Staphylococcus aureus secreted lipases do not inhibit innate immune killing mechanisms

Background: Staphylococcus aureus causes an array of diseases in both humans and livestock. Pathogenesis is mediated by a plethora of proteins secreted by S. aureus, many of which remain incompletely characterised. For example, S. aureus abundantly secretes two isoforms of the enzyme lipase into the extracellular milieu, where they scavenge upon polymeric triglycerides. It has previously been suggested that lipases may interfere with the function of innate immune cells, such as macrophages and neutrophils, but the impact of lipases on phagocytic killing mechanisms remains unknown. Methods: We employed the epidemic S. aureus clone USA300 strain LAC and its lipase deficient isogenic mutant, along with recombinant lipase proteins, in in vitro experimental infection assays. To determine if lipases can inhibit innate immune killing mechanisms, the bactericidal activity of whole blood, human neutrophils, and macrophages was analysed. In addition, gentamycin protection assays were carried out to examine the influence of lipases on S. aureus innate immune cell escape. Results: There were no differences in the survival of S. aureus USA300 LAC wild type and its lipase-deficient isogenic mutant after incubation with human whole blood or neutrophils. Furthermore, there was no detectable lipase-dependent effect on phagocytosis, intracellular survival, or escape from both human primary and immortalised cell line macrophages, even upon supplementation with exogenous recombinant lipases. Conclusions: S. aureus lipases do not inhibit bacterial killing mechanisms of human macrophages, neutrophils, or whole blood. These findings broaden our understanding of the interaction of S. aureus with the innate immune system.


Introduction
The Gram-positive bacterium Staphylococcus aureus is the cause of an array of nosocomial and community-acquired infections. To be a successful pathogen, S. aureus must evade killing by the innate immune system which it does via a large number of secreted factors. Within the Staphylococcus genus, a lipaseencoding gene (lip1) is present in at least 12 species, and a second lipase gene is present in S. aureus (lip2) and S. epidermidis (gehD) 1,2 . S. aureus lipases are glycerol-ester hydrolases that cleave triglyceride lipids, resulting in the release of glycerol derivatives and free fatty acids 1 . Lipase 1 has an affinity for short-chain fatty acids, whereas lipase 2 has no bias towards chain length 1 . Transcription of lipase genes is regulated by the accessory gene regulator (agr) two component system, leading to the expression of a pre-pro-lipase precursor that is secreted into the extracellular milieu 1,3 . The catalytic activity of lipases is regulated through downstream processing by the secreted zinc metalloprotease, Aur, which proteolytically cleaves the prepro precursor enzyme resulting in the mature, active form of the enzyme 4 . The activity of the mature lipase is governed by a catalytic triad, which cleaves glycerol-ester bonds through a serine hydrolase mechanism 1,2 . Lipases have been reported to account for approximately 20% of the total S. aureus secretome, but our understanding of the role of lipases in hostpathogen interactions is limited 5 . It has been shown that 80% of clinical isolates from both systemic and localised S. aureus infections exhibit lipolytic activity, and patients typically test positive for anti-lipase IgG in serum 6,7 . Lipases have further been attributed to the formation of biofilm, which subsequently confers resistance to toxic polyamines thus promoting bacterial persistence 3,8,9 . Previous studies have further demonstrated that lipases can produce free-fatty acids from host lipid metabolites, such as low-density lipoproteins, which subsequently are incorporated into the lipid moieties of S. aureus 10 . The incorporation of lipoprotein particles has been shown to render the bacterium resistant to the antimicrobial drug triclosan, which is commonly used in the treatment of S. aureus infection 10 .
In a previous study, human granulocytes were treated with S. aureus lipases resulting in the loss of microvilli, projections, and pseudopodia on their surface suggesting a potential impact on phagocytosis or neutrophil extracellular trap (NET) formation 11,12 . More recently, it was demonstrated that lipase 2 interferes with macrophage signalling, which subsequently diminishes the downstream pro-inflammatory response 13 . Specifically, lipase 2 inactivates S. aureus secreted lipoproteins, which are a major pattern-associated molecular pattern recognised by Toll-like receptor 2 (TLR2) in response to S. aureus infection 13 .
Macrophages are equipped with an array of pathogen recognition receptors and, alongside a role in modulation of cellular signalling, are professional phagocytes that aid in the clearance S. aureus 14,15 . However, studies have shown that once entrapped within the macrophage phagolysosome, S. aureus can subvert killing mechanisms and persist for several days 16,17 . The subsequent death of the macrophage through membrane blebbing and caspase-3 activation results in the release of viable bacteria, promoting intra-host dissemination in a Trojan horse-like system 14,17,18 .
Here, we tested the hypothesis that lipases can interfere with the antibacterial activity of whole blood, neutrophils and macrophages. We report that, despite their abundant secretion, lipases have no effect on killing, phagocytosis, intracellular survival or escape of S. aureus USA300 LAC.

Methods
Bacterial growth conditions 40% (v/v) glycerol stocks of both a wild type (S. aureus USA300 WT) and an isogenic mutant (S. aureus USA300 Δlip1/Δlip2) of the CC8 epidemic clone S. aureus USA300 LAC generated in a previous study 19 were stored at -80°C. When required, stocks were sub-cultured onto tryptone soy agar (TSA, Oxoid CM131B) or cultured into tryptone soy broth (TSB, Oxoid CM129B) overnight at 37°C with agitation (200 rpm). The culture was diluted 1 in 100 in TSB and incubated, until exponential phase (OD 600 =0.6-0.8), as measured using an Amersham Biosciences Ultrospec 2100 pro spectrophotometer. For infection protocols, bacteria were washed in cell culture media and suspended at the required OD 600 .
Hexa-histidine tagged proteins were purified by immobilised metal affinity chromatography as described previously 19 . Western blot analysis confirmed the presence of hexa-histidine tagged proteins at 76 kDa (1 in 10,000 monoclonal anti-poly His, α-diagnostics HISP12-HRP, in 8% (w/v) skimmed milk (Sigma, 70166-500G) in sterile phosphate buffered saline (PBS)). Primary antibody binding was detected using enhanced luminolbased chemiluminescent (ECL) western blotting substrate (GE Healthcare, RPN2232). For lipopolysaccharide (LPS) removal, 1 ml of Pierce high capacity endotoxin removal resin (Thermo Fisher Scientific, 88271) was used according to the manufacturer's instructions and proteins were quantified using a bicinchoninic acid (BCA) assay (Merck millipore, 71285-3). To analyse recombinant protein lipolysis, a turbiometric assay was used following the methodology outlined previously 20 . For each of the following experiments, 200 nM of recombinant lipase 1 (rLip1) or 2 (rLip2) was used, according to previous estimates of lipase secretion levels by S. aureus 21 .

Ethics statement
Human blood was obtained from healthy volunteers in syringes treated with anticoagulant citrate dextrose. Ethical approval for the collection of blood from anonymous donors was granted by the University of Edinburgh Research Ethics Committee. This study was reviewed by the University Of Edinburgh College Of Medicine Ethics Committee (2009/01) and subsequently renewed by the Lothian Research Ethics Committee (11/AL/0168). Written informed consent was received from all volunteers participating in the study.
Bacterial killing by whole blood 75 µl of whole blood was infected with 25 µl of 1. 5 × × 10 5 CFU of S. aureus USA300 WT and S. aureus USA300 Δlip1/Δlip2 in the presence or absence of 200 nM rLip1, rLip2 or both in a 96 well Cellstar U bottomed plate for 1, 2 and 4 h at 37°C, with shaking at 200 rpm. Blood was lysed in 0.1% (v/v) TritonX-100 (Sigma), viable bacteria counts were determined with 10 µl of ten-fold bacterial dilutions in PBS onto TSA using a modified Miles-Misra technique 22 and incubated overnight at 37°C.

Isolation of CD14 + monocytes
Monocytes were isolated from human whole blood following centrifugation at 1200 x g (no break) for 20 min. Buffy coats were combined and diluted with PBS and subsequently slowly pipetted over 15 ml of 1.199 g/mol ficoll paque plus (Sigma). A gradient was generated by centrifugation for 45 min at 200 x g (no break), in which the mononuclear cell layer was subsequently removed. Ficoll was removed by centrifugation for 10 min with 300 x g, and resuspension in PBS. CD14 + monocytes were collected using a MAC-LS column as per the manufacturer's instructions (Miltenyi Biotec, 130-042-401).
Gentamycin-protection assay THP1 macrophages and blood-monocyte derived macrophages were infected at an MOI of 1 with bacteria suspended in fresh media (RPMI-1640 (Sigma), 10% (v/v) heat-inactivated foetal bovine serum (Gibco) and 1% (v/v) GlutaMAX (Gibco). Cells were centrifuged at 400 x g for 5 min and incubated for 1 h at 37°C, 5% CO 2 . For analysing internalised bacteria (phagocytosis), cells were subsequently incubated with 100 µg/ml gentamycin (Sigma, G1397-10ML) in cRPMI for 30 min. To analyse intracellular survival, cells were subsequently left in 20 µg/ml gentamycin in media and were incubated for a further 24 h at 37°C, 5% CO 2 . Finally, to analyse the escape of intracellular bacteria, cells were incubated for 24 h in antibiotic-free media at 37°C, 5% CO 2 . At each time point, corresponding to the degree of phagocytosis, bacterial intracellular survival, and bacterial escape from the macrophage, cells were lysed in 0.1% Triton X-100 in PBS for 5 min at room temperature, and viable cell counting by plating onto TSA as described above.

Statistical methods
Statistical analysis was performed with GraphPad Prism 8 software (GraphPad, USA).

Results
Lipases do not inhibit S. aureus survival in human whole blood Peripheral whole blood contains an array of innate immune components involved in the direct killing of S. aureus [23][24][25][26][27] . To evaluate if lipases can promote S. aureus survival in blood, human whole blood was incubated with S. aureus USA300 LAC (S. aureus USA300 WT) or its isogenic mutant deficient in both lipase 1 and lipase 2 production (S. aureus USA300 Δlip1/Δlip2) for 1, 2, and 4 h at 37°C. Concurrently, S. aureus USA300 Δlip1/Δlip2 was also incubated with 200 nM of functionally active rLip1 and rLip2 (Extended Figure 1 28 ). There was a 10-fold reduction in the number of recoverable bacteria in the first hour post-infection, followed by a stabilisation of the number of viable bacteria recovered up to 4 h, but there was no difference between the S. aureus USA300 WT and the lipase-deficient mutant or strains supplemented with recombinant lipase (Figure 1 28 ). Overall, these data indicate that lipases do not inhibit killing of S. aureus USA300 LAC in human whole blood.

S. aureus lipases do not inhibit neutrophil bactericidal activity
It was previously demonstrated that purified S. aureus lipases alter the phenotype of granulocytes, suggesting a possible impact on their function 11,12 . To establish if lipases can inhibit neutrophil killing of S. aureus, human neutrophils were isolated from fresh whole blood and incubated with opsonised S. aureus USA300 WT or S. aureus USA300 Δlip1/Δlip2 for 30 min. As with whole blood, there was a 10-fold reduction in the number of viable bacteria after incubation with neutrophils, but viability between the S. aureus USA300 WT and the lipases-deficient strain did not differ (Figure 2 28 ). In addition, neutrophils were incubated with S. aureus USA300 Δlip1/Δlip2 in the presence of exogenous recombinant lipases and there were no differences in the number of recovered viable bacteria between the tested conditions (Figure 2 28 ). Taken together, these data indicate that lipases do not inhibit neutrophil-mediated killing of S. aureus USA300 LAC.
Lipases do not influence phagocytosis, intracellular survival or escape of S. aureus from macrophages Recently, it was demonstrated that lipolysis of S. aureus lipoproteins by lipase 2 facilitated the survival of S. aureus through the manipulation of macrophage cellular signalling 13 . In addition, S. aureus can interfere with macrophage phagolysosomal killing, enabling intracellular persistence 16 . To examine the capacity for S. aureus lipases to influence phagocytosis, intracellular survival, and escape from macrophages, primary human monocytederived macrophages were incubated with S. aureus USA300 WT or S. aureus USA300 Δlip1/Δlip2 in the presence or absence of rLip1 or rLip2 (Figure 3a 28 ). Considerable variation in the number of recovered bacteria was observed between technical replicates due to donor variability, but no significant lipase-dependent differences were observed (Figure 3b 28 ). To further explore the effect of lipases on macrophage function, an immortalised cell line derived from human peripheral blood monocytes (THP1) cells was employed 29 . PMA induces THP1 monocyte differentiation into adherent macrophages which represent a model of human monocyte-derived macrophages 30 . S. aureus USA300 LAC infection of THP1 macrophages exhibited less variation between  replicates when compared to primary cultures but no lipasedependent differences in the number of bacteria recovered was observed (Figure 3c 28 ). Together, these data indicate that lipases do not affect phagocytosis, survival or escape of S. aureus from human macrophages.

Discussion
The importance of neutrophils in the initial response to S. aureus infection is well established 24,31 . Previously, Rollof et al., demonstrated, using scanning electron microscopy, that supernatant-purified S. aureus lipases altered granulocyte morphology by denuding surface projections 11 . As neutrophil phagocytosis is reliant on pseudopod extensions for ingesting bacteria, it was hypothesised that this phenotype could inhibit bactericidal activity 24 . Furthermore, the release of extracellular DNA into the environment, through NETosis, could be impacted by lipase-mediated changes to the cellular membrane which could influence bacterial killing.
Here, we demonstrate that lipases do not inhibit direct killing of S. aureus mediated by human neutrophils, macrophages or whole blood in vitro. The findings are consistent with the findings of Nguyen et al., who did not observe any differences in bacterial burden in the heart and liver in an in vivo murine sepsis model 24 h after infection with S. aureus USA300 WT LAC or an isogenic lipase-deficient mutant 3 . These data suggest that lipases do not interfere with the initial clearance of S. aureus from the blood.
A recent study by Chen et al., reported that lipases have no direct effect on initial bacterial clearance in the early stages of infection. However they demonstrated that after 48 h, there was an indirect effect of lipase 2 resulting in reduced pro-inflammatory cytokine release by macrophages 13 . The authors found that S. aureus lipase 2 mediates cleavage of S. aureus lipoproteins, which are well characterised TLR2 ligands, resulting in increased bacterial burden by thwarting macrophage responses.
Previously it has been shown that S. aureus virulence factors regulated by the agr quorum-sensing system are required for survival and escape of S. aureus from macrophages, including the zinc metalloprotease Aur which is responsible for the downstream activation of the catalytically active lipases 16,32,33 . Here, we report that the agr-regulated lipases do not influence the survival of S. aureus in human monocyte-derived macrophages, although considerable donor specific variation was observed with primary cells. Data obtained using the THP1 macrophage cell line further support the finding that S. aureus lipases do not affect phagocytosis, intracellular survival or escape of S. aureus from human macrophages. The lack of an observable effect of lipases may reflect the fact that bacterial capture by macrophages is dependent on dynamic actin-rich protrusions, with negligible involvement of triglyceride lipids in the process 34 .

Conclusion
Overall, we report that S. aureus lipases do not directly impact on the killing mechanisms of neutrophils and macrophages. These data add to our understanding of S. aureus interactions with the innate immune system and the role of lipases in the pathogenesis of S. aureus disease.

Data availability
Underlying data Edinburgh Datashare: Staphylococcus aureus secreted lipases do not inhibit innate immune killing mechanisms: Extended It was observed that both lipase 1 and 2 were functionally active enzymes which were able to cleave Tween-20 over a broad scope of concentrations. Indeed, it was also observed that lipase 2 was much more kinetically active in comparison to lipase 1, which could be attributed to its broader substrate range.

Meera Unnikrishnan
Division of Biomedical Sciences, University of Warwick, Coventry, UK This is a clearly written manuscript, investigating the role of the S. aureus lipase 1 and 2 in bacterial killing in whole blood, neutrophils and macrophages. This work is important as lipases are generally considered to impact innate cell functions, although effects on bacterial killing have not been directly assessed yet.
In general experiments have been done well, and the main conclusions are valid. However a few technical clarifications are required.
For the macrophage internalisation assays, when following bacterial replication by CFU, usually it is good to look at an earlier time point as by 24h there can be significant cell lysis induced by the WT. Have the authors checked the morphologies or the states of the macrophages during their experiments?
The bacterial 'escape' measurements from macrophages need some clarification. It is not clear if after the 1h infection the macrophages were treated with gentamicin to kill all the extracellular bacteria first prior to adding the antibiotic-free medium. This is essential to do in order to measure the escape of intracellular bacteria. It is also not clear if escape was quantified from the culture supernatants or from the the cell lysates. Quantifying bacterial escape is quite tricky, and should be preferably done at multiple time points after infection to get a clear picture. Counting from supernatants are not accurate as most bacteria settle down to the well by 24h forming microcolonies, and by 24h there is substantial cell lysis, so hard to deduce the 'escaped' population by measuring intracellular counts (by cell lysis). Finally, the authors may want to comment on why the 'escape' numbers are so high between the THP1 vs the primary cells.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound? For measuring S. aureus escape we followed the procedure outlined in the schematic shown in Figure 3a. Cells were infected for 1 hour at an MOI of 1 followed by treatment with 100 µg/ml of gentamycin for 30 minutes to kill any extracellular bacteria. The cells were washed and then subsequently left for 24h in a low concentration of gentamycin (20 µg/ml) which does not penetrate the cells, however, was sufficient for killing any bacteria which had escaped due to cell lysis in this time. After 24 hours, the cells were washed again to remove any residual gentamycin and incubated in the antibiotic-free media for the 24 hour "escape" period. In preliminary data not shown, we analysed bacterial escape over a range of time points (ranging from 3-24 hours post removal of gentamycin media) and observed no differences in the recovered viable CFU in the presence and absence of lipase, and thus 24 hours was chosen as a representative time point. To quantify escape, 0.1% (v/v) TritonX-100 was added on top of both the supernatant and the cells (at a 1:1 ratio). Both micro-colonies and substantial cell lysis was observed and thus wells were scraped to incorporate both the supernatant and lysate for escape CFU quantification.
Finally, the authors may want to comment on why the 'escape' numbers are so high between the THP1 vs the primary cells.
There are multiple potential reasons why there was variation between the recovered escape CFU from primary and the THP1 macrophages. Importantly, considerable differences exist between the immortalised THP1 macrophage cell-line and the primary macrophages. For example, a broader repertoire of chemokine and cytokine responses in primary macrophages could contribute to more efficient killing by the primary cells and thus lower viable recovered CFU.
these lipases may interfere with the function of innate immune cells but no previous articles have investigated the direct role of the lipases. This article describes the phenotypic analysis of a S. aureus double lipase mutant and recombinant lipase proteins. There were no differences in the survival between the wild type and mutant strains with human whole blood or neutrophils and no effect on phagocytosis, intracellular survival, or escape from both human primary and immortalised cell line macrophages. Therefore, it was concluded that S. aureus lipases do not inhibit bacterial killing mechanisms of human macrophages, neutrophils, or whole blood. The article gives a very clear background of the literature of S. aureus lipases in innate immunity. The study design is logical and uses a series of well-established phenotypic assays that are well described and use appropriate controls and are shown to be statistically reproducible. There are no major issues to address.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes