Exosome-dependent immune surveillance at the metastatic niche requires BAG6 and CBP/p300-dependent acetylation of p53

Extracellular vesicles released by tumor cells contribute to the reprogramming of the tumor microenvironment and interfere with hallmarks of cancer including metastasis. Notably, melanoma cell-derived EVs are able to establish a pre-metastatic niche in distant organs, or on the contrary, exert anti-tumor activity. However, molecular insights into how vesicles are selectively packaged with cargo defining their specific functions remain elusive. Methods: Here, we investigated the role of the chaperone Bcl2-associated anthogene 6 (BAG6, synonym Bat3) for the formation of pro- and anti-tumor EVs. EVs collected from wildtype cells and BAG6-deficient cells were characterized by mass spectrometry and RNAseq. Their tumorigenic potential was analyzed using the B-16V transplantation mouse melanoma model. Results: We demonstrate that EVs from B-16V cells inhibit lung metastasis associated with the mobilization of Ly6Clow patrolling monocytes. The formation of these anti-tumor-EVs was dependent on acetylation of p53 by the BAG6/CBP/p300-acetylase complex, followed by recruitment of components of the endosomal sorting complexes required for transport (ESCRT) via a P(S/T)AP double motif of BAG6. Genetic ablation of BAG6 and disruption of this pathway led to the release of a distinct EV subtype, which failed to suppress metastasis but recruited tumor-promoting neutrophils to the pre-metastatic niche. Conclusion: We conclude that the BAG6/CBP/p300-p53 axis is a key pathway directing EV cargo loading and thus a potential novel microenvironmental therapeutic target.


160
Representative microscopic images of macrophages after 7 days are shown.

163
Other supplement files related to Figure 3: 164

178
(C) Analysis of the EV release by NTA from WT or p53 siRNA knock down (kd) HEK293 179 cells that were either non-treated or treated with 100 nM doxorubicin or LBH for 16h.

180
Immunoblot for p53 and actin as a loading control is shown.

(A) Immunoblot analysis of HEK293 BAG6KO cells transfected with full-length BAG6
193 (+WT BAG6) or with a N-terminal deleted BAG6 mutant (+nucBAG6), detected by 194 using a myc-tag specific antibody. Probing for GAPDH was done as a loading control.

195
The blot represents one out of three independent experiments.

198
Transfected BAG6 was visualized by a GFP-tag, the nucleus and cell membrane 199 were visualized by staining with DAPI and PKH, respectively, and merged images are

372
Cell lysates or in vitro expressed proteins were precipitated using specific antibodies against 373 BAG6, p53, CBP/p300, HRS, ubiquitin or acetyl-lysine. A minimum of 1 µg of antibody were 374 used for 100 µg of total protein and Protein A magnetic beads were used for pull-down.
To confirm BAG6 interaction with the ESCRT protein TSG101, we used Gold Yeast two hybrid 377 system. BAG6 and TSG101 were cloned into pGBT9 and pGADT7 expression vectors.         Table S5.

460
The western blots were developed using X-ray films or by detection with a CCD camera. For

508
Mass spectrometry bioinformatics and statistical analysis 509 All mass spectrometric raw data were processed with Maxquant using default parameters.

510
Briefly, MS2 spectra were searched against the Uniprot MOUSE database, including a list of 511 common contaminants. False discovery rates on protein and PSM level were estimated by the 512 target-decoy approach to 1% (Protein FDR) and 1% (PSM FDR), respectively. The minimal 513 peptide length was set to 7 amino acids and carbamidomethyolation at cysteine residues was 514 considered as a fixed modification. Oxidation (M) was included as variable modification. The 515 match-between runs option was enabled. LFQ quantification was enabled using default 516 settings. Downstream data processing was conducted within the Perseus computational 517 platform. Briefly, protein groups flagged as "reverse", "potential contaminant" or "only identified 518 by site" were removed from the data. LFQ data were log2 transformed. Statistical analysis of 519 differentially regulated proteins was performed using a two-sided t-test (fudge factor s0 was 520 adjusted to 0.1). Resulting p values were corrected for multiple testing using a permutation-