ABSTRACT
This meeting report summarizes the highlights of the II International Frontiers in Oncology meeting “The present and future of the RAS pathway: from function and genomics to inhibition” (RAS Frontiers) organized by the Center for Applied Medical Research (CIMA; Pamplona, SPAIN), the Clinic of the University of Navarra (CUN; Pamplona, SPAIN) and the Stanford Cancer Institute (SCI; Stanford University, CA, USA) in Pamplona (October 5–7, 2015).
The RAS Frontiers meeting gathered together scientists from all over the world and featured the latest advances in the study of the RAS pathway covering aspects from basic research to translational and clinical investigation. Among the topics presented were novel mouse models that recapitulate human carcinogenesis and serve as preclinical platforms for drug testing, cutting-edge approaches for the identification of novel vulnerabilities and regulators of the RAS pathway, as well as current inhibitory strategies for the treatment of human RAS-driven cancers.
Ras frontiers: A scientific report
Multiple elements of the RAS pathway are often found mutated in human cancer. Thus, the RAS pathway is the focus of an intense basic, translational and clinical research in the Oncology field. An outstanding overview of the latest advances in the RAS field from both recently published and unpublished work was discussed across several multidisciplinary sessions at the II International Frontiers in Oncology meeting “The present and future of the RAS pathway: from function and genomics to inhibition” (RAS Frontiers).
1. Strategies to target anomalous ras signaling: Preclinical models
In the plenary session Mariano Barbacid (Spanish Cancer Research Center -CNIO-, Madrid, SPAIN) and Julian Downward (The Francis Crick Institute, London, UK) outlined the current approaches to target RAS oncogenes. These included 1) direct inhibition of RAS oncoproteins, 2) inhibition of immediate downstream effectors, 3) targeting of distal effectors and 4) targeting synthetic lethal interactions. Mariano Barbacid described the use of sophisticated genetically-engineered mouse models (GEMM) that combine Cre and Flippase recombinases, to discover that deletion of the immediate RAS effector CRAF causes a strong deleterious effect on established lung adenocarcinoma (LAC). However, using a novel knock-in GEMM strain expressing an inactivating mutant CRAF protein, which would mimic the effects of treatment with CRAF inhibitors, his lab did not found any therapeutic effect of CRAF inactivation in established LAC. These results argue that GEMM models can provide clinically relevant insight into the likely therapeutic responses to pharmacological inhibitors.
Julian Downward described his lab's work on targeting downstream effectors of RAS, such as the type I phosphoinositide 3-kinases (PI3Ks). Disruption of the interaction of PI3K p110α with oncogenic mutant KRAS was shown to prevent tumor development in a mouse model of KRAS driven lung cancer and also to cause partial regression of established tumors. Downward also described loss-of-function genomics screens aimed at identifying unique dependencies of RAS mutant tumor cells. His data suggested that it would be worth re-examining proteasome inhibition in KRAS mutant tumors. Downward also described the existence of molecular mechanisms whereby RAS signaling pathways can promote local immune suppression in the tumor. Given the increasing interest in how best to combine immune checkpoint therapies with targeted agents to the RAS pathway, his lab is currently developing new GEMMs that more closely recapitulate human tumors by knocking-out MAD2 to increase aneuploidy, knocking-out the expression of mutation-correcting enzymes or increasing the mutational rate by overexpressing Apobec3. It is anticipated that these models will elicit a stronger immune response and thus serve as better models of the human disease.
Karen Cichowski (Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA) described 3 different strategies to target tumors with abnormal RAS signaling. First, Cichowski showed that inhibition of single bona fide therapeutic targets, such as MNK1/2 using the pharmacological inhibitor XL184, has a strong antitumor effect in mouse peripheral nerve sheath sarcoma tumors (MPNST). Second, she described various strategies to combine pharmacological inhibitors with suboptimal effect as single agents to target Ras-driven cancers. In one of these instances, Cichowski described that, in Nf1-deficient malignancies and Kras/p53 mutant lung cancer, agents that enhance proteotoxic stress, including the HSP90 inhibitor IPI-504, induce tumor regression when combined with rapamycin. Lastly, Cichowski reported that SUZ12 inactivation impaired polycomb repressive complex 2 (PRC2) thus sensitizing RAS-driven cells to bromodomain (BRD) inhibitors. In this context, cotargeting RAS tumors with RAS effector (MEK) inhibitors and BRD inhibitors has a strong antitumoral effect.
Claudia Wellbrock (Manchester Cancer Research Center, The University of Manchester, Manchester, UK) showed that the melanoma-specific transcription factor MITF, which displays heterogeneous expression and marks MITF-high and MITF-low cells within a tumor, confers cell-autonomous resistance to MAP-kinase pathway therapy. Importantly, in the majority of patients treated with BRAF and MEK inhibitors MITF expression was up-regulated within the first weeks. Wellbrock reported that this is a direct consequence of effective inhibition of the MAP-kinase pathway. With the relevance of MITF to melanoma biology and response to targeted therapy, Wellbrock's lab performed a high throughput drug screen using a FDA approved drug library and identified a novel approach for combination therapies in melanoma, which not only improves responses in BRAF mutant, but also NRAS mutant melanoma.
Jens Siveke (DKFZ/DKTK partner site Essen, West German Cancer Center, University Hospital Essen, Essen, GERMANY) presented a new pharmacological strategy for the treatment of PDAC. Siveke showed that the bromodomain inhibitor JQ1 impairs PDAC development in mice by inhibiting both MYC and inflammatory signals. Siveke described that JQ1 treatment synergizes with the histone deacetylase inhibitor SAHA in advanced PDAC and detailed that the molecular mechanism underlying PDAC regression was through de-repression of p57 (also known as CDKN1C). These data suggest that therapeutic strategies targeting epigenetic-based mechanisms could provide effective in one of the most lethal KRAS-driven cancers.
Xosé Bustelo (Salamanca Cancer Research Center, CSIC-University of Salamanca, Salamanca, SPAIN) presented work on the role in cancer of R-Ras2 (also known as TC21), a GTPase that shows high structural similarity with the 3 members of the Ras subfamily. Using loss-of-function approaches, he showed that the elimination of endogenous R-Ras2 can abate breast cancer primary tumorigenesis and the metastatic dissemination of cancer cells, effects that are even seen when such an inactivation takes place in cancer cells bearing oncogenic mutations in classical Ras proteins. Interestingly, the contribution of R-Ras2 to this process seemed to be through a PI3K-dependent route that mediates efficient protein synthesis rather than distal transcriptional events. Perhaps more importantly, Bustelo described data using a new inducible R-Ras2 knock-in mouse strain that the somatic replacement of endogenous wild-type R-Ras2 by a mutant version sporadically found in human tumors leads to the development of aggressive T cell acute lymphoblastic leukemias, which exhibit both constitutive Notch1 and PI3K signaling, in addition to noninvasive marginal zone splenic lymphomas, Harderian gland adenomas, skin papillomas, and several subtypes of genitourinary tumors. Collectively, his data indicate that R-Ras2 promotes signaling events that are overlapping, but not identical, to those usually triggered by the 3 Ras subfamily proteins and demonstrate, for the first time, that genetic alterations in R-Ras2 probably act as oncogenic drivers for specific types of hematological and solid tumors.
2. Regulation of ras function
Data on the regulation of RAS oncoprotein expression and function revealed the involvement of complex and unanticipated molecular mechanisms. Chris Counter (Duke University Medical Cancer, Durham, NC, USA) presented work from his laboratory showing that KRAS proto-oncogene is enriched for rare codons, leading to decreased translation and therefore decreased protein expression compared to other Ras family members. To investigate the consequences of rare codons in tumorigenesis, introduction of synonymous mutations in exon 3 of the Kras gene in mice to convert rare codons into common (optimized) codons was carried out. These experiments revealed that mice with at least 1 copy of the “codon optimized” allele developed fewer tumors upon carcinogen treatment. In addition, this “codon optimized” allele was less frequently mutated. Additional results showed that the reduction in tumorigenesis was partially due to the higher expression of the “codon optimized” allele which induced growth arrest.
Piero Crespo (Biomedicine and Biotechnology Institute of Cantabria –IBBTEC-, Santander, SPAIN) talked about the distribution of HRAS proteins in different plasma-membrane microdomains and endomembranes. His data indicated that activation of endogenous RAS at the Golgi complex through palmytoylation blocks ERK phosphorylation, a mechanisms dependent on PTP Kappa expression. In this regard, the acyl protein thioesterase 1 (APT1) inhibitor, a bona fide depalmitoylating enzyme, was found to block RAS at the Golgi complex to induce apoptosis. Finally, Crespo also reported data indicating that physiological signals inducing differentiation or proliferation avoid activating RAS at the Golgi.
Extensive sequencing of cancer genomes has revealed a list of non-canonical mutations that can activate KRAS. Kevin Haigis (Beth Israel Deaconess Medical Center, Boston, MA, USA) presented data demonstrating that KRAS mutations at codon 146 (A146T) induce a different phenotype to that caused by G12D mutations, including different epithelial morphology and increased senescence. Of note, cells expressing KrasG12D and KrasA146T were differentially sensitive to MEK inhibition. Data in mouse pancreas showed that expression of KrasA146T did not elicit PanIN development in contrast to robust PanIN development in mice expressing KRASG12D. Quantification of proteome changes in human colorectal cancer cell lines and colons of genetically-engineered mice using quantitative mass spectroscopy (MS) unveiled a complex picture including lack of phosphorylation of many KRAS effectors (i.e. MAPK and PI3K), suggesting that molecular mechanisms triggered by KRAS are highly influenced by the type of mutation.
Eugenio Santos (Salamanca Cancer Research Center, CSIC-University of Salamanca, Salamanca, SPAIN) presented data on the GTPase exchange factors (GEFs) families Sos and Grf. His lab undertook a systematic approach to compare side-by-side the effect of WT, single knockout (KO) and double knockout (DKO) genotypes of Sos1/2 and Grf1/2 genes in different biological settings. Analysis of WT, Sos1 KO, Sos2 KO and Sos1/2 DKO MEFs showed that Sos1 KO and Sos1/2 DKO MEFs exhibited distinct flat morphology, enlarged cell perimeter and altered cytoskeletal organization that were not observed in the WT and Sos2 KO counterparts and also displayed significant accumulation of autophagosomes containing degraded mitochondria. In vivo labeling with specific fluorophores uncovered increased levels of oxidative stress consistent with a mitophagic phenotype. These experiments uncovered a direct mechanistic link between Sos1 and control of intracellular oxidative stress, and demonstrated functional prevalence of Sos1 over Sos2 with regards to cellular proliferation and viability. Next, comparisons among WT, Grf1 KO, Grf2 KO and Grf1/2 DKO mouse retinas uncovered the presence of misplaced, “ectopic” cone photoreceptor nuclei located in the photoreceptor segment (PS) area of retinas from Grf2 KO and Grf1/2 DKO, but not of WT or Grf1 KO mice. Accumulation of ectopic nuclei in the PS area of Grf2-devoid retinas was highest between P10 and P15, indicating that abnormal postnatal cone nuclear migration is a novel, specific Grf2-dependent phenotype. These data identified defective cone nuclear migration as a novel phenotype in Grf2 KO mouse retinas, supporting a critical role of Grf2 in control of the nuclear migration processes required for proper postnatal development of retinal cone photoreceptors.
3. Functional genomics approaches to identify novel ras vulnerabilities
Several speakers described functional genomics strategies, including gain- and loss-of-function approaches as well as gene- and protein-expression profiling with the overall goal of better understanding the RAS signaling network and identifying potential RAS vulnerabilities. Alejandro Sweet-Cordero (Stanford University Medical Center, Stanford, CA, USA) described his lab's work identifying novel Kras synthetic lethal interactions. Previously, his laboratory identified Wt1 (Wilm's tumor 1) as a gene whose loss triggers senescence in cells expressing oncogenic KRAS but which is dispensable in cells expressing wild-type KRAS. More recent work from his lab using proteomic approaches suggests possible novel protein interaction networks that may explain the molecular function of Wt1 in Kras mutant tumors. His lab has also carried out a novel shRNA screen in “3D” cultures in order to identify KRAS dependencies and also to identify genes that allow tumor-propagating cells to become chemoresistant. Work elucidating the effects of oncogenic Kras in cells deprived of glutamine was also presented.
William Hahn (Dana Farber Cancer Institute) described his lab's work as part of the Achilles Project, a large effort to identify Kras synthetic lethal interactions using both genome-wide shRNA screens and genome-wide CRISPR screens. His work suggests that a significant number of shRNA seeds actually are off-target but that these off-target effects can be identified using computational tools.
In KRAS-driven tumors, the challenge is to identify authentic intervention targets for development of an appropriately diverse cohort of therapies to contend with disease heterogeneity together with molecular enrolment biomarkers that specify patient populations responsive to those therapies. Michael White (UTSouthwestern Medical Center, Dallas, TX, USA) unveiled a series of nuclear export proteins that are specifically expressed in KRAS-driven cells irrespectively of proliferative status. White then showed data on the deleterious effect of inactivating the nuclear export protein XPO1 in mutant human LAC KRAS cells and provided evidence that the mechanism involves addiction of mutant KRAS cells to NFkB signaling.
Reinhold Schäfer (Charité Comprehensive Cancer Center, Berlin, GERMANY) presented data on the analysis of the effects on the transcriptome in CRC cell lines, in which signaling via MAPK, PIK3CA and RAL pathways were selectively blocked by pharmacological inhibition or RNA interference. Schäfer described the regulatory roles of the upregulated transcription factors HMGA2 and FOSL1 as well as further components of the transcription factor network. These factors appeared to play important roles in the dysregulation of the genetic program and particularly in executing neoplastic properties: HMGA2, FOSL1, JUNB and KLF6 controlled anchorage-independent proliferation, a hallmark of tumorigenicity, of RAS-transformed cells. RELA, OTX1 and GFI1 genes controlled cell cycle progression and proliferation in monolayer culture. In parallel to these studies, his lab integrated mathematical modeling of the MAPK signaling network and experimental validation to identify approaches for clinically relevant, combinatorial therapeutic targeting. In addition, Schäfer embarked on a systematic, time-resolved analysis of transcriptional (miRNA, mRNA) and translational effects of inducible RAS in human embryonic kidney cells to show that protein phosphorylation accompanies enhanced expression of HMGA2 and FOSL1 at the mRNA level. In addition, miRNAs targeting HMGA2 and FOSL1 mRNA were downregulated in a MAPK-dependent manner. These findings indicate several layers of complex regulation downstream of the RAS signaling system at the transcriptional and post-translational level.
4. Inhibition of the ras pathway in the clinic
The RAS Frontiers meeting also featured various sessions oriented to learn about the latest advances in the clinic for the treatment of the deadliest KRAS-driven tumors, colorectal cancer (CRC), PDAC and LAC. Josep Tabernero (Vall D'Hebron Institute of Oncology -VHIO-, Barcelona, SPAIN) explained the current therapeutic approaches for the treatment of CRC with mutations in elements of the RAS pathway. Tabernero showed that 1) not all KRAS mutations have the same value as negative predictive biomarker of anti-EGFR response and, in fact, patients with KRASG13D mutations may respond to EGFR inhibitors and 2) the combined action of BRAF and EGFR inhibitors has antitumor effect in BRAFV600E xenograft models, and that these therapeutic approaches has extended to currently ongoing clinical trials in human patients. He also went on to describe the mechanisms of resistance to MEK inhibitors in CRC patients, which partially occur through transcriptional induction of ERBB3. Lastly, Tabernero reported that HER2 amplified CRC patients can benefit from a combined strategy including the HER2 inhibitor Trastuzumab and the dual HER2 and EGFR inhibitor Lapatinib.
Egbert Smit (VU University Medical Center, Amsterdam, HOLLAND) first reviewed his own as well as others data reflected in the literature addressing the lack of a clear predictive or prognostic role of KRAS mutations among patients diagnosed with NSCLC. However, Smit presented the results from a phase II trial in which KRAS-mutated NSCLC patients treated with Sorafenib, a multi-tyrosine kinase inhibitor with inhibitory activity over the Ras-Raf pathway, showed significant antitumour activity with a disease control rate over 50%. Finally, additional information from another phase II trial combining Metformin and Sorafenib in the same clinical setting was explained.
Luis Paz-Ares (Hospital 12 de Octubre, Madrid, SPAIN), presented a general overview of the current treatment options for specific gene alterations-driven non-small cell lung cancer (NSCLC) patients. More specifically, he showed positive data from the EURTAC clinical trial addressing the clinical benefit of first-line Erlotinib in EGFR mutant NSCLC Caucasian patients leading to the approval of Erlotinib by the regulatory agencies in that clinical setting. He also presented the results from the ARCHER 1009 trial in which Erlotinib was compared to Dacomitinib, an irreversible EGFR tyrosine-kinase inhibitor, for the treatment of an unselected population of advanced NSCLC patients including KRAS wild-type-harbouring tumors. Both drugs showed to be similar in terms of progression-free survival. In addition, he addressed the treatment rationale and the study design of the JUNIPER study, a randomized study of Abemaciclib, a CDK4/CDK6 inhibitor, with best supportive care (BSC) versus Erlotinib with BSC in patients with stage IV NSCLC whose tumors have detectable KRAS mutations and whose disease has progressed after platinum-based chemotherapy and one other previous therapy, or who are not eligible for further chemotherapy. This trial is currently recruiting patients and has been designed to provide a new alternative third-line treatment option for patients with NSCLC whose tumors show detectable KRAS mutations. Finally, Paz-Ares summarized the most relevant results obtained from 2 recent clinical trials showing the superior efficacy of second-line Nivolumab, a PDL1 inhibitor in the treatment of both, squamous and non-squamous NSCLC patients, when compared to Docetaxel.
Summary
The RAS Frontiers was a successful meeting driven by a group of outstanding speakers, who discussed recently published as well as novel data, and was highly enjoyed by all the participants. The innovative meeting format, which combined sessions that covered topics related to basic science as well as translational and clinical investigation, was highly appreciated by the attendees and provided an exceptional forum for potential collaborations between basic scientists and oncologists. We are grateful to all public and private institutions that contributed toward this initiative and we look forward to have you all again in Pamplona soon.
Disclosure of potential conflicts of interest
No potential conflicts of interest were disclosed.