LZTR1 is a melanoma oncogene that promotes invasion and suppresses apoptosis

Abstract

Leucine zipper like transcription regulator 1 (LZTR1) is amplified in acral melanomas, is required for melanocytes and melanoma cell proliferation, and it induces anchorage-independent growth, by yet unknown mechanisms. We therefore performed comprehensive studies to identify its activity in melanomas employing proximity biotinylation and co-immunoprecipitation combined with LC-MS/MS proteomics and molecular characterization. The results show that LZTR1 regulates the ubiquitin proteasome system in melanoma cells and also associates with actin-related proteins and actin cytoskeleton organization. Its downregulation suppresses the protective effect of the autophagy-initiating ULK1 and AMBRA1, regulators of normal cell survival and proliferation, and upregulates the sequestosome 1 (SQSTM1/p62), an autophagic cargo adapter which mediates selective degradation of ubiquitinated proteins. In contrast, overexpression of LZTR1 provides growth advantage under environmental stress, enhancing cell invasion, by activating ERBB3 receptor and its downstream targets PYK2 and SRC tyrosine kinases that regulate the cytoskeleton, actin organization, cell spreading, cell migration and adhesion. LZTR1 is a "safeguard" for melanoma cells under stress and its downregulation can be exploited for melanoma therapy.

Fig. 1. LZTR1-associated proteins in melanoma cells identified with proximity biotinylation.

A Normal human melanocytes expressing GFP-TurboID-HA or LZTR1-TurboID-HA, as indicated. B–D Proteomic analysis of biotinylated proteins (see also Supplementary Table S2). LZTR1-TurboID or GFP-TurboID (control) were expressed in YUHIMO, YUHEF and YUSIK melanoma cells, followed by biotinylation, streptavidin enrichment and quantitative proteomics. B Venn diagram indicating the number of biotinylated proteins enriched (p < 0.05) over control in each cell line. Three biological replicates were analyzed for each cell line tested. C Selected Gene Ontology (GO) terms (Cellular Components and Biological Processes) enriched in YUHIMO LZTR1-TurboID proteomics data. For complete Gene Ontology analysis see Supplementary Table S2. D Volcano plots for enrichments (difference log₂ LFQ, label-free quantitation value) versus significance (p value, Student’s t test) for experimental compared to control in each melanoma cell line tested. X and y axes for all three plots are on the same scale. Examples of proteins enriched in all three cell lines (LZTR1, GIT1, GOLGA3, PAK2, TRIM24), or in 1–2 cell lines (USP47, VCPIP1/VCIP135, SQSTM1) are highlighted in red. E, F Western blots validating LZTR1 streptavidin-proximal proteins. Melanoma cells expressing LZTR1-TurboID or GFP-TurboID in the absence or presence of doxycycline to induce expresion were incubated with biotin for 24 h (E) or for 2 and 18 h (F), and bead-bound proteins (IP), as well as whole cell lysates (WCL) were separated by SDS-PAGE and the membrane were probed with antibodies to the indicated proteins. Cells from acral or sun-exposed melanoma are marked blue and red, respectively.

Fig. 2. LZTR1 association with the proteosome.

A Volcano plot showing enrichments of LZTR1 co-immunoprecipitated proteins (difference log₂ LFQ, label-free quantitation value). B Western blot analysis validating LZTR1 assciated proteins involved in ubiquitin-protein ligase, proteasome and autophagosomes (CUL4B, SQSTM1, PSMD2, BAG3 and BAG6, and PIK3C3/VPS34), protein stablization (USP9X) and actin cytoskeleton organization (Myosin IIa) composed of MYH9, MYH10, MYH14. The top panel shows precipitated LZTR1-HA. The RAS panel confirmed lack of association with LZTR1. Non-induced YUHIMO cells were used as control.

Fig. 3. USP9X regulates LZTR1 stability.

A USP9X shRNAs GAGAGTTTATTCACTGTCTTA and CGCCTGATTCTTCCAATGAAA (shUSP9X1 and shUSP9X2, respectively), inhibited YUHIMO acral melanoma cell proliferation compared to shRNA SCH002 vector pLKO.1 as control (shControl). **** indicates p < 0.0001 for comparison of each treated group to control using one-way ANOVA with Dunnett correction for multiple hypothesis testing. B Downregulation of USP9X in YUHEF sun-exposed melanoma cells suppressed the expression of endogenous LZTR1, as well as inducible LZTR1-TurboID, that were recovered with velcade (bortezomib). Cells were collected before (0) and at different time points after incubation with doxycycline (Dox) as indicated, and cell lysates were probed for USP9X and LZTR1, using actin as a control. Arrow points at LZTR1-BioID protein. C Cell proliferation in response to the DUB inhibitor EOAI3402, also known as G9. YUHIMO cells were the most sensitive to the drug (IC50 226 nM). D Normal human melanocytes (NBMEL) were treated with shRNA SCH002 (Control) or with USP9X shRNA (USP9X) for 6 days; velcade was added 21 h before harvest (+) to rescue proteins from degradation. The Western blot of cell lysates revealed downregulation of LZTR1, BIRC5, and RAC1 in response to USP9X knockdown that were recovered from degradation by velcade (+), validating the protecting role of USP9X from degradation of critical selected proteins. In contrast, there were no changes in the expression of CRKL, CUL3, NRAS and vinculin.

Fig. 4. LZTR1 regulates apoptosis and autophagy.

A, B Western blots showing changes in expression of BIRC3, BIRC5 (survivin), SQSTM1, ULK1 and AMBRA1 in response to depletion of LZTR1 (shLZTR1) compared to shRNA SCH002 (shControl) in melanoma cells from acral (blue) or sun-exposed (red) melanomas. C Histogram of proteasome activity in YUZEST-shLZTR1 (LZTR1), compared to YUZEST-SCH002 (Control), the background (no cells, probe only) was subtracted. MFI, mean fluorescent intensity. **** indicate significant change (p < 0.0001 by two-tailed t-test). Top shows downregulation of LZTR1 protein. D cell proliferation and E gene expression of non-melanocytic cell lines U-87MG, HEK293 and SCC-25 in response to LZTR1 knockdown compared to YUZEST melanoma cells. LZTR1 and control represent shLZTR1 and shRNA SCH002, respectively. **** indicates p < 0.0001, *** indicates p < 0.001, and * indicates p < 0.05 for comparison of each treated group to control using two-tailed t-test. The cell proliferation of the other melanoma cell lines was presented in Fig. 3 of reference [8].

Fig. 5. LZTR1 enhances cell invasion and migration under mechanical compression.

A Schematic representation of the transwell piston adapted from Tse et al. [19]. B Histograms of YUHIMO-LZTR1-HA melanoma cell invasion in the absence or presence of mechanical stress and ±200 ng/ml doxycycline (Dox). Weights (4.3 g or 7.0 g) were added to half of the transwells (as indicated). Each condition represents average of triplicate wells. *** indicates p < 0.001, ** indicates p < 0.01 for comparison of each dox treated group to untreated using unpaired two-tailed t-test. ns not significant. C Western blot analysis of YUHIMO-LZTR1-HA cell extracts. The panels show LZTR1-HA induction, activation of ERBB3, PYK2 and SRC in response to mechanical stress. D Histogram of YUHIMO-LZTR1-HA melanoma cell invasion, ±100 ng/ml doxycycline (Dox) and ±the ERBB inhibitor afatinib (ERBBi, 0.5 µM) for 6 days. Values are average of triplicates transwell and *** indicates significant change (p < 0.001) by ANOVA test with Tukey’s correction for multiple comparisons. NS Not Significant. E Western blots showing that mechano-compressions downregulated some proteins that were not affected by LZTR1 expression. These blots were performed with the same samples used in (C) (left) and thus share actin levels.