Abstract

   Polybromo 1 (PBRM1) inactivating mutations are associated with clinical
   benefit from immune checkpoint inhibitor treatments in clear cell renal
   cell carcinoma (ccRCC). However, whether targeting PBRM1 has the
   potential to enhance immunotherapy efficacy in patients with wild‐type
   PBRM1 and the upstream pathways that regulate PBRM1 protein stability
   remain unclear. Here, it is demonstrated that PBRM1 knockdown induced
   M1 macrophage polarization and infiltration, which enhanced the
   efficacy of anti‐PD‐1 immunotherapy in RCC. Meanwhile, CDC20 catalyzes
   K27 ubiquitination of PBRM1 and promotes its degradation via
   p62‐mediated selective autophagy. A bicyclic peptide (PB1‐p62) is
   designed and constructed to target PBRM1 and p62, thereby promoting the
   degradation of PBRM1. As a result, the efficacy of anti‐PD‐1
   immunotherapy is enhanced, leading to improved overall survival rates
   in syngeneic mouse tumor models. Overall, this finding suggest the
   clinical application of PB1‐p62 and provide a novel approach for
   enhancing the effectiveness of immunotherapy in RCC patients with
   wild‐type PBRM1.

   Keywords: CDC20, Immunotherapy, p62, PBRM1, RCC
     __________________________________________________________________

   This study found that Polybromo‐1 (PBRM1) sensitizes renal cell
   carcinoma (RCC) to immunotherapy through regulating
   macrophage‐associated chemokines and increasing M1 macrophage
   infiltration. Furthermore, the APC^CDC20 E3 ligase is identified as the
   upstream regulator responsible for p62‐dependent selective autophagy
   degradation of PBRM1. The inhibition of PBRM1 through a synthesized
   peptide resulted in increased sensitivity of cancer cells to
   immunotherapy.

   graphic file with name ADVS-12-2412967-g007.jpg

1. Introduction

   Renal cell cancer (RCC), as one of the most common malignant tumors,
   has a high incidence and fatality rate worldwide, and it is a major
   malignant disease that threatens human health.^[ [52]^1 ^] Accumulating
   evidence indicates that the gene mutations are responsible for RCC
   initiation and progression. Polybromo 1 (PBRM1, also known as PB1 or
   BAF180), which has 9 major spliceosome types, is a subunit of the
   SWI/SNF transcription‐modulating chromatin remodeling complex that
   controls DNA accessibility for transcription. PBRM1 has been found to
   be mutated in up to 41% of clear cell renal cell carcinoma (ccRCC),
   making it the second most commonly mutated gene after VHL in ccRCC.^[
   [53]^2 ^] Additionally, PBRM1 mutation is also observed in several
   other malignancies, including chordomas (11%–59%), cholangiocarcinomas
   (12%–23%), and non‐small cell lung cancers (≈3%). In general, PBRM1 was
   considered as a tumor suppressor, which regulates divergent processes
   including cell cycle and DNA damage.^[ [54]^3 ^] Growing evidence
   suggests a close association between PBRM1 and immunity. Recently,
   immunotherapy has been considered as a promising approach for advanced
   ccRCC, the patients in this stage were too late to surgery and had poor
   prognosis.^[ [55]^4 ^] Previous clinical research has demonstrated that
   the loss of PBRM1 increases the sensitivity of renal cancer to T
   cell‐mediated cytotoxic killing.^[ [56]^5 ^] PBRM1 status has been
   significantly associated with the response to anti‐PD‐1 therapy, with a
   higher proportion of patients with PBRM1 mutations showing a
   significant benefit in both progression‐free survival (PFS) and overall
   survival (OS) than patients with PBRM1 wild‐type.^[ [57]^6 ^] What we
   are interested, however, is to enable wild‐type PBRM1 patients to
   benefit from immunotherapy. Thus, the upper regulatory mechanism of
   PBRM1 stability is critical for improving the efficiency of
   immunotherapy in PBRM1 wild‐type patients.

   Cell division cycle 20 homologue (CDC20), a well‐known regulator of the
   cell cycle, controls the correct segregation of chromosomes during
   mitosis.^[ [58]^7 ^] CDC20 and CDC20‐like protein 1 (CDH1) are
   co‐activators of the anaphase‐promoting complex/cyclosome (APC/C), a
   multifunctional E3 ubiquitin ligase that participates in the cell
   cycle, metabolism, DNA damage and repair, autophagy, apoptosis,
   senescence, and tumorigenesis.^[ [59]^8 ^] APC/C and CDC20 play
   important roles in the transition from metaphase to anaphase during
   mitosis, ensuring proper chromatid segregation.^[ [60]^9 ^] In
   addition, previous studies have shown that CDC20 plays a key role in
   different cell types and participates in specific biological processes
   such as cilia disassembly,^[ [61]^10 ^] brain development,^[ [62]^11 ^]
   tissue homeostasis in human keratinocytes,^[ [63]^12 ^] genomic
   stability,^[ [64]^13 ^] aging,^[ [65]^14 ^] and autophagy.^[ [66]^15 ,
   [67]^16 ^] APC/C^CDC20 also participates in cellular processes
   associated with tumorigenesis by regulating REV1,^[ [68]^17 ^]
   Wnt/β‐catenin,^[ [69]^18 ^] SOX2,^[ [70]^19 ^] MCL‐1,^[ [71]^20 ^]
   PC‐PLC,^[ [72]^21 ^] BCL‐2,^[ [73]^22 ^] and BIM.^[ [74]^23 ^] Recent
   research has revealed that the expression level of CDC20 plays a
   crucial role in regulating the anti‐tumor immune system. In
   hepatocellular carcinoma (HCC), CDC20 expression showed a significant
   correlation with immune infiltration of B cells, neutrophils,
   macrophages, and myeloid dendritic cells.^[ [75]^24 , [76]^25 ^]
   Moreover, the CDC20 protein has been associated with the tumor mutation
   burden, immune checkpoint molecules, tumor microenvironment, and
   immunological infiltration in ccRCC.^[ [77]^26 ^] However,
   APC‐dependent molecular mechanism of CDC20 in controlling RCC tumor
   immunity remains largely unknown.

   Autophagy refers to the degradation of various material components in
   cells through lysosomes, and it can be divided into 3 categories:
   macroautophagy, microautophagy, and chaperone‐mediated autophagy
   (CMA).^[ [78]^27 ^] Macroautophagy can also be divided into selective
   and non‐selective forms. After the degradation substrate is labeled
   during selective autophagy, it is recognized by selective autophagy
   receptors, of which the most canonical is p62/SQSTM1, and isolated into
   the autophagosome, where it is degraded when the autophagosome binds to
   the lysosome to form the autophagolysosome.^[ [79]^28 , [80]^29 ^] p62
   is the most extensively studied receptor in selective autophagy, but
   the regulatory mechanism of p62 has not yet been elucidated.^[ [81]^30
   ^] As p62 is the first and most important autophagy receptor of
   selective autophagy, the binding and degradation of p62 to specific
   substrates will greatly affect target protein level, which also
   provides new ideas and research directions for the targeted therapy of
   some clinical diseases.

   In our study, we validated that CDC20 catalyzes PBRM1 K27
   ubiquitination and degradation via p62‐mediated selective autophagy,
   and low‐level of PBRM1 enhances the efficacy of immunotherapy through
   promoting M1 macrophages and CD8^+ T cells infiltration. Therefore, a
   peptide named PB1‐p62 was designed to target PBRM1 and p62, resulting
   in the degradation of most of PBRM1 via autophagy and notably ensuring
   the efficacy of immunotherapy for renal cancer. Thus, our study shows a
   new research direction for improving the immune efficacy of RCC in
   PBRM1 wild‐type patients and greatly improves the possibility of using
   targeted peptides in the clinical treatment of PBRM1 wild‐type
   patients.

2. Results

2.1. CDC20 was Identified as an Interactor of PBRM1

   To investigate the interactions, we overexpressed PBRM1 by transferring
   the PBRM1‐Flag plasmid into the 293T cell line and then detected the
   interacting proteins of PBRM1 via liquid chromatography mass
   spectrometry (LC‐MS/MS). Enrichment analysis resulted in the discovery
   of proteins that were enriched in the cell cycle signaling pathway
   (Figure [82]1A), with key cell cycle component CDC20 being among the
   identified cell cycle‐related proteins (Figure [83]1B). Further
   co‐immunoprecipitation (Co‐IP) assays confirmed the exogenous and
   endogenous interactions between PBRM1 and CDC20 (Figure [84]1C,D;
   Figure [85]S1A, Supporting Information). Meanwhile, a direct
   interaction between PBRM1 and CDC20 was demonstrated by a GST pull‐down
   assay (Figure [86]1E). As most APC CDC20 substrates contain a
   destruction box (D‐box) motif RXXL (where R represents Arginine, L
   represents Leucine, and X represents an arbitrary amino acid),
   examination of the PBRM1 sequence revealed two putative D‐boxes
   (Figure [87]1F). We mutated all the arginine (R) and leucine (L) to
   alanine (A) and generated three PBRM1 mutant constructs (35RLA, 622RLA
   and B‐RLA). PBRM1 single mutation (35RLA or 622RLA) did not affect the
   interaction between PBRM1 and CDC20, whereas simultaneous mutation
   (B‐RLA) of these sites significantly weakened the interaction between
   PBRM1 and CDC20 (Figure [88]1G). Structurally, CDC20 contains a WD40
   repeat domain at its C‐terminus that mediates substrate recognition. As
   shown in Figure [89]S1B,C (Supporting Information), PBRM1 specifically
   interacted with the WD40‐repeat domain of CDC20, supporting the
   hypothesis that PBRM1 is a putative CDC20 substrate. In addition, we
   also found a weak interaction between PBRM1 and CDH1 (Figure [90]S1D,
   Supporting Information). However, protein purification and GST
   pull‐down revealed that PBRM1 and CDH1 did not interacted directly
   (Figure [91]S1E, Supporting Information). These results may be due to
   CDC20 and CDH1 being coactivators of APC/C and their similar protein
   structures.

Figure 1.

   Figure 1
   [92]Open in a new tab

   CDC20 directly interacts with and modulates PBRM1. A) anti‐Flag
   immunoprecipitates derived from 293T cells transfected with Flag‐tagged
   PBRM1 plasmids coupled with liquid chromatography mass spectrometry
   (LC‐MS/MS) to identify PBRM1‐interacting proteins in 293T. IPs‐MS
   results were subjected to enrichment analysis. B) The proteins were
   enriched in the “cell cycle” categories. C) Immunoblot (IB) analysis of
   whole cell lysates (WCL) and anti‐CDC20 or anti‐PBRM1
   immunoprecipitates derived from 786‐O and 769‐P cells. D) IB analysis
   of WCL and anti‐Flag immunoprecipitates derived from 293T cells
   transfected with HA‐CDC20 and Flag‐PBRM1. E) IB analysis of WCL and GST
   pull‐down products derived from 293T cells transfected with HA‐CDC20
   and indicated constructs of GST‐PBRM1. F) Schematic representation of
   the conserved sites in PBRM1 matching the optimal CDC20 substrate
   motif. The sequence alignment of from various species as well as a
   schematic representation of the D‐box mutant using is shown. G) IB
   analysis of WCL and anti‐Flag immunoprecipitates derived from 293T
   cells transfected with HA‐CDC20 and indicated constructs of V5‐PBRM1.
   H) IB analysis of WCL derived from 786‐O cells synchronized by double
   thymidine block, following by releasing back into the cell cycle for
   the indicated times. I) Quantification of CDC20 and PBRM1 blot
   intensity using the ImageJ software, according to Figure [93]1H. Data
   are statistics of three independent experiments. CDC20 and PBRM1 bands
   were normalized to vinculin. J) IB analysis of WCL derived from 786‐O
   cells synchronized by nocodazole block, following by releasing back
   into the cell cycle for the indicated times. K) Quantification of CDC20
   and PBRM1 blot intensity using the ImageJ software, according to
   Figure [94]1J. Data are statistics of three independent experiments.
   CDC20 and PBRM1 bands were normalized to Vinculin. L) IB analysis of
   WCL derived from 786‐O cells synchronized by serum starvation (S.S.),
   following by releasing back into the cell cycle for the indicated
   times. M) Quantification of CDC20 and PBRM1 blot intensity using the
   ImageJ software, according to Figure [95]1L. Data are statistics of
   three independent experiments. CDC20 and PBRM1 bands were normalized to
   vinculin. N) Cell lysates derived from 786‐O and 769‐P cells stably
   expressing shCDC20 or Scr. Scr, Scramble. O) Cell lysates derived from
   786‐O and 769‐P cells transfected with siCDC20 or NC. NC, negative
   control. P) IB analysis of WCL derived from wild‐type (WT) or CDC20
   knockdown 786‐O cells synchronized by double thymidine block, following
   by releasing back into the cell cycle for the indicated times. Scr,
   Scramble. Q) IB analysis of WCL derived from wild‐type (WT) or CDC20
   knockdown 786‐O cells synchronized by nocodazole block, following by
   releasing back into the cell cycle for the indicated times. R,S) IB
   analysis and quantification of cell lysates of wild‐type (WT) or CDC20
   knockdown 786‐O cells treated with cycloheximide (CHX, 20 µg ml^−1) at
   indicated time points. T–U) IB analysis and quantification of cell
   lysates derived from 293T cells transfected with indicated plasmids and
   treated with cycloheximide (CHX, 20 µg ml^−1) at indicated time points.
   (V–W) Representative images(V) and quantifications(W) of ccRCC patient
   samples stained for PBRM1 and CDC20 by IHC assay. Scale bar, 20 µm.

2.2. CDC20 Deficiency Upregulates Protein Stability of PBRM1

   CDC20 plays an important role in mitosis and its protein levels
   fluctuate during the cell cycle.^[ [96]^31 , [97]^32 ^] The PBRM1
   protein abundance fluctuated during the cell cycle in cells
   synchronized by double thymidine treatment, displaying a dramatic
   reduction when CDC20 was most active (Figure [98]1H,I;
   Figure [99]S1F,G, Supporting Information). Consistent with this
   finding, an inverse correlation between CDC20 activity and the
   abundance of PBRM1 was also observed in cells synchronized by
   nocodazole (Figure [100]1J,K) and serum starvation block
   (Figure [101]1L,M; Figure [102]S1H,I, Supporting Information).

   Previous results demonstrated that paclitaxel (Taxol) and nocodazole
   can induce G2/M arrest and activate SAC to suppress CDC20. Thus,
   nocodazole and taxol were used to block the cell cycle, and PBRM1
   protein level was upregulated while CDC20 protein level was decreased
   (Figure [103]S1J,K, Supporting Information). Furthermore, we found that
   CDC20 deficiency resulted in the upregulation of PBRM1 as well as other
   identified CDC20 substrates, including Cyclin B1 and p21
   (Figure [104]1N,O), whereas CDH1 deficiency had no effect on PBRM1
   (Figure [105]S1L, Supporting Information). In keeping with these
   findings, depletion of CDC20 resulted in a rather stable pattern of
   PBRM1 expression throughout the cell cycle (Figure [106]1P,Q;
   Figure [107]S1M,N, Supporting Information). Meanwhile, the half‐life of
   PBRM1 was remarkably prolonged in the CDC20 deficiency group
   (Figure [108]1R,S; Figure [109]S1O,P, Supporting Information), whereas
   CDH1 deficiency did not affect the half‐life of PBRM1
   (Figure [110]S1Q,R, Supporting Information). In the PBRM1 mutant
   protein, the half‐life of 35RLA and B‐RLA were significantly prolonged
   when treated with CHX (Figure [111]1T,U). Furthermore, to investigate
   the relationship between PBRM1 and CDC20 in RCC patients, we collected
   91 tumor tissues from patients in our hospital, and
   immunohistochemistry was performed to detect the expression of PBRM1
   and CDC20 by using IHC score calculations. These results confirmed that
   the expression of PBRM1 was negatively correlated with CDC20 expression
   (Figure [112]1V,W). Collectively, these results indicate that CDC20
   negatively regulated the protein stability of PBRM1.

2.3. APC/C^CDC20 Functions as E3 Ubiquitin Ligase and Promotes K27
Ubiquitination of PBRM1

   As a key regulator of mitosis and co‐activator of the E3 ubiquitin
   ligase APC/C, APC/C^CDC20 might regulate the stability of PBRM1 by
   mediating its ubiquitination. To verify our hypothesis, the Ni‐NTA
   protein purification system was used to validate the ubiquitination of
   PBRM1 promoted by CDC20. Indeed, ubiquitination assays showed that
   CDC20, rather than CDH1, could increase PBRM1 protein ubiquitination
   (Figure [113]2A). Moreover, CDC20 overexpression upregulated PBRM1
   ubiquitination, whereas CDC20 knockdown decreased PBRM1 ubiquitination
   (Figure [114]2B–D). We further found that ubiquitination of PBRM1 was
   remarkably decreased in the 35RLA and B‐RLA mutants, but not in the
   622RLA mutant, indicating that the promotional effect of CDC20 on PBRM1
   ubiquitination was up to the^[ [115]^35 ^] RRRL^[ [116]^38 ^] domain at
   position 35 of PBRM1 (Figure [117]2E). To further evaluate the
   ubiquitination mechanism of PBRM1, we constructed different
   ubiquitin‐independent expression plasmids (K6, K11, K27, K29, K33, K48,
   and K63) and different ubiquitin mutant plasmids (K6R, K11R, K27R,
   K29R, K33R, K48R, K63R, and 7KR). Further experiments found that CDC20
   promoted K27‐linked polyubiquitination of PBRM1 (Figure [118]2F,G). To
   determine the lysine residue(s) of PBRM1 ubiquitinated by CDC20, we
   applied mass spectrometry analysis of 293T cells transfected with
   Flag‐tagged PBRM1 in combination with HA‐tagged CDC20 and ubiquitin.
   Ubiquitination at lysine residues 414, 616, 1208, and 1293 in PBRM1 was
   detected by mass spectrometry (Figure [119]S2A–D, Supporting
   Information). We then mutated the major individual lysine residues to
   alanine (K414A, K616A, K1208A and K1293A), and we found that mutation
   of K1208 (but not K414A, K616A or K1293A) decreased CDC20‐mediated
   polyubiquitination of PBRM1 (Figure 2H; Figure [120]S2E, Supporting
   Information). Thus, we identified the formation of a polyubiquitin
   chain on PBRM1 K1208 mediated by CDC20. Consistent with this finding,
   compared with PBRM1 WT, the PBRM1 K1208 mutant was resistant to
   CDC20‐mediated degradation (Figure [121]S2F–I, Supporting Information).

Figure 2.

   Figure 2
   [122]Open in a new tab

   CDC20 catalyzes the Lys27‐linked ubiquitination of PBRM1. A)
   Immunoblotting (IB) analysis of WCL and Ni‐NTA pull‐down lysates
   derived from 293T cells transfected with indicated plasmids. B) IB
   analysis of WCL and Ni‐NTA pull‐down lysates derived from 293T cells
   transfected with Flag‐PBRM1, HA‐CDC20, and His‐Ub plasmids. C) IB
   analysis of WCL and Ni‐NTA pull‐down lysates derived from WT and CDC20
   knockdown 293T cells transfected with Flag‐PBRM1 and His‐Ub plasmids.
   D) IB analysis of WCL and Ni‐NTA pull‐down lysates derived from 293T
   cells transfected with siCDC20, Flag‐PBRM1 and His‐Ub plasmids. NC,
   negative control. E) IB analysis of WCL and Ni‐NTA pull‐down lysates
   derived from WT and CDC20 knockdown 293T cells transfected with His‐Ub,
   HA‐CDC20, and indicated constructs of V5‐PBRM1 plasmids. F) IB analysis
   of WCL and Ni‐NTA pull‐down lysates derived from 293T cells transfected
   with HA‐CDC20, V5‐PBRM1 and indicated KR (Lys to Arg) ubiquitin
   mutants. G) IB analysis of WCL and Ni‐NTA pull‐down lysates derived
   from 293T cells transfected with HA‐CDC20, V5‐PBRM1 and K‐only
   ubiquitin mutants. H) IB analysis of WCL and Ni‐NTA pull‐down lysates
   derived from WT and CDC20 knockdown 293T cells transfected with His‐Ub,
   HA‐CDC20 and indicated constructs of Flag‐PBRM1 plasmids.

2.4. CDC20 Promotes p62‐Mediated Selective Autophagic Degradation of PBRM1

   To distinguish which degradation system dominantly regulates the
   degradation of PBRM1 promoted by CDC20, we further examined the protein
   stability of PBRM1 using MG132, which inhibits ubiquitin proteasome
   function, and ammonium chloride (NH[4]Cl), an inhibitor of autophagy.
   We found that NH[4]Cl increased the protein level of PBRM1 and
   attenuated CDC20 knockdown induced PBRM1 upregulation (Figure [123]3A;
   Figure [124]S3A, Supporting Information), indicating that PBRM1 protein
   stability was mostly controlled by the autophagy‐lysosome pathway. The
   protein level of PBRM1 was upregulated when autophagy was gradually
   inhibited by NH[4]Cl or chloroquine (CQ, Figure [125]3B,C), two
   well‐known autophagy inhibitors. Furthermore, we observed that
   autophagic inducer, glucose or serum starvation, decreased the protein
   level of PBRM1 (Figure [126]3D,E). In addition, PBRM1 was upregulated
   in ATG5 knockdown (KD) cells with or without serum starvation treatment
   (Figure [127]3F). Taken together, these results indicated that CDC20
   promoted PBRM1 ubiquitination and degradation via autophagy.

Figure 3.

   Figure 3
   [128]Open in a new tab

   Autophagy selectively degrades PBRM1 via the autophagic receptor p62 in
   a CDC20‐dependent manner. A) IB analysis of WCL lysates derived from WT
   or CDC20 knockdown 786‐O and 769‐P cells treated with NH[4]Cl (20 mM,
   24 h) before harvesting. B) IB analysis of WCL lysates derived from
   786‐O and 769‐P cells treated with NH[4]Cl (20 mM) for indicated hours.
   C) IB analysis of WCL lysates derived from 786‐O and 769‐P cells
   treated with chloroquine (CQ, 50 µM) for indicated hours. D) IB
   analysis of WCL lysates derived from 786‐O and 769‐P cells treated with
   glucose starvation (G.S.) for indicated hours. E) IB analysis of WCL
   lysates derived from 786‐O and 769‐P cells treated with serum
   starvation (S.S.) for indicated hours. F) IB analysis of WCL lysates
   derived from wild‐type (WT) or ATG5 knockdown 786‐O and 769‐P cells
   treated with serum starvation (S.S.) for indicated hours before
   harvesting. G) IB analysis of WCL and anti‐HA immunoprecipitates
   derived from 293T cells transfected with HA‐p62 and V5‐PBRM1. H) IB
   analysis of WCL and anti‐Flag immunoprecipitates derived from 293T
   cells transfected with Flag‐p62 and V5‐PBRM1. I) IB analysis of WCL
   lysates derived from shScr or p62 knockdown 786‐O and 769‐P cells using
   shRNA. J) IB analysis of WCL lysates derived from NC or p62 knockdown
   786‐O cells using siRNA. NC, negative control. K,L) Immunoblotting
   analysis (K) and quantification (L) of cell lysates of WT and p62
   knockdown 786‐O cells treated with cycloheximide (CHX, 20 µg ml^−1) at
   indicated time points. M) IB analysis of WCL and GST‐pull‐down products
   derived from 293T cells transfected with GST‐p62 and V5‐PBRM1. 24 h
   post transfection, cells were treated with serum starvation (S.S.) for
   24 h before harvesting the cells. N) IB analysis of WCL and
   GST‐pull‐down products derived from 293T cells transfected with
   GST‐p62, HA‐CDC20 and V5‐PBRM1. O) IB analysis of WCL and GST‐pull‐down
   products derived from WT and CDC20 knockdown 293T cells transfected
   with GST‐p62 and V5‐PBRM1. P) Schematic representation of truncated
   constructs of p62 for mapping the interaction domain with PBRM1. Q) IB
   analysis of WCL and GST‐pull‐down products derived from 293T cells
   transfected with V5‐PBRM1 and indicated constructs of GST‐p62 plasmids.
   R) Schematic representation of truncated constructs of p62 for mapping
   the interaction domain with PBRM1. S) IB analysis of WCL and
   GST‐pull‐down products derived from 293T cells transfected with
   V5‐PBRM1 and indicated constructs of GST‐p62 plasmids.

   To further analyze the mechanism by which PBRM1 is degraded by the
   autophagy‐lysosome pathway, we examined whether p62, a cargo receptor
   deliver cargoes to the autophagosome for selective degradation,
   participates in CDC20‐mediated PBRM1 stabilization. Our result
   demonstrated that PBRM1 interacted with p62 (Figure [129]3G,H) rather
   than HSC70 chaperone protein (Figure [130]S3B, Supporting Information),
   confirming that PBRM1 was degraded by selective autophagy mediated by
   p62 rather than by chaperone‐mediated autophagy (CMA). Moreover,
   knockdown of p62 upregulated the protein expression of PBRM1 in 786‐O
   and 769‐P cells (Figure [131]3I,J), and the half‐life of PBRM1 was
   prolonged in the p62 deficiency group (Figure [132]3K,L;
   Figure [133]S3C,D, Supporting Information). We also found that the
   association between PBRM1 and p62 increased considerably in response to
   serum starvation treatment (Figure [134]3M). Meanwhile, we found that
   CDC20 enhanced the interaction between PBRM1 and p62
   (Figure [135]3N,O), indicating that CDC20 promoted autophagic
   degradation of PBRM1 through assisting PBRM1‐p62 interaction. To
   further confirm the specific regions of interaction between p62 and
   PBRM1, we constructed truncated plasmids with different p62 domains
   (Figure [136]3P). Further Co‐IP assays revealed the involvement of the
   region (104‐221) for the interaction with PBRM1, but not to other
   domains (Figure [137]3Q). Similarly, we constructed truncated plasmids
   within p62 104–221 and abbreviated the domain of p62 interacting with
   PBRM1 within 134–197 (Figure [138]3R,S). In conclusion, our results
   revealed that CDC20 promotes PBRM1 degradation through p62‐mediated
   selective autophagy.

2.5. Inhibition of PBRM1 Sensitizes RCC to Immunotherapy Through Regulating
Macrophage‐Associated Chemokines and Increasing M1 Macrophage Infiltration

   Previous study has confirmed that the mutation and loss of function of
   PBRM1 promote immune cell infiltration, which may affect the efficacy
   of immunotherapy for RCC.^[ [139]^33 ^] Thus, we aimed to evaluate the
   effects and regulatory mechanisms of wild‐type PBRM1 on immunotherapy
   of RCC. We first constructed PBRM1 knockdown cell lines using the shRNA
   (Figure [140]S4A, Supporting Information) and performed RNA‐seq
   analysis. Enrichment analysis of transcriptome data suggested that
   immune‐related signaling pathways were significantly enriched in PBRM1
   knockdown cells (Figure [141]4A). Gene set enrichment analysis (GSEA)
   showed enrichment of up‐regulated genes in immune‐related pathways,
   including TNF signaling pathway and IL2‐STAT5 signaling pathways
   (Figure [142]4B). Moreover, data derived from RNA‐seq analysis revealed
   that downstream genes of immune‐related pathways, especially a variety
   of chemokines, were significantly up‐regulated after PBRM1 knockdown
   (Figure [143]4C). Given the pivotal roles of chemokines in macrophage
   recruitment and activation, we hypothesized that PBRM1 knockdown exerts
   immunomodulatory effects by regulating macrophage recruitment. The
   findings demonstrated that depletion of PBRM1 resulted in enhanced
   macrophage infiltration (Figure [144]4D). Furthermore, we found that
   knockdown of Ccl7 or Ccl17 exhibited negligible impact on macrophage
   recruitment, while knockdown of Ccl5 or Ccl20 significantly attenuated
   macrophage recruitment induced by PBRM1 depletion (Figure [145]4E). In
   addition, the transcription levels of the markers of M1 macrophages,
   TNFα, IL12, and IL1b, were significantly upregulated after co‐culturing
   with PBRM1 knockdown tumor cell culture medium (Figure 4F;
   Figure [146]S4B, Supporting Information). Moreover, we observed the
   presence of infiltrating macrophages in Pbrm1 WT or knockdown (KD)
   Renca xenograft tumors. The results indicated that PBRM1 knockdown
   could stimulate the infiltration and differentiation of M1 macrophages
   in vivo (Figure 4G; Figure [147]S4C–E, Supporting Information). The
   functional activity loss of the PBRM1 mutation has been clinically
   validated to potentially enhance the efficacy of targeted therapy.^[
   [148]^2 , [149]^34 , [150]^35 ^] To investigate the hypothesis that
   PBRM1 knockdown may impact immunotherapy sensitivity in ccRCC, we
   administered anti‐PD‐1 mAb to BALB/c mice with Renca cells expressing
   either wild‐type or knockdown Pbrm1. Our findings demonstrate that
   Pbrm1 knockdown suppressed tumor growth and improved survival rates
   when combined with PD‐1 monoclonal antibody treatment (Figure 4H;
   Figure [151]S4F, Supporting Information). To investigate the role of
   macrophages in PBRM1 knockdown‐mediated enhancement of
   immunotherapeutic effects, we utilized clodronate liposomes to deplete
   tumor‐infiltrating macrophages and assessed tumor volume and burden.
   The results demonstrated that depletion of macrophages abolished the
   augmenting immunotherapeutic effects mediated by PBRM1 knockdown in
   vivo (Figure [152]4I–K and Tables [153]S2 and [154]S3, Supporting
   Information), highlighting the crucial involvement of macrophage
   recruitment in PBRM1‐mediated immune evasion. Consistently, silencing
   PBRM1 significantly increased CD8^+ tumor‐infiltrating T lymphocytes
   (TILs), which could be reversed by clearance of macrophages
   (Figure [155]4L–N; Figure [156]S4G, Supporting Information).
   Furthermore, immunohistochemistry was performed on microarrays of RCC
   samples to assess the expression levels of iNOS (a marker for M1
   macrophages), CD8 (a marker for CD8^+ T cells), and PBRM1. Our findings
   indicate a significant inverse correlation between PBRM1 expression and
   the infiltration of M1 macrophage cells (Rho  =  −0.3138, P  = 0.0026)
   as well as CD8^+ T cells (Rho  =  −0.2662, P  =  0.0112,
   Figure [157]4O–Q). These results revealed that PBRM1 inhibition could
   promote the infiltration of M1 macrophages into RCC, thereby improving
   the efficacy of anti‐PD‐1 immunotherapy and prolonging the survival of
   tumor‐bearing mice.

Figure 4.

   Figure 4
   [158]Open in a new tab

   Inhibition of PBRM1 enhanced efficacy of PD‐1 blockade through
   regulating macrophage‐associated chemokines and increasing M1
   macrophage infiltration. A) Immune‐related pathway enrichment analysis
   of wild‐type (WT) and PBRM1 knockdown Renca cells. n = 3. B) Gene set
   enrichment plots of TNF signaling pathway and IL2‐STAT5 signaling
   pathways according to RNA‐seq data in Figure [159]4A. C) Immune‐related
   differentially expressed genes (DEGs) for wild‐type (WT) and PBRM1
   knockdown Renca cells according to RNA‐seq data in Figure [160]4A. D)
   Photomicrographs of the migratory RAW264.7 following coculture with
   wild‐type (WT) or Pbrm1 knockdown cells for 24 h in 24‐well transwell
   lower chambers. E) Photomicrographs of the migratory RAW264.7 following
   coculture with indicated genes knockdown Renca cells for 24 h in
   24‐well transwell lower chambers. F) TNFα, IL‐12, and IL‐1β mRNA
   expression of RAW264.7 after treatment with conditioned medium derived
   from wild‐type (WT) or Pbrm1 knockdown Renca cells was determined by
   qTR‐PCR. n = 3. G) Proportions of M1 macrophage infiltrated in
   wild‐type (WT) or Pbrm1 knockdown Renca syngeneic tumors. n = 6. H)
   Kaplan–Meier survival curves of Balb/c mice bearing Renca syngeneic
   tumor, according to tumor growth data in Figure [161]S4F (Supporting
   Information). n = 10. I) Renca cells stably expressing shPBRM1 were
   subcutaneously injected into Balb/c mice to establish xenograft model
   and treated with anti‐PD‐1 mAb (2mg per mice, three times) or
   Clodronate liposome (0.5 mg per mice, three times). Statistical
   analysis of the tumor volumes which were measured every three days and
   plotted individually, n = 5 J) Tumor sizes in Balb/c mice bearing Renca
   renal carcinoma with indicated treatment. n = 5. K) Tumor burdens in
   Balb/c mice bearing Renca renal carcinoma with indicated treatment. n =
   5. L–N) Representative dot plots and the proportions of CD4^+ and CD8^+
   tumor‐infiltrating lymphocytes (TILs) in wild‐type (WT) and Pbrm1
   knockdown Renca syngeneic tumors. n = 5. O–Q) Representative images and
   quantifications of primary lung carcinoma patient samples stained for
   PBRM1, iNOS and CD8 expression by IHC assay. Scale bar, 20 µm.

2.6. Design of a Peptide Targeting PBRM1 Enhances the Infiltration of M1
Macrophages and Efficacy of Anti‐PD‐1 Immunotherapy in RCC

   We proposed that the regulation of PBRM1 by p62 may depend on both
   interaction and lysosomal sorting. To further analyze the interaction
   interface between p62 and PBRM1 at the structural level, we first
   examined the substrate recognition structure of p62 (Figure [162]5A)
   and the protein structure of PBRM1 (Figure [163]5B). Using ZDOCK, we
   performed docking and interaction interface analysis of the p62‐PBRM1
   protein complex. As shown in Figures [164]S5A,B, Supporting
   Information, the substrate recognition structure of p62 interacts with
   PBRM1, with the most likely binding interface located at positions
   1386–1451 of PBRM1. Co‐immunoprecipitation (Co‐IP) experiments at the
   cellular level also demonstrated the interaction between positions
   139–162 of p62 and PBRM1 (Figure [165]S5C, Supporting Information).
   Based on the interaction interface analysis between p62 and PBRM1, we
   designed novel degradation drugs targeting PBRM1 (Figure [166]5C). By
   identifying a peptide sequence that can bind to PBRM1 through its
   interaction interface with p62, we introduced the REE sequence at the
   N‐terminus of the recognition sequence signal of p62. This approach
   transforms a PBRM1 binding sequence into a PBRM1‐targeted degradation
   peptide sequence.

Figure 5.

   Figure 5
   [167]Open in a new tab

   The PB1‐p62 drug effectively induces PBRM1 degradation and boost the
   impact of anti‐PD‐1 therapy. A) The protein crystal structure depicts
   amino acids 139 to 162 of p62 (obtained from PDB: 6MJ7). B) Alphafold
   prediction illustrates the full‐length structure of the PBRM1 protein.
   C) Schematic diagram and process of designing bicyclic peptide
   degradation drugs targeting PBRM1, based on the predicted protein
   sequence of p62 binding to PBRM1. D,E) Affinity detection of PB1‐p62 by
   ITC (D) and FP (E). F) IB analysis of PBRM1 in Renca cells after 24 h
   of treatment with the PB1‐p62 drug. G) IB analysis of WCL and anti‐HA
   immunoprecipitates derived from 293T cells transfected with HA‐p62 and
   V5‐PBRM1. 24 h post transfection, cells were treated with PB1‐p62 drug
   (5 µM) and chloroquine (CQ, 50 µM) for 24 h before harvesting the
   cells. H,I) IB analysis of cell lysates of Renca cells treated with
   PB1‐p62 drug (5 µM, 24 h) before treated with cycloheximide (CHX,
   20 µgml^−1) at indicated time points. J) Kaplan–Meier survival curves
   of Balb/c mice bearing Renca syngeneic tumor, according to tumor growth
   data in Figure [168]S6F,G (Supporting Information). n = 12. K) Renca
   cells were subcutaneously injected into Balb/c mice to establish
   xenograft model and treated with anti‐PD‐1 mAb (2 mg per mice, three
   times) or PB1‐p62 (2 mg per mice, seven times). Statistical analysis of
   the tumor volumes which were measured every three days and plotted
   individually, n = 6. L) Tumor sizes in Balb/c mice bearing Renca renal
   carcinoma with indicated treatment. n = 6. M) Tumor burdens in Balb/c
   mice bearing Renca renal carcinoma with indicated treatment. n = 6. N)
   The proportions of M1 macrophage in Renca syngeneic tumors. n = 6. O–Q)
   Representative dot plots and the proportions of CD4^+ and CD8^+
   tumor‐infiltrating lymphocytes (TILs) in Renca syngeneic tumors. n = 6.

   Peptide drugs offer advantages such as high affinity to target proteins
   and strong specificity. However, their clinical application is limited
   by their short half‐life and membrane penetration capabilities. The
   bicyclic peptide chemical method, which has garnered significant
   attention in recent years, can induce linear peptides to form bicyclic
   peptide structures, thus addressing the stability and membrane
   penetration issues of peptide drugs. In designing bicyclic peptides
   targeting PBRM1, we employed BiBr3 to induce the simultaneous reaction
   of three cysteines. We incorporated cysteines at the C‐terminus,
   N‐terminus, and the middle of the PBRM1 binding sequence and p62
   recognition sequence, ultimately achieving the design of bicyclic
   peptide degradation drugs targeting PBRM1. We obtained linear
   PBRM1‐targeted degradation peptides through solid‐phase synthesis,
   which were subsequently converted into bicyclic peptide drugs using
   BiBr3 to induce bicyclic peptide formation. Mass spectrometry data
   confirmed that BiBr3 effectively facilitated the formation of bicyclic
   peptide structures (Figure [169]S5D, Supporting Information). Upon
   completion of the bicyclic peptide synthesis, we employed ZDOCK to
   assess the binding affinity of both linear and bicyclic peptides to
   PBRM1. The results demonstrated that the bicyclic peptide PB1‐p62
   exhibited stronger binding to PBRM1 compared to its linear counterpart.
   Specifically, both the ΔiG and ΔG values for PB1‐p62 were lower,
   indicating a more favorable binding interaction and higher affinity
   (Figure [170]S5E,F, Supporting Information). The chemical structures of
   the linear and cyclic peptides are shown in Figure [171]S5G,H
   (Supporting Information). The membrane‐penetrating ability of peptide
   drugs is critical for their efficacy. To evaluate this, we labeled
   PB1‐p62 with rhodamine and used flow cytometry to assess its cellular
   uptake. The flow cytometry results indicated that PB1‐p62 effectively
   penetrated the cell membrane and entered the cells within 6 h
   (Figure [172]S5I, Supporting Information). Next, we evaluated the high
   affinity of the PB1‐p62 for PBRM1 using isothermal titration
   calorimetry (ITC) (Figure [173]5D) and fluorescence polarization (FP)
   (Figure [174]5E). The binding affinity between PB1‐p62 and the p62
   protein was also measured by ITC (Figure [175]S5J, Supporting
   Information). Serum stability experiments demonstrated that the
   bicyclic peptide significantly extended the half‐life of the peptide
   drug (Figure [176]S5K, Supporting Information), thereby enhancing its
   potential for therapeutic applications.

   As a result, the fusion constructs PB1‐p62 significantly promoted the
   degradation of PBRM1 (Figure 5F; Figure [177]S5L–N, Supporting
   Information). Meanwhile, proteasome inhibitor MG132 could also exhibit
   the ability to enhance the protein level of PBRM1 (Figure [178]S5O,
   Supporting Information), suggesting that PB1‐p62 facilitates the
   degradation of PBRM1 through autophagy and proteasome pathways. PB1‐p62
   markedly enhanced the interaction between PBRM1 and p62
   (Figure [179]5G) and decreased the half‐life of PBRM1 protein
   (Figure [180]5H,I; Figure [181]S5P,Q, Supporting Information). To
   evaluate the druggability potential of PB1‐p62, we conducted in vivo
   toxicity assessments. The results revealed no significant alterations
   in alanine transaminase (ALT), aspartate transaminase (AST), creatinine
   (Cr), and blood urea nitrogen (BUN) levels between the PB1‐p62 peptide
   treatment groups and the control group (Figure [182]S6A, Supporting
   Information). Histological examination of vital organs, including
   heart, liver, spleen, lung, and kidney, also demonstrated an absence of
   toxicity (Figure [183]S6B, Supporting Information). Furthermore, our
   investigation into the impact of PB1‐p62 on tumor progression confirmed
   its dose‐dependent inhibitory effect on tumor growth (Figure [184]S6C–E
   and Tables [185]S4 and [186]S5, Supporting Information). To verify the
   effect of PB1‐p62 on renal cancer immunotherapy, we performed Renca
   subcutaneous tumor formation experiments in BALB/c mice. PB1‐p62 in
   combination with the PD‐1 monoclonal antibody significantly prolonged
   the survival time of tumor‐bearing mice, suppressed tumor
   proliferation, and reduced tumor burden when compared with PD‐1
   monoclonal antibody alone, suggesting that PB1‐p62 can enhance the
   efficacy of the PD‐1 monoclonal antibody (Figure [187]5J–M;
   Figure [188]S6F,G and Tables [189]S6 and [190]S7, Supporting
   Information). In addition, we confirmed that PB1‐p62 in combination
   with the PD‐1 monoclonal antibody may enhance the therapeutic effect of
   the PD‐1 monoclonal antibody by increasing the infiltration of immune
   cells. The results indicated that PB1‐p62 increased the infiltration of
   M1 macrophages into mouse RCC tissues, whereas no significant
   difference was detected in the infiltration of M2 macrophages among the
   groups (Figure 5N; Figure [191]S6H,I, Supporting Information). We also
   found that PB1‐p62 significantly improved the infiltration of CD8^+
   TILs in the subcutaneous tumors of mice induced by the PD‐1 monoclonal
   antibody (Figure [192]5O–Q). These findings together suggest that
   PB1‐p62 enhances the anti‐PD‐1 immunotherapy through both increasing
   the infiltration of M1 macrophages and CD8^+ T cells.

3. Discussion

   Somatic mutations of PBRM1, which are especially prevalent in clear
   cell renal cell carcinoma (ccRCC), always co‐occurred with deletion of
   the non‐mutated PBRM1 allele, consequently resulting in the loss of
   polybromo‐1 function. Previous studies have found that PBRM1 mutation
   was a promising biomarker for immunotherapy in ccRCC. Here we
   demonstrated that inhibition of wild‐type PBRM1 could also sensitizes
   RCC to immunotherapy through regulating macrophage‐associated
   chemokines and increasing M1 macrophage infiltration. Nonetheless, we
   aimed to improve the effects of immunotherapy through downregulating
   PBRM1 in PBRM1 wild‐type RCC. In this study, we demonstrated that CDC20
   interacts with PBRM1 and promotes K27 ubiquitination at the K1208 site,
   which weakens the stability of PBRM1. In particular, we validated that
   CDC20‐mediated PBRM1 degradation was via p62‐dependent selective
   autophagy. In addition, we synthesized a peptide to promote the
   degradation of PBRM1 by enhancing the interaction between PBRM1 and
   p62, providing a promising tumor therapeutic approach with PBRM1
   inhibition and PD‐1 blockade (Figure [193]6 ).

Figure 6.

   Figure 6
   [194]Open in a new tab

   A schematic diagram to show CDC20‐mediated selective autophagy
   degradation of PBRM1 and regulates tumor immune evasion in renal cell
   carcinoma.

   Many studies have demonstrated that the differentiation of macrophages
   and ratio of M1/M2 macrophages are related to tumor therapy, in which
   the M1 macrophage phenotype is classified to have an anti‐tumor effect
   and M2 has the opposite effect.^[ [195]^36 , [196]^37 , [197]^38 ^] We
   previously showed that the status of PBRM1 may be related to RCC
   immunotherapy.^[ [198]^33 ^] Our data demonstrate that alterations in
   PBRM1 particularly promote polarization of M1 macrophages, thereby
   inducing remodeling of the immune microenvironment and augmenting
   immunotherapeutic effects. Macrophages, as pivotal immune cells within
   the tumor microenvironment (TME), play intricate roles in driving tumor
   progression. M1 macrophage polarization has been confirmed to exert
   antineoplastic effects,^[ [199]^39 ^] which partly explains the reason
   why PBRM1 knockdown enhanced the response to immune checkpoint
   inhibitors. Moreover, our data confirmed that Ccl5 or Ccl20 act as the
   key chemokines involved in PBRM1 knockdown‐mediated macrophage
   recruitment, although the specific mechanism leading to Ccl5 or Ccl20
   up‐regulation remains to be defined. Furthermore, we identified CDC20
   as the upstream regulator of PBRM1, which provides a novel viewpoint
   for CDC20 to influence tumor immunity. In addition, the tissues of RCC
   patients indicated that lower level of PBRM1 was related to increased
   infiltration of M1 macrophages and CD8^+ T cells. Thus, the results of
   our study demonstrated that PBRM1 may be a new target for RCC patients
   with wild‐type PBRM1 to improve the clinical efficacy of immunotherapy.

   More importantly, we designed a peptide named PB1‐p62 to promote the
   autophagic degradation of PBRM1 by enhancing the interaction between
   PBRM1 and p62. A continuous amino acid sequence was truncated at the
   protein‐protein interaction interface between p62 and PBRM1 as the
   binding moiety of PBRM1. To achieve the targeted degradation of PBRM1,
   the p62 feature recognition REEE sequence^[ [200]^40 ^] was added to
   the N‐terminus of PBRM1 binding peptide and served as a p62 recognition
   signal. In addition, a series of assays were conducted to prove that
   PB1‐p62 was efficient in promoting the interaction between PBRM1 and
   p62 and enhancing the degradation of PBRM1. Our study further validated
   that PB1‐p62 improves immune efficacy by increasing the infiltration of
   M1 macrophages and T cells, which prolonged the survival of PBRM1
   wild‐type tumor‐bearing mice. In fact, many therapeutic targets of
   cancer and other diseases are selective autophagy substrates, such as
   cGAS and PD‐L1, among others. Thus, it is of interest to develop novel
   lysosome‐targeting chimeras against these disease‐related proteins in
   future studies.

   In summary, our work identifies CDC20 as a regulator of PBRM1 lysosomal
   degradation, providing new insight into the mechanisms that control the
   homeostasis of PBRM1. CDC20 regulates PBRM1 K27‐type ubiquitination and
   promotes p62‐mediated selective autophagic degradation of PBRM1. Thus,
   the rationalized design of chimeric peptides targeting the degradation
   process may greatly reduce the level of PBRM1 in renal cancer and
   improve the efficacy of immunotherapy. This will provide a new research
   direction for improving the immune efficacy of PBRM1 wild‐type patients
   in the clinic and greatly increase the possibility of targeted peptides
   in the clinical treatment of PBRM1 wild‐type patients.

4. Experimental Section

Antibodies, Plasmids, and Reagents

   Antibodies against PBRM1 (91 894), Vinculin (13 901), Cyclin D1 (2978),
   Cyclin B1 (12 231), p62 (5144), V5(13202), ATG5(12994), hemagglutinin
   (HA; 3724), and Flag (2368) were purchased from Cell Signaling
   Technology. CDC20 (10252‐1‐AP), CDH1 (16368‐1‐AP), His(66005‐1‐Ig),
   GST(10000‐0‐AP) and LC3(14600‐1‐AP) were purchased from Proteintech.
   Cyclin E (MA5‐14336) was purchased from Invitrogen. pCMV‐8 x His‐Ub
   (107392) and pcDNA6.2/N‐EmGFP‐PBRM1‐V5 (65387) were purchased from
   Addgene. pcDNA3.1‐PBRM1‐3xFlag was purchased from Biokeeper.
   pcDNA3‐HA‐CDC20, pcDNA3‐HA‐CDH1, pcDNA3‐HA‐Hsc70, pcDNA3‐HA‐p62,
   pcDNA3‐Flag‐p62, pCMV‐GST‐p62, and pCMV‐GST‐PBRM1 were constructed in
   our laboratory, and pcDNA6.2/N‐EmGFP‐PBRM1 35RLA, 622RLA, B‐RLA, and
   pCMV‐GST‐p62 domain plasmids were amplified and cloned according to
   standard protocols (KOD‐Plus‐Mutagenesis Kit; SMK‐101). Nocodazole
   (487928), double thymidine (T1895), and taxol (580555) were purchased
   from Sigma–Aldrich. Chloroquine (CQ; T8689), MG‐132 (T2154), NH[4]Cl
   ([201]T64755), Cycloheximide (CHX; T1225) were purchased from
   Targetmol. Different mutants were generated by site‐directed
   mutagenesis PCR reaction using platinum PWO SuperYield DNA polymerase
   (Roche, Basel, Switzerland) according to the product manual.

Cell Culture, Transfection, and Construction of Knockdown Cell Lines

   The cell lines 786‐O, 769‐P, MB49, HEK293T, Raw264.7, and Renca used in
   this study were purchased from ATCC (Manassas, VA, USA) and provided by
   the Institute of Urology, First Affiliated Hospital of Xi’ an Jiaotong
   University. All cells were cultured in RPMI‐1640 or DMEM medium (Gibco)
   with 10% fetal bovine serum (Gibco) and maintained at 37 °C in 5%
   CO[2]. The cells were subjected to a double thymidine block assay,
   wherein they were incubated with 2 mM double thymidine for 18 h during
   the second round and then release. Cells synchronized with
   nocodazole‐arrest (100 ng mL^−1, 18 h) and release were collected at
   the indicated time points.

   PEI Prime linear polyethylenimine (PEI; 919 012) was purchased from
   Sigma–Aldrich and dissolved in DMSO, according to the standard
   protocol. PEI is suitable for gene delivery, and it was used for
   plasmid transfection in this study. Knockdown of PBRM1, CDC20, CDH1,
   and p62 was performed using shRNA, which were packaged with pSPAX2 and
   pMD2.G in HEK293T cells. All these components formed a lentiviral
   vector, which was collected to infect the target cell line. Puromycin
   (1 µg mL^−1) was used to select the infected cells for 2 weeks, and the
   protein was extracted to validate target knockdown efficacy. All
   knockdown cell lines were cultured as described previously and used for
   assays within 6 months.

Western Blot

   Cells were washed 3 times with ice‐cold phosphate‐buffered saline
   (PBS). The appropriate RIPA or IP buffer was added, which was mixed
   with protease and phosphatase inhibitors in advance. After
   centrifugation, the cell lysate supernatant was collected and checked
   using the BCA qualification system. The appropriate protein‐loading
   buffer was added to the protein lysates and denatured in a boiling
   water bath for 5 min. Each sample contained 30 µg protein, which was
   loaded onto 8% or 10% SDS‐PAGE gels and transferred to a polyvinylidene
   fluoride (PVDF) membrane. Then, the PVDF membranes were immunoblotted
   in fresh milk for 1 h at room temperature. Primary antibodies of the
   targets were added to the PVDF membranes and incubated overnight at
   4 °C. The cells were rinsed 3 times with TBST buffer, followed by
   incubation with peroxidase‐conjugated antibodies at room temperature
   for 1 h. After washing, the PVDF membranes were analyzed and visualized
   using an ECL system (Bio‐Rad, Hercules, CA, USA).

Total RNA Extraction and Real‐Time RT‐PCR Analysis

   Total cell RNA was extracted using TRIzol reagent (Invitrogen; Thermo
   Fisher Scientific) and the cDNA was synthesized using Primer Script RT
   reagent kit (Takara Bio). cDNA was diluted in SYBR Green Master Mix
   (Takara Bio) and then real‐time RT‐PCR was conducted following cycling
   conditions: 95 °C for 15 s, 55 °C for 30 s, and 72 °C for 30 s.
   Relative mRNA levels were evaluated using the 2^−ΔΔ CT method. The
   primer sequences were listed in Table [202]S1 (Supporting Information).

Immunoprecipitation and Co‐Immunoprecipitation Assay

   Suitable protease and phosphatase inhibitors were added to the IP lysis
   buffer (50 mM Tris‐HCl [pH 7.4], 120 mM NaCl, 5 mM EDTA, and 0.5%
   NP‐40) was used to lyse cells, and the proteins were extracted as
   described previously. According to the experimental design, 0.5 mL
   protein solution (1 mg) was collected from the top of each sample, and
   10 µL primary targets or Flag/HA antibody coupled with agarose
   particles was added to each sample. After sealing, the samples were
   incubated in a refrigerator for 4–6 h. At the end of the reaction, the
   EP tube was removed, the supernatant was discarded after centrifugation
   at 15 000 × g for 30 s, and 1 mL of precooled IP lysate was added and
   used several times for rinsing. An appropriate amount of IP lysate
   containing 1× loading buffer was added to each group of samples and
   placed in a boiling water bath for ≈10 min and then removed and placed
   in a −20 °C refrigerator. The experimental results were by performing
   the western blotting experiment the next day.

GST Pull‐Down Assay

   In our study, pGEX‐4T‐1 (Addgene, 129 567), pGEX‐4T‐1‐GST‐PBRM1, and
   pGEX‐4T‐1‐GST‐CDC20/CDH1 vectors were constructed and expressed in
   Escherichia coli BL21. The transfected bacteria were collected, and a
   suitable IP buffer was added for ultrasonic lysis. Targets with GST
   tags were purified using Glutathione Sepharose 4B beads (17 075 605; GE
   Healthcare). pcDNA3‐HA‐CDC20/CDH1 was expressed in HEK293T cells, and
   the proteins were purified using HA‐tag beads (Sigma–Aldrich, A2220).
   Purified GST‐PBRM1 and HA‐CDC20/CDH1 were incubated together at 4 °C
   for GST pull‐down assays. The system should spin softly for no more
   than 3 h. Finally, the beads were washed twice with IP buffer and
   maintained at 95 °C for 5 min, and the proteins were collected and
   analyzed with western blotting, as described previously.

Ni‐NTA Pull‐Down Assay

   Flag‐PBRM1/GFP‐PBRM1‐V5, HA vector, HA‐CDC20, and His‐Ub were
   transfected and expressed in HEK293T cells. After washing with
   pre‐cooled PBS, the appropriate cells were lysed with IP buffer for the
   whole cell lysis group. The remaining cells were lysed with buffer A (6
   M guanidine‐HCl, 0.1 M Na[2]HPO[4]/NaH[2]PO[4], and 10 mM imidazole [pH
   8.0]). After sonication for 10 s, the proteins were collected by
   centrifugation and incubated with the appropriate
   nickel‐nitrilotriacetic acid (Ni‐NTA) beads (Qiagen) for 2–3 h at room
   temperature. Then, the Ni‐NTA beads were washed 2 times with buffer A
   and buffer A/TI (1:3 of volume buffer A and buffer TI) and once with
   buffer TI (25 mM Tris‐HCl, 20 mM imidazole [pH 6.8]). Finally, a
   suitable IP buffer and protein‐loading marker were added to the washed
   beads, which were then placed in boiling water for 5 min to denature
   the proteins. The pull‐down proteins were separated and analyzed with
   western blotting.

Flow Cytometry Analysis (FACS)

   Flow cytometry was used to analyze immune cells infiltrated in TME.
   Briefly, immune cells were extracted from xenograft tumors, and washed
   with MACS buffer (PBS, 2%FBS, 1 mM EDTA). Surface staining was
   performed by adding the indicated antibodies to the cell suspension:
   anti‐CD8 (Biolegend: 126 614), anti‐CD4 (Biolegend: 100 432), and
   anti‐CD3 (Biolegend: 100 206), F4/80 (BioLegend: 157 304), CD11c
   (BioLegend: 117 326), CD206(BioLegend: 141 708). After 30 min
   incubation, cells were washed with MACS buffer and analyzed on BD
   Fortessa flow cytometer. The results were analyzed with Flowjo
   software.

Immunohistochemical Assay

   Ninety‐four RCC specimens were obtained from the First Affiliated
   Hospital of Xi'an Jiaotong University (Shaanxi, China). Informed
   consent was obtained for the human RCC specimens and the usage of these
   specimens was approved by the Ethic Committee of First Affiliated
   Hospital of Xi'an Jiaotong University (NO. LLSBPJ‐2023‐059). The tumor
   tissues from the RCC patients were fixed with 10% formaldehyde for
   24 h, embedded in paraffin, and cut into 5‐µm‐thick sections. After
   deparaffinization, rehydration, endogenous peroxidase blocking, and
   antigen retrieval, the sections were blocked with 5% BSA for 0.5 h and
   incubated with primary antibodies at 4 °C overnight. The next day, the
   tumor tissue sections were incubated with secondary antibodies for 1 h
   and visualized using a DAB kit, according to the manufacturer's
   instructions. Photographs were obtained and analyses were performed
   using a microscope (Olympus Optical Co., Tokyo, Japan). Three random
   fields (×100) of each tumor tissue in each section were photographed
   and scored for further analysis. The grading of staining intensity was
   as follows: no staining (score 0), light brownish‐yellow (score 1),
   brownish‐yellow (score 2), and dark brownish‐brown (score 3). The
   grading of positive cell count was as follows: less than 25% positive
   cells were graded as score 1, 26% to 50% were graded as score 2, 51% to
   75% were graded as score 3, and ≥75% were graded as score4. The final
   score was calculated by multiplying the staining intensity score with
   the positive cell count score.

Xenograft Animal Model for RCC

   All animal experiments were authorized by the Health Science Center
   (HSC) of Xi'an Jiaotong University (NO. XJTUAE2023‐285), and the
   animals were killed according to standard ethical guidelines. Forty
   BALB/c mice were purchased for each experiment, and these mice were 5
   weeks of age and weighed ≈20 g. Renca cells were harvested and
   suspended in serum‐free medium containing Matrigel, and 200 µL of Renca
   cells (1 × 10^6 cells) was subcutaneously injected into the right lower
   limbs of each mouse. After the mice were fed for 1 week and the tumors
   grew to ≈100–150 mm^3 in size, they were randomly divided into 4
   groups, and each group was intraperitoneally injected with placebo,
   PB1‐p62 (2 mg kg^−1, or knockdown of PBRM1), or anti‐PD‐1 monoclonal
   antibody with or without PB1‐p62 (or knockdown of PBRM1) every 3 days.
   The anti‐PD‐1 monoclonal antibody (BE0146), named InVivoMAb anti‐mouse
   PD‐1 (CD279), was purchased from Bio X Cell and used according to the
   manufacturer's instructions. The longest and shortest tumor diameters
   were measured at the same frequency. After 3 weeks, the tumors were
   harvested and ground to separate and extract immune cells, and the
   results were analyzed using flow cytometry. Survival assays were
   performed until the mice died or the tumor reached its setting volume.
   Tumor volume was calculated using the following formula: tumor volume =
   (length × width^2) × 0.5. We also assessed it by tumor inhibitory rate
   (TIR). TIR was calculated as follows: TIR  =  [1 − (Vt / Vc)] × 100% or
   TIR  =  [1 − (Wt / Wc)] × 100%, where Vt is the final mean tumor volume
   of the tested groups, Vc is the final mean tumor volume of the control
   group, Wt is the final mean tumor weight of the tested groups, and Wc
   is the final mean tumor weight of the control group.

Statistical Analysis

   All experiments were performed at least 3 times, and the data were
   represented as mean ± standard deviation (SD) values. Differences
   between groups were analyzed using one‐way ANOVA. Only 2 groups were
   compared with Student's t‐test, and the Mann–Whitney U test was used
   for IHC analysis. ImageJ software (NIH, Bethesda, MD, USA) was used to
   quantify the densities of the bands in the western blot. p < 0.05 was
   used as the statistically significant standard in all cases.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [203]ADVS-12-2412967-s001.docx^ (17MB, docx)

Acknowledgements