Staphylococcus aureus biofilm properties and chronic rhinosinusitis severity scores correlate positively with total CD4+ T‐cell frequencies and inversely with its Th1, Th17 and regulatory cell frequencies

Chronic rhinosinusitis (CRS) represents chronic inflammation of the sinus mucosa characterised by dysfunction of the sinuses' natural defence mechanisms and induction of different inflammatory pathways ranging from a Th1 to a Th2 predominant polarisation. Recalcitrant CRS is associated with Staphylococcus aureus dominant mucosal biofilms; however, S. aureus colonisation of the sinonasal mucosa has also been observed in healthy individuals challenging the significance of S. aureus in CRS pathogenesis. We aimed to investigate the relationship between CRS key inflammatory markers, S. aureus biofilm properties/virulence genes and the severity of the disease. Tissue samples were collected during endoscopic sinus surgery from the ethmoid sinuses of CRS patients with (CRSwNP) and without (CRSsNP) nasal polyps and controls (n = 59). CD3+ T‐cell subset frequencies and key inflammatory markers of CD4+ helper T cells were determined using FACS analysis. Sinonasal S. aureus clinical isolates were isolated (n = 26), sequenced and grown into biofilm in vitro, followed by determining their properties, including metabolic activity, biomass, colony‐forming units and exoprotein production. Disease severity was assessed using Lund–Mackay radiologic scores, Lund–Kennedy endoscopic scores and SNOT22 quality of life scores. Our results showed that S. aureus biofilm properties and CRS severity scores correlated positively with total CD4+ T‐cell frequencies but looking into CD4+ T‐cell subsets showed an inverse correlation with Th1 and Th17 cell frequencies. CD4+ T‐cell frequencies were higher in patients harbouring lukF.PV‐positive S. aureus while its regulatory and Th17 cell subset frequencies were lower in patients carrying sea− and sarT/U‐positive S. aureus. Recalcitrant CRS is characterised by increased S. aureus biofilm properties in relation to increased total CD4+ helper T‐cell frequencies and reduced frequencies of its Th1, Th17 and regulatory T‐cell subsets. These findings offer insights into the pathophysiology of CRS and could lead to the development of more targeted therapies.


INTRODUCTION
Chronic rhinosinusitis (CRS) is a persistent inflammation of the nasal cavity and paranasal sinuses that afflicts up to 10% of the general population, imposing a substantial direct and indirect burden on healthcare systems and economies worldwide [1,2]. In contrast to the conventional phenotypic classification of CRS into those with (CRSwNP) and without (CRSsNP) nasal polyps [3], endotyping defines the disease variants by their pathophysiologic mechanisms [4]. CRS endotypes are commonly described based on patterns of inflammatory cells, specifically CD4+ T helper (Th) cells which regulate the expression of various chemokines and cytokines. Endotypes of CRS mainly fall into two categories: type 2 (Th2) and non-type 2 (Th1 and Th17) [1]. Th2 is associated with CRSwNP in Caucasian patients and enhanced secretion of IL-4, IL-5, IL-13, IgE antibodies and tissue eosinophilia [3,5,6]. Non-Th2 is associated with CRSsNP, but recent studies have shown geographical variation [7,8]. Th1 is predominantly characterised by increased neutrophils linked to myeloperoxidase and elevated secretion of IFN-γ, IL-2 and TNF-α. Th17 is mainly dominant in Asian patients with CRSwNP and is associated with increased expression of IL-17 and IL-22 cytokines [5]. However, patients do not always fit neatly into these endotypes, making classification difficult [9]. Diverse T-cell polarisations affect the choice of treatment strategies for CRS [10,11]. Furthermore, regulatory T cells (Tregs) are a specialised subpopulation of T cells that suppress the immune response, maintain homeostasis and self-tolerance and prevent autoimmune disease. The immunosuppressive properties of Tregs are achieved by downregulating effector T cells [12]. The FoxP3 transcription factor is generally used as a marker for identifying these cells [13].
Despite appropriate medical therapy and surgical intervention, 10% of CRS patients suffer from recalcitrant disease. Those patients commonly have nasal polyps with eosinophilia, Staphylococcus aureus dominant mucosal biofilms, comorbid asthma and a severely compromised quality of life [14][15][16][17]. The remarkable success of S. aureus as a pathogen might be due to its numerous measures to protect itself against the host's immune system, including the biofilm mode of existence. Bacteria in biofilms express different genes and proteins from their planktonic counterparts [18,19] and are more resistant to antimicrobial therapy and host defences [20].
Despite the vast knowledge of inflammatory endotypes in CRS and the significance of S. aureus biofilm in this disease, little is known about the relationship between CRS inflammatory markers, S. aureus biofilm properties and the severity of the disease. Here, S. aureus was isolated from CRS patients and non-CRS controls, grew into biofilm in vitro, and its biofilm properties were characterised in relation to inflammation and CRS severity. We further investigated the relationship between S. aureus virulence genes and inflammatory markers in patients' sinonasal tissue. A better understanding of S. aureus biofilm properties and their contribution to CRS pathogenesis will be critical for enhancing the prognosis of patients suffering from CRS.

Patients and clinical data
Ethics approval for the collection, storage and use of clinical isolates and patient samples from CRS and non-CRS control patients was granted by The Queen Elizabeth Hospital (TQEH) Human Research Ethics Committee, South Australia (HREC/15/TQEH/132), and all patients had signed written informed consent. A prospective study was conducted from 2019 to 2020 with patients recruited from the Queen Elizabeth Hospital, Memorial Hospital and Calvary Hospital, Adelaide, Australia. The diagnostic criteria for CRSwNP and CRSsNP were contented by the American Academy of Otolaryngology and Head and Neck Surgery and the European Position Statement on CRS [15]. Eligibility for the study included age ≥18 years and having CRS requiring endoscopic sinus surgery (ESS). The control subjects underwent endoscopic skull base surgery or septoplasty with no clinical or radiologic evidence of sinus disease. Exclusion criteria included using antibiotics or oral corticosteroids the month before surgery. The severity of CRS was measured based on the completion of the Lund-Mackay (LM) [21], Lund-Kennedy (LK) [22,23] and patient-reported 22-item Sino-Nasal Outcome Test (SNOT-22) questionnaire [24]. A self-reported questionnaire was used to assess the status of asthma, aspirin sensitivity, gastro-oesophageal reflux disease and diabetes mellitus. Sinonasal polyp or mucosal tissue samples were collected from the ethmoid sinuses of CRS patients and the ethmoid sinuses and  middle turbinate of non-CRS control subjects during ESS. Nasal swabs were also collected from patients' middle meatus. All samples were immediately transported to the laboratory for processing.

FACS analysis
Fresh sinonasal polyps or mucosal samples were processed into single-cell suspension by enzymatic digestion. For cell surface and intracellular staining, the single-cell suspensions (4 Â 10 6 cells/mL) were stimulated for 4 h at 37 C. Subsequently, the cells were stained with Fixable Viability Dye eFluor ® 780 (BD Bioscience). Next, the cells were incubated with Fc Block (BD Bioscience) and fluorochrome-conjugated surface antibodies. True-Nuclear Transcription Factor Buffer Set (Biolegend) was used according to the manufacturer's protocol. The cells were then stained with fluorochromeconjugated intracellular antibodies. Finally, the cells were washed and resuspended in autoMACS Running Buffer (Milteny Biotec, Germany). Multi-colour flow cytometry was performed using a BD FACS Canto II instrument (BD bioscience) and FACS Diva software (BD Bioscience). For each sample, at least 500 000 events were collected. Data were analysed using Flowjo software (FlowJo LLC, Ashland, USA). A summary of immune T cells analysed, and reagents used for FACS is illustrated in Tables S1 and S2.

Isolation of S. aureus
Clinical isolates of S. aureus were isolated from the nasal swabs of CRS patients and control subjects using mannitol salt agar plates (Oxoid, Basingstoke, UK). The species-level identification was performed using the API Staph (bioMerieux, Australia) and Staphylase test (Oxoid) kits. The isolates were screened for MRSA using a super sensitive and specific chromogenic MRSA selective agar (CHROMID ® MRSA SMART, bioMerieux, Australia). The S. aureus isolates were stored at À80 C in tryptone soy broth (TSB, Oxoid) plus 20% (v/v) sterile glycerol until further analysis.

Evaluation of S. aureus biofilm biomass
Forty-eight-hour biofilms in clear 96-well plates were washed with PBS, air-dried and stained with 180 μL of 0.1% (v/v) crystal violet solution per well for 15 min. Subsequently, the crystal violet was removed, and the plates were rinsed three times with sterile Milli-Q water. Next, 180 μL/ well of 30% acetic acid was added and incubated on a plate shaker until the crystal violet was solubilised. The absorbance was measured at 595 nm using the FLUOstar OPTIMA microplate reader (BMG LABTECH). A sterility control containing uninoculated nutrient broth was used to determine the background optical density. The experiment was performed three times in six replicates. The values of S. aureus optical densities were normalised to values for ATCC 25923 and reported as normalised optical densities.

Evaluation of S. aureus biofilm CFU
Forty-eight-hour biofilms in 6-well cell plates were washed with PBS. Subsequently, 1 mL of PBS was added to each well and resuspended thoroughly by pipetting for 3 min to detach the biofilm. Ten-fold serial dilutions of the liquid bacterial cultures were prepared and spot-plated onto TSA plates in 3 Â 20 μL spots. The plate was then incubated overnight. The dilutions which gave easy-to-count colonies were selected to count the number of colonies in all three spots. Finally, CFU/mL were calculated using CFU/mL = the average number of colonies Â 50 Â dilution factor. The values of S. aureus CFU were normalised to values for the ATCC 25923 strain and reported as normalised CFU.

Collection and quantification of S. aureus exoprotein
The 48-h S. aureus biofilm supernatants from 6-well cell culture plates were collected and centrifuged at 1500 Âg for 10 min at 4 C.
Assemblies were quality-controlled using QUAST, v 5.0.2 [28]. All genomes were MLST typed and grouped into Clonal Complexes using the MLST [29]

Statistical analysis
Statistical calculations were performed using Graph Pad Prism v 9.0.0 (121) (GraphPad Software, San Diego CA, USA). Statistical differences among groups were evaluated using one-way ANOVA or Kruskal-Wallis test depending on the normality of the variables' distribution. The Independent Samples t-test was used to assess the statistical differences between the means of two groups. The Pearson or Spearman's rank correlation coefficient was used for the correlation analysis where r-value: 0.00-0.19 = very weak, 0.20-0.39 = weak, 0.40-0.59 = moderate, 0.60-0.79 = strong and 0.80-1.0 = very strong. The results were considered statistically significant when the p-value was <0.05. *p < 0.05; **p < 0.01; ***p < 0.001 and ****p < 0.0001.  Table S3. The severity of CRS disease in CRSwNP, CRSsNP and control subjects is illustrated in Figure S1.

Demographic data and clinical details
Th1 and Th17 cell frequencies are elevated in CRSsNP, Th2 cell frequencies are elevated in CRSwNP, but Treg frequencies are decreased in both groups compared to control The flow cytometric analysis of fresh sinonasal polyp or mucosal samples from CRSwNP (n = 23), CRSsNP (n = 19) and non-CRS control subjects (n = 17) was conducted, and a multiparametric flow cytometry approach was applied for the gating strategy, detailed in Figure S2.

S. aureus from CRSwNP is characterised by high metabolic activity, high biomass, high exoprotein production and high CFU counts
In vitro-grown S. aureus biofilm properties such as metabolic activity, biomass, exoprotein production and CFU were evaluated. As shown in Figure S5, the metabolic activity was variable among different isolates at different time points, but the highest variability was observed at 1 h after incubation with alamarBlue. The quantification of S. aureus biofilm metabolic activity in the patient cohorts showed higher metabolic activity in CRSwNP-derived clinical isolates than those from CRSsNP ( p = 0.0010) and controls (p = 0.0012) at 1 h after incubation with alamarBlue ( Figure 4a). In addition, CRSwNP-derived S. aureus biofilms showed higher biomass compared to CRSsNP ( p = 0.0073) and controls (p = 0.0276; Figure 4b). Higher CFU/mL were also detected in S. aureus biofilms derived from CRSwNP patients than those from controls (p = 0.0431; Figure 4c). Furthermore, the CRSwNPderived clinical isolates exhibited higher exoprotein production than CRSsNP-derived clinical isolates (p = 0.0241; Figure 4d).
CRS severity scores correlate positively with CD4+ and CD4+CD8+ T-cell frequencies, but inversely with Th1, Th17 and Tregs Spearman rank correlation coefficient was calculated to determine whether CRS patients' severity scores and the frequency of T-cell subsets and T helper inflammatory cytokines were related. Weak to moderate positive correlations were observed between CRS severity scores (LM, LK and SNOT22) and the frequencies of CD3+ T cells, CD4+ T cells and CD4+CD8+ T cells (Figure 5a-c). There were inverse correlations between CRS severity scores and the frequencies of CD4ÀCD8À, Th1 (CD4+ IFN-γ+), Th17 (CD4+ IL-17a+) and Tregs (CD4+ FOXP3+; Figure 5d-g).

DISCUSSION
This is the first study to show the correlation of CD3+ cell subsets in the sinonasal tissue of CRS and non-CRS control patients with CRS severity scores, in vitrogrown biofilm properties and virulence genes of the corresponding patient-derived S. aureus. Our results show significant immunophenotypic differences in CD3+ cell subsets of CRS patients in relation to CRS disease severity, suggesting that these cells might play a critical role in CRS pathogenesis. Furthermore, the positive correlation of S. aureus biofilm properties and virulence genes with CD4+ T-cell frequencies and an inverse correlation with Th1 and Th17 cell frequencies, provides a novel understanding of how straindependent variation in S. aureus biofilm properties and genomic content marks disease severity and relates to immune polarisation.   In the current study, the FACS analysis of CD3+ Tcell subsets and key inflammatory markers of CD4+ helper T cells (Th1, Th2, Th17 and Tregs) and CD8+ cytotoxic T cells (Tc1, Tc2, Tc17 and Tregs) was carried out. Elevated CD4+ T-cell frequencies in CRSwNP patient tissue agree with previous studies supporting the contribution of those cells in CRS pathogenesis [30,31]. Numerous studies have also reported an increased frequency of CD4+CD8+ double-positive T cells with ageing and in the context of infectious and autoimmune diseases, as well as cancer [32][33][34][35], suggesting a phenotypic profile linked to memory T cells capable of producing cytokines and lytic enzymes. However, their exact role in CRS is yet to be defined. CD4ÀCD8Àdoublenegative T-cell frequencies were significantly reduced in CRSwNP patients compared to control. These cells have been reported to have a key role in regulating immune responses in the context of transplant rejection, graft-versus-host disease, autoimmune and infectious diseases [36,37]. It has also been shown that these cells have a   regulatory function similar to Tregs and possess a unique array of cytokines, including IL-4, IL-17 and IFN-γ [38]. However, their function in CRS warrants further investigation. The higher frequency of Th1 and Th17 cells in CRSsNP and Th2 cells in CRSwNP suggests a non-Th2 skewed inflammation in CRSsNP patients and a Th2 polarisation in our CRSwNP patients. Since these cytokines are not entirely restricted to a specific inflammatory pathway and their ratio determines the ultimate consequence on immune activation [39], their ratio analysis was conducted to confirm the type of inflammation. Higher IL-4/IL-17a (Th2/Th17) and IL-4/ IFN-γ (Th2/Th1) in CRSwNP patient tissue demonstrated a Th2-polarised inflammatory milieu in those patients, while higher IFN-γ/IL-4 (Th1/Th2) confirmed Th1 dominance in CRSsNP. Furthermore, decreased Tregs in CRS patients are in line with previous studies [40,41], suggesting poor recruitment of those cells in CRSwNP [12,42] potentially due to their impaired migratory function, which ultimately contributes to inflammation [43].
The positive correlation of CD3+, CD4+ and CD4+-CD8+ T cells with LM, LK and SNOT22 severity scores in CRS implies that the substantial expansion of those cells might explain the severity of inflammation. In contrast, the inverse correlation of Th1 and Th17 cell frequencies with CRS severity scores suggests the induction of non-Th2 skewed inflammation in mild CRS cases. This is supported by previous findings where Th2-skewed inflammation is mainly observed in severe CRSwNP, often in association with concomitant asthma [44]. The negative correlation of Tregs and CD4ÀCD8À doublenegative T cells, both having regulatory properties, with disease severity, suggests that impaired regulatory function is associated with severe disease in CRS. Those cells might serve as potential biomarkers of disease severity. Interestingly, some inflammatory T cells were closely correlated with S. aureus biofilm properties. The moderate to strong correlation of CD4+ cell frequencies with S. aureus biofilm metabolic activity, biomass, CFU counts and exoprotein production is strongly suggestive of the greater immunogenicity of high biofilm-producing S. aureus that can lead to CD4+ T-cell expansion in  sinonasal tissues/polyps of CRS patients. The inverse correlation of S. aureus biofilm properties with Th1, Th17 and CD4ÀCD8À T-cell frequencies implies that low biofilm-producing S. aureus might induce non-Th2 inflammation with impairment of CD4ÀCD8À cells. By linking the outcome of the correlations, our study supports the hypothesis that high biofilm-producing S. aureus might induce a severe disease by skewing the immune response towards a Th2 inflammation. S. aureus, as a highly successful pathogenic bacterium, produces a wide range of surface components and extracellular products, including toxins and enzymes that contribute to the pathogenesis of infection [45] by promoting tissue adhesion, immune evasion and host cell damage [46]. Here, we demonstrated higher CD4+ T cells in patient tissue harbouring the isolates carrying the lukF.PV gene. Panton-Valentine Leukocidin, as an important leukotoxin, lyses cells of the leukocytic lineage and destroys neutrophils [47], stimulating the release of proinflammatory cytokines, which might favourably dispose the host to resist infection [48]. Previous studies have shown higher levels of IgG antibodies against LukF-PV in CRS patients, indicating that these specific S. aureus antigens affect the pathogenesis of CRS [49]. Staphylococcal superantigens, such as enterotoxins (SEA), are highly mitogenic exotoxins that trigger an enormously powerful stimulatory activity for T lymphocytes, leading to a substantial release of T-cell mediators and proinflammatory cytokines [50,51], intensifying the Th2 response in the tissue and diminishing the immunosuppressive activity of Tregs, which finally leads to inflammation [52]. This is supported by our findings showing the lower Treg frequencies in patient tissue harbouring sea-carrying strains. Furthermore, higher Th17 in patient tissue with sarT/sarU-lacking strains indicates that non-Th2 skewed inflammation, which is normally observed in mild CRS cases, might occur due to lack of certain virulence genes such as sarT and sarU. These genes' products regulate the synthesis of cell-surface proteins such as protein A and exoproducts such as hemolysins by controlling the agr locus [53].
In conclusion, stratifying CRS patients based on their inflammatory markers in relation to disease severity and S. aureus biofilm properties offers insights into a potential pathogenic link between severe disease and bacterial biofilms, which could lead to the development of more targeted therapies.