The autophagic response to Staphylococcus aureus provides an intracellular niche in neutrophils

Abstract

Staphylococcus aureus is a major human pathogen causing multiple pathologies, from cutaneous lesions to life-threatening sepsis. Although neutrophils contribute to immunity against S. aureus, multiple lines of evidence suggest that these phagocytes can provide an intracellular niche for staphylococcal dissemination. However, the mechanism of neutrophil subversion by intracellular S. aureus remains unknown. Targeting of intracellular pathogens by macroautophagy/autophagy is recognized as an important component of host innate immunity, but whether autophagy is beneficial or detrimental to S. aureus-infected hosts remains controversial. Here, using larval zebrafish, we showed that the autophagy marker Lc3 rapidly decorates S. aureus following engulfment by macrophages and neutrophils. Upon phagocytosis by neutrophils, Lc3-positive, non-acidified spacious phagosomes are formed. This response is dependent on phagocyte NADPH oxidase as both cyba/p22phox knockdown and diphenyleneiodonium (DPI) treatment inhibited Lc3 decoration of phagosomes. Importantly, NADPH oxidase inhibition diverted neutrophil S. aureus processing into tight acidified vesicles, which resulted in increased host resistance to the infection. Some intracellular bacteria within neutrophils were also tagged by Sqstm1/p62-GFP fusion protein and loss of Sqstm1 impaired host defense. Together, we have shown that intracellular handling of S. aureus by neutrophils is best explained by Lc3-associated phagocytosis (LAP), which appears to provide an intracellular niche for bacterial pathogenesis, while the selective autophagy receptor Sqstm1 is host-protective. The antagonistic roles of LAP and Sqstm1-mediated pathways in S. aureus-infected neutrophils may explain the conflicting reports relating to anti-staphylococcal autophagy and provide new insights for therapeutic strategies against antimicrobial-resistant Staphylococci.Abbreviations: ATG: autophagy related; CFU: colony-forming units; CMV: cytomegalovirus; Cyba/P22phox: cytochrome b-245, alpha polypeptide; DMSO: dimethyl sulfoxide; DPI: diphenyleneiodonium; EGFP: enhanced green fluorescent protein; GFP: green fluorescent protein; hpf: hours post-fertilization; hpi: hours post-infection; Irf8: interferon regulatory factor 8; LAP: LC3-associated phagocytosis; lyz: lysozyme; LWT: london wild type; Map1lc3/Lc3: microtubule-associated protein 1 light chain 3; NADPH oxidase: nicotinamide adenine dinucleotide phosphate oxidase; RFP: red fluorescent protein; ROS: reactive oxygen species; RT-PCR: reverse transcriptase polymerase chain reaction; Sqstm1/p62: sequestosome 1; Tg: transgenic; TSA: tyramide signal amplification.

Keywords: Autophagy; NADPH oxidase; ROS; Staphylococcus aureus; lc3-associated phagocytosis (LAP); neutrophil; zebrafish.

Figure 1.

Different kinetics of the Lc3-mediated response in macrophages and neutrophils infected by S. aureus. (A) Confocal photomicrographs are shown as maximum intensity projections of fixed CMV:GFP-Lc3 embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus at 1 (top panel) and 4 hpi (bottom panel). Phagocytes are seen containing bacteria with (+) or without (-) Lc3 aggregates. Images shown are representative of three independent experiments. Scale bars: 10 µm. (B) Quantification of Lc3 associations with intracellular S. aureus in fixed CMV:GFP-Lc3 embryos at 1, 2, and 4 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (5–6 larvae per experiment per group). For 1 hpi, 175 infected phagocytes in 18 larvae were analyzed. For 2 hpi, 198 infected phagocytes in 17 larvae were analyzed. For 4 hpi, 204 infected phagocytes in 17 larvae were analyzed. One-way ANOVA with Bonferroni’s posttest was used. *** P < 0.001. (C) Confocal photomicrographs are shown as maximum intensity projections of fixed CMV:GFP-Lc3 transgenic embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus. Embryos were fixed at 1 (top panel) and 6 hpi (bottom panel) and chemically stained for Mpx activity (TSA, magenta). TSA-negative macrophages are seen containing bacteria with (M+) or without (M-) Lc3 aggregates as well as TSA-positive neutrophils containing bacteria with (N+) and without Lc3 aggregates (N-). The images shown are representative of three independent experiments. Scale bars: 10 µm. (D) Quantification of Lc3 associations with intracellular S. aureus within macrophages (black bars) and neutrophils (gray bars) of fixed CMV:GFP-Lc3 transgenic embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus at 1, 2, 4 and 6 hpi. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (5–6 larvae per experiment per group). For 1 hpi, 151 infected macrophages and 60 infected neutrophils in 18 larvae were analyzed. For 2, hpi 146 infected macrophages and 64 neutrophils in 18 larvae were analyzed. For 4 hpi, 159 infected macrophages and 56 neutrophils in 17 larvae were analyzed. For 6 hpi, 161 infected macrophages and 70 infected neutrophils in 18 larvae were analyzed. Two-way ANOVA with Bonferroni’s posttest was used. **** P < 0.0001, ns – not significant

Figure 2.

Generation of a lyz:RFP-GFP-Lc3 transgenic line in zebrafish confirms the Lc3-mediated response to S. aureus within neutrophils. (A) Schematic of the pDEST(lyz:RFP-GFP-Lc3) construct encoding the fusion RFP-GFP-Lc3 protein under the neutrophil-specific lyz promoter. In addition, the heart marker myl7-driven GFP is used to facilitate the screening of positive larvae. (B) Confocal images at maximum projection of the Lc3-mediated response at 1 hpi in live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus. Lyz-positive neutrophils are seen containing bacteria with (N+) or without (N-) Lc3 aggregates. Lyz-negative macrophages are also seen containing bacteria (M). The images shown are representative of three independent experiments. Arrows indicate spacious Lc3-positive vesicles, whereas arrowheads show tightly wrapped Lc3-associated bacteria. Scale bars: 10 µm. (C) Quantification of Lc3 associations with intracellular S. aureus within infected neutrophils of live lyz:RFP-GFP-Lc3 embryos at 1 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (6 larvae per experiment). 79 infected neutrophils were analyzed in 18 larvae

Figure 3.

NADPH oxidase is required for the Lc3-mediated response to S. aureus infection. (A) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response in control (top panel) and cyba knockdown (bottom panel) of fixed CMV:GFP-Lc3 embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus. Embryos were fixed at 1 hpi and chemically stained for Mpx activity (TSA, magenta). Scale bars: 10 µm. (B) Quantification of Lc3 associations with intracellular S. aureus within infected phagocytes of fixed CMV:GFP-Lc3 control and cyba knockdown embryos at 1, 2, and 8 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (5–6 larvae per experiment per group). For 1 hpi, 205 infected phagocytes in 18 larvae analyzed. For 2 hpi, 188 infected phagocytes in 18 larvae were analyzed. For 8 hpi, 161 infected phagocytes in 17 larvae were analyzed. Two-way ANOVA with Bonferroni’s posttest was used. **** P < 0.0001, ** P < 0.01, ns – not significant. (C) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response at 1 hpi in control (top panel) and cyba knockdown live lyz:RFP-GFP-Lc3 embryos infected with mCherry-labeled S. aureus. Scale bars: 10 µm. (D) Quantification of Lc3 associations with intracellular S. aureus within infected neutrophils of live lyz:RFP-GFP-Lc3 embryos at 1 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (6 larvae per experiment per group). 96 infected neutrophils in 18 control larvae were analyzed. 92 infected neutrophils in 18 cyba knockdown larvae were analyzed. Unpaired two-tailed t-test was used. **** P < 0.0001)

Figure 4.

Formation of NADPH oxidase-dependent Lc3-positive vesicles containing S. aureus in neutrophils is detrimental for the infected host. (A and B) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response at 1 hpi in irf8-only (A) and irf8 + cyba knockdown (B) fixed CMV:GFP-Lc3 embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus. Scale bars: 10 µm. (C) Quantification of Lc3 associations with intracellular S. aureus within infected neutrophils of irf8-only and irf8 + cyba knockdown fixed CMV:GFP-Lc3 embryos at 1 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (5–6 larvae per experiment per group). 174 infected neutrophils in 16 irf8-only knockdown larvae were analyzed. 210 infected neutrophils in 17 irf8 + cyba knockdown larvae were analyzed. Unpaired two-tailed t-test was used. **** P < 0.0001. (D and E) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response at 1 hpi in control (DMSO) (D) and DPI-treated (E) irf8 knockdown live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of mCherry-labeled S. aureus. The images shown are representative of three independent experiments. Scale bars: 10 µm. (D’ and E’). Zoomed-in fragments of photomicrographs d and e. (F) Quantification of Lc3 associations with intracellular S. aureus within infected neutrophils of control (DMSO) and DPI-treated irf8 knockdown live lyz:RFP-GFP-Lc3 embryos at 1 hpi with approximately 1500 CFU. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments. 252 infected neutrophils in 18 DMSO-treated larvae were analyzed. 208 infected neutrophils in 17 DPI-treated larvae were analyzed. Unpaired two-tailed t-test was used.**** P < 0.0001. (G) Survival of irf8-only or irf8 + cyba knockdown zebrafish larvae following intravenous injection with approximately 1500 CFU of S. aureus at 30 hpf (25 larvae per group). This result is representative of three independent experiments. Survival curves were compared using a log-rank (Mantel-Cox) statistical test. ** P < 0.01. (H and I) The CFU counts of the irf8-only or irf8 + cyba knockdown larvae infected intravenously with approximately 1500 CFU of S. aureus at 1 and 3 hpi (H) or 28 hpi (I). At each timepoint, larvae were sacrificed, homogenized, and the recovered staphylococci were enumerated by serial dilutions. The red line represents the level of the initial inoculum, whereas the green lines represent the mean value of each group. Data are obtained from 3 independent experiments (n of larvae per timepoint ≥24). One-way ANOVA with Bonferroni’s posttest was used for (H) and an unpaired two-tailed t-test was used for (I). * P < 0.05, *** P < 0.001

Figure 5.

The Lc3-mediated response in neutrophils occurs to both live and heat-killed S. aureus, but spacious Lc3-positive vesicles are formed only with live bacteria. (A and B) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response at 1 hpi in live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of live (A) or heat-killed (B) mCherry-labeled S. aureus. Scale bars: 10 µm. (C) Quantification of Lc3 associations with intracellular S. aureus at 1 hpi within neutrophils of live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of live or heat-killed (HK) mCherry-labeled S. aureus. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments. 80 infected neutrophils in 16 larvae injected with live bacteria were analyzed. 72 infected neutrophils in 16 larvae injected with heat-killed bacteria were analyzed. Unpaired two-tailed t-test was used. ns – not significant. (D) Quantification of neutrophils with spacious S. aureus-containing phagosomes at 1 hpi within live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of live or heat-killed (HK) mCherry-labeled S. aureus. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments. 46 infected Lc3-positive neutrophils in 16 larvae injected with live bacteria were analyzed. 44 infected neutrophils Lc3-positive in 16 larvae injected with heat-killed bacteria were analyzed. Unpaired two-tailed t-test was used. **** P < 0.0001

Figure 6.

Loss of Cyba leads to increased acidification of neutrophil-ingested S. aureus. (A and B) Confocal photomicrographs shown as maximum intensity projections of the control (A) and cyba knockdown (B) live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of S. aureus stained with pHrodo Red and Fluorescein pH-indicating dyes. Green bacteria indicate that they are localized in neutral pH, whereas red bacteria are acidified. Scale bars: 10 µm. (C) Quantification of acidification rates at 1.5, 2, and 3 hpi of intracellular S. aureus within control and cyba knockdown neutrophils of live lyz:RFP-GFP-Lc3 embryos infected with approximately 1500 CFU of S. aureus. Data are shown as mean ± standard (SD) obtained from three independent experiments (6 larvae per group). For 1.5 hpi, 170 infected neutrophils in 18 control larvae, and 168 infected neutrophils in 18 larvae were analyzed. For 2 hpi, 188 infected neutrophils in 18 control larvae and 214 infected neutrophils in 18 cyba larvae were analyzed. For 3 hpi, 154 infected neutrophils in 18 control larvae and 204 neutrophils in 18 cyba larvae were analyzed. Two-way ANOVA with Bonferroni’s posttest was used. *** P < 0.001, **** P < 0.0001. (D) Examples of control neutrophils (indicated in panel a) with LAPosomes containing non-acidified S. aureus at 2 hpi. Contrast was enhanced equally for both channels to visualize LAPosomes. (E) A rare example of a control neutrophil with a LAPosome containing acidified S. aureus at 3 hpi. (F) Quantification of non-acidified LAPosomes within control neutrophils of live lyz:RFP-GFP-Lc3 embryos at 1.5, 2 and 3 hpi. For 1.5 hpi, 142 of LAPosomes were analyzed. For 2 hpi, 157 of LAPosomes were analyzed. For 3 hpi, 132 of LAPosomes were analyzed

Figure 7.

S. aureus within neutrophils is targeted by selective autophagy receptor Sqstm1. (A and B) Confocal photomicrographs of Sqstm1-mediated response at 1 hpi in live lyz:GFP-Sqstm1 embryos infected with mCherry-labeled S. aureus. The fusion GFP-Sqstm1 protein colocalizes with intracellular bacteria (A) or with apparent vesicles containing the bacteria (B). The images shown are representative of three independent experiments. Scale bars: 10 µm. (C) Quantification of Sqstm1 associations with intracellular S. aureus at 2 hpi within neutrophils of live lyz:GFP-Sqstm1 embryos infected with approximately 1500 CFU of live or heat-killed mCherry-labeled S. aureus. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments (5–6 larvae per group per experiment). 109 infected neutrophils in 18 larvae injected with live bacteria were analyzed. 101 infected neutrophils in 17 larvae injected with heat-killed bacteria were analyzed. Unpaired two-tailed t-test was used. **** P < 0.0001

Figure 8.

Diminished Sqstm1 recruitment to S. aureus in LAP-deficient neutrophils. (A) Confocal photomicrographs of Sqstm1-mediated response at 2 hpi in live control (top panel) and cyba knockdown (bottom panel) lyz:GFP-Sqstm1 embryos infected with mCherry-labeled S. aureus. The images shown are representative of three independent experiments. Scale bars: approximately 10 µm. (B) Quantification of Sqstm1 associations with intracellular S. aureus at 2 hpi within neutrophils of live lyz:GFP-Sqstm1 infected with mCherry-labeled S. aureus. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments. 66 infected neutrophils in 17 control larvae were analyzed. 90 infected neutrophils in 19 cyba knockdown larvae were analyzed. Unpaired two-tailed t-test was used. **** P < 0.0001

Figure 9.

Loss of Sqstm1 leads to increased susceptibility to S. aureus infection. (A and B) Confocal photomicrographs are shown as maximum intensity projections of the Lc3-mediated response at 1 hpi within infected macrophages and neutrophils of control (A) and sqstm1 knockdown (B) fixed CMV:GFP-Lc3 embryos infected with mCherry-labeled S. aureus. Embryos were fixed at 1 hpi and chemically stained for Mpx activity (TSA, magenta). TSA-negative macrophages are seen containing bacteria with (M+) or without (M-) Lc3 aggregates. TSA-positive neutrophils contain bacteria with (N+) and without Lc3 aggregates (N-). The images shown are representative of three independent experiments. Scale bars:10 µm. (C) Quantification of Lc3 associations with intracellular S. aureus at 1 hpi within infected macrophages and neutrophils of control and sqstm1 knockdown fixed CMV:GFP-Lc3 embryos. Data are shown as mean ± standard deviation (SD) obtained from three independent experiments. 174 infected macrophages and 72 neutrophils were analyzed in 18 control larvae. 165 infected macrophages and 68 neutrophils were analyzed in sqstm1 knockdown larvae. One-way ANOVA with Bonferroni’s posttest was used. ns – not significant. (D) Survival of irf8-only or irf8 + sqstm1 knockdown zebrafish larvae following intravenous injection with S. aureus at 30 hpf (≥69 larvae per group). This result is obtained from three independent experiments. Survival curves were compared using a log-rank (Mantel-Cox) statistical test. ** P < 0.01

Figure 10.

Proposed model of fate of S. aureus within neutrophils. Staphylococci are internalized by neutrophils and trapped within LC3-associated phagosomes (LAPosomes) triggered by NADPH oxidase. LAPosomes do not get acidified and provide a replication niche for internalized S. aureus, which eventually will damage the phagosomal membrane and escape into the cytoplasm. This is sensed by Sqstm1/p62 – a member of the selective autophagy machinery, which could lead to the formation of autophagosomes containing staphylococci and leading to pathogen inactivation