MEK/ERK‐mediated oncogenic signals promote secretion of extracellular vesicles by controlling lysosome function

Abstract Cancer cells secrete large amounts of extracellular vesicles (EVs) originating from multivesicular bodies (MVBs). Mature MVBs fuse either with the plasma membrane for release as EVs, often referred as to exosomes or with lysosomes for degradation. However, the mechanisms regulating MVB fate remain unknown. Here, we investigated the regulators of MVB fate by analyzing the effects of signaling inhibitors on EV secretion from cancer cells engineered to secrete luciferase‐labeled EVs. Inhibition of the oncogenic MEK/ERK pathway suppressed EV release and activated lysosome formation. MEK/ERK‐mediated lysosomal inactivation impaired MVB degradation, resulting in increased EV secretion from cancer cells. Moreover, MEK/ERK inhibition prevented c‐MYC expression and induced the nuclear translocation of MiT/TFE transcription factors, thereby promoting the activation of lysosome‐related genes, including the gene encoding a subunit of vacuolar‐type H+‐ATPase, which is responsible for lysosomal acidification and function. Furthermore, c‐MYC upregulation was associated with lysosomal gene downregulation in MEK/ERK‐activated renal cancer cells/tissues. These findings suggest that the MEK/ERK/c‐MYC pathway controls MVB fate and promotes EV production in human cancers by inactivating lysosomal function.


| INTRODUC TI ON
Cancer cells aberrantly secrete large amounts of extracellular vesicles (EVs) with cargo molecules different from those secreted by healthy cells 1,2 ; therefore, EVs have attracted widespread attention as a novel diagnostic marker that can be easily obtained through liquid biopsy. 3 EVs produced by cancer cells induce phenotypic changes in other cells in both the local and distant tumor microenvironment via the activity of their cargo, thereby supporting the establishment of the premetastatic niche and promoting organspecific metastasis. [4][5][6] Furthermore, EVs play a crucial role in the maintenance of cellular homeostasis as an additional mechanism for disposal of waste molecules. 7 Therefore, biogenesis of EV may be a promising target for cancer therapy; however, the mechanisms underlying the upregulation of EV biogenesis in cancer cells is largely unknown, and this lack of knowledge has hampered the development of EV-targeting therapeutic strategy. 1,8,9 Endosome-originated EVs known as exosomes are ~40-150 nm in diameter and contain various proteins, nucleic acids, and lipids. 7,10 During the maturation process of endosomes, invagination of endosomal membranes results in the formation of multivesicular bodies (MVBs) encapsulating a large number of intraluminal vesicles (ILVs).
Matured MVBs face 2 alternative fates: fusion with lysosomes or autophagosomes that are driven to degradation within the cell, whereas fusion with the plasma membrane leads to the release of ILVs into the extracellular space as EVs. 8,11,12 The balance between these 2 fates is pivotal to regulate the secretion of MVB-derived EVs in cancer cells; however, its mechanisms, such as the signaling pathways responsible for promoting MVB fusion with the plasma membrane for EV formation, still need to be defined.
In cancer cells, various signaling pathways are frequently deregulated, thereby contributing to cancer development, such is the case of the mitogen-activated protein kinase kinase (MEK) and mitogenactivated protein kinase (ERK) 13,14 that are aberrantly activated by the small G proteins Ras and Raf kinase, which are frequently mutated in several cancers. 15,16 The mechanisms underlying the regulation of the MEK/ERK pathway and the signaling circuits that drive the different features of cancers, including tumor growth, invasion, and metastasis, have been extensively studied since their discovery. 17,18 Nevertheless, very few studies have analyzed the effects of the MEK/ERK pathway on the phenotypic characteristics of cancer cells, including the upregulation of EVs. 19 In this study, the effects of known cellular signal inhibitors on EV secretion encouraged us to explore potential regulators of the fate of MVBs in cancer. Previously, we developed a system using engineered cancer cells that secrete luminescent EVs by the expression of a tetraspanin (such as CD63) fused with luciferase. 20 Detailed evaluation of these cells demonstrated that the intensity of luciferase luminescence in the cell culture medium correlated well with the number of EVs, demonstrating that this experimental system is convenient to estimate the quantity of secreted EVs including exosomes. Here, this system was used to study the molecular mechanisms underlying the upregulation of EV biogenesis in cancer cells. Based on the results, we propose a MYC-MiT/TFE-mediated mechanism by which MEK/ERK activation in cancer cells suppress lysosome function, causing the upregulation of EV secretion, which is, in turn, shown to be crucial for the growth of MEK/ERK-activated and lysosome-inhibited cancer cells. We further demonstrate that

| Acidic lysosome detection
Cells on a glass-bottomed dish were cultured in medium containing Lysotracker Red DND-99 (50 nM) (Thermo Fisher Scientific) for 30 min, and then the medium was replaced with Lysotracker-free medium, followed by observation of the cells using a ZEISS LSM 800 with Airyscan confocal microscope (Carl Zeiss).

| Quantitative RT-PCR
Quantitative RT-PCR was performed as previously described. 21 Relative gene expression levels were calculated using 18S ribosomal RNA or GAPDH sequence as control, as per the 2 −ΔΔC t method.
Primer sequences are described in Table S1.

| Immunofluorescence staining
Immunocytochemistry was performed as described previously. 22 Fluorescence was observed using a ZEISS LSM 800 with Airyscan confocal microscope (Carl Zeiss).

| Gene expression and gene silencing
Ectopic gene expression and site-directed mutagenesis were performed using pCX4 retroviral vectors. 22 Each gene was PCR amplified and subcloned into the pCX4 vector. Virus production and infection were performed as previously reported. 23 Lentiviral vectors carrying each gene were purchased from Sigma-Aldrich.

| Gene expression analysis of several cancer samples
Data analysis was performed using FireBrowse (http://fireb rowse. org/), which provides access to data stored in The Cancer Genome Atlas (TCGA) database.

| Quantification of EVs
NanoLuc luciferase assay was performed as described previously 20 using a Nivo multiplate reader (PerkinElmer). For EV preparation, cells were seeded onto a 150-mm culture dish at the density of 5 × 10 6 cells, and cultured for 24 h. After washing with 20 ml PBS 2 times, the culture medium was replaced with 13 ml 1% EV-depleted FBS containing medium. EV-depleted FBS was prepared by ultracen- (I) HT29 cells treated with DMSO, or 1 nM BafA1, and/or GW at 1 or 5 μM were used for the soft-agar colony formation assay. Relative colony numbers obtained from 3 independent experiments are shown. Data are represented as means ± standard deviation of 3 independent measurements. Statistical analysis was performed using one-way ANOVA. **p < 0.01, ***p < 0.001; n.s., not significant to confirm no growth differences arose among the different conditions or cells. For nanoparticle tracking analysis (NTA), cell supernatant was centrifuged at 2000 g for 10 min at 4°C to remove cells and cellular debris, and then filtered through a 0.22μm filter (Millipore Sigma) and ultracentrifuged at 110,000 g for 70 min at 4°C (SW41Ti rotor; Beckman Coulter). Apolipoprotein A1, a non-EV marker, in the sample was checked to be below the detection limit. The size distribution and concentration of the EVs were determined by NTA using a NanoSight LM10 (Malvern Panalytical). 24

| Immunohistochemical analysis
Kidney carcinoma with matched kidney tissue array (KD244 and  For normal samples, we selected a sample for each patient if it was annotated with "Solid Tumor Normal," which resulted in 72 samples.

| Gene set variation analysis
To estimate the variation of gene set enrichment through the samples of an expression data set, gene set variation analysis (GSVA), 25 a nonparametric unsupervised method, was performed on TCGA data sets using R software to obtain GSVA enrichment scores for lysosome-related gene sets in MSigDB 7.1. 26,27 For GSVA, FPKM-UQ gene expression values were transformed with log 2 after addition of 1.

| Statistical analysis
Data are presented as mean ± standard deviation. Statistical significance was calculated using Student t test or one-way ANOVA with Dunnett's post hoc analysis using XLSTAT for Microsoft Excel. Test results were reported as two-tailed p-values, where p < 0.05 was considered statistically significant.

| MEK/ERK pathway promotes EV secretion via suppression of lysosomal function
The effect of inhibitors on EV secretion in human colon cancer HT29 cells expressing CD63-Nluc was evaluated ( Figure 1A). U0126, a MEK inhibitor, decreased the luminescence of the culture medium of HT29/CD63Nluc cells in a dose-dependent manner (Figures 1A and S1A), concurrent with the inhibition of downstream ERK activation (pERK) ( Figure S1B). CD63-positive EVs from HT29 cells were 100-150 nm in size and verified to contain MVB-associated proteins, such as Alix and Tsg101, which were suppressed upon MEK inhibition ( Figures 1B and S1C). An approved MEK-targeting drug trametinib also inhibited EV secretion ( Figure S1D). As inhibitors potentially have multiple effects, we confirmed the effect of MEK activation on EV secretion. Consistently, EV secretion significantly   Figure S1F). Therefore, the impact of MEK inhibition on lysosome function was evaluated next. Levels of LAMP1, a lysosomal biomarker, and lysosome acidity, visualized by Lysotracker staining, were found to be considerably increased in HT29 cells upon MEK inhibition ( Figure 1C). Moreover, the expression of lysosomerelated genes was upregulated in HT29 cells treated with MEK inhibitors ( Figure 1D).
To verify the significance of lysosome activation on suppression of EV release, we examined the effect of bafilomycin A1. The levels of cellular CD63 protein, as a marker of MVBs, were downregulated in U0126-treated HT29 cells and restored by the addition of the lysosome inhibitor bafilomycin A1, although CD63 expression remained unaltered (Figures 1E and S1G). U0126-induced suppression of EV secretion was significantly restored by addition of bafilomycin A1 ( Figure 1F) without activation of ERK ( Figure S1H).
Therefore the upregulation of EV secretion by bafA1 treatment

| Nuclear localization of TFE3 regulates ATP6V1B1-mediated lysosomal acidification under MEK/ERK pathway
As alteration of V-ATPases expression might impair the ability of lysosomes to maintain low pH, 30 , 10 μm (B, H). Data are represented as mean ± standard deviation of 3 independent measurements. Statistical analysis was performed using one-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001 indicated that MEK/ERK-mediated exclusion of TFE3 from the nucleus suppresses lysosome activity, accompanied by the upregulation of EV secretion from cancer cells.

| MEK/ERK pathway regulates MYC expression to control the localization of TFE3
The subcellular localization and activity of the MiT/TFE transcription factor family is known to be regulated through phosphorylation by the mammalian target of rapamycin (mTOR) or ERK2 35,36 : phosphorylated MiT/TFE is retained in the cytoplasm, whereas dephosphorylated MiT/TFE is translocated into the nucleus. 34,37 In addition, c-MYC was recently reported to be involved in regulating lysosomal genes. 38 The presence of the mTOR inhibitor rapamycin did not induce lysosome acidification unlike U0126 or trametinib ( Figure S3E).
Moreover, TFE3 mutants with putative phosphorylation sites recognized by ERK or mTOR 39 ( Figure S3F introduced cell after treatment of trametinib was not marked, it was sufficient to restore cytoplasmic localization of TFE3 in these cells ( Figure 3F,G). To confirm the contribution of TFE3 translocation to the reduction of EV secretion in MYC-downregulated cancer cells, EV secretion was compared between MYC knockdown and MYC/ TFE3 knockdown HT29 cells. As shown in Figure S4E,F, decrease of EV secretion by MYC knockdown was completely restored by TFE3 knockdown, suggesting that the regulation of EV secretion by MYC is dominantly caused via TFE3. These findings suggest that MEK/ERKmediated MYC expression plays a crucial role in the regulation of EV secretion from cancer cells through controlling TFE3 translocation.

| MEK/ERK-MYC-lysosomal genes axis in human cancer tissues
To address the functional relevance of the MEK/ERK-MYC-lysosomal genes pathway to human cancers, expression of MYC and ATP6V1B1 was examined in various types of cancer and match-paired adjacent normal tissues using publicly available data from TCGA database. MYC was highly expressed in several cancers, including colon, rectum, and colorectal adenocarcinoma, glioblastoma, and glioma, kidney cancers, lung squamous cell carcinoma, prostate adenocarcinoma, and skin cutaneous melanoma ( Figure S5A). Analysis of ATP6V1B1 expression in these particular cancers revealed that it was much lower in renal tumors than in the match-paired normal tissues ( Figure S5B). Reassessment of renal cancer carcinoma (RCC) from the TCGA RNA-seq data confirmed that MYC expression was significantly higher and ATP6V1B1 expression was lower in tumor tissues than in their normal counterparts ( Figure 4A). Similarly, other RCC data sets from the Gene Expression Omnibus database also showed an inverse correlation between MYC and ATP6V1B1 expression ( Figure S5C-E). These findings highlighted a potential correlation between MYC expression and lysosome function; therefore, a GSVA was performed in RCC samples. Consistently, an inverse correlation between the expression of MYC and lysosome-related genes was observed when analyzing a large data set of renal cancers from the Kyoto Encyclopedia of Genes and Genomes database (p = 0.000323; Reactome, p = 0.0003729; Figure 4B). Furthermore, immunohistochemical analysis of the 24 primary tumors revealed that pERK (indicative of MEK/ERK activation) was greatly increased  Currently, it is difficult to experimentally determine the origin of the secreted EVs and, therefore, whether they are so-called exosomes or do not need to be investigated in the future. Therefore, the general term EV is used in this article in accordance with the recommendation of the international society for EVs. 42 It is noted that we present here the simplest hypothesis in which the observed changes in the secretion of EVs are directly attributable to changes in MVB.
Considering that the MEK/ERK pathway is activated under the control of various oncogenic signaling pathways, including Ras and EGFR, 15,16,43 c-MYC-mediated upregulation of EV in cancer cells might be explained by similar mechanisms. Indeed, MEK/ERKmediated upregulation of c-MYC and EV secretion in HRas-and KRas-transformed cells was observed ( Figure S6A-C). A previous report indicated that the stability of c-MYC protein is regulated by ERK-mediated phosphorylation 44 ; however, in the present study, MYC expression was regulated at the mRNA level ( Figure S4C).
Further extensive analysis will be necessary to elucidate the detailed mechanism of MEK/ERK-mediated MYC expression; whereas the apparent positive correlation between MYC expression and MEK/ ERK activity in human cancers would provide a missing link between c-MYC, MEK/ERK, and EV secretion. Mechanisms driving the upregulation of EV secretion, which are activated by Src oncogenic signals, were also demonstrated to promote ILV formation in cancer cells. 24,45 Here it was shown that the MEK/ERK pathway does not change the number of ILV in MVBs ( Figure 1G) and inhibition of Src does not induce activation of lysosome function ( Figure S3E), suggesting that the mechanisms underlying the upregulation of EV secretion would differ between oncogenic signals. Very recently, it has been reported that EV secretion is enhanced by MYC overexpression, but can be attenuated by MEK in triple-negative breast cancer (TNBC) cells. 46 We found that MEK inhibition decreased MYC expression in in a type of TNBC cells, MDA-MB-231 ( Figure S7D). We found that c-MYC prevents the localization of MiT/TFE proteins into the nucleus ( Figure 3B). A recent study demonstrated that cyclin-dependent kinase CDK4/6 phosphorylates MiT/TFE in the nucleus, promoting their shuttling to the cytoplasm. 47 Considering that CDK4/6 is a transcriptional target of c-MYC, CDK4/6 downregulation by MYC disruption might prevent MiT/TFE phosphorylation, thereby retaining them in the nucleus and resulting in the MiT/TFE-mediated transcription of lysosomal genes. 48 Additional extensive analysis will be necessary to elucidate the precise mechanism underlying regulation of MYC-mediated MiT/TFE subcellular localization. The present data demonstrating that repression of MYC suppresses EV secretion via upregulation of lysosomal genes provide the first firm evidence of MYC-mediated EV secretion. As MYC overexpression is frequently observed in human cancers, 49,50 this mechanism is expected to contribute to EV upregulation in several cancers. Our findings also suggest that targeting MYC expression/function would effectively prevent EV secretion in several types of cancers, regardless of the status of the MEK/ERK pathway (Figures 3 and S7A-C).
Of note, overaccumulation of MVBs in cells was found to suppress the growth of cancer cells ( Figure 1H,I). In agreement, it was reported that reduction of EV secretion by targeting Alix or Rab27b, or in the presence of GW4869, suppresses cell transformation, suggesting that overaccumulation of MVBs due to dysfunctional EV secretion suppresses the growth of cancer cells. 24 Several studies have shown that cancer cells use EVs to excrete waste molecules, such as unfolded proteins and damaged DNA. 1,51 These lines of evidence also suggest that the turnover of MVBs is important to support the growth of cancer cells, possibly by discarding harmful molecules via lysosomal digestion or extracellular excretion. In conclusion, this study demonstrates that the MEK/ERK-MYC-MiT/TFE-lysosomal protein axis has a crucial role in the regulation of MVB fate for secretion or degradation, and consequently on EV secretion in cancer cells and cancer cell growth. Although the cellular context-specific contribution of this axis should be further investigated, our findings suggest that the regulatory machinery of lysosome activity may offer new opportunities for therapeutic interventions targeting EV secretion in cancer.

ACK N OWLED G M ENTS
This work was supported by the Japan Science and Technology Agency, PRESTO program (JP1005457 to CO) and Grant-in-Aid Scientific Research (B) (20H03456 to CO).

D I SCLOS U R E
The authors have no conflict of interest.