RAB7 counteracts PI3K-driven macropinocytosis activated at early stages of melanoma development

Abstract

Derailed endolysosomal trafficking is emerging as a widespread feature of aggressive neoplasms. However, the oncogenic signals that alter membrane homeostasis and their specific contribution to cancer progression remain unclear. Understanding the upstream drivers and downstream regulators of aberrant vesicular trafficking is distinctly important in melanoma. This disease is notorious for its inter- and intra-tumoral heterogeneity. Nevertheless, melanomas uniformly overexpress a cluster of endolysosomal genes, being particularly addicted to the membrane traffic regulator RAB7. Still, the underlying mechanisms and temporal determinants of this dependency have yet to be defined. Here we addressed these questions by combining electron microscopy, real time imaging and mechanistic analyses of vesicular trafficking in normal and malignant human melanocytic cells. This strategy revealed Class I PI3K as the key trigger of a hyperactive influx of macropinosomes that melanoma cells counteract via RAB7-mediated lysosomal degradation. In addition, gain- and loss-of-function in vitro studies followed by histopathological validation in clinical biopsies and genetically-engineered mouse models, traced back the requirement of RAB7 to the suppression of premature cellular senescence traits elicited in melanocytes by PI3K-inducing oncogenes. Together, these results provide new insight into the regulators and modes of action of RAB7, broadening the impact of endosomal fitness on melanoma development.

Keywords: endosomes; macropinocytosis; oncogene-induced senescence; oncogenic stress; small GTPases.

Figure 1. The pan-PI3K inhibitor LY294002 inhibits constitutive RAB7-regulated endosomal trafficking in melanoma cells

A. Confocal fluorescence images of the incorporation of Lucifer Yellow (green) into RAB7-coated enlarged vesicles (pink) in SK-Mel-103 melanoma cells. Endogenous RAB7 protein was visualized by immunofluorescence (IF) staining. Phalloidin (red) staining outlines cytoskeletal actin. B. Snapshots of live imaging of SK-Mel-103 melanoma cells constitutively expressing RAB7 shRNAs, showing an active formation of large vacuolar structures ( > 1 μm diameter) from ruffling regions of the plasma membrane that accumulate in the perinuclear area. C. Bright field micrographs showing the reversion of shRAB7-driven cytosolic vacuolization by the pan-PI3K inhibitor LY294002 (10 μM, 8 h) in SK-Mel-103 melanoma cells transduced with three different RAB7 shRNAs. D. Quantification of the inhibitory effect of 10μM LY294002 on RAB7 shRNA-driven vacuolization of SK-Mel-103 assessed 24 h after treatment. Pooled data show means ± SEM of two independent experiments performed in duplicate. E. Bright field, fluorescence and merged micrographs showing the uptake of 70 kD Rhodamine(Rhd)-Dextran (8h) by SK-Mel-103 melanoma cells incubated in the absence or presence of 10μM LY294002. F. Confocal-based quantification of Lucifer Yellow uptake per cell (30 minute time frame acquisition), estimated in a minimum of 100 cells and expressed as arbitrary fluorescence units, AFU, with respect to non-treated cells.

Figure 2. Class I PI3K signaling drives constitutive RAB7-regulated macropinocytosis in melanoma cells

A. K_i,app values (in nM) of the indicated Class I PI3K inhibitors (see Methods for additional detail). B. Western blot analysis of SK-Mel-103 melanoma cells treated for the indicated times with Class I PI3K inhibitor GDC-0941 at the indicated concentrations, blotted for total and phosphorylated (p) AKT (Ser473). Vinculin is included as loading control. C. Representative bright field micrographs of shControl or shRAB7 SK-Mel-103 cells treated with DMSO (left) 0.5μM GDC-0941 (middle) or 0.5μM ETP-46992 (right) for 7h. The corresponding quantification of the impact of these treatments on cytosolic vacuolization is shown in D. E. Bright field micrographs showing the reversion of cytosolic vacuolization of SK-Mel-103 expressing RAB7 shRNAs by treatment with the indicated Class I PI3K inhibitors for 48h. F. Bright field, fluorescence and merged micrographs of the basal 8h-uptake of 70 kD Rhodamine(Rhd)-Dextran by SK-Mel-103 melanoma cells incubated in the absence or presence of the indicated Class I PI3K inhibitors for 48h.

Figure 3. Recruitment of RAB7 to PI3K-driven macropinosomes upon oncogene activation in human melanocytes

A. Representative IF staining for RAB7 (green) in primary human melanocytes expressing empty vector or oncogenic HRAS^G12V. Nuclei are counterstained with DAPI. B. Transmission electron microscopy image of primary human melanocytes expressing empty vector or HRAS^G12V. Note the presence of large single membrane macropinosomes in HRAS^G12V-expressing melanocytes, absent in the empty vector-expressing counterparts. C. Representative fluorescence, bright field and merged micrographs showing the activation of macropinocytic uptake (visualized by 70kD Rhodamine(Rhd)-Dextran uptake) in primary human melanocytes expressing oncogenic HRAS^G12V. D. Representative bright field micrographs showing senescence-associated β-Galactosidase (β-Gal) stainings of melanocytes transduced with empty HRAS^G12V-encoding vectors and treated with 10μM LY294002, 10 μM U0126 or vehicle control. Inhibitors were added one day post-transduction and were refreshed every 24h. Pictures were taken at day 6 post-transduction. E. Pooled quantification of vacuolized and β-Gal-positive cells from two independent experiments in vector- or HRAS^G12V- expressing melanocytes treated as indicated. Unless otherwise indicated, scale bars correspond to 10 μm.

Figure 4. RAB7 counteracts PI3K-driven oncogenic stress

A. Immunoblot analyses of total cell extracts isolated from melanocytes co-expressing the indicated oncogenes and wild-type (WT) or dominant negative (T22N) GFP-RAB7, or their corresponding empty vector controls. B. Representative micrographs showing the effect of RAB7 wild-type (WT) or dominant negative (T22N) overexpression on SA-β-gal staining and cytoplasmic vacuolization in primary human melanocytes expressing the indicated oncogenes. Scale bars, 10μm. C. Dot plot showing the impact of RAB7 wild-type (WT) or dominant negative (T22N) overexpression in the size of HRAS^G12V-induced vacuoles in primary human melanocytes (vacuoles of ≥1μm in diameter were individually measured). D. Quantification of SA-β-gal positive cells from B.. Data are presented as means ± SEM of three independent experiments.

Figure 5. In vivo visualization of RAB7-positive macropinosomes

A. Co-staining of RAB7 (red) and S100 (green) in paraffin-embedded sections of the indicated human melanocytic skin lesions. B. Visualization of RAB7-decorated endosomes (red) in paraffin-embedded melanocytic lesions (marked in green by S100) generated in indicated mouse models. Note high vesicular-patterned staining of RAB7 in S100-positive (S100+) melanocytes compared to S100- skin cells.