Cooperation between Noncanonical Ras Network Mutations

Abstract

Cancer develops after the acquisition of a collection of mutations that together create the cancer phenotype. How collections of mutations work together within a cell and whether there is selection for certain combinations of mutations are not well understood. We investigated this problem with a mathematical model of the Ras signaling network, including a computational random mutagenesis. Modeling and subsequent experiments revealed that mutations of the tumor suppressor gene NF1 can amplify the effects of other Ras pathway mutations, including weakly activating, noncanonical Ras mutants. Furthermore, analyzing recently available, large, cancer genomic data sets uncovered increased co-occurrence of NF1 mutations with mutations in other Ras network genes. Overall, these data suggest that combinations of Ras pathway mutations could serve the role of cancer "driver." More generally, this work suggests that mutations that result in network instability may promote cancer in a manner analogous to genomic instability.

Figure 1. Modeling predicts that weakly activating Ras mutants can appear strong within the NF1-deficient context

(A) (upper) Simulations of the Ras network model with all wild-type Ras, a canonical oncogenic mutant (RasG12D and RasG12V), or a noncanonical, non-oncogenic mutant (RasF28L) for both NF1-WT and NF1-deficient conditions. (lower) The net change in predicted RasGTP levels going from NF1-WT to NF1-deficient conditions for the cases above. (B) Histogram from our ‘computational mutagenesis’ displaying the number of Ras mutants with varying levels of RasGTP signal in the context of NF1-WT (blue) and NF1-deficient (red) conditions. Dashed black line shows level of RasGTP for a network with all RasWT (no mutation present); dashed brown line shows level of RasGTP for a network with a RasG12V mutation. Histogram is binned into 0.1% intervals. (C) One million random mutants were simulated in the NF1-WT and NF1-deficient states, and the resulting levels of RasGTP are plotted for both conditions. Filled circles indicate a network containing a Ras mutant: RasG12D (black), RasG12V (white), RasF28L (red); or no Ras mutant (yellow). Green lines indicate 25% total RasGTP. Any random Ras mutant falling above the dashed gold line shows a greater net change in percent RasGTP in the NF1-deficient network compared to the NF1-WT network.

Figure 2. Weak Ras mutants in mammalian cells can behave as strong activators of Ras pathway signaling under the Nf1 deficient conditions

(A) Immunoblots of Nf1+/+, Nf1+/−, and Nf1−/− mouse embryo fibroblasts (MEFs) for expression of neurofibromin, p120 Ras GAP, and phosphorylated ERK. (B) MEFs of the Nf1+/+, Nf1+/−, and Nf1−/− genotype were analyzed by qPCR for Ras GAP genes Rasa1 (p120GAP), Rasa4 (CAPRI), DAB2IP, and Nf1. Error bars represent standard deviation from three independent experiments from three different RNA extractions/preparations. (C) (left) Histograms present p-ERK profiles within Nf1+/+, Nf1+/−, and Nf1−/− MEF cells. Data presented are representative of at least 6 similar experiments. (right) (D) Nf1+/+ and Nf1−/− MEFs transfected with HA-tagged H-RasWT, H-RasF28L, or H-RasG12D with HA-tag expression and pERK signal quantified by multi-color flow cytometry. (E) Immunoblots showing expression of HA-tagged H-RasF28L and V5-tagged NF1-GRD in Nf1−/− and Nf1+/+ MEF cells following transfection, either alone or together. (F) MEFs transfected with HA-tagged H-RasF28L or with HA-tagged H-RasF28L and NF1-GRD with HA-tag expression and pERK signal quantified by multi-color flow cytometry.. Nf1(−/−) + F28L, red; Nf1(−/−) + F28L + NF1-GRD, green; Nf1(+/+) + F28L, blue; Nf1(+/+) + F28L + NF1-GRD, black. Higher HA, solid; lower HA, dashed.

Figure 3. Mathematical model of the Ras network predicts the NF1-deficient Ras network is generally more sensitive to perturbations

(A) The components of the Ras network considered are Ras, Ras GEFs, Ras GAPs, and Ras effectors, along with the modeled reactions and their biochemical parameters. (B) Sample steady-state RasGTP and net change in steady-state RasGTP levels for a range of fold-changes in model parameters. Ras expression level is shown here; all parameters are presented in Figure S3. ΔRasGTP levels are normalized to the total amount of Ras. (C) The magnitude of the immediate rate of change in RasGTP for a change in parameter is a measure of the sensitivity of the Ras network to that parameter. (D) The ratio of the sensitivities determined in C.

Figure 4. Co-occurrence of noncanonical Ras mutants with NF1, Ras GAP, and Ras GEF mutants

(A) Percentage of canonical and noncanonical KRAS, NRAS, and HRAS mutants that co-occur with an NF1 mutation within the CCLE dataset (Barretina et al., 2012) or (B) within the TCGA dataset (Kandoth et al., 2013). (C) Percentage of canonical and noncanonical KRAS mutant samples from the CCLE and TCGA that also harbor at least one GAP or GEF mutant. (D) Mutation frequency for Ras network genes within NF1 mutant and NF1 WT subsets of the CCLE dataset (Barretina et al., 2012). (E) Mutation frequency for Ras network genes within NF1 mutant and NF1 WT subsets of the TCGA dataset (Kandoth et al., 2013). N.S., not significant. The p-value for all panels is by Fisher’s exact test.