PPARγ inhibited tumor immune escape by inducing PD‐L1 autophagic degradation

Abstract

Blockade of the programmed death 1 (PD‐1)/ programmed death ligand 1 (PD‐L1) immune checkpoint could increase antitumor immunotherapy for multiple types of cancer, but the response rate of patients is about 10%–40%. Peroxisome proliferator activated receptor γ (PPARγ) plays an important role in regulating cell metabolism, inflammation, immunity, and cancer progression, while the mechanism of PPARγ on cancer cell immune escape is still unclear. Here we found that PPARγ expression exhibits a positive correlation with activation of T cells in non‐small‐cell lung cancer (NSCLC) by clinical analysis. Deficiency of PPARγ promoted immune escape of NSCLC by inhibiting T‐cell activity, which was associated with increased PD‐L1 protein level. Further analysis showed that PPARγ reduced PD‐L1 expression independent of its transcriptional activity. PPARγ contains the microtubule‐associated protein 1A/1B‐light chain 3 (LC3) interacting region motif, which acts as an autophagy receptor for PPARγ binding to LC3, leading to degradation of PD‐L1 in lysosomes, which in turn suppresses NSCLC tumor growth by increasing T‐cell activity. These findings suggest that PPARγ inhibits the tumor immune escape of NSCLC by inducing PD‐L1 autophagic degradation.

(1) PPARγ acts as an autophagic receptor for PD‐L1 autophagic degradation. (2) Agonist enhanced PD‐1 antibody against NSCLC tumor immune escape. (3) PPARγ agonist could be a therapic strategy for antitumor immunotherapy of NSCLC. (1) PPARγ induced PD‐L1 autophagic degradation. (2) Loss of PPARγ led to NSCLC tumor immune escape. (3) PPARγ acts as autophagic receptor for PD‐L1 autophagic degradation

1. INTRODUCTION

Programmed death 1 (PD‐1)/ programmed death ligand 1 (PD‐L1) is an important immune checkpoint signal, and PD‐L1 is highly expressed on many types of cancer cells that bind to PD‐1 on T cells, resulting in inhibition of T‐cell activity and proliferation. ¹ , ² Blockade of the PD‐1/PD‐L1 immune checkpoint could facilitate antitumor immunotherapy for multiple types of cancer, including in non‐small‐cell lung cancer (NSCLC), melanoma, breast cancer, and gastric cancer. ³ However, the response rates of patients with PD‐1/PD‐L1 blockade immunotherapy are less than 40%, with an unclear mechanism. ⁴ Although PD‐L1 is aberrantly expressed by activation of mammalian target of rapamycin (mTOR), Nuclear factor‐κB (NFκB), signal transducer of activators of transcription (STAT), mitogen‐activated protein kinase (MAPK), and cellular‐myelocytomatosis viral oncogene (c‐Myc) pathways, ⁵ , ⁶ PD‐L1 undergoes ubiquitination and degradation by multiple pathways such as STIP1 homology and U‐box containing protein 1 (STUB1), ⁷ Cullin3^SPOP, ⁸ ER‐associated protein degradation (ERAD), ⁹ β‐transducin repeat‐containing protein (β‐TrCP), ¹⁰ , ¹¹ and cyclinD‐CDK4/SPOP/Cdh1. ⁸ In addition to proteasomal‐dependent degradation, huntingtin interacting protein 1 related (H1P1R) ¹² and CKLF like MARVEL transmembrane domain containing 6 (CMTM6) ¹³ could induce PD‐L1 autophagic degradation, therefore induction of PD‐L1 degradation could enhance antitumor immunotherapy. ¹⁴ As one of peroxisome‐proliferator‐activated receptors (PPARs), PPARγ regulates cell metabolism, inflammation, immunity, and cancer progression. ¹⁵ , ¹⁶ , ¹⁷ Activation of PPARγ could promote immune escape of muscle‐invasive bladder cancer, ¹⁸ melanoma, ¹⁹ and breast cancer. ²⁰ In contrast, PPARγ agonist enhances PD‐L1 antibody blockade therapy in microsatellite stable (MSS) colorectal cancer. ²¹ These findings suggest that the role of PPARγ in cancer cell immune escape is contradictory and the mechanism is unclear. Here we found that deficiency of PPARγ promoted NSCLC immune escape by increasing PD‐L1 protein stability, which is independent of PPARγ transcription activity.

2. RESULTS

2.1. PPARγ is a positive regulator of T cells

To detect the role of PPARγ in T‐cell activity, bioinformatic analysis was performed using TIMER and TISIDB software. The results show that there is a positive correlation between PPARγ expression and T‐cell infiltration (Figure 1A) or T‐cell activation (Figure 1B) in lung adenocarcinoma (LUAD) but not in lung squamous cell carcinoma (LUSC) (Figure 1A,B). TCGA database analysis showed that the gene expression of PPARγ was low compared to normal in LUAD and LUSC (Figure 1C). Consistently, immune staining analysis using tumor tissues showed that the PPARγ protein level was low compared to normal in LUAD and LUSC (Figure 1D). Further analysis showed that a high expression of PPARγ corresponding to low expression of PD‐L1 and a high level of T cells was observed in LUAD tumor patient tissues (Figure 1E). These findings suggest that PPARγ is a positive regulator of CD⁺8 T cells.

FIGURE 1.

PPARγ is a positive regulator of T cells. (A) T‐cell infiltration was assayed using TIMER2.0 (cistrome.org). (B) Activation of T‐cell abundance was assayed using TISIDB (hku.hk). (C) PPARγ expression in LUAD or LUSC was assayed using ualcan.path.uab.edu/home. (D) Expression of PPARγ in LUAD or LUSC patient tissues was assayed by immunohistochemical analysis (n = 48). (E) Immunohistochemical staining of PPARγ, PD‐L1, and T cells using LUAD tumor (n = 48). The correlation of PPARγ with PD‐L1 or CD⁺8 T was assayed. Results are expressed as mean ± SEM. *p < 0.05.

2.2. Loss of PPARγ promoted cancer cell immune escape by increasing the PD‐L1 protein level

As an important regulator, PPARγ exhibits an antitumor role. ¹⁵ , ¹⁶ Consistently, induction of PPARγ expression could inhibit tumor growth (Figure S1A,B). However, the effect of PPARγ on tumor immune escape is still unclear. Clinical analysis suggests that PPARγ is a positive regulator of CD⁺8 T cells, while high expression of PD‐L1 could inhibit T‐cell activity. ¹⁴ To further detect the effect of PPARγ on the PD‐L1 protein level, Western blot analysis was performed using the deficiency of PPARγ in H520 and H1975 cells. The results showed that loss of PPARγ led to increased PD‐L1 protein level (Figures 2A and S2) since the binding of PD‐L1 on cancer cells to PD‐1 on T cells inhibits T‐cell activity. ¹⁴ To detect the effect of PPARγ on T‐cell activity, co‐cultured analysis of Jurkat T cells with PPARγ‐deleted H520 cells was analyzed. The results showed that deleted PPARγ reduced T cell IL‐2 production (Figure 2B). To further detect the effect of PPARγ on cancer cell immune escape, a syngeneic tumor model was assayed using PPARγ gene knockout lewis lung carcinoma (LLC) cells. The results showed that PPARγ gene knockout promoted tumor growth (Figure 2C), which was consistent with increased PD‐L1 protein level (Figure 2D). Flow cytometry analysis using isolated tumor cells showed that deleted PPARγ significantly reduced cytotoxic T‐cell activity (Figure 2E,F), suggesting that loss of PPARγ promoted cancer cell immune escape by inhibiting T‐cell activity, which was associated with increased PD‐L1 protein level.

FIGURE 2.

Loss of PPARγ promoted cancer cell immune escape by increasing the PD‐L1 protein level. (A) Western blot analysis of WT or PPARγ^−/− H520 cells. (B) IL‐2 production was detected using a medium of co‐cultured Jurkat cells and cancer cells (WT or PPARγ^−/− H520 cells). Results are expressed as mean ± SEM (n = 3). *p < 0.05. (C) WT or PPARγ^−/− LLC cells were inoculated subcutaneously into C57BL/6 mice. Tumor volume was detected. Results are expressed as mean ± SEM, n = 6. *p < 0.05. (D) Relative cell surface PD‐L1 expression shown in tumor cells. MFI, median fluorescence intensity. Results are expressed as mean ± SEM (n = 6). *p < 0.05. (E and F) Flow cytometry analysis of intracellular IFN‐γ or perforin in T cells from isolated tumor cells. Results are expressed as mean ± SEM (n = 6). *p < 0.05.

2.3. PPARγ induced PD‐L1 autophagic degradation independent of its transcription activity

The above results show that loss of PPARγ led to increased PD‐L1 protein level. In contrast, overexpression of PPARγ reduced the PD‐L1 protein level without affecting its gene expression (Figures 3A and S3A–C). As deleted the nuclear location signal (NLS) blocked PPARγ transcription activity ¹⁶ (Figure 3B), and Western blot analysis showed that both PPARγ and PPARγ/ΔNLS could reduce the PD‐L1 protein level (Figure 3C), suggesting that PPARγ reduced the PD‐L1 level independent of its transcription activity. To detect the effect of deleted PPARγ on PD‐L1 protein stability, the half‐life of PD‐L1 was assayed. The results showed that loss of PPARγ increased PD‐L1 protein stability (Figure 3D). These findings suggest that PPARγ induces PD‐L1 protein degradation. Further analysis showed that overexpression of PPARγ significantly reduced the PD‐L1 protein half‐life (Figure 3E), suggesting that PPARγ induced PD‐L1 degradation since both the ubiquitin–proteasome system (UPS) and autophagy can mediate protein degradation. ²² To address the mechanism of PPARγ‐induced PD‐L1 degradation, H520 cells were treated with carbobenzoxy‐Leu‐Leu‐leucinal (MG132) (proteasome inhibitor) or chloroquine (CQ) (lysosome inhibitor). The results showed that cells treated with chloroquine (CQ) rather than MG132 inhibited PPARγ‐mediated PD‐L1 degradation (Figure 3F), suggesting that PPARγ facilitated PD‐L1 degradation by autophagy. PPARγ‐mediated PD‐L1 autophagic degradation was further determined using autophagy‐related 7 (ATG7) gene knockout H520 cells. The results showed that ATG7 gene knockout reduced the accumulation of LC3b‐II and inhibited PPARγ‐mediated PD‐L1 degradation (Figure 3G). These findings suggest that PPARγ induced PD‐L1 protein autophagic degradation independent of its transcription activity.

FIGURE 3.

PPARγ induced PD‐L1 autophagic degradation independent of its transcription activity. (A) H520 cells were transfected vector or Flag‐PPARγ plasmids for 48 h. Cell lysates were subjected to Western blot. (B) Luciferase analysis of PPARγ or PPARγ/ΔNLS transcription activity in H520 cells. Results are expressed as mean ± SEM (n = 3). *p < 0.05. (C) Western blot analysis of PD‐L1 protein level in PPARγ or PPARγ/ΔNLS overexpressing H520 cells. (D) Half‐life analysis of PD‐L1 protein in WT or PPARγ^−/− H520 cells. Cells were treated with cycloheximide (30 μg/mL) as indicated time course. The relative PD‐L1 protein level was quantified. Results are expressed as mean ± SEM (n = 3). *p < 0.05. (E) Half‐life analysis of PD‐L1 protein using Flag‐PPARγ overexpressed H520 cells. Cells were treated with cycloheximide (30 μg/mL) as indicated time course. The relative PD‐L1 protein level was quantified. Results are expressed as mean ± SEM (n = 3). *p < 0.05. (F) H520 cells were transfected vector or PPARγ plasmids for 48 h. Cells were treated with CQ (30 μM), MG132 (20 μM), or DMSO as indicated for another 4 h before cell lysis. Cell lysates were subjected to Western blot. (G) ATG7^−/− or WT H520 cells were transfected vector or Flag‐PPARγ plasmids as indicated for 48 h. Cell lysates were subjected to Western blot.

2.4. The interaction of PPARγ with PD‐L1

Under basal conditions, PPARγ bound to PD‐L1 by immunoprecipitation analysis (Figures 4A and S4). Confocal analysis showed that PPARγ was co‐localized with PD‐L1 in H520 cells (Figure 4B). Ni‐NTA pull‐down analysis showed that PPARγ bound to the cytoplasmic fragment of PD‐L1 (Figure 4C). To further assay the specific domain of PD‐L1 binding to PPARγ, The ligand binding domain (LBD) of PPARγ was deleted. In vitro glutathione‐S transferase (GST) pull‐down analysis showed that the deleted LBD of PPARγ did not bind to PD‐L1 (Figure 4D), suggesting that the LBD was required for PPARγ binding to PD‐L1.

FIGURE 4.

The interaction of PPARγ with PD‐L1. (A) H520 cell lysates were subjected to immunoprecipitation and Western blot. (B) Confocal analysis of H520 cells. Scale bar: 25 μm. (C) H520 cells were transfected plasmids as indicated, and Ni‐NTA pull down and Western blot analysis of cell lysates. (D) Schematic diagram of PPARγ. DBD, DNA binding domain; LBD, ligand binding domain. In vitro binding analysis of the interaction of PPARγ with PD‐L1 as described in the methods section.

2.5. The LIR motif of PPARγ induced PD‐L1 degradation

The LC3 interacting region (LIR) motif contains a core consensus sequence ([W/F/Y] XX [L/I/V]) that is essential for selective autophagy. ²³ Alignment analysis showed that PPARγ contains an LIR motif sequence (VDF70SSIST) (Figure 5A,B). The PPARγ/F70A mutant inhibited the binding of PPARγ to LC3 by immunoprecipitation analysis (Figure 5C), suggesting that the LIR motif was required for PPARγ binding to LC3, which was further demonstrated by in vitro binding analysis (Figure 5D).

FIGURE 5.

LIR motif of PPARγ was required for PD‐L1 autophagic degradation. (A) Alignment of PPARγ with LIR contained motif proteins. (B) Schematic diagram of the PPARγ‐LIR motif. (C) H520 cells were transfected plasmids as indicated for 48 h, and immunoprecipitation and Western blot analysis of cell lysates. (D) In vitro binding analysis of the interaction of PPARγ with LC3 as described in the methods section. (E) H520 cells were transfected plasmids as indicated for 48 h, and Western blot analysis of cell lysates. (F) ATG7^−/− or WT H520 cells were transfected plasmids as indicated for 48 h. Cell lysates were subjected to Western blot. (G) H520 cells were transfected plasmids as indicated for 48 cells. Cells were collected and co‐localization of proteins was detected by confocal analysis. Scale bar: 25 μm.

To further analyze the effect of the PPARγ‐LIR motif on PD‐L1 degradation, H520 cells were transfected with PPARγ or PPARγ/F70A mutant plasmids. The results showed that the PPARγ/F70A mutant did not reduce the protein level (Figure 5E) and half‐life of PD‐L1 (Figure S5). In contrast, ATG7 gene knockout resulted in inhibition of autophagy and PPARγ‐mediated PD‐L1 degradation (Figure 5F). Immunofluorescent analysis showed that PPARγ enhanced the co‐localization of PD‐L1 with lysosome, but the PPARγ/F70A mutant reversed this event (Figure 5G). These findings suggest that the LIR motif of PPARγ plays an important role in regulating PD‐L1 degradation by autophagy.

2.6. The LIR motif of PPARγ is required for antitumor immune escape

The above results suggest that the LIR motif of PPARγ was required for PPARγ‐mediated PD‐L1 lysosomal‐dependent degradation. To detect the effect of the PPARγ‐LIR motif on T‐cell activity, co‐cultured Jurkat T cells with PPARγ or PPARγ/F70A mutant H520 cells were performed. The results showed that PPARγ but not PPARγ/F70A mutant increased T‐cell IL‐2 production (Figure 6A). To detect the effect of PPARγ on tumor immune escape, a syngeneic tumor model was assayed using PPARγ or PPARγ/F70A stable expressing LLC cells. The results showed that PPARγ inhibited tumor growth and reduced tumor weight, but F70A reversed this event (Figure 6B,C), which was consistent with a reduced PD‐L1 protein level (Figure 6D). Similarly, deleting the LIR motif mutant of PPARγ abolished the inhibition of PPARγ on tumor growth (Figure S6). To further detect the effect of the PPARγ‐LIR motif on T‐cell activity in tumors, flow cytometry analysis was performed using tumor cells. The results showed that PPARγ significantly increased cytotoxic T‐cell activity, but the F70A mutant reversed this event (Figure 6E–G), suggesting that the PPARγ‐LIR motif is required for increased T‐cell activity. The PPARγ‐LIR motif therefore plays an important role in inhibiting NSCLC immune escape, which is associated with reduced PD‐L1 level and increased T‐cell activity.

FIGURE 6.

The PPARγ‐LIR motif was required for inhibition of tumor immune escape. (A) IL‐2 production was detected using a medium of co‐cultured Jurkat cells and cancer cells (PPARγ or F70A mutant expressing H520 cells). Results are expressed as mean ± SEM (n = 3). *p < 0.05. (B and C) PPARγ or PPARγ/F70A stabling expressed LLC cells were inoculated subcutaneously C57BL/6 mice. Tumor volume and weight were detected. Results are expressed as mean ± SEM, n = 6. *p < 0.05. (D) Relative cell surface PD‐L1 expression shown in tumor cells. MFI, median fluorescence intensity. Results are expressed as mean ± SEM (n = 6). *p < 0.05. (E–G) Flow cytometry analysis of intracellular IFN‐γ, Granzyme B, and perforin in T cells from isolated tumor cells. Results are expressed as mean ± SEM (n = 6). *p < 0.05.

3. DISCUSSION

PPARγ is a member of the PPAR family, which plays an important role in the regulation of inflammation, cancer, and metabolism. ¹⁶ , ²⁴ Several reports demonstrate that activation of PPARγ increases bladder, renal pelvis, colon, prostate, and liver cancer, ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ while the antitumorigenic effect of PPARγ by multiple signaling pathways has been reported. ¹⁵ , ¹⁶ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ As a ligand binding and activated receptor, some ligands regulate cancer progression independent of PPARγ activity, ³⁸ , ³⁹ , ⁴⁰ suggesting that the contradictory role of PPARγ on cancer progression may be derived from the cancer types, ligands, and conditions. Similarly, it exhibits the contradictory role of PPARγ on tumor immunotherapy by inhibiting ²¹ or promoting ¹⁸ , ¹⁹ , ²⁰ immune escape. Constitutive activation of PPARγ/retinoid X receptor alpha (RXRα) heterodimer promotes immune escape of muscle‐invasive bladder cancer by inhibiting T‐cell activity. ¹⁸ Consistently, inhibition of PPARγ by its antagonist GW9662 enhances antitumor immunotherapy. ¹⁹ , ²⁰ Combined PPARγ antagonist GW9662 with PD‐L1 antibody effectively enhances immunotherapy in B16 melanoma in females but not males, ¹⁹ and GW9662 decreases PD‐L1 expression in adipose and increases PD‐L1 antibody immunotherapy of breast cancer. ²⁰ Conversely, PPARγ agonist GW9662 induces PD‐L1 expression in microsatellite stable colorectal cancer cells, which in turn enhances PD‐L1 antibody blockade therapy. ²¹ However, GW9662 could be the antagonist of PPARδ and PPARγ. ⁴¹ Increasing evidence suggests that reduced a PD‐L1 protein level enhance antitumor immunotherapy. ¹⁴ In agreement with this, we found that PPARγ induced PD‐L1 autophagic degradation, subsequently inhibiting tumor immune escape. Another study shows that PPARγ agonist GW9662 induces PD‐L1 gene expression and increases PD‐L1 antibody efficiency by blockade therapy. ²¹ However, GW9662 is both of the antagonist of PPARδ and PPARγ. ⁴¹ To determine the direct role of PPARγ on PD‐L1 expression, PPARγ gene knockout cells were developed. The results showed that deficient PPARγ led to increased PD‐L1 expression, and overexpression of PPARγ reduced the PD‐L1 protein level, suggesting that PPARγ directly regulates PD‐L1 expression.

Although PPARγ induces the targeted proteins degradation by proteasome, ¹⁵ , ¹⁶ cells treated with the proteasome inhibitor had no effect on PPARγ‐mediated PD‐L1 degradation, suggesting that PPARγ‐mediated PD‐L1 degradation was not involved in proteasomal‐dependent pathway. As both UPS and autophagy can degrade the targeted proteins, ¹⁶ , ²² , ⁴² further analysis showed that PPARγ induced PD‐L1 autophagic degradation, but the mechanism is unclear.

Autophagy is an intracellular component degrading process in response to stress such as nutrient starvation and hypoxia, which is also nutrients recycling. ²² , ⁴² , ⁴³ Autophagy plays a critical role in regulating cancer immunotherapy by inducing immune checkpoint protein degradation and release of proinflammatory cytokines. ¹⁴ , ⁴⁴ In contrast, autophagy induction promotes PD‐L1 degradation in lysosome, which in turn enhances T‐cell activity and T‐cell killing to cancer cells. ¹⁴ Our results showed that autophagy inhibition by AGT7 deficiency had no significant effect on the PD‐L1 protein level in H520 cells, but overexpression of PPARγ or its agonist could facilitate PD‐L1 lysosomal‐dependent degradation, suggesting that PD‐L1 does not undergo autophagic degradation under base conditions in NSCLC.

Autophagy is a nonselective degradation process, but increasing evidence suggests that autophagy can be selective. ²² , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ The LIR motif containing proteins plays an important role in selective autophagy. ²³ PPARγ contains LIR motif (70‐FSSI‐72). This LIR motif was determined using immunoprecipitation and GST pull‐down analysis. These results show that PPARγ rather than F70A mutant bound to LC3, suggesting that PPARγ is an LIR‐containing protein, which mediates PD‐L1 selective degradation by autophagy. Although PD‐L1 exhibits resistance of tumor immunotherapy, ¹⁴ , ⁴⁴ PPARγ mediated it lysosomal‐dependent degradation, which in turn increased antitumor immunotherapy.

Taken together, PPARγ contains the LIR motif, which induces PD‐L1 degradation in lysosome and subsequently inhibits tumor immune escape.

4. MATERIALS AND METHODS

4.1. Cells, reagents, and plasmids

NCH1975, NCH520, LLC, and HEK293T cell lines were maintained in DMEM with 10% fetal bovine serum (FBS; Gibco). Jurkat cells were maintained in RPMI 1640 with 10% FBS. Cells were cultured in an incubator with 5% CO₂ at 37°C. His‐tag PD‐L1 cDNA was cloned into pcDNA3 vector. Flag‐PPARγ, Flag‐PPARγ/LBD, and Flag‐PPARγ/ΔNLS plasmids were described previously. ¹⁶ Flag‐PPARγ‐F70A or his‐PD‐L1‐N was mutated by the site‐directed mutagenesis method. Human PPARγ cDNA was cloned into plenti‐Tet‐on vector. Mice PPARγ cDNA was cloned into plenti‐CMV vector, and PPARγ‐F70A or PPARγ/ΔLIR (70‐FSSI‐72) was mutated by the site‐directed mutagenesis method. All the plasmids were identified by DNA sequencing. Protease inhibitor cocktail was obtained from Sigma. Geneticin (G418 sulfate) and doxycycline were purchased from CSN Pharm.

4.2. Gene knockout cell lines

pYSY‐CMV‐Cas9‐U6‐PPARγ (human)‐sgRNA1‐EFla‐neo, pYSY‐CMV‐Cas9‐U6‐PPARγ (mice)‐sgRNA1‐EFla‐neo, and pYSY‐CMV‐Cas9‐U6‐ATG7 (human)‐sgRNA1‐EFla‐neo plasmids were provided by YST BioTech (China). PPARγ (human) sgRNA: CAACTTTGGGATCAGCTCCG; PPARγ (mice) sgRNA: AGTGGTCTTCCATCACGGAG; ATG7 (human) sRNA: GAAGCTCCCAAGGACATTAA. Plasmids were transfected into HEK293T cells, and collected lentiviral particles infected H520 or LLC cells, and selected using G418. Gene knockout signal clone cells were picked and confirmed by gene sequencing (Figure S7).

4.3. Antibodies, immunoprecipitation, and Western blot

PPARγ (16643‐1‐AP), Tubulin (11224‐1‐AP), PD‐L1 (66248–1), ATG7 (10088‐2‐AP), LC3B (14600‐1‐AP), GST (66001‐2), Lamp1(21997‐1‐AP), and Flag (66008‐4) antibodies were obtained from Proteintech. Secondary antibodies were obtained from Jackson Immunoresearch. Cells were lysed in lysis buffer (50 mM Tris–HCl pH 7.4, 250 mM NaCl, 0.5% Triton X100, 10% glycerol, 1 mM DTT, protease inhibitor cocktail). Protein concentration in the supernatant was determined by the Pierce BCA Protein Assay Kit (Thermo). For immunoprecipitation, cell lysates were precleared with protein A/G magnetic bead (Cat: B23202; Bimake). Precleared lysates were subjected to immunoprecipitation using the indicated antibodies with protein A/G magnetic beads as described previously. ⁴⁹ , ⁵⁰ The samples were subjected to SDS‐PAGE and immunoblotted with indicated antibodies. Data are triplicates from three independent experiments. Blots were quantified by Image J.

4.4. Immunofluorescent analysis

Cells were fixed with 3.7% paraformaldehyde and washed, then were permeabilized with 0.5% Triton‐X100 and washed. After that, cells were blocked in 10% BSA, then incubated with primary antibodies (PPARγ, PD‐L1, or Lamp1). Subsequently, cells were incubated with secondary antibodies (Jackson Immunoresearch). Stained cells were viewed by confocal microscope.

4.5. In vitro binding analysis

Human PD‐L1 or LC3B cDNA was cloned into pET28a vector. PPARγ cDNA was cloned into PGEX‐6P. PGEX‐6P‐F70A or PGEX‐6P‐ΔLBD was mutated by the site‐directed mutagenesis method and identified by DNA sequencing. GST‐PPARγ, GST‐F70A, GST‐ΔLBD, his‐LC3B, and his‐PD‐L1 were expressed in E. coli strain BL21(DE3) (pAPlacIQ). The recombinant proteins were purified using glutathione beads or Ni‐NTA beads (Thermo Fisher). For in vitro binding of PPARγ to PD‐L1, GST‐PPARγ or GST‐ΔLBD (5 μg) fusion protein was immobilized on glutathione‐agarose beads in buffer (25 mM HEPES pH 7.5, 0.2% NP‐40, 6 mM NaCl) for 1 h at 4°C, then the same amount of recombinant his‐PD‐L1 was added. For in vitro binding of PPARγ to LC3, GST‐PPARγ or GST‐F70A (5 μg) fusion protein was immobilized on glutathione‐agarose beads in buffer (25 mM HEPES pH 7.5, 0.2% NP‐40, 6 mM NaCl) for 1 h at 4°C, then the same amount of recombinant his‐LC3B was added. These reactions were incubated for another 1 h. Adsorbates to glutathione‐conjugated beads were analyzed by Western blot.

4.6. Immunohistochemical staining

The human LUAD or LUSC tumor tissue and adjacent normal tissue sections were incubated with primary antibodies as indicated, then sections were stained by DAB (3,3′‐diaminobenzidine). After that, they were counterstained by hematoxylin. The relationship between PPARγ and PD‐L1 or CD8⁺ protein expression was assayed using Pearson's correlation coefficient test. P < 0.05 was accepted as being statistically significant. LUAD or LUSC tissue chips were purchased from WLLBIO (China).

4.7. Jurkat co‐culture and IL‐2 assay

Jurkat cells were activated using phorbol 12‐myristate 13‐acetate (PMA) (25 ng/mL) and phytohemagglutinin (PHA) (1 μg/mL) for 24 h. Then, Jurkat cells were added to cancer cells at a ratio of 4:1 (Jurkat:cancer cells) for 24 h. The cell culture media was collected. The level of secreted IL‐2 was detected using an IL‐2 ELISA kit (MLBIO, China).

4.8. Tumor model

C57BL/6 or Nude mice were purchased from the Experimental Animal Center of Jiangsu University. Plenti‐PPARγ, Plenti‐PPARγ/F70A or Plenti‐PPARγ/ΔLIR plasmids were transfected into HEK293T cells, and collected lentiviral particles infected LLC cells and developed stable gene expression LLC cells. PPARγ^−/− LLC cells were developed by the CRISPR/Cas9 method. Then, 2 × 10⁵ LLC cells were injected subcutaneously into 5‐week‐old C57BL/6 male mice. Tumor volume was measured with a digital caliper. Tumor volume = 1/2 (length × width²).

Plenti‐Tet‐PPARγ plasmids were transfected into HEK293T cells, and collected lentiviral particles infected H520 cells and developed stable gene expression H520 cells. Then, 1 × 10⁷ H520 cells were injected subcutaneously into 5‐week‐old Nude male mice. Mice were treated without or with doxycycline in drinking water (100 μg/mL). Tumor volume was measured with a digital caliper. Tumor volume = 1/2 (length × width²). All studies were carried out with the approval of the Jiangsu University Animal Care Committee.

4.9. T‐cell activity analysis of tumors

Tumors were cut into pieces and digested in DMEM with collagenase (2 mg/mL; Sigma) and DNase (10 μg/mL) for 1 h at 37°C. Cells were collected by centrifuge and filtered through a 70‐μm strainer in DMEM. Cell pellets were suspended and lysed in red blood cell lysis buffer for 5 min. The cells were filtered by a 40‐μm strainer in PBS with 2% BSA. Then, 1 × 10⁶ cells were incubated with antibody against PD‐L1 (APC conjugated; Proteintech) and the corresponding isotype IgG control (APC conjugated; Proteintech). To detect T‐cell activity in tumors, cells were fixed in PBS with 4% paraformaldehyde for 20 min, and then permeabilized in PBS with 0.1% saponin and 2% BSA. After that, cells were washed and co‐stained with CD8 (PE conjugated; Proteintech), IFN‐γ (APC conjugated, Biolgend), granzyme B (APC conjugated; Proteintech), perforin (APC conjugated), and corresponding isotype IgG control at room temperature for 30 min. Cells were washed by PBS with 2% BSA and analyzed by flow cytometry.

4.10. Quantitative real‐time PCR

Total RNA was isolated by RNeasy kit (Sangon Biotech) and assayed using a Real‐Time PCR assay kit (Takara). The relative mRNA expression level was normalized against β‐actin. The fold change over control was determined according to the ΔCt method. The PD‐L1 primers used were forward 5‐ATGGAG AGGAAGACCTGAAGGTTCA‐3 and reverse 5′‐GGGGCATTGACTTTCACAGTAATTCGC‐3′, and the β‐actin primers were forward: 5′‐GGTGGGCATGGGTCAGAAGGAT‐3′ and reverse 5′‐CACACGCAGCTC ATTGTAGAAGGT‐3′.

4.11. Luciferase assay

H520 cells were transfected with vector, PPARγ, or PPARγ/ΔNLS together with PPRE₃‐lu and Ptk‐RL for 48 h. Cell lysates were assayed using a dual luciferase reporter assay system (YPH Biotech).

4.12. Statistical analysis

Data are expressed as the mean ± SEM. Differences between two dependent groups were evaluated with the paired Student's t‐test. P < 0.05 was accepted as being statistically significant.

AUTHOR CONTRIBUTIONS

Qian Gou, Suning Che, Mingjun Chen, and Huiqing Chen performed the experiment analysis. Yongzhong Hou designed the experiments. Juanjuan Shi and Yongzhong Hou corrected the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interest.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Reviewer Board: The human LUAD and the LUSC tissue study was approved by the Research and Ethical Committee of Tenths People's Hospital of Tongji University.

Informed Consent: N/A.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies: All studies were carried out with the approval of the Jiangsu University Animal Care Committee.

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.

Figure S7.

Gou Q, Che S, Chen M, Chen H, Shi J, Hou Y. PPARγ inhibited tumor immune escape by inducing PD‐L1 autophagic degradation. Cancer Sci. 2023;114:2871‐2881. doi: 10.1111/cas.15818