Shipping Out MEK Inhibitor Resistance with SHP2 Inhibitors

RAS proteins are small GTPases that switch between an inactive GDP-bound state and an active GTP-bound state. Transition between these two states is precisely regulated by the coordinated action of two sets of regulatory enzymes: guanine nucleotide exchange factors (GEF), which promote the GTP-bound state, and GTPase-activating proteins (GAP), which promote the GDP-bound state. RAS proteins coordinate signals from cell-surface receptors. When activated, RAS triggers several signaling cascades, including the RAF–MEK–ERK, the PI3K–AKT, and the RAC pathways, which promote cell growth, proliferation, survival, migration, and metabolic adaptation (Fig. 1). Consequently, oncogenic RAS proteins are critical cancer drivers, and alterations in the RAS genes are common abnormalities in human malignancies. In humans, there are three members of the RAS family: HRAS, NRAS, and KRAS (1). KRAS is the most frequently altered RAS member in cancer. It is mutated in 95% of pancreatic adenocarcinomas (PDAC), 50% of colorectal cancers, and 30% of non–small cell lung cancers (NSCLC; mainly adenocarcinomas; ref. 2). In addition, KRAS amplifications are found in gastric, ovarian, and endometrial cancers (2). Multiple attempts to therapeutically target RAS directly or pathways downstream of RAS have been explored.

Although small-molecule inhibitors targeting the KRAS^G12C mutant have recently been developed (3), most mutant forms of RAS have been undruggable. MEK inhibitors have been widely used to treat RAS-altered cancers. However, clinical results have fallen short of expectations due to the disruption of negative feedback regulatory mechanisms that lead to the emergence of adaptive resistance. Treatment with MEK inhibitors enhances the activation of receptor tyrosine kinases (RTK), which in turn can lead to a rebound of the MEK–ERK axis or increased signaling through other pathways, such as the PI3K–AKT cascade (ref. 4; Fig. 1). Consequently, combinations of MEK inhibitors with RTK or PI3K inhibitors have been explored; however, dose-limiting toxicities and diverse RTK resistance mechanisms limit the clinical usefulness of these combinations (4).

Now, four independent studies (5–7), including one in this issue by Fedele and colleagues (8), report a potential therapeutic strategy to prevent adaptive resistance to MEK inhibitors and broadly treat RAS-dependent tumors by cotargeting the protein tyrosine phosphatase SHP2 (encoded by the gene PTPN11). SHP2 represents a common node downstream of RTKs that is required for RAS activation. Notably, RTK-driven cancers are sensitive to SHP2 inhibition (9). To study the therapeutic potential of the combination of MEK and SHP2 inhibitors, Fedele and colleagues use KRAS-mutant pancreatic cancer and NSCLC cell culture and mouse models to demonstrate the efficacy of combining MEK and SHP2 inhibitors. The authors validate SHP2 as an essential target through rescue experiments in cultured cells with drug-resistant SHP2 mutants. Furthermore, depleting cells of SHP2 in combination with MEK inhibitor treatment synergistically impairs cancer cell growth, increases apoptosis, and promotes senescence (Fig. 1). Importantly, the authors show that the combination of a MEK inhibitor and a SHP2 inhibitor also prevents adaptive resistance in difficult-to-treat wild-type RAS tumor cells, such as triple-negative breast cancer (TNBC) and high-grade serous ovarian cancer. Consistent with these results, Wong and colleagues (7) report efficacy of the combination in wild-type and KRAS-amplified gastric cancers.

Biochemically, Fedele and colleagues show that SHP2 inhibition prevents MEK–ERK rebound following MEK inhibitor treatment by limiting MEK inhibitor–induced RAS-GTP loading (Fig. 1). Importantly, the sensitivity to the allosteric SHP2 inhibitor SHP099 correlated with the GTPase activity of the different RAS mutants, an observation corroborated by Mainardi and colleagues (6). This is an extremely important finding, with translational implications. Patients with cancer who are homozygous for codon 61 RAS mutations (such as Q61R) will likely be refractory to MEK/SHP2 dual-inhibitory therapy, because this mutant has the lowest intrinsic GTPase activity (6, 8).

Previous research places SHP2 upstream of RAS, although the mechanisms by which SHP2 contributes to RAS activation have not been completely elucidated. Both Fedele and colleagues and Wong and colleagues show that SHP2 acts upstream of the RAS-GEF SOS. Expression of the SOS catalytic domain rescues SHP2 inhibition, and depletion of SOS1 and the related protein, SOS2, phenocopies the effect of the SHP2 inhibitor. In addition, several lines of evidence are presented to show that the phosphatase activity of SHP2 is required to promote adaptive resistance to MEK inhibitors. Specifically, Mainardi and colleagues demonstrate that, unlike wild-type SHP2, a phosphatase-dead mutant of SHP2 (C459S) does not enable MEK inhibitor resistance in cells lacking SHP2. Furthermore, Ruess and colleagues (5) demonstrate that the SHP2 inhibitor GS-493, which targets the catalytic site of SHP2, also works synergistically with MEK inhibitors. Because different mechanisms of resistance to allosteric or catalytic site SHP2 inhibitors could develop, both could have clinical value if resistance emerges.

Using a combination of cell line and patient-derived xenografts, the four studies demonstrate the efficacy of combining MEK and SHP2 inhibitors in KRAS-mutant PDAC, NSCLC, KRAS-amplified gastric cancer models, wild-type RAS TNBC, and high-grade serous ovarian cancer models. Fedele and colleagues show that in the wild-type RAS cancer models, SHP2 is essential for RAS activation only after MEK inhibitor treatment. One striking observation is that SHP099 had no activity in two-dimensional cell cultures of KRAS-mutant cell lines or KRAS-amplified gastric cancer cells as a single agent. However, Fedele and colleagues show efficacy of SHP099 alone in the xenografted mice, where it induced significant tumor shrinkage in NSCLC and pancreatic cancer. This difference between in vitro and in vivo effectiveness is supported by Mainardi and colleagues and Wong and colleagues, who report similar findings in KRAS-mutant NSCLC models and KRAS-amplified gastric cancer cell line mouse xenografts, respectively. This apparent discrepancy might reflect different dependencies for activation of the MEK–ERK pathway in relation to different environmental contexts and nonautonomous effects that include limiting vascularization and stimulating an immune response.

Indeed, efficacy of the combination in vivo might be partially due to modulation of the surrounding stroma. Fedele and colleagues also observed reduced tumor vasculature in animals treated with MEK and SHP2 inhibitors. Mainardi and colleagues report that mice treated with the combination show increased tumor infiltration by T cells. The induction of senescence following treatment with MEK and SHP2 inhibitors, as well as an accompanying senescence-associated secretory phenotype, could promote activation of the immune system and clearance of the senescent cancer cells. Although the four groups used immunodeficient mouse strains, studies with immunocompetent syngeneic or genetically engineered models will reveal the spectrum of effects that combining MEK and SHP2 inhibitors has on the tumor microenvironment and the immune system.

Inhibitors of other cell signaling pathways are likely to synergize with SHP2 inhibitors. Additional combinations have already been tested, including combining SHP2 inhibitors with ALK inhibitors in EML4–ALK-resistant lung cancers (10) or with PI3K inhibitors in NSCLC and pancreatic cancer cell line cultures (5). Both combinations demonstrated promising results, hinting that we are just beginning to scratch the surface of the clinical utility for SHP2 inhibitors. Despite remaining questions, these studies provide a strong rationale to start assessing the efficacy of MEK and SHP2 inhibitor combination therapy in KRAS-driven cancers and shed light on new ways to target these difficult-to-treat cancers.

No potential conflicts of interest were disclosed.