S. pneumoniae have been reported, the knowledge of the relationship between these factors and clinical features of pneumococcal infection is limited. To clarify factors leading from non-invasive to invasive disease, we analyzed virulence factor genes of S. pneumoniae isolates from patients and evaluated relationship between presence or absence of the genes and clinical features. Methods Pneumococcal surface protein (Psp)A (pspA and pspA PspC and Pilus-2 virulence factor were measured by PCR using 13 isolates from patients with invasive pneumococcal disease (IPD) and 111 from patients with non-invasive pneumococcal diseases (NIPD) in a community hospital in Japan during 2016. We also tested serotype of isolates. Statistical analysis was performed using multivariable logistic regression analysis for calculating adjusted odds ratio (AOR).


Background
Streptococcus pneumoniae, one of the most prevalent human pathogens worldwide, constitutes a major causative agent of both non-invasive (e.g., otitis media and pneumonia) and invasive disease (e.g., bacteremia and meningitis), leading to the death of a large number of young children and the elderly in particular [1][2][3]. In Japan, pneumococcal conjugate vaccine (PCV13) and pneumococcal polysaccharide vaccine (PPSV23) have been approved for routine vaccination of children and the elderly, respectively.
S. pneumoniae colonises the nasopharynx of many healthy young children asymptomatically [ 4 ]. However, despite S. pneumoniae being one of the most extensively studied microorganisms and the identification of many virulence factors, little is known regarding the factors that contribute to cause of disease and invasion. The pneumococcal surface protein A (PspA), a cell-wall-associated protein, interferes with complement deposition on the bacterial surface [ 5 ] and binds human lactoferrin [ 6 ]. PspA is variable at the DNA level, and can be classified into three families by polymerase chain reaction (PCR) [ 7 ], with PspA family 1 and 2 as the major alleles. PspC, also termed choline binding protein A (CbpA), constitutes a major pneumococcal adhesin. PspC binds to the polymeric immunoglobulin receptor (pIgR) in human nasopharyngeal cells and promotes translocation across a mucosal barrier. PspC also binds secretory IgA (sIgA), platelet activating factor receptor and more specifically complement proteins such as C3 [ 8 ]. PspC can inhibit C3b deposition but also binds the host complement inhibitor factor H, leading to inhibition of alternative pathway activation. Dieudonné-Vatran et al.
demonstrated that the pneumococcus exploits host C4b-binding protein (C4BP) for complement evasion in a PspC group 4-dependent manner [ 9 ]. Pilli have been identified on several gram-positive bacteria; two different pilus islets have been described in S. pneumoniae that encode for two different types of pilli, Pilus-1 and Pilus-2 [ 10 , 11 ]. Pilus-1 is encoded by a pathogenicity islet including genes for three structural proteins, rrgABC, three sortases, srtBCD, and a regulator, rlrA. Pilus-1 was reported to mediate host-bacterial interactions as an adhesin, and a proinflammatory stimulus [ 10 , 12 ]. In particular, an isolate is defined as pilus-1 positive if a PCR for the rrgC gene is positive [ 13 ]. In turn, Pilus-2, detected via primers for the signal peptidase-like protein (SipA) gene [ 14 ], mediates adhesion of S. pneumoniae to eukaryotic cells [ 11 ].
To clarify bacterial factors which foresee invasive diseases, we analysed five virulence factor genes, pspA F1, pspA F2, pspC group 4 ( pspC.4), rrgC, and sipA, of clinical isolates and evaluated the relationship between gene possession and clinical features in this study.

Clinical isolates
A total of 511 clinical S. pneumoniae isolates were collected at Saiseikai Yokohamashi Tobu Hospital, a regional core hospital in Kanagawa prefecture, Japan, during January-December 2016. Those patients from whom multiple pneumococcal isolates were obtained within a 30-day period, they were regarded as a single episode. In the 331 episodes (187 males and 144 females), 124 episodes were confirmed pneumococcal diseases based on both bacteriological isolation and consistent clinical symptom, signs and laboratory findings with S. pneumoniae infection. Table 1 shows the distribution of patient's age, type of specimen, clinical diagnosis, and serotype. Thirteen episodes were diagnosed as invasive pneumococcal diseases (IPD) according to that S. pneumoniae had been isolated from normally sterile body sites in patients. Patients who had been diagnosed with pneumonia as well as bacteremia were designated as having IPD. Thirteen isolates from normally sterile body sites were defined as IPD isolates; 11 isolates from blood, and 2 from cerebrospinal fluid. We defined the other 111 episodes as non-invasive pneumococcal diseases (NIPD), although NIPD might have contained IPD because all episodes did not have blood culture test. The first isolate per NIPD episode subjected to genetic analysis; 97 isolates were from sputum, 10 from bronchial wash, 3 from nasopharyngeal swab, and 1 from otorrhea.

DNA extraction and serotyping
Isolates were cultured on 5% blood agar plates at 37 °C with 5% CO 2 . DNA was extracted using the Cica Geneus DNA Extraction Reagent (Kanto Chemical Co, Tokyo, Japan) according to manufacturer instruction. Capsular serotypes were determined via sequential multiplex PCR analysis [ 15 ] using QIAGEN Multiplex PCR Kits (QIAGEN, Hilden, Germany) or pneumococcal capsule-specific antisera (Statens Serum Institut, Copenhagen, Denmark) per manufacturer instruction. Strains whose serotypes could not be determined by PCR or the Analysis of PspA family 1 and 2 PspA family classification was performed by PCR using DNA extracted from isolates with primers reported by Hollingshead, et al. [ 16 ] as follows: LSM12 and SKH63 for PspA family 1 (pspA F1), and LSM12 and SKH52 for PspA family 2 (pspA F2) ( Table 2). PCR reactions were carried out using Quick Taq HS DyeMix (TOYOBO, Osaka, Japan). The PCR conditions were 95 ℃ for 3 min; then 30 cycles of 95 ℃ for 1 min, 62 ℃ for 1 min, 72 ℃ for 3 min, and finally 72 ℃ for 10 min. PCR products were loaded onto 1% agarose gels, electrophoresed at 100 V for 30 min, and stained with 0.5 μg/ml ethidium bromide. The isolates that were not initially amplified were further processed with the same cycling pattern at an annealing temperature of 58 ℃, or, if that also failed, of 55 ℃.

Analysis of pspC.4, rrgC, and sipA genes
The presence of pspC.4 was investigated by PCR using primers LU9 and LU10 ( min. PCR products were electrophoresed in 2% agarose gels and stained.
To determine whether Pilus-2 was present, primers sipA-up and sipA-dn (

Serotype distribution of isolates
Serotype distribution of the isolates showed that serotype 11A/11E was the most common serotype, followed by 15A, among isolates from total pneumococcal diseases including IPD (Table 1). IPD distribution did not follow the total serotype distribution. Among IPD isolates, serotype 7F was the most common, and serotype 24B and 38 were the only one isolates in total studied (Table 1). It was noted that three of five 7F isolates caused IPD (bacteremia), and two of the three 7F isolates that caused bacteremia were isolated from patients with no underlying diseases in age group of 5-64 years.
Virulence factor gene distribution in clinical pneumococcal isolates Figure 1 shows the proportions of the pspA F1, pspA F2, pspC. 4

Discussion
Pneumococci express multiple virulence factors, which include, for example, the polysaccharide capsule, pneumolysin, pneumococcal surface proteins, and pilli. It is considered that they can cause diseases because they possess efficient complement evasion strategies and resist opsonophagocytosis. Complement resistance constitutes a major contributor to pneumococcal virulence and pathogenesis. In particular, the capsule reduces the amount of bound C3b and restricts the access of phagocytes to cell-bound C3b, which hampers opsonophagocytosis [ 18 ]. Pneumolysin quenches complement away from the pneumococcal surface [ 19 ].
PspA is present on almost all strains of S. pneumoniae [ 16 ] and PspA families 1 and 2 (fusion PspA) are considered as promising candidate antigens for pneumococcal vaccines [ 7 , 20 , 21 ]. Our data showed that all isolates from patients with IPD and 96.0% of total isolates in this study contained PspA families 1 or 2 and that relationship between clinical invasiveness and each of them were not significant in this study.
PspC proteins contribute to virulence in colonisation and systemic mouse models [ 22 , 23 ]. PspC binds to the secretory component of pIgR to promote the adherence and invasion of epithelial cells [ 24 ] and also binds soluble host factors such as sIgA and IgM as well as C3 and complement inhibitor factor H ( [ 8 , 25 ]). PspC exhibit high variability at the sequence level, with allelic variants of PspC being divided into 11 groups. The presence of PspC group 4 was found to be correlated with the ability to bind C4BP [ 9 ]. C4BP retains its inhibitory function when bound to the bacteria, although the pneumococcus binds via the same site as C4b. Our study showed that isolates from IPD exhibited significantly higher rate of pspC.4, suggesting that pspC.4 is related to invasiveness. It is also suggested that the serotype distribution of IPD isolates did not follow that of total isolates because proportions of pspC.4 positive isolates were serotype dependent. Seven of isolates from IPD were serotypes, over half of which shows pspC.4 positive regardless of serotype frequency. Whereas, the serotype 3 from a patient with pneumococcal bacteremia was the only pspC.4-positive isolate among the 22 serotype 3 isolates. However, no serotype 15A isolate including two from IPD carried pspC.4, further supporting that pspC.4 constitutes one of the factors related to invasiveness but not related with invasiveness in certain serotypes.
The presence of rrgC had no appreciable difference in frequency between IPD and NIPD, as reported previously [ 26 ]. In comparison, rrgC was significantly associated with penicillin susceptibility (data not shown).
All sipA positive isolates in IPD were serotype 7F. Notably, 7F was the most frequent serotype in IPD in this study and all of them carried both pspC.4 and sipA. Although we studied just five among many potential virulence factors, the results may relate to the observation that 7F is among the serotypes with highest invasive disease potential [ 27 ].
The pathogenicity of S. pneumoniae has been reported to differ between clones and even isolates of the same clone [ 28 ]. Our study suggested, although factors responsible for pathogenicity vary depending on serotypes and isolates, bacterial virulence factors are associated with clinical invasive aspects.
A limitation of this study is that it was performed in a single center for one year and sample number is limited. There are concerns regarding geographical and time variation, and it is not known whether our data are representative of that in the world.

Conclusions
Our results suggested that the presence of genes for PspC group 4 is associated with clinical invasiveness and the results will promote the understanding of virulence factors in pneumococcal diseases. Further studies including more virulence factors are required to elucidate the pathogenic mechanisms of S. pneumoniae.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.