key: cord-0301266-jn68rbvb
authors: Grousd, Jennifer A; Riesmeyer, Abigail M.; Cooper, Vaughn S.; Bomberger, Jennifer M.; Richardson, Anthony R.; Alcorn, John F.
title: Novel Requirement for Staphylococcal Cell Wall-Anchored Protein SasD in Pulmonary Infection
date: 2022-04-02
journal: bioRxiv
DOI: 10.1101/2022.04.01.486802
sha: 016cc766fb0d1ad8a771d8e5a201b9ec2626f291
doc_id: 301266
cord_uid: jn68rbvb

Staphylococcus aureus can complicate preceding viral infections, including influenza virus. A bacterial infection combined with a preceeding viral infection, known as super-infection, leads to worse outcomes compared to single infection. Most of the super-infection literature focuses on the changes in immune responses to bacteria between homeostatic and virally infected lungs. However, it is unclear how much of an influence bacterial virulence factors have in super-infection. Staphylococcal species express a broad range of cell wall-anchored proteins (CWAs) that have roles in host adhesion, nutrient acquisition, and immune evasion. We screened the importance of these CWAs using mutants lacking individual CWAs in vivo in both bacterial pneumonia and influenza super-infection. In bacterial pneumonia, lacking individual CWAs led to varying decreases in bacterial burden, lung damage, and immune infiltration into the lung. However, the presence of a preceding influenza infection partially abrogated the requirement for CWAs. In the screen, we found that the uncharacterized CWA S. aureus surface protein D (SasD) induced changes in both inflammatory and homeostatic lung markers. We further characterized a SasD mutant (sasD A50.1) in the context of pneumonia. Mice infected with sasD A50.1 had decreased bacterial burden, inflammatory responses, and mortalty compared to wildtype S. aureus. Mice also had reduced levels of IL-1β compared with wildtype, likely derived from macrophages. Reductions in IL-1β transcript levels as well as increased macrophage viability implicate altered macrophage cell death pathways. These data identify a novel virulence factor for S. aureus that influences inflammatory signaling within the lung. Importance Staphylococcus aureus is a common commensal bacteria that can cause severe infections, such as pneumonia. In the lung, viral infections increase the risk of staphylococcal pneumonia, leading to combined infections known as super-infections. The most common virus associated with S. aureus pneumonia is influenza, and super-infections lead to worse patient outcomes compared to either infection alone. While there is much known about how the immune system differs between healthy and virally infected lungs, the role of bacterial virulence factors in super-infection is less understood. The significance of our research is identifying new bacterial virulence factors that play a role in the initiation of infection and lung injury, which could lead to future therapies to prevent pulmonary single or super-infection with S. aureus.

Introduction expression via qPCR (p=0.0710) (data not shown). The other mutant in the low inflammation 155 cluster was sdrD::Tn (serine aspartate repeat containing protein D), which had higher levels of 156 expression of type 1 and type 2 cytokines compared to srtA::Tn. The mutants found in the mixed 157 inflammation cluster typically had higher levels of innate immunity cytokines and chemokines, 158 but lower levels of type 1, 2, and 17 cytokines. The high inflammation cluster, which contained 159 the WT strain as well as most Clf and Sdr members, had the highest levels of cytokines. During The literature suggests that increased inflammation and tissue damage lead to increased adhesion 241 within the lung, contributing to increases in bacterial burden(9, 10). However, to our knowledge, 242 there has been no specific testing of bacterial adhesion components in vivo during single or viral 243 super-infection in the lung. Most studies that have investigated S. aureus virulence factors in the 244 lung have focused on secreted toxins, such as the alpha toxin (19, (27) (28) (29) (30) . While toxin-mediated 245 damage contributes to lung pathology, the alpha toxin has been shown to decrease adhesion to 246 lung epithelial cells(31). Thus, we wanted to determine if proteins with known adhesion 247 properties influenced the outcomes of single or super-infection. 248

Our data supports the finding that changes due to influenza infection are the primary driver of 250 super-infection, with influenza increasing bacterial burden, immune recruitment, and acute lung 251 injury seen in the model. Interestingly, regardless of what CWA was removed, influenza 252 appeared to "level the playing field" for the mutants, with endpoints being much higher and 253 tighter grouped in super-infection than in bacterial pneumonia alone. S. aureus strains seen in 254 super-infected individuals are less virulent and more closely related to nasal colonizing strains 255 than those strains found in bacterial pneumonia patients(32). This is likely due to the increased 256 inflammation and damage within lung as well as a more dysregulated immune response during 257 super-infection leading to less aggressive colonizing strains taking hold in the lung. However, 258 viral-bacterial synergism is likely adding to this phenomenon, as influenza can increase both 259 internalization and adhesion of bacteria within the lung(33, 34). This is not specific to influenza, 260 as the same phenomenon is seen in rhinovirus-S. aureus super-infections(1, 35). 261

We saw more variability in the endpoints studied during bacterial pneumonia, likely because 263 adhesion in the lung is more difficult in a homeostatic state. SasG has a known role in biofilm 264 formation(36, 37), which may explain the decrease in burden seen in bacterial pneumonia. SasG 265 has also been shown to adhere to human desquamated nasal epithelial cells via an unknown 266 ligand (38) CWAs are known to bind to several proteins within the host such as fibrinogen and 283 fibronectin(48). In this study we did not explore bacterial adhesion to specific ligands, but it is 284 likely a combination of several ligands, as described at other host sites such as the nose (49). 285

CWAs also have overlapping ligands, such as ClfA, ClfB, FnbA, and FnbB all binding 286 fibrinogen(48). Because we only looked at single CWA mutants, some of the functions of these 287 proteins in bacterial pneumonia and super-infection could be masked. 288

Even though the CWA mutants had more clear phenotypes in bacterial pneumonia compared to 290 super-infection, the cytokine signature in both settings appears to be driven by the expression of 291 these CWAs. The mutants found in each cluster were consistent in both bacterial pneumonia and 292 super-infection, with the exception of sasG::Tn. This suggests that while a majority of the 293 inflammation in the lung is driven by influenza, at least some part of the immune response is 294 shaped by the presence of these CWAs on the cell surface of the bacteria. As SasG has a known 295 role in biofilm formation and influenza is known to induce dissemination of S. aureus 296 biofilms (15, 36, 37) , this effect could influence how the immune system reacts to this mutant. A 297 majority of the MSCRAMM proteins (clfA::Tn, clfB::Tn, sdrC::Tn, sdrE::Tn) cluster together in 298 the high inflammation cluster. This is what we expected to find, as these proteins have similar 299 domains used for ligand binding and this may influence the immune response(22). ClfA has been 300 shown to be a T cell activator driving Th1 and Th17 activation(50). While we did not see any 301

significant changes in IL-2 or IFNγ, we did see a nearly significant decrease in IL-17A (p=0.0571) with the clfA::Tn mutant. Unsurprisingly, the Sortase A mutant, which lacks all 303

CWAs on the cell surface, had the lowest expression of cytokines. It is important to note that the 304 Sortase A mutant still makes all the CWAs, but they are secreted into the environment instead of 305 covalently attached to the cell wall. However, it does suggest that the influence on immune 306 signaling is greatest when the CWAs are still attached to the bacteria. However, more testing 307 would be needed in defining the portions of each CWA responsible for altering immune 308 shown that IL-33 induction of type 2 responses is protective in lethal models of S. aureus sepsis 331 and pneumonia by counterbalancing pro-inflammatory responses(57, 58). While we did not see 332 any differences in IL-33 (data not shown) or gross pathology at 24 hours post infection, we did 333 see a reduction in type 17 cytokines and neutrophils, which has been shown to be protective in 334 patients with S. aureus infection(58, 59). Thus, the reduction in inflammation or alteration of 335 inflammatory cell ratios could help explain the delayed mortality seen in mice. 336

Since we saw a change in IL-1β production both early and late during infection, we decided to 338 examine the inflammasome. S. aureus is known to prime and activate the NLRP3 inflammasome 339 via pore-forming toxins, such as the alpha toxin(26). The NLRP3 inflammasome activates 340 caspase 1, which cleaves pro-IL-1β(30). We did see a significant downregulation of il-1β and 341 nlrp3 transcripts but not the more common ASC (pycard) component, suggesting that potentially 342 the priming step of the NLRP3 inflammasome expression may be reduced. Priming of the 343 NLRP3 inflammasome is thought to be due to sensing of S. aureus lipoproteins and toll-like 344 receptor (TLR) 2 and 4 signaling(26, 60). While we did not see changes in expression in TLR-2 345 or -4 in macrophages (data not shown), we cannot rule out the possibility that SasD may be 346 involved in the sensing of S. aureus. When infected with sasD A50.1, RAW264.7 cells had a 347 reduction in il-1β and tnfα without a significant change in bacterial burden or bacterial phagocytosis. In BMDMs, we saw increased viability when infected with sasD A50.1 compared 349 to WT S. aureus. While we did not see any differences in pro-IL-1β, caspase 1, or caspase 1 350 cleavage at 3 hours post infection, there may be other cell death pathways involved such as 351 necroptosis. Blocking necroptosis has been shown to reduce bacterial burden and damage during 352 S. aureus pneumonia (29) Table 2 . Strain sasD A50.1 378 was generated via phage 11 transduction of sasD::Tn lysate into the wildtype JE2 strain, selected 379 with 5 μg/ml erythromycin and confirmed by PCR (Table 2) For RAW264.7 experiments, cells were infected for 1 hour in the absence of antibiotics, media 412 was replaced with antibiotic-and serum-free media with and without gentamicin (100 ug/ml) for 413 1 hour, then replaced with antibiotic-free media for an additional hour. At collection, cells were 414 lysed with 1% Triton X-100 at room temperature for 10 minutes and 50 μl was collected for CFU 415 determination. Phagocytosis was calculated by the equation ((CFU+gentamicin)/(CFU-416 gentamicin))*100. RLT (Qiagen) was added to the wells and collected and ran through a 417

Qiashredder and frozen at -80 o C until RNA extraction. For BMDM experiments, cells were 418 rested overnight, treated with 10 ng/ml IFNγ (R&D Systems) for 24 hours. BMDMs were 419 infected for 3 hours, washed and resuspended in antibiotic free RPMI media. BMDM viability 420 was determined by trypan blue (Gibco) staining and the Countess 3 automatic cell counter 421 (Invitrogen). BMDMs wells were combined and incubated in RIPA buffer (25mM Tris, 150 mM 422 NaCl, 1% NP-40, 0.1% SDS, 5 mM EDTA, 0.5% sodium deoxycholate) for 30 minutes at 4C 423 with agitation, centrifuged at 10,000 rpm for 10 minutes at 4C, and frozen at -80C until 424 Western Blot. Primary antibodies were rabbit anti-IL-1β (Abcam 254360), rabbit anti-caspase 1 425 (Abcam 138483), rabbit anti-caspase p20 (Invitrogen PA5-99390), and mouse anti-β-actin (Cell 426

Signaling 8H10D10). Samples were thawed, proteins were quantified using BCA protein assay 427 (Pierce), boiled in Laemmli buffer (Bio-Rad), and loaded on a 4-20% gel (Bio-Rad). Proteins 428 were transferred to a PVDF membrane using the Trans-Blot Turbo transfer system (Bio-Rad). 429

Blots were probed with primary antibodies and donkey anti-mouse or goat anti-rabbit secondary 430 antibodies conjugated to IRDye 800CW or 680RD flurophores (LI-COR). Blots were imaged 431 using the Odyssey CLx and analyzed using Image Studio (LI-COR). Relative protein expression 432 is normalized to beta-actin levels in each sample. 433 RNA extraction and qPCR. RNA was extracted from mouse lungs using the Absolutely Total 435 RNA Purification Kit (Agilent). RNA extraction from cell culture experiments were performed 436 using the Qiagen RNeasy kit (Qiagen). RNA was quantified and converted to cDNA using 437 iScript™ cDNA Synthesis Kit (Bio-Rad). Quantitative PCR was performed using SsoAdvanced 438

Universal Probes Supermix (Bio-Rad) and TaqMan primer-probe sets (ThermoFisher Scientific) 439 listed in Table 3 

Staphylococcus aureus colonization and non-influenza respiratory viruses: 465 Interactions and synergism mechanisms

Clinical outcomes in patients co-467 infected with COVID-19 and Staphylococcus aureus: a scoping review

Vital Signs: Epidemiology and Recent Trends in Methicillin-472

Methicillin-Susceptible Staphylococcus aureus Bloodstream Infections -473

United States

Global mortality associated with seasonal 476 influenza epidemics: New burden estimates and predictors from the GLaMOR Project

Critical illness from 2009 pandemic 480 influenza A virus and bacterial coinfection in the United States

Severe coinfection with seasonal influenza A (H3N2) virus and 483 Staphylococcus aureus--Maryland

Bacterial and viral co-492 infections complicating severe influenza: Incidence and impact among 507 U.S. patients

Injury and Sepsis Investigator's Network and the National Heart Ln, and Blood Institute 498

Critically ill children during the 2009-2010 499 influenza pandemic in the United States

The co-pathogenesis of influenza viruses with bacteria in the lung

Port d'Entrée for Respiratory Infections -Does the Influenza A 504 Virus Pave the Way for Bacteria? Front Microbiol

Vogel 506 SN. 2019. Influenza "Trains" the Host for Enhanced Susceptibility to Secondary 507 Bacterial Infection

Immune dysfunction and bacterial coinfections following 509 influenza

The immunology of influenza virus-associated 511 bacterial pneumonia

Influenza and Bacterial Superinfection: 513 Illuminating the Immunologic Mechanisms of Disease

Physiologic Changes Induced by Influenza A Virus Lead to Staphylococcus aureus

Biofilm Dispersion and Transition from Asymptomatic Colonization to Invasive Disease

Influenza-Induced Priming and Leak of 520

Human Lung Microvascular Endothelium upon Exposure to Staphylococcus aureus

Airway epithelial repair

regeneration, and remodeling after injury in chronic obstructive pulmonary disease

Role of neuraminidase in lethal synergism between 526 influenza virus and Streptococcus pneumoniae

The remarkably multifunctional fibronectin binding proteins of 596 Staphylococcus aureus

Increased virulence of a 598 fibronectin-binding protein mutant of Staphylococcus aureus in a rat model of 599 pneumonia

Staphylococcus aureus Virulence and Survival in Blood

Surface proteins that promote adherence of 604

Staphylococcus aureus to human desquamated nasal epithelial cells

Differential expression and roles of Staphylococcus aureus virulence determinants 608 during colonization and disease

Surface Proteins of Staphylococcus aureus

Staphylococcus aureus Nasal Colonization

The Staphylococcus aureus Cell Wall-Anchored Protein Clumping Factor A Is an 614

Characterization of novel LPXTG-containing proteins of 617

Staphylococcus aureus identified from genome sequences

Signal peptides direct surface 620 proteins to two distinct envelope locations of Staphylococcus aureus

Activation of inflammasome signaling mediates pathology 623 of acute P. aeruginosa pneumonia

IL-1β activation in response to Staphylococcus aureus lung 625 infection requires inflammasome-dependent and independent mechanisms

Neutrophil activation and acute lung injury

Eosinophils in innate immunity: an evolving 630 story

IL-33-mediated Eosinophilia Protects against Acute Lung Injury

Protection against Staphylococcus aureus 636 bacteremia-induced mortality depends on ILC2s and eosinophils

Distinct T-helper cell responses to Staphylococcus 639 aureus bacteremia reflect immunologic comorbidities and correlate with mortality

Orchestration of human macrophage NLRP3 642 inflammasome activation by

Necroptosis Promotes Staphylococcus aureus Clearance by 645

A 647 genetic resource for rapid and comprehensive phenotype screening of nonessential 648 Staphylococcus aureus genes

Immunologic recognition of influenza virus-infected cells. I

Generation of a virus-strain specific and a cross-reactive subpopulation of cytotoxic T 651 cells in the response to type A influenza viruses of different subtypes

Signaling Regulates Macrophage Phenotype During Influenza and Bacterial Super-656 Infection

SasD is Required for S. aureus Bacterial Pneumonia. A-E. Mice were infected with 697 1x10 8 CFU WT MRSA or MRSA lacking SasD (sasD A50.1) for 24 hours. A. Bacterial burden 698 in mice infected with MRSA for 24 hours

Cell differentials (C) and neutrophil cell ratios (D) of BAL cells. E. Total protein in the 700

Mice were infected with a lethal dose (2x10 8 CFU) of WT or MRSA lacking SasD 701 (sasD A50.1). G. Competitive index of WT and mutant sasD A50.1 MRSA in the lung

Whole lungs were collected in 2 ml PBS and homogenized and plated for CFU with and without 704 antibiotic selection. Competitive index is calculated as the ratio of Mutant

Two-way ANOVA with Sidak's 706 multiple comparisons (C), log-ranked Mantel Cox test (F), one sample T-test with H 0 set to 1 707 (1:1 ratio of mutant:WT) (G). * p<0.05, ** p<0.01,**** p<0.0001. N=4-8, combination of 708 several experiments

SasD is Required for Inflammation in Mice Infected with MRSA. Mice were 712 infected with 1x10 8 CFU WT MRSA or MRSA lacking SasD (sasD A50.1) for 24 hours

Cytokine protein levels in lung homogenate. B. Gene expression levels of cytokines and 714 inflammasome components relative to average WT levels in the lung

05, *** p<0.001. N=4, combination of several experiments

SasD Impacts Initiation of Host Defense against MRSA in Bacterial Pneumonia

Mice were infected with 1x10 8 CFU WT MRSA or MRSA lacking SasD (sasD A50.1) for 721

Bacterial burden in mice infected with MRSA for 6 hours. B. Total cells in the 722 bronchoalveolar lavage. C-D. Percentage (C) and total number (D) of macrophages in the BAL

Competitive index of WT 724 and mutant sasD MRSA in the lung. Male and female mice were infected with a 1:1 ratio of 725 WT:sasD A50.1 for a total dose of 1x10 8 CFU for 6 hours

PBS and homogenized and plated for CFU with and without antibiotic selection. Competitive 727 index is calculated as the ratio of Mutant CFU:WT CFU at 6 hpi. Statistics tested by student 728

Two-way ANOVA with Sidak's multiple comparisons (E), one 729 sample T-test with H 0 set to 1 (1:1 ratio of mutant:WT) (G)

0001. N=8, combination of several experiments, data graphed as mean  SEM

SasD is Required for Early Inflammation During Infection with MRSA. Mice 734 were infected with 1x10 8 CFU WT MRSA or MRSA lacking SasD (sasD A50.1) for 6 hours

Cytokine protein levels in lung homogenate. B. Gene expression levels of cytokines and 736 inflammasome components relative to average WT levels

05,**p<0.01, *** p<0.001. N=8, combination of several experiments, data 738 graphed as mean  SEM

SasD Increases Macrophage Inflammation and Decreases Survival. A-B

RAW264.7 macrophages infected with WT or sasD A50.1 MRSA for 3 hours at an MOI of 10

Macrophages were infected for one hour in the absence of antibiotics, media was then replaced 743 with antibiotic-and serum-free media with or without gentamicin for 1 hour, and changed to 744 antibiotic free media. CFU and transcript graphs show without gentamicin conditions

% Phagocytosed 746 bacteria is calculated by the following equation: ((average CFU with gentamicin)/(average CFU 747 without gentamicin))*100. B. Gene expression in RAW264.7 macrophages infected with WT or 748 sasD A50.1 MRSA for 3 hours. C-E. Bone marrow-derived macrophages (BMDMs) were 749 infected with WT or sasD A50.1 MRSA for 3 hours at an MOI of 50 in the absence of 750 antibiotics. C. Viability measured by trypan blue staining of BMDMs 3 hours post infection

Representative images (D) and quantification of western blot analyses (E) of BMDM levels of 752

One to three wells were combined per sample and protein levels are 753 normalized to beta actin in each sample. Arrows denote which band was used for quantification

N=4-7, combination of several 755 experiments, data graphed as mean  SEM

Growth rate (μ) was calculated off at least two 765 independent experiments using the equation A t =A t-1 *e μt (see methods). The μ max was calculated 766 as the average of the three highest μ rates. C. Relative expression of influenza PR8 M protein via 767 qPCR normalized to the average F-WT values. Statistics tested by Kruskal-Wallis with Dunn's 768 multiple comparisons correction (B-C). * p<0.05, ****p<0.0001. N=2-6, combination of several 769 experiments

clfA/B: clumping factor A/B, sdrC/D/E: serine-aspartate repeat containing protein C/D/E, isdB: 771 iron-regulated surface determinant B

Supplemental Figure 2: SasD Increases Inflammatory Cytokine and Decreases Lung 774

Homoestatic Gene Expression. A. Lung homogenate protein levels of cytokines in mice 775 infected with WT or sasD::Tn MRSA during bacterial pneumonia. B. Gene expression of lung 776 epithelial markers in mice infected with WT or sasD::Tn MRSA during bacterial pneumonia

Statistics tested by unpaired t test. * p<0.05, **p<0.01. N=2, combination of several 778 experiments, data graphed as mean  SEM

Supplemental Figure 3: Characterization of SasD during Pneumonia. A. WT or mutant 782

MRSA (see graphs) were grown overnight in tryptic soy broth and diluted 1:200 in a 96-well 783 microtiter plate in sexaplicate

Growth rate (μ) was calculated off at least two 787 independent experiments using the equation A t =A t-1 *e μt (see methods). The μ max was calculated 788 as the average of the three highest μ rates. C. Gene expression of neutrophil markers relative to 789 the average WT values. D. Histology scoring of H&E-stained lung sections. E. Gene expression 790 of lung epithelial markers relative to the average WT values. Statistics done by Mixed-effects 791 model with Dunnett's multiple comparisons correction (A), Kruskal-Wallis Test

N=2-6, combination of multiple experiments, data graphed as mean  SEM

Supplemental Figure 4: Characterization of early SasD infection. A. Histology scoring of 796

Gene expression of lung epithelial (B) and neutrophil (C) 797 markers relative to the average WT values. Statistics done by Two-way ANOVA with Sidak's 798 multiple comparisons correction (A), unpaired T test (B-C). * p<0.05, n=4-6, combination of 799 multiple experiments