DISCUSSION
Our findings suggest that there are differences in virulence gene carriage between S. aureus cultured from control subjects and patients with CRSsNP or CRSwNP. In the case of CRSwNP isolated S. aureus these differences may translate into the ability to localise and replicate intracellularly, thus conferring a bacterial survival advantage.
Phenol soluble modulins such as delta haemolysin gene (hld ) are specific to S. aureus bacteria and induce neutrophil and erythrocyte lysis.23 We noted its absence in 50% of CRSwNP isolates but presence in all other isolates. Deletion ofhld has been shown to nearly abolish transcription of the accessory gene regulator (agrA ) transcription and downstream virulence genes including alpha toxin, beta haemolysin and enterotoxin B and gamma haemolysin.24 Given that mutations in the accessory gene regulator locus have been shown to create senescent bacteria known as small colony variants with reduced virulence gene expression and superior intracellular translocation in epithelial cells, the absence of hld may allow for similar pathogenic behaviour thus establishing an intracellular source of infection.25-27
Immune modulators including aureolysin which have been shown to cleave the C3 protein of the classical and alternative complement pathways were well preserved within the groups. Staphylococcal complement inhibitor (scn) was less frequently carried by the CRSwNP strain. Complement lowers the threshold for B-cell activation and antibody formation; hence the absence of staphylococcal complement inhibitor could lead to greater antibody formation towards S. aureus as seen in CRSwNP.3, 28
Exfoliative virulence factors involved in pore formation in host cells including bi-component gamma haemolysin components (hlgA, hlgB HlgC ) were well conserved throughout all isolate groups. However, leukocidin E/D (LukE/D ), which is not under control of the accessory gene regulator locus, was carried by 40% of control and 75% of CRSwNP isolates but no CRSsNP isolates. Both toxins have been shown to create pores within target neutrophils and erythrocytes while leukocidin E/D has been shown to induce calcium channel activation in neutrophils independent of its pore forming function leading to cell death.29 Furthermore, leukocidin E/D can deplete the adaptive immune response by lysing polymorphic neutrophils leading to failure of local immunity and was essential for bacterial growth and seeding in a mouse model of bacteraemia.30 Often found in invasive S. aureus disease, serine proteases (splA, splB and splE) have been shown to cleave the mucin 16 glycoprotein which forms a mucosal barrier. There carriage was abundant in the CRSwNP group (75%) but not seen in the CRSsNP group and only in 40% of the control group.31 Furthermore, serine proteases (splA, splB) and leukocidins E/D (lukE/D ) were absent in presence of the enterotoxin gene cluster and may suggest different phenotypic behaviours between these strains. This data would be in keeping with the findings of Nepal et al who recently demonstrated greater number of CRSwNP isolates carrying the Sa3int prophage which contains lukE/D genes, when compared with controls and CRSsNP patients.17
Superantigens including the enterotoxin gene cluster sei, sem, sen, seo seu were carried by 60% of control isolates, 25% of CRSwNP isolates and only minimally present in a CRSsNP isolates. These enterotoxins are associated with long term colonisation in the nasal airways, cystic fibrosis lung and atopic dermatitis wounds, and were found at a similar carriage frequency in our control group.32,33Furthermore, they have been shown to stimulate T cell proliferation, but conversely act like decoy targets for antibodies protecting S. aureus from attack.33
As greater numbers of intracellular bacteria have been observed within CRSwNP than CRSsNP tissue, we selected a strain from the control group and CRS group which represented the virulence gene pattern observed.7 When investigating intracellular survival, we confirmed that a representative CRSwNP S. aureus strain carrying serine proteases (splA, splB) and leukocidins E/D without the enterotoxin gene cluster was better able to internalise, survive and replicate within the LAD2 cell line than a representative control strain carrying the enterotoxin gene cluster with absent serine protease (splA, splB) and leukocidinE/D genes. Confocal microscopy confirmed the intracellular localisation of the bacteria and demonstrated a large number of apoptotic mast cells with a number of live bacteria residing within them when compared to the control group. Hence the CRSwNPS. aureus phenotype may further support its own survival by using dead cell bodies to protect itself from the effects of antibiotics and the immune system.
There may be a survival advantage for S. aureus in a CRSwNP environment to shed the hld gene and enterotoxin gene cluster allowing a more senescent phenotype with reduced virulence factor production and increased intracellular survival. Furthermore, the carriage of LukE/D may well promote the Th2 environment as this has been shown to skew the immune response, depleting memory T lymphocytes thereby reducing the number of IL-17 and IFN-gamma producing cells, and impairing bacterial clearance.13,34 The CRSsNP group S. aureus genomes did not carry lukE/D orsplA/B genes, with only one isolate demonstrating minimal gene carriage from the enterotoxin gene cluster. This may lead to lower levels of IgE raised towards S. aureus enterotoxins, and could explain why CRSsNP patients have lower serum IgE levels towards S. aureus enterotoxins than CRSwNP patients.
Sequencing large numbers of virulence factors from genomic samples is likely to demonstrate significant heterogeneity which makes statistical significance testing difficult without large cohorts. Nevertheless, our findings complement existing sequencing data and further our understanding of mechanisms of enhanced pathogenicity for S. aureus in CRS.