Activation of autophagy by inflammatory signals limits IL-1β production by targeting ubiquitinated inflammasomes for destruction

Abstract

Autophagosomes delivers cytoplasmic constituents to lysosomes for degradation, whereas inflammasomes are molecular platforms activated by infection or stress that regulate the activity of caspase-1 and the maturation of interleukin 1β (IL-1β) and IL-18. Here we show that the induction of AIM2 or NLRP3 inflammasomes in macrophages triggered activation of the G protein RalB and autophagosome formation. The induction of autophagy did not depend on the adaptor ASC or capase-1 but was dependent on the presence of the inflammasome sensor. Blocking autophagy potentiated inflammasome activity, whereas stimulating autophagy limited it. Assembled inflammasomes underwent ubiquitination and recruited the autophagic adaptor p62, which assisted their delivery to autophagosomes. Our data indicate that autophagy accompanies inflammasome activation to temper inflammation by eliminating active inflammasomes.

Figure 1

Induction of inflammasomes induces autophagy. (a) Induction of GFP dots in differentiated THP-1 cells stably expressing GFP-LC3 following transfection with 1.5 μg/ml poly(dA-dT) for 6h and exposed to 3-MA (5mM), antibodies that neutralize IL-1β, or control antibodies. Top, quantitation of GPF-LC3 dots. Bottom, representative images of individual cells. NS-not significant, ** - p < 0.01, *- p< 0.05. Scale bars, 10 μm. (b) AIM2 and Beclin-1 expression in cells transfected with AIM2, Beclin-1 and control siRNA. The ratio of targeted to control is shown below. (c) Induction of GFP dots in differentiated THP-1 cells stably expressing GFP-LC3 and treated with indicated siRNAs, quantitated 6h after transfection with poly(dA-dT). (d) LC3-I and LC3-II expression and IL-1β production in Aim2^−/− or wild-type (WT) BMDMs transfected or not with poly(dA-dT) for 6 hrs and immunoblotted for LC3. The ratio shown is the induced LC3-I/LC3-II ratio divided by the basal LC3-I/LC3-II ratio. IL-1β production was measured in the supernatant following overnight culture. Data shown is representative of four repeats using two separate BMDM preparations. (e) LC3-I and LC3-II expression in Asc^−/−, Casp1^−/− or wild-type BMDMs primed with LPS (20 ng/ml) and then transfected with poly(dA-dT) or treated with nigericin for 4 h. The ratio is the induced LC3-I/LC3-II ratio divided by the basal LC3-I/LC3-II ratio. Shown is one representative LC3-I and LC3-II immunoblot of four repeats using two separate BMDM preparations. Data are from three separate experiments (a, b; means +/− s.d.).

Figure 2

Inflammasome activation leads to partial co-localization of autophagosomes and inflammasomes. (a) Confocal microscopy of differentiated THP-1 GFP-LC3 cells transfected with Cy3 labeled 1.5 μg/ml poly(dA-dT) for 3h. The yellow dots (Merge panel) indicate co-localization between green (LC-3) and red (Cy3) signals. (b) Confocal microscopy of differentiated THP-1 GFP-LC3 cells transfected with Cy3-poly(dA-dT) imaged every 15 minutes starting 1h after transfection. Coalescence between the Cy3 and GFP signal is shown in 60 and 75 minute panels. (c) Confocal microscopy of differentiated THP-1 GFP-LC3 cells transfected with poly(dA-dT) or sham transfected, fixed, and immunostained for ASC (red). The insert to the right is a 5X zoom of the indicated areas. (d) Confocal microscopy of differentiated THP-1 cells transfected with poly(dA-dT) and immunostained for Lamp-1 and ASC. Below each image is a z-axis projection. (e) Confocal microscopy of differentiated THP-1 GFP-LC3 cells treated with LPS (500ng/ml) +ATP (3mM) for 2h, immunostained for ASC (red), and imaged. The 1^st and 2^nd panels are 3D volume renderings reconstructed from z-stack images (2^nd panel is rotated 90^O and zoomed 3x). The same images are shown as a surface segmentation model in the 3^rd and 4^th panels. Scale bars shown are 10 μm. (f) Electron microscopy of differentiated THP-1 cells stimulated with LPS + ATP or transfected with poly(dA-dT) for 2h and immunostained for ASC visualized with 15 nm gold conjugated protein A. ASC immunostaining is indicated with red arrows. Scale bars are 500 nm.

Figure 3

The inflammasome sensor partial co-localizes with autophagosomes following inflammasome activation. (a) Confocal microscopy of differentiated GFP-LC3 expressing THP1 cells transfected with 1.5 μg/ml poly(dA-dT) for 2h, and then immunostained for endogenous AIM2. (b) Confocal microscopy of differentiated THP-1 cells transfected with poly(dA-dT) for 3h prior and immunostained for Lamp-1 and AIM2. (c) Confocal microscopy of differentiated THP-1 cells treated with LPS (500 ng/ml) and ATP (3mM) for 2 hrs, or not, prior to immunostaining for endogenous LC3 and NLRP3. Scale bars shown are 10 μm. Each experiment repeated a minimum of three times.

Figure 4

Manipulating autophagy affects inflammasomes. (a) Comparison of cell lysates from differentiated THP-1 cells, transfected or nor, with 1.5 μg/ml poly(dA-dT), treated or not with the indicated reagents, separated into an inflammasome enriched fraction (top) or an autophagosome enriched fraction (middle). Those fractions along with the total cell lysates (bottom) were immunoblotted for the indicated proteins. (b) Comparison of cell lysates and cell supernatants prepared from differentiated THP-1 cells, transfected or not, with poly(dA-dT) and immunoblotted for the indicated proteins. (c) Analysis of cell lysates and supernatants from mouse BMDMs primed with LPS (20 ng/ml) and treated with Alum, exposed to live M. tuberculosis or dead M. tuberculosis for 5h for LC3 expression IL-1β production. To detect IL-1β primed mouse BMDMs were infected with live TB overnight in the presence or absence of 3-MA (2mM). LPS+ATP treatment served as a positive control. Cell supernatants were collected and IL-1β levels immunoblotted. All experiments were repeated at least twice.

Figure 5

Beclin-1 and p62 are linked to inflammasome regulation (a) ASC immunoprecipitates from differentiated THP-1 cells transfected with 1.5 μg/ml poly(dA-dT) for 2 h, or not, were immunoblotted for Beclin-1 and ASC, stripped and re-immunoblotted for p62. (b) Confocal microscopy of similarly treated differentiated THP-1 cells immunostained for p62 (green) and ASC (red). Arrows indicate co-localized proteins. (c) Confocal microscopy of differentiated THP-1 cells treated with LPS (500 ng/ml) and ATP (3mM) for 2h, immunostained for p62 (green) and ASC (red). The left two images show a merged channel volume and surface rendering of a 3D reconstruction. The four right images show co-localization in a zoomed structure. Part of the signal was removed to uncover red (ASC) signal inside green (p62) stained area (arrows). (d) Immunoblotting of cell lysates from cells twice transfected with scrambled, Beclin-1, or p62 siRNAs for the indicated proteins. The ratio of targeted to control expression is shown. (e) Comparison of inflammasome and autophagosome fractions to total cell lysates from differentiated THP-1 cells transfected with scrambled, Beclin-1, or p62 siRNAs X2. The second transfection contained poly(dA-dT). The cell lysates prepared 4h after the last transfection were fractionated or not and immunoblotted for the indicated proteins. (f) Immunoblotting of supernatants and cell lysates from differentiated THP-1 cells treated as in part e. Supernatants were collected from cells cultured 6h in serum free media. Scale bars shown are 10 μm. All experiments were performed at least twice.

Figure 6

Induction of the AIM2 inflammasome triggers polyubiquitination of ASC. (a) ASC immunoprecipitates from differentiated THP1 cells transfected with 1.5 μg/ml poly(dA-dT) for 2h were immunoblotted for ubiquitin, stripped, re-blotted for K63-linked ubiquitin, stripped, and re-blotted for ASC. Experiment performed three times. (b) ASC immunoprecipitates from differentiated THP1 cells treated as above were prepared using a lysis buffer containing 1% (v/v) SDS, and heated for 95°C for 5 minutes to dissociate the proteins. The samples were diluted and ASC immunoprecipitated (24 kDa). The indicated proteins were immunoblotted. Experiment performed twice. (c) Confocal microscopy of differentiated THP-1 cells transfected with poly(dA-dT) for 2h immunostained for ASC (red) and polyubiquitin (green). Scale bars shown are 10 μm.

Figure 7

Inflammasome activity in primary human monocytes/macrophages can be modulated by autophagy. (a) Immunoblot of cell lysates prepared from human monocytes/macrophages prime with LPS (10 ng) and subsequently transfected with 1.5 μg/ml poly(dA-dT) or treated with nigericin (4 μM) for 6 hrs. The ratio shown is the induced LC3-I/LC3-II ratio divided by the basal LC3-I/LC3-II ratio. (b) Analysis of supernatants from monocytes/macrophages treated as indicated for 6h for IL-1β. Supernatants were immunoblotted or assayed for IL-1β by ELISA, p values < 0.01 (**). (c) Similar experiment as part b except nigericin was used to stimulate inflammasome activation. (d) Confocal microscopy of LPS primed human monocyte/macrophages transfected with poly(dA-dT) for 3 hours, treated with nigericin for 3h, or non-stimulated immunostained for p62 (green) and ASC (red). Arrows indicate the co-localized proteins. The second panel from the top is a 3D volume rendering reconstructed from confocal z-stack images showing p62 (green) and ASC structures (red). Scale bars shown are 10 μm. All experiments performed at least twice.

Figure 8

Involvement of RalB in inflammasome triggered autophagy. (a) Immunoblot of GTP-bound RalB collected by RALBP1-agarose affinity purification from cell lysates from LPS primed (20 ng/ml) mouse BMDMs exposed to LPS (500 ng/ml) and ATP (3 μM),1.5 μg/ml poly(dA-dT), or uric acid crystals (50 μg/ml). Immunoblotting visualized RalB-GTP in the pull downs and RalB in cell lysates. (b) Immunoblot of lysates and supernatants prepared from mouse BMDM transfected with RalB or control siRNA on 2 sequential days, primed with LPS and exposed to 1.5 μg/ml poly(dA-dT) for 6 h for RalB, LC3, actin, and IL-1β expression. To collect cell supernatants the cells were cultured overnight in serum free media. (c) Immunoblot of cell lysates from mouse BMDMs transfected with AIM2 or control siRNA on 2 sequential days, primed with LPS and 2h later exposed for 1h to poly(dA-dT) for the indicated proteins. A similar experiment using Aim2^−/− BMDMs is shown. (d) ASC immunoprecipitates from LPS primed mouse BMDM transfected with poly(dA-dT) for 1h, treated with LPS+ATP for 30 minutes, or non-stimulated immunoblotted as indicated. (e) Confocal microscopy of LPS primed mouse BMDMs transfected with poly(dA-dT) for 30 minutes were immunostained for RalB and AIM2 (first three panels); RalB and ASC (4^th and 5^th panels); and AIM2 and Beclin-1 (6^th and 7^th panel). Areas subjected to 3X electronic zoom are outlined by white squares. Scale bars shown are 10 μm. All experiments performed at least twice.