For decades, mutated KRAS has been recognized as an attractive drug target for treating multiple types of cancer, but the development of targeted drugs has not been as successful as anticipated. The difficulties in targeting KRAS are due to several factors: (1) the broad scope of KRAS’s activity, including its essential role in many normal cellular functions, meaning drugs that directly inhibit KRAS may have significant toxicity and strong side effects; (2) KRAS’s primary functional domain involves a pocket that binds to GDP or GTP. Unlike protein kinases, which have a weak affinity for ATP, KRAS’s binding to GTP or GDP is extremely strong, with an affinity coefficient on the picomolar (10–12) level, while the concentration of GDP and GTP in normal cells is on the micromolar (10–6) level. This means that finding a small molecule compound with a binding ability to KRAS that is equivalent to GDP or GTP is extremely challenging; (3) designing a drug that selectively inhibits the activity of mutated KRAS protein while minimizing the impact on normal KRAS activity requires a compound with good selectivity for mutated KRAS, which is another difficult challenge in drug design; and (4) indirect strategies for targeting KRAS are also fraught with challenges, including the fact that KRAS signaling pathway is a necessary pathway for normal cell growth and survival, and targeting essential pathways is often associated with significant toxicity that reduces the therapeutic window to the point where it may be absent, compensatory escape mechanisms, and feedback and redundancy resulting from strict regulation.

Direct targeting drugsKRASG12D inhibitors

Wang et al. developed MRTX1133 (Fig. 2), a potent, selective, non-covalent KRASG12D inhibitor with picomolar binding affinity, through a series of structurally-based optimizations on the KRASG12C inhibitor adagrasib. Firstly, the compound has a pyrido[4,3-d]pyrimidine scaffold, and three substituents were searched to interact widely with KRASG12D protein. The C4 position is a [3.2.1]bicyclic diamino substituent to achieve optimal interaction with Asp12 and Gly60 in the mutant. The C2 position is modified with the pyrrolizidine with a 2-fluoro substituent which forms a strong ionic interaction with the negatively charged Glu62 carboxylate salt. Finally, the C7 substituent is an optimized 7-fluoro and 8-ethynyl group that lies well within the hydrophobic pocket of KRASG12D protein and forms a well-organized hydrogen-bonding network. The interaction of hydrogen bonds allows the terminal ethynyl group to effectively bridge the lipophilic and polar regions of the KRASG12D protein switch II pocket. Experimental validation demonstrated that MRTX1133 inhibits KRASG12D signal transduction in cells and in vivo, and its anti-tumor effects have been confirmed in a mouse model, showing robust in vivo efficacy and potential for targeted therapy against this “undruggable” target [30].

Fig. 2: MRTX1133.

The chemical structure of the KRASG12D inhibitor MRTX1133.

Mao et al. used a strategy based on strong interactions (salt bridges) between the alkylamine moiety and Asp 12 on the inhibitor to design a series of potent inhibitors (TH-Z816, TH-Z827, and TH-Z835) that can form salt bridges with the Asp 12 residue of KRASG12D, using the G12C inhibitor MRTX 22 as a scaffold and characterized their in vitro and in vivo activity. ITC experiments showed that these salt bridge-forming inhibitors bound to both GDP and GTP-bound KRASG12D and effectively disrupted KRAS-CRAF interaction, but did not bind to wild-type or G12C mutant KRAS. These molecules also disrupted the activation of MAPK and PI3K/mTOR signaling in different cancer cells and displayed anti-proliferative and anti-tumor effects. This study demonstrated proof-of-concept for the strategy of targeting KRASG12D by inducing adaptation pockets via salt bridge formation [3].

Meanwhile, Zhou et al. discovered an effective, selective, biologically stable, and cell-permeable peptide drug, NKTP-3, which targets NRP1 and KRASG12D. NKTP-3 first binds to NRP1 on the cancer cell membrane and is then delivered into the cell. Once inside the cell, it binds to KRASG12D and significantly inhibits downstream signaling, including AKT and ERK phosphorylation, leading to anti-tumor effects. Strong anti-tumor activity of the NRP1/KRASG12D dual-targeting cyclic peptide NKTP-3 was demonstrated in xenografts derived from A427 cells and primary lung cancer model driven by KRASG12D, with no obvious toxicity. These findings suggest that NKTP-3 may be a potential drug for treating KRASG12D-driven lung cancer [31].

In 2017, a synthetic cyclic peptide, KRpep-2d, was discovered as the first selective inhibitor of KRASG12D. The two Cys residues in the peptide are essential for its cyclic structure and control of its binding and inhibitory activity, but the bond is cleaved under intracellular reducing conditions, limiting application. Sakamoto et al. generated KS-58, a KRpep-2d derivative identified as a bicyclic peptide with a non-proteinogenic amino acid structure. KS-58 enters cells and exerts anti-cancer effects by blocking two pathways: RASGDP-SOS1 interaction (i.e., GDP-GTP exchange on RAS) and RASGTP-BRAF interaction. KS-58 was shown to selectively bind to KRASG12D and inhibit the in vitro proliferation of both the A427 human lung cancer cell line and the PANC-1 human pancreatic cancer cell line that expresses KRASG12D. However, the pharmacokinetic properties and high doses required for the treatment of this drug still need improvement. Nevertheless, KS-58 is an attractive lead molecule for developing new cancer drugs that target KRASG12D [32].

Pan-KRAS inhibitors

The broad-spectrum KRAS inhibitor is defined as a non-covalent inhibitor that exhibits high affinity for the inactive state of KRAS and can block nucleotide exchange to prevent the activation of wild-type KRAS and a wide range of KRAS mutants [33]. Using the selective KRASG12C inhibitor BI-0474 as a starting point, Kim et al. designed a broad-spectrum KRAS inhibitor, BI-2865, with potent non-covalent inhibitory activity. Experimental evidence showed that this inhibitor demonstrated similar efficacy to BI-0474 against KRASG12C mutant cells while also significantly inhibiting cell proliferation in G12D or G12V mutant cells. The inhibitor functions by preferentially targeting the inactive state of KRAS to prevent its reactivation through nucleotide exchange. Additionally, the research group conducted experiments with BI-2865 to inhibit NRAS and HRAS mutant cells, revealing that the inhibitor’s ability to inhibit nucleotide exchange in HRAS or NRAS is several orders of magnitude lower than that in KRAS, and this difference is attributed to direct and/or indirect constraints imposed by three residues in the G domain. Consequently, in cells with wild-type KRAS, the use of this inhibitor leads to increased activation of other RAS homologs, thereby limiting its antiproliferative effects [33].

SOS1 is a key guanine nucleotide exchange factor (GEF) for KRAS, which binds to KRAS protein at its catalytic binding site and promotes the exchange of GDP for GTP, thereby activating the KRAS protein. In addition to its catalytic site, SOS1 can also bind to GTP-bound KRAS at an allosteric site, forming a positive feedback regulation mechanism [34]. Hofmann et al. reported the discovery of a highly efficient, selective, and orally bioavailable small molecule SOS1 inhibitor BI-3406, which binds to the catalytic domain of SOS1, thereby preventing its interaction with KRAS. Experimental evidence suggests that BI-3406 reduces the formation of GTP-loaded KRAS and restricts the growth of most tumor cells driven by KRAS variants at positions G12 and G13. Furthermore, BI-3406 can weaken the feedback reactivation induced by MEK inhibitors, thereby enhancing the sensitivity of KRAS-dependent cancers to MEK inhibition. Thus, the development of clinical SOS1 compounds in combination with MEK inhibitors and potentially other RTK/MAPK pathway inhibitors holds promise for significant clinical benefits [35]. Hillig et al. designed a pan-KRAS inhibitor, BAY-293, using a dual-screening approach and structure-guided design. The study demonstrated that this inhibitor binds to a surface pocket on SOS1, preventing the formation of the KRAS-SOS1 complex. This pocket is located adjacent to the KRAS binding site and thus blocks the reloading of KRAS with GTP, leading to anti-proliferative activity. BAY-293 also exhibits synergistic effects with covalent inhibitors of KRASG12D, highlighting the potential of combined therapy targeting both KRAS and SOS1 [36].

Indirectly targeted drugsMEK inhibitor

Due to the difficulty of directly targeting KRASG12D drugs, targeting the KRAS signaling pathway has always focused on downstream targets, one of which is MEK. Drugs targeting MEK downstream in the MAPK cascade via inhibition of signal transduction pathways are less effective in treating KRAS-mutant NSCLC in multiple experiments. For example, Pasi et al. found that the addition of the MEK inhibitor trametinib to docetaxel did not improve progression-free survival in advanced KRAS-mutant non-small cell lung cancer patients compared to docetaxel alone [37]. The main reason is that although targeting MEK blocks MAPK cascade signaling, other downstream pathways of KRAS (such as PI3K-AKT, RAL, etc.) are strengthened. Lee et al. demonstrated synergistic effects of combination therapy with MEK inhibitor cobimetinib and immunotherapy for the treatment of KRAS-mutant NSCLC, showing anti-tumor effects and improved survival in a mouse model [38].

GRP78 inhibitor

Studies have shown that newly synthesized KRAS is cytoplasmic and inactive and undergoes a series of translation and post-translational modifications on the cytoplasmic surface of the endoplasmic reticulum (ER), which are mediated by enzymes that act as transmembrane ER proteins. Therefore, ER is the main site of KRAS maturation, and perturbation of ER homeostasis and protein quality control may be detrimental to KRAS-driven LUAD. The 78-kDa glucose-regulated protein GRP 78/BiP is a critical chaperone protein in the ER and a major pro-survival effector in the unfolded protein response (UPR). The loss of GRP78 induces UPR and apoptotic markers, which are associated with the loss of cell viability in lung cancer cell lines carrying the same KRAS mutation [39]. Ha et al. targeted GRP78 with small molecule inhibitors (such as HA15 and YUW70) with anti-cancer activity, which consistently reduced the levels of oncogenic KRAS protein in the tested cell lines. They also found that GRP78 deficiency can inhibit PI3K, AKT, TGF-β, and CD44 signaling pathways, as well as many other signaling pathways. Combined with ER stress-induced cell apoptosis and autophagy, this will provide a strong defense against the development of cancer cell resistance before cancer cells are eliminated [40].

NFkB activating kinase inhibitor

Preclinical studies have provided evidence that both classical and non-classical NF-κB pathways are co-activated in LUAD-carrying KRAS mutation. The specificity of IKK (NFkB activation kinase) synergistically induces tumorigenesis with mutant KRAS in an autocrine manner, providing survival advantages for mutant cells in vitro and in vivo. The NCT01833143 phase II single-center clinical trial of bortezomib subcutaneously administered to patients with advanced NSCLC carrying KRASG12D mutations or without a previous smoking history at Memorial Sloan Kettering Cancer Center showed some anti-tumor activity, especially in a unique subtype of lung adenocarcinoma, invasive mucinous adenocarcinoma (IMA), while boron-tazoxime is inactive in most patients with advanced KRASG12D mutant lung adenocarcinomas. Therefore, novel inhibitors of the NF-κB pathway need to be explored [41].

HSP inhibitor

Inhibiting heat shock proteins has been identified as another potential therapeutic strategy for KRAS-mutant NSCLC. Molecular chaperone Hsp 90 is essential for protein stability and maturation and prevents protein degradation by the proteasome. Vreka et al. found that IKKα is a partner of KRAS non-oncogene addiction, and specifically synergizes with mutant KRAS to induce tumorigenesis, providing survival advantages for mutant cells in vitro and in vivo. The Hsp 90 inhibitor 17-DMAG can block IKK function and have better efficacy against KRASG12D-mutant lung adenocarcinoma, opening up a new way to prevent/treat KRAS-mutant LUAD [42].

ERBB inhibitors

Previous research has shown that EGFR mutations and KRAS mutations rarely occur together, and the use of EGFR-targeted drugs alone to treat KRAS-mutant lung adenocarcinoma has not shown significant clinical benefits. However, recent experimental results suggest that the independence of mutated KRAS from upstream signaling pathways may not be absolute. Kruspig et al. demonstrated through experiments that the initiation and progression of KRAS-driven lung tumors require the involvement of ERBB family receptor tyrosine kinases (RTKs), and inhibition of the ERBB network weakens the activation of a series of downstream signaling proteins (such as pERK, STAT3, etc.), while transient pharmacological inhibition of the ERBB network enhances the therapeutic benefits of MEK inhibitors in the autologous tumor environment. Multiple ERBB inhibitors almost completely inhibit the formation of KRASG12D-driven lung tumors and enhance the benefits of MEK inhibition in tumor therapy [26].

SHP2 inhibitors

SHP2, encoded by the PTPN11 gene, plays an important role in signal transduction downstream of growth factor receptors by mainly regulating cell survival and proliferation through activation of the RAS-ERK signaling pathway. Ruess et al. found that PTPN11 gene deletion significantly inhibited tumor development in KRAS-driven pancreatic ductal adenocarcinoma and non-small cell lung cancer mouse models, providing evidence for the critical dependence of mutant KRAS on SHP2 in the process of carcinogenesis [43]. Chen et al. found that SHP2 is activated by peptides and proteins containing appropriately spaced phosphotyrosine residues, which bind the N-terminal and C-terminal SH2 domains in a bidentate manner, releasing it from the self-inhibitory interface, and make the active site available for substrate recognition and turnover. They used this natural regulatory mechanism to screen out the SHP2 inhibitor SHP099, which locks SHP2 in an autoinhibitory conformation and directly targets the inhibition of MAPK signaling and proliferation in RTK-dependent cells. This provides a feasible strategy for targeting RTK for cancer treatment [44]. Nichols et al. found through the treatment with another SHP2 inhibitor, RMC-4550 (a small molecule allosteric inhibitor), that it reduces oncogenic RAS-RAF-MEK-ERK signaling and cancer growth by disrupting SOS1-mediated RAS-GTP loading, highlighting SHP2 inhibition as a promising molecular therapeutic strategy for nucleotide cycling oncogenic KRAS in cancer [45].

Immune therapies

The inhibition and rewiring of the immune system play a crucial role in the onset and development of tumors. Immune therapies aim to reactivate anti-tumor immune cells and overcome the tumor’s immune escape mechanisms. Tumor immunotherapy, represented by immune checkpoint blockade and adoptive cell transfer, has made enormous clinical successes by inducing long-term remission of some tumors that are difficult to treat with all other therapies. Among them, immune checkpoint blockade therapy represented by PD-1/PD-L1 inhibitors (nivolumab) and CTLA-4 inhibitors (ipilimumab) has shown encouraging therapeutic effects in the treatment of various malignancies, such as non-small cell lung cancer (NSCLC) and melanoma, etc. [46].

Immune checkpoint inhibitors (ICIs), mainly represented by PD-1/PD-L1 inhibitors (nivolumab), have been widely used in the treatment of non-small cell lung cancer (NSCLC). ICIs are currently used as single-agent therapy or in combination with other treatments for first-line and subsequent therapy of metastatic NSCLC. In addition, ICI in the neoadjuvant and adjuvant treatment setting has shown efficacy for resectable disease patients, highlighting the potential of ICIs to improve outcomes for this patient group [47]. However, research has shown that most tumors have immune inhibitory mechanisms that limit the effectiveness of immunotherapy, including PD-1 expression on tumor-infiltrating T cells and the accumulation of inhibitory T cells, such as CD4+Foxp3+ regulatory T (Treg) cells, in the tumor microenvironment, which hinders anti-tumor immune responses. Studies have shown that KRASG12D mutation correlates with reduced TMB, and KRASG12D/TP53 co-mutation has a significant effect on reducing TMB and PD-L1 expression and reducing immune cell infiltration. KRASG12D mutation, especially in combination with TP53 co-mutation, maybe a negative predictive biomarker for PD-1/PD-L1 immune checkpoint inhibitors in NSCLC patients [48]. Eliminating these inhibitory mechanisms in tumors can pave the way for more effective anti-tumor responses.

Martinez-Usatorre et al. improved the efficacy of PD-1/PD-L1 inhibitors by modulating the tumor microenvironment in a KRASG12D/+;TP53−/− genetically engineered mouse model. They induced vascular normalization and facilitated T cell trafficking through inhibition of angiogenic factors VEGFA and ANGPT2 with A2 V, which improved maturation and antigen presentation of tumor-associated macrophages (TAMs). They also used CSF1R inhibitor 2G2 and platinum-based chemotherapy to deplete TAMs and reduce Treg cell numbers, respectively, to enhance the response of KRAS tumors to A2 V and anti-PD-1 dual therapy. The combination of these three agents induced an immune-infiltrated tumor microenvironment with increased CD4+ and CD8+ T cells, and decreased TAMs and Treg cells, leading to improvement in the response of KP tumors to checkpoint inhibitors [49]. Adeegbe et al. improved the response of KRAS-mutant NSCLC to immune therapy by combining JQl (a BET family inhibitor containing a bromodomain) with anti-PD-1 treatment. Bromodomain proteins are epigenetic regulators that cause growth inhibition and/or cell cytotoxicity in tumor cells. Treatment with JQl alone induced a decrease in Treg cell numbers, while combination with anti-PD-1 enhanced activation of infiltrated T cells in the tumor bed and improved effector function, leading to increased expression of Th1 cytokine profile, which is consistent with the persistent antitumor response observed with this novel treatment combination [48]. In addition, Lee et al. improved the prognosis of KRAS-driven lung cancer by combining MEK inhibitors with immunomodulatory anti-PD-1 and anti-PD-L1 antibodies. Low-dose MEKi (trametinib) and anti-PD-L1 combined therapy in KRAS-mutant NSCLC enhanced tumor microenvironment and T cell infiltration, reduced Ly6Ghigh PMN-MDSCs (myeloid-derived suppressor cells, a type of immune inhibitory cells that mainly suppress natural killer cells and effector CD8+ T cells) in tumor tissue, and inhibited tumor cell proliferation, leading to tumor cell apoptosis. MEKi acted as a sensitizer for KRAS-mutant tumors that were previously unresponsive to immune therapy [38].