Genetic and functional characterization of putative Ras/Raf interaction inhibitors in C. elegans and mammalian cells

Abstract

Background: Activation of the mammalian Ras-Raf-MEK-ERK MAPK signaling cascade promotes cellular proliferation, and activating Ras mutations are implicated in cancer onset and maintenance. This pathway, a therapeutic target of interest, is highly conserved and required for vulval development in C. elegans. Gain-of-function mutations in the Ras ortholog lead to constitutive pathway signaling and a multivulva (Muv) phenotype. MCP compounds were identified in a yeast two-hybrid screen for their ability to disrupt Ras-Raf interactions. However, this had not been confirmed in another system, and conflicting results were reported regarding selective MCP-mediated blockade of Ras- and Raf-mediated biological activities in mammalian cells. Here we used the easily-scored Muv phenotype as an in vivo readout to characterize the selectivity of MCP110 and its analogs, and performed biochemical studies in mammalian cells to determine whether MCP treatment results in impaired interaction between Ras and its effector Raf.

Results: Our genetic analyses showed significant dose-dependent MCP-mediated reduction of Muv in C. elegans strains with activating mutations in orthologs of Ras (LET-60) or Raf (LIN-45), but not MAP kinases or an Ets-like transcription factor. Thus, these inhibitors selectively impair pathway function downstream of Ras and upstream of or at the level of Raf, consistent with disruption of the Ras/Raf interaction. Our biochemical analyses of MCP110-mediated disruption of Ras-Raf interactions in mammalian cells showed that MCP110 dose-dependently reduced Raf-RBD pulldown of Ras, displaced a fluorescently-tagged Raf-RBD probe from plasma membrane locations of active Ras to the cytosol and other compartments, and decreased active, phosphorylated ERK1/2.

Conclusions: We have effectively utilized C. elegans as an in vivo genetic system to evaluate the activity and selectivity of inhibitors intended to target the Ras-Raf-MAPK pathway. We demonstrated the ability of MCP110 to disrupt, at the level of Ras/Raf, the Muv phenotype induced by chronic activation of this pathway in C. elegans. In mammalian cells, we not only demonstrated MCP-mediated blockade of the physical interaction between Ras and Raf, but also narrowed the site of interaction on Raf to the RBD, and showed consequent functional impairment of the Ras-Raf-MEK-ERK pathway in both in vivo and cell-based systems.

Figure 1

MCP110 and MCP116 but not MCP146 inhibit the Ras/LET-60-induced Muv phenotype. A. Representative images of untreated wild-type animals (WT phenotype with normal vulva (arrowhead) but no ventral protrusions); animals harboring constitutively activated Ras treated with vehicle only (Muv phenotype; arrows mark ventral protrusions formed by pseudovulvae); the same strain following treatment with MCP110 (20 μM) (WT phenotype, as indicated by lack of ventral protrusions). B. Animals harboring constitutively activated Ras as in Panel A were treated with either vehicle (DMSO) or MCP110, MCP116, MCP146 or negative control MCP122 (3, 10 or 20 μM; higher concentrations precipitated out of solution). The Y-axis indicates the percentage of MCP-treated animals with a Muv phenotype, normalized to vehicle-treated animals. Data were analyzed by three-way ANOVA. (***) and (*) indicate p values of < 0.0001 and < 0.05, respectively.

Figure 2

MCP110 and derivatives do not inhibit the Muv phenotype induced by activation of the Ras-Raf-MAPK pathway downstream of Raf. Animals were treated, data were analyzed and results are presented as described in the legend to Figure 1B. A. Ets and B. activated MEK/ERK.

Figure 3

MCP110 inhibits the Raf-AA Muv phenotype. Animals expressing a constitutively activated Raf/LIN-45 protein were treated with vehicle (DMSO) or MCP110 as in Figure 1B. Bar graphs show the percentage of MCP-treated animals with a Muv phenotype, normalized to vehicle-treated animals. (***) and (**) indicate p-values of < 0.0001 and < 0.01, respectively.

Figure 4

MCP110 inhibits Ras/Raf interaction and signaling to ERK in mammalian cells. A. Pulldown assay for Ras-Raf interaction. Pulldowns of active Ras bound to Raf-RBD were done using GST-Raf-RBD as described in Methods. NIH 3T3 cells were treated with vehicle or MCP110 (3, 10 and 30 μM) before and after lysis. Ras was detected by immunoblotting with anti-Ras antibody. Both Raf-RBD-bound Ras (Ras-GTP, upper panel) and total Ras in the lysates (lower panel) are shown. MCP110 disrupted the Ras/Raf interaction in a dose-dependent manner; numbers shown below panels indicate quantitation of Ras pulldown by densitometry, normalized to vehicle control. B. Western blot analysis for active, phospho-ERK1/2. The same lysates from cells depicted in panel A above were immunoblotted for phospho-ERK1/2 (P-ERK, upper panel) and for total ERK1/2 (total ERK, lower panel). Numbers shown below PERK panel indicate quantitation by densitometry, normalized to vehicle control. MCP110 decreased phospho-ERK1/2 in a dose-dependent manner.

Figure 5

MCP110 impairs recruitment of Raf-RBD to subcellular locations of active Ras. A. Recruitment of YFP-Raf-RBD to sites where Ras is localized, as detected by fluorescence microscopy. NIH3T3 cells transiently expressing both HA-tagged active Ras and YFP-Raf-RBD were treated with vehicle or MCP110. Shown are representative images of cells quantitated in Panel B below. In the absence of active Ras (v.o.), YFP-Raf-RBD (green) was localized diffusely throughout the cytoplasm and nucleus, whereas in the presence of active Ras, the Raf-RBD probe was recruited to the plasma membrane and, like Ras (red), was nuclear-excluded ("vehicle" panels). Increasing concentrations of MCP110 increasingly shifted the YFP-Raf-RBD probe from the plasma membrane to the cytosol and to internal membranes and finally to both cytosol and nucleus, whereas Ras remained membrane-associated and nuclear-excluded. These results indicate dose-dependent disruption of the Ras/Raf-RBD interaction by MCP110. B. Quantification of the distribution of YFP-Raf-RBD subcellular localization. Cells treated and analyzed as described in Methods and depicted qualitatively in Panel C were binned according to whether the YFP-Raf-RBD probe accumulated primarily in the cytosol, or cytosol + internal membranes, or at the plasma membrane and was nuclear-excluded. MCP110 dose-dependently disrupts the ability of Ras to recruit Raf-RBD.