Dawn is dawning—the way to attack KRAS G12C star inhibitor (part 1)

The mutation of ras oncogene is the most common activation mutation in human cancer, which occurs in 30% of human tumorsAlthough it is one of the earliest oncogenes, three decades of efforts have failed to identify clinically feasible KRAS protein inhibitorsThere are two main reasons: (1) KRAS binds to GDP and GTP with picomolar affinity, which seriously hinders the efforts to develop nucleotide competitive inhibitors; (2) KRAS protein lacks other deep surface hydrophobic pockets, It hinders the search for high affinity allosteric inhibitors [1] [2] [3]< br / > there was a major breakthrough in 2013Shokat et al [4] reported a new strategy to overcome these challenges, which uses covalent inhibitors to target the active cysteine of krasg12cThe crystal structure of the combination of wild KRAS and GDP is shown in the left figure of Figure 3 [5]Crystallographic studies show that a new pocket is formed under the switch-ii region where the effector bindsIt is worth noting that the mutant cys12 is located near the nucleotide pocket and switching region related to the interaction of the effector (as shown in the right figure of Figure 1 below), The fragment screening method based on disulfide was used to screen the library with 480 chain compounds under the condition of GDP bindingThe fragments 6h05 and 2e07 were obtained, and the two fragments did not react with wild-type K-ras containing 3 natural cysteine residues< br / >This completely formed pocket is not obvious in other published structures of RAS, although grooves can be seen in some casesPrevious studies have shown that there are allosteric sites in this regionThe composite binding area is called switch-ii pocket (s-iip) [4]< br / > < br / > < br / >The hydrophobic dichlorophenyl of 6 forms several hydrophobic contactsGlu99 and gly60 form direct hydrogen bonds to 6 (as shown on the right of Figure 3 below)However, switch-ii has an obvious rearrangement effect on the formation of s-iip, and the configuration of switch-i has not changed from GDP binding state [4] < br / > < br / > < br / > It is also proved that compound 12 does not react with the cysteine of wild type KRAS Covering multiple eutectic structures shows that the compound follows a similar trajectory through the pocket and projects the functional group into the (o -) and (P -) subpackages Although there are considerable differences in terminal benzene rings, these compounds satisfy similar hydrophobic interactions, supporting the key role of this region of s-iip (as shown in Figure 4 below) [4] The catalytic effect of < br / > < br / > < br / > EDTA on nucleotide exchange shows that although compounds may subtly affect metal binding, leading to changes in nucleotide affinity, Mg2 + cannot be excluded even when s-iip is occupied Structural analysis also predicts that the function of the exchange factor SOS may be affected by its combination with s-iip Using compounds 8 or 12 to block the SOS catalyzed nucleotide exchange, K-ras (G12C) is treated As mentioned above, these compounds will not damage the EDTA catalyzed GDP exchange The specific results are shown in Figure 5 [4] < br / > < br / > The role of compound 12 was studied in a small number of lung cancer cell lines with gene phenotypes As expected, cell lines with G12C mutations (h1792, H358, H23, and Calu-1) showed reduced survival and increased apoptosis after treatment compared to groups without G12C mutations (h1437, H1299, and A549), as shown in Figure 6 (left) below Highly sensitive h1792 cells showed low levels of k-rasgtp, consistent with the preferred binding of inhibitors to K-ras GDP, as shown in Figure 6 (middle) They are highly K-ras dependent as shown in Figure 6 (right) below It is worth noting that K-Ras-dependent (A549) and non-dependent (H1299) cell lines lacking G12C are not sensitive to compound 12 The EC50 of compound 12 in h1792 cells was 0.32 ± 0.01 μ m [4] < br / > < br / > < br / > in general, cell data provide a conceptual validation for the genotype specific use of the s-iip binding complex in k-rasg12c driven cancers < br / > Matthew P Patricelli et al [6] continued to explore on the basis of compound 12, and found that compound 12 did not show substantial krasg12c CO binding in nci-h358 (H358) cells containing krasg12c mutation, even after treatment with 100 μ mol / L compound for 6 hours In order to solve the possible reason of lacking cell activity of compound 12, we need more effective switch II pocket inhibitors We designed a covalent krasg12c targeting reagent based on iterative structure, and tested their activity to purified recombinant krasg12c and their ability to bind to KRAS G12C in cells Ars-853, which gives great hope, is shown in Figure 7 below In biochemical assay, the binding rate constant of ars-853 and krasg12c was 76m-1s-1, 600 times higher than that of compound 12, and the cell binding IC50 was 1.6 μ mol / L in 6 hours In the presence of GDP, the high-resolution crystal structure of the ligand binding krasg12c determines the binding site of ars-853 as the switch II pocket described above In this structure, ars853 is covalently connected to C12 and extends to the switch II pocket area between the central β - fold and α 2 and α 3 helices of KRAS Compared with the structure of other published switch II binding compounds, ars853 induced the rotation of α 2-helix, accompanied by the displacement of M72, to adapt to the ligands in different hydrophobic pockets The hydrophobic pocket is occupied by the aromatic ring of ars-853 Chlorine and methylcyclopropyl substituent provide close van der Waals contact, while the phenolic hydroxyl forms hydrogen bond with D69 The carbonyl group of ars-853 acrylamide forms hydrogen bond with the conserved K16 and a water molecule coordinated with the usual magnesium ion, and occupies a similar position with the terminal phosphoric acid of GTP binding form of KRAS In general, several features of the structure indicate that krasg12c combined with ars-853 represents the inactive state of KRAS See Figure 8 [6] for details < br / > < br / > < br / > < br / > despite the above breakthroughs, revealing previously unknown Ras binding pockets, researchers have also carried out a large number of screening tests, but they only lead to limited proof, and at the same time, there is a serious lack of convincing evidence of in vivo response to targeted mechanism of action [7] [8] Matthew R Janes et al [9] Based on the structure of drug design, continue to improve the efficacy of mutation targeting and KRAS specific inhibitors and drug-like properties It was found that the main disadvantages of ars-853 series compounds were short plasma metabolic stability (T1 / 2 < 20min) and poor oral bioavailability (f < 2%), which made them unable to be used for further in vivo study The limited structure activity relationship also limits the further improvement of ars-853 series The emphasis is shifted to different scaffolds of different design structures, which properly locate the conformation and trace of acrylamide, and allow the best hydrophobic binding part to be properly placed in s-iip We assume that the flexible 2-amino-1 - (π - piperazine-1-yl) ethane-1-one linker of ars-853 series can be shortened and replaced by a more rigid double ring scaffold, and is suitable for the free region between the S-II and A3 helices repackaged by the previously reported ars-853 eutectic structure [6] After extensive stent optimization, we determined that the core of quinazoline is a multifunctional stent, which can overcome the SAR limitation of ars-853 and has better drug like properties See Figure 9 below The progress of < br / > < br / > < br / / > led to the quinoxaline based series, and after the systematic optimization of the substituents around the scaffold, it led to a significant improvement in the covalent binding activity of ADME / PK and KRAS, as shown in Figure 10 [9] < br / > < br / > < br / > < br / > < br / > the eutectic structure of ars-1620 combined with krasg12c shows the unique binding mode and binding track from the initial s-iip krasg12c inhibitor to cys-12, and reveals the acquisition of additional key interactions with His-95, thus obtaining a more rigid and favorable covalent reaction conformation than ars-853 series compounds (as shown in Figure 12 and figure 13) [9] < br / > < br / > < br / / > in biochemical experiments, the observation rate of ars-1620 covalently modified krasg12c was 10 times higher than that of ars-853, which proved that the modification effect of ars-1620 on KRAS was significantly higher than that of the previous compounds Ars-1620 preferentially interacts with the GDP binding state of krasg12c, lacking detectable reactivity to any residue of wt-kras protein In addition, KRAS G12C covalently bound to ars-1620 lacks the ability of nucleotide exchange mediated by SOS and EDTA At the same time, the inhibitory activity of ars-1620 and the selectivity of krasg12c in pan-ras were confirmed at the cellular level [9] < br / > a major risk of covalent inhibitors is the possibility of non-specific reactions with non target proteins There are 200 kinases with reactive cysteine near the ATP pocket [10], which also extend to non kinase protein with reactive cysteine [11] If targeted, it may cause unnecessary specific toxicity In order to better define the potential Miss susceptibility and specificity of ars-1620, unbiased chemical proteomic screening was used to evaluate the covalent reaction activity in cells The resolution covered 8501 cysteine residues of 3012 annotation proteins, and finally confirmed the specificity of ars-1620 target [9] < br / > ars-1620 shows excellent oral bioavailability and sufficient blood stability in mice, as shown in Figure 14 (left), which enables quantitative measurement of G12C target occupancy in vivo It is not clear whether ars-1620 can satisfy enough covalent and pharmacokinetic characteristics to make the s-iip G12C targeting method successful in vivo In order to show whether ars-1620 has the ability of target occupancy in vivo, ars-1620 was taken orally in the established xenotransplantation model with KRAS p.g12c After a single oral dose or 5 consecutive daily doses, the average tumor peak concentration generated by ars-1620 was 1.5 μ m (50 mg / kg) and 6.5 μ m (200 mg / kg), respectively, which made KRAS G12C target occupy a significant (≥ 70% g12c-te at 200 mg / kg) for more than 24 hours, as shown in Figure 14 (right) Under these exposures, ars-1620 provided significant G12C target occupancy (75% to 90% g12c-te), as well as ras-gtp and downstream signal suppression [9] < br / > < br / > As shown in Figure 15 [9] < br / > < br / > < br / / > the systematic evaluation of KRAS dependence reveals the different sensitivity of tumor models to KRAS inhibition in vitro and in vivo, and also shows that 3D ellipsoid system can well predict the antitumor activity in vivo As shown in Figure 16 [9] At last, ars-16 was proved in PDX tumor model