Rewatee H. Gokhale,
Ashley M. Rosenberg,
Cathie M. Pfleger
Published: June 19, 2020
?This is an uncorrected proof.
AbstractDysregulation of the Ras oncogene in development causes developmental disorders, “Rasopathies,” whereas mutational activation or amplification of Ras in differentiated tissues causes cancer. Rabex-5 (also called RabGEF1) inhibits Ras by promoting Ras mono- and di-ubiquitination. We report here that Rabex-5-mediated Ras ubiquitination requires Ras Tyrosine 4 (Y4), a site of known phosphorylation. Ras substitution mutants insensitive to Y4 phosphorylation did not undergo Rabex-5-mediated ubiquitination in cells and exhibited Ras gain-of-function phenotypes in vivo. Ras Y4 phosphomimic substitution increased Rabex-5-mediated ubiquitination in cells. Y4 phosphomimic substitution in oncogenic Ras blocked the morphological phenotypes associated with oncogenic Ras in vivo dependent on the presence of Rabex-5. We developed polyclonal antibodies raised against an N-terminal Ras peptide phosphorylated at Y4. These anti-phospho-Y4 antibodies showed dramatic recognition of recombinant wild-type Ras and RasG12V proteins when incubated with JAK2 or SRC kinases but not of RasY4F or RasY4F,G12V recombinant proteins suggesting that JAK2 and SRC could promote phosphorylation of Ras proteins at Y4 in vitro. Anti-phospho-Y4 antibodies also showed recognition of RasG12V protein, but not wild-type Ras, when incubated with EGFR. A role for JAK2, SRC, and EGFR (kinases with well-known roles to activate signaling through Ras), to promote Ras Y4 phosphorylation could represent a feedback mechanism to limit Ras activation and thus establish Ras homeostasis. Notably, rare variants of Ras at Y4 have been found in cerebellar glioblastomas. Therefore, our work identifies a physiologically relevant Ras ubiquitination signal and highlights a requirement for Y4 for Ras inhibition by Rabex-5 to maintain Ras pathway homeostasis and to prevent tissue transformation.
Ras and Ras-related genes play important roles in how an organism develops in addition to maintaining health as an adult. Failure to properly regulate Ras proteins can lead to a range of diseases including cancer. Our study identifies an important aspect of how Ras is recognized by one of its inhibitors. This knowledge could be a first step in developing new anti-Ras therapeutics for Ras-associated cancers.
Citation: Washington C, Chernet R, Gokhale RH, Martino-Cortez Y, Liu H-Y, Rosenberg AM, et al. (2020) A conserved, N-terminal tyrosine signal directs Ras for inhibition by Rabex-5. PLoS Genet 16(6):
https://doi.org/10.1371/journal.pgen.1008715Editor: Jessica Esther Treisman, Skirball Institute of Biomolecular Medicine – New York University Medical Center, UNITED STATESReceived: August 22, 2019; Accepted: March 13, 2020; Published: June 19, 2020Copyright: © 2020 Washington et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Data Availability: All relevant data are within the manuscript and its Supporting Information files.Funding: CMP R01GM12995, National Institutes of Health/National Institute of General Medical Sciences, https://www.nigms.nih.gov/. CMP R21AA025722, National Institutes of Health/National Institute on Alcohol Abuse and Alcoholism, https://www.niaaa.nih.gov/. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Competing interests: The authors have declared that no competing interests exist.
IntroductionIn Drosophila, Ras proteins are represented by Ras1/Ras85D and Ras2/Ras64B. Drosophila Ras1/Ras85D corresponds to mammalian H-Ras, N-Ras, and K-Ras and is distinct from the closely related Ras2/Ras64B protein represented by R-Ras in mammals. Using the convention that predominates in the literature, here we refer to Drosophila Ras1/Ras85D as Ras; we refer collectively to mammalian H-Ras, N-Ras, and K-Ras as Ras or individually to specific isoforms as H-Ras, N-Ras, and K-Ras as appropriate; and we refer to Ras2/Ras64B as Ras2 for the Drosophila protein and R-Ras for the mammalian protein. Drosophila Ras and mammalian H-Ras N-Ras and K-Ras share sequence identity in their N-termini but diverge in a C-terminal region called the HyperVariable Region or HVR [1; for review, 2–3] (depicted schematically in S1 Fig). It is well accepted that one of the roles of the HVR is to direct the membrane association of Ras in part by the C-terminal CAAX box (cysteine, aliphatic amino acid, aliphatic amino acid, any amino acid) [1; for review, 2–3].
Ras signaling regulates cell proliferation, growth, differentiation, and cell survival by signaling to a range of downstream effectors including Raf/ERK/MAPK, RalGDS, and PI3K among others [1–12]. Consequently, Ras dysregulation in development alters patterning and causes developmental disorders collectively called “Rasopathies” [for review, 4–9]. Mutational activation and amplification of Ras in differentiated tissues are implicated in cancer [for review, 10–12]. Therefore, mechanisms of attenuating Ras activity are crucial for proper development and to prevent disease.
Rabex-5 (also called RabGEF1), an A20-like E3 ubiquitin ligase, promotes inhibitory mono- and di-ubiquitination of Drosophila Ras and mammalian H-Ras and N-Ras to restrict signaling to downstream effectors [13–16]. Rabex-5 inhibits both wild-type Ras and also the constitutively active oncogenic mutant RasG12V (also referred to as RasV12 in the literature) [13–16].
No signal in Ras has been reported to direct its inhibition by Rabex-5, and no ubiquitination targeting motif has been ascribed to Rabex-5 or the A20 family of ubiquitin ligases. We mapped a ubiquitination signal in Drosophila Ras; we report here that Rabex-5 inhibition of Drosophila Ras requires Ras N-terminal tyrosine 4 (Y4). Phenylalanine substitution mutants of Ras at Y4 (to prevent phosphorylation) were insensitive to Rabex-5-mediated ubiquitination in S2 cells and showed Ras gain-of-function phenotypes in vivo. Glutamic acid substitution mutants of Ras at Y4 (to mimic the charge of phosphorylation) showed increased Rabex-5-mediated ubiquitination in S2 cells, and glutamic acid substitution mutants of RasG12V at Y4 suppressed oncogenic Ras phenotypes in vivo, dependent on the presence of Rabex-5. JAK2 and SRC kinases are capable of promoting phosphorylation of RasWT at Y4, whereas JAK2, SRC, and EGFR can promote phosphorylation of an oncogenic form of Ras, RasG12V, at Y4 as measured by recognition by anti-pY4 antibodies.
Results and discussion
An N-terminal tyrosine-based signal directs Ras for mono- and di-ubiquitination
To elucidate the molecular mechanism of Ras inhibition by Rabex-5 and to advance our understanding of the A20 family of E3 ubiquitin ligases, we mapped a signal in Drosophila Ras responsible for Rabex-5 mediated ubiquitination with a deletion strategy (S1A Fig). In our previous work, we used a double FLAG-His6 tag on full length Ras [13, 16]. Deletion constructs were tagged with a triple tag of GFP-FLAG-His6 so that smaller constructs (corresponding to larger deletions) would be large enough to eliminate concerns of peptide instability. Ubiquitin conjugates of deletion constructs were isolated from Schneider S2 cells using nickel purification (to isolate the His6 tag). Visualization of the conjugates was achieved using the FLAG tag (to visualize Ras) and HA (to visualize ubiquitin which was expressed from an HA-Ub plasmid) as done previously [13, 16]. Ras is also regulated by the E3s Nedd4 , βTRCP , and LZTR1 [19–20]. Therefore, to map a Rabex-5 ubiquitination signal but without excluding Nedd4, βTRCP, or LZTR1 signals, our initial deletion strategy followed ubiquitination of Ras in Schneider S2 cells without Rabex-5 supplementation (S1A–S1D Fig). As noted, Ras membrane association is directed by a C-terminal CAAX signal which is represented by the amino acids CKML in Drosophila Ras. To properly localize N-terminal constructs, we tagged each deletion construct at its C-terminus with the Drosophila Ras CAAX box CKML (depicted schematically in S1A Fig; sequences listed in the methods section).
Previous work by Jura et al. reported the importance of the HVR for Ras ubiquitination. Inhibitory ubiquitination of H-Ras and N-Ras but not K-Ras was reported in mammalian cells . Notably, replacing the K-Ras HVR with the H-Ras HVR conferred ubiquitination onto K-Ras . This could have reflected the requirement for specific sequences in the H-Ras HVR not present in the K-Ras HVR; alternatively, this could have reflected the importance of the HVR in directing the localization of each Ras isoform to a compartment where the ubiquitination occurs. We report here that the Drosophila Ras HVR was neither sufficient nor required for Ras ubiquitination. C-terminal constructs were not ubiquitinated (S1A and S1B Fig). After narrowing the region sufficient for ubiquitination to the N-terminal 20 amino acids of Ras, we tested the ability of the N-terminal 20 amino acids to serve as a competitive inhibitor. Expressing GFP-Myc-tagged 1-20CKML peptides in excess prevented the formation of Ras-ubiquitin conjugates of full-length FLAG-His6 tagged RasWT isolated on nickel beads and detected by anti-HA antibodies (S1E Fig), whereas GFP-myc tagged peptides of a different 20 amino acid region in excess had no effect on RasWT ubiquitin conjugates (S1E Fig). We further narrowed the region sufficient to confer Ras ubiquitination in S2 cells to the N-terminal 10 amino acids of Ras (Fig 1A, S1A Fig, S1F and S1G Fig). Co-transfecting cells with Rabex-5 increased the ubiquitination of this region (Fig 1A, S1F and S1G Fig) but not of other small regions of Ras (S1F Fig).
Fig 1. Ras Tyrosine 4 is required for Rabex-5-mediated Ras ubiquitination.(A) Flag-His6-GFP tagged Ras (RasWT) or FLAG-His6-GFP tagged Ras N-terminal fragments tagged with C-terminal localization signal CKML (1–10 CKML, 1–20 CKML) were co-transfected into Schneider S2 cells with HA-Ub with or without Rabex-5 and purified on nickel beads as done previously [13, 16]. The N-terminal 10 amino acids of Ras contain a signal sufficient to confer ubiquitination onto GFP in the pattern of full length Ras and to support Rabex-5-mediated ubiquitination (image of entire gel in S1G Fig). The bands recognized by both anti-FLAG (the tag on Ras) and anti-HA (the tag on ubiquitin) antibodies represent ubiquitinated species of Ras and are marked by an asterisk, *. Other bands in the anti-HA gel reflect non-Ras, co-purifying ubiquitinated proteins. (B-D) Localization in S2 cells of FLAG-His6 tagged RasWT (B), RasY4E (C), and RasY4F (D) visualized by staining for FLAG. Boxes represent 20 μm square regions. (E) Western blot of FLAG-His6 RasWT and Y4 mutants purified from S2 cells on nickel beads. Ubiquitin conjugates (anti-HA antibodies, upper blot) and total Ras (anti-FLAG antibodies, lower blot) show an increase in basal ubiquitination for phoshomimic Ras, RasY4E, compared to RasWT (lane 3 compared to lane 1) and Rabex-5-mediated ubiquitination (increase in lane 4 compared to lane 3 versus the increase in lane 2 compared to lane 1). Non-phosphorylatable Ras, RasY4F, shows decreased basal ubiquitination (lane 5 compared to lane 1) and less responsiveness to Rabex-5 (lane 6 compared to lane 5). Quantification of these experiments shows the percent of Ras conjugated to ubiquitin (graph in E’) and the relative Rabex-5 mediated ubiquitination (graph in E”). *** indicates p