DJ-1 binds to Rubicon to Impair LC-3 Associated Phagocytosis

Abstract

The ability to effectively clear infection is fundamental to host survival. Sepsis, defined as dysregulated host response to infection, is a heterogenous clinical syndrome that does not uniformly clear intact bacterial or sterile infection (i.e., lipopolysaccharide). These findings were further associated with increased survival in DJ-1 deficient animals exposed to intact bacteria relative to DJ-1 deficient challenged with lipopolysaccharide. We analyzed bacterial and lipopolysaccharide clearance in bone marrow macrophages (BMM) cultured ex vivo from wild-type and DJ-1 deficient mice. Importantly, we demonstrated that DJ-1 deficiency in BMM promotes Rubicon-dependent increase in L3C-associated phagocytosis, non-canonical autophagy pathway used for xenophagy, during bacterial but not lipopolysaccharide infection. In contrast to DJ-1 deficient BMM challenged with lipopolysaccharide, DJ-1 deficient BMM exposed to intact bacteria showed enhanced Rubicon complexing with Beclin-1 and UVRAG and consistently facilitated the assembly of complete autophagolysosomes that were decorated with LC3 molecules. Our data shows DJ-1 impairs or/and delays bacterial clearance and late autophagolysosome formation by binding to Rubicon resulting in Rubicon degradation, decreased L3C-associated phagocytosis, and decreased bacterial clearance in vitro and in vivo - implicating Rubicon and DJ-1 as critical regulators of bacterial clearance in experimental sepsis.

Fig. 1. DJ-1 deficient mice are differentially protected in sterile inflammation versus bacterial models through altered macrophage responses.

A Percent survival of wild-type (WT) and DJ-1 deficient (DJ-1^−/−) mice at 7 days after intraperitoneal administration of LPS (15 mg/kg) (n = 5 saline, n = 11-12 LPS, *p ≤ 0.05, **p ≤ 0.01, Log-rank/Mantel–Cox analysis). B Percent survival of WT and DJ-1^−/− mice at 7 days after intratracheal (i.t.) E. coli (7 × 10⁷ CFU) infection (n = 5 mice each). C Survival studies determined DJ-1 deficiency in myeloid cells conferred survival advantage (n = 12–15, *p ≤ 0.05, Log-rank/Mantel–Cox analysis) in a cecal ligation and puncture sepsis animal model. D Peritoneal and (E) bronchoalveolar lavage fluid (BALF) total cell counts from WT and DJ-1^−/− mice and exposed to vehicle or LPS (15 mg/kg) and (F) BALF total cell counts from WT and DJ-1^−/− chimeric mice following CLP surgery. The bars are mean ± SEM of total cell counts. G Measurement of phagocytic function of WT and DJ-1^−/− BMMs using E. coli pHrodo Bioparticles. Assays were conducted in 96 well plates in triplicate and repeated twice. Data were normalized to baseline values at time 0 and presented as % fluorescence intensity. Results are presented as means ± SEM (*p ≤ 0.05, one-way ANOVA). Each group had peritoneal cells pooled from four animals and separated into four replicates each.

Fig. 2. DJ-1 deficient bone marrow macrophages have impaired p62 flux with LPS treatment but not E. coli bacteria.

A Representative Western blots showing protein expression of p62, LC3-II, DJ-1 and GAPDH in WT and DJ-1^−/− BMMs following 24 h of saline or LPS (1 µg/mL) treatment alone, or in the presence of control and DJ-1 siRNA (siR) or control adenovector and DJ-1 overexpressing adenovector. B Representative Western blot showing time dependent protein expression of p62 and GAPDH protein expression in response to LPS (1 µg/mL). C Representative immunofluorescence images of LC3 (red) and nucleus (DAPI, blue) following 24 h of saline or LPS (1 µg/mL) treated BMMs, scale bar = 50 µm. D Representative immunofluorescence images of p62 expression (green) co-stained with lysosomal associated membrane protein 2 A (LAMP-2A) in WT and DJ-1^−/− BMMs following 2 h of saline, LPS (1 µg/mL) or E. coli bacteria (EB) (7 × 10⁷ CFU). E Representative immunofluorescence images showing protein expression of LC3, Cathepsin D co-stained with nuclear DAPI in WT and DJ-1^−/− BMMs infected with 15 min of E. coli bacteria, scale bar = 50 μm. F Representative Western blot showing protein expression of p62 and LC3-I, II in WT and DJ-1^−/− BMMs in vehicle and 1 h pre-treatment with Bafilomycin A, followed by 24 h exposure to LPS (1 µg/mL) or EB (7 × 10⁷ CFU).

Fig. 3. DJ-1 deficient bone marrow macrophages challenged exclusively with lipopolysaccharide fail to assemble complete auto phagolysosomes.

A, B Representative TEM images of WT and DJ-1^−/− BMMs exposed to saline or LPS (1 µg/mL) for 90 min at 11,500x magnification, scale bar = 3 µ. Quantification of different autophagic vesicles from WT and DJ-1^−/− BMMs following saline or LPS treatment. Data represent quantification from 15 cells per group, *p ≤ 0.05. C Representative TEM images of wild-type (WT) and DJ-1 deficient (DJ-1^−/−) BMMs treated with E. coli (7 × 10⁷ CFU; 11,500x magnification) for 30 min, scale bar = 3 μ. Quantification of phagocytic vacuoles, number of E. coli phagocytosed per cell, vacuoles with peri bacterial space or apposed membraned and intact versus degraded bacteria, *p < 0.05. D Representative TEM images of wild-type (WT) and DJ-1 deficient (DJ-1^−/−) BMMs treated with S. aureus (1 × 10⁷ CFU; 11,500x magnification) for 30 min. Quantification of phagocytic vacuoles, number of S. aureus phagocytosed per cell, vacuoles with peri bacterial space or apposed membraned and intact versus degraded bacteria. Data are presented as means ± SEM (*p ≤ 0.05 compared with WT, n = 20–25 cells each). E Representative immunofluorescence images showing protein expression of LC3, Cathepsin D co-stained with nuclear DAPI in WT and DJ-1^−/− BMMs infected with 15 min of zymosan, scale bar = 50 μm. Quantification of total number of zymosan phagocytosed per cell and LC3 + encapsulated zymosan per cell. Data are presented as individual cells ± SEM from 25 cells. F Representative quantification of zymosan uptake in acidic lysosomes as assessed by flow cytometry of Alexa Fluor 488 pHrodo zymosan WT (black) and DJ-1^−/− (red) BMMs 30 min after incubation. Right panel – Average mean fluorescence intensity (MFI) of three replicates of WT and DJ-1^−/− BMMs with Alexa Fluor 488 pHrodo zymosan lysosomal uptake.

Fig. 4. Silencing of Rubicon in DJ-1^−/− BMMs reverses enhanced bacterial clearance through LAP.

A Representative Western blot showing protein expression of Rubicon, DJ-1 and GAPDH in BMMs treated with LPS (1 µg/mL), E. coli bacteria (EB; 7 × 10⁷ CFU) or H₂O₂ (100 μM). B Representative Western blots showing protein expression of Rubicon and p22 in WT and DJ-1^−/− BMMs following 24 h of vehicle (saline), LPS (1 µg/mL) or live E. coli (EB) (7 × 10⁷ CFU). C Immunoprecipitation assay (Ubiquitin pull down) in BMMs isolated from WT and DJ-1^−/− mice and treated with vehicle, LPS (1 µg/mL), E. coli bacteria (EB; 7 × 10⁷ CFU) for 24 h. D Representative Western blot showing protein expression of Rubicon and β-actin in WT BMMs following 24 h exposure to vehicle (saline), LPS (1 µg/mL) or live E. coli (EB) (7 × 10⁷ CFU) alone or in combination with the proteasome inhibitor MG 132 (10 µg/mL). E Immunoprecipitation assay (DJ-1 and Rubicon pull down) in BMMs isolated from WT and DJ-1^−/− mice and treated with vehicle, LPS (1 µg/mL), E. coli bacteria (EB; 7 × 10⁷ CFU) for 24 h. F Immunoprecipitation assay (DJ-1 and Rubicon pull down) shows Rubicon binds to DJ-1 in THP-1 cells following exposure to vehicle, LPS (1 µg/mL) or E. coli bacteria (EB; 7 × 10⁷ CFU) for 24 h. G Immunoprecipitation assay (DJ-1 and maltose-binding fusion protein (MBP)-Rubicon pull down) of pure protein extracts grown in and harvested from E. coli bacteria. MBP fused to the His-Rubicon N-terminus (RNT) or Hi-Rubicon C-terminus (RCT). H Immunoprecipitation assay (Beclin and UVRAG pull down) in WT and DJ-1^−/− BMMs treated with vehicle, LPS (1 µg/mL), E. coli bacteria (EB; 7 × 10⁷ CFU) for 24 h. I Quantification of E. coli and S. aureus bacteria in WT and DJ-1^−/− BMMs transfected with scramble control siRNA (ctrl siRNA) or Rubicon siRNA. Data are presented as means ± SEM (n = 3–6, *p ≤ 0.05).