Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Aug 9;9(1):3169.
doi: 10.1038/s41467-018-05707-2.

Small Molecule Inhibitors of RAS-effector Protein Interactions Derived Using an Intracellular Antibody Fragment

Affiliations
Free PMC article

Small Molecule Inhibitors of RAS-effector Protein Interactions Derived Using an Intracellular Antibody Fragment

Camilo E Quevedo et al. Nat Commun. .
Free PMC article

Abstract

Targeting specific protein-protein interactions (PPIs) is an attractive concept for drug development, but hard to implement since intracellular antibodies do not penetrate cells and most small-molecule drugs are considered unsuitable for PPI inhibition. A potential solution to these problems is to select intracellular antibody fragments to block PPIs, use these antibody fragments for target validation in disease models and finally derive small molecules overlapping the antibody-binding site. Here, we explore this strategy using an anti-mutant RAS antibody fragment as a competitor in a small-molecule library screen for identifying RAS-binding compounds. The initial hits are optimized by structure-based design, resulting in potent RAS-binding compounds that interact with RAS inside the cells, prevent RAS-effector interactions and inhibit endogenous RAS-dependent signalling. Our results may aid RAS-dependent cancer drug development and demonstrate a general concept for developing small compounds to replace intracellular antibody fragments, enabling rational drug development to target validated PPIs.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Competition SPR identifies RAS-binding compounds. a, b The SPR screen involves a differential binding approach to identify compounds binding to activated, GTP-bound mutant RAS (HRASG12V), but not GDP-bound HRAS, indicated in (a) and (b), respectively. c Schematic representation of the cSPR approach. Anti-GST polyclonal goat antibody was captured on a CM5 SPR chip, and GST-RAS proteins were captured with anti-GST. Compounds that bind to a target protein (in this case HRAS) can be challenged with binding to the target protected by the high-affinity antibody fragment. If the binding regions coincide, the compound will not bind to the target. d A chemical fragment library of 656 compounds was initially screened as single points at 200 mM using in the four channels (Fc) of a Biacore T100: Fc1: reference cell; Fc2: red diamond GST-HRASG12V-GTPγS (active form of HRAS target); Fc3: green diamond GST-HRAS wild-type protein-GDP (inactive form of HRAS); Fc4: blue diamond recombinant GST only. eg The RAS-binding compound Abd-1 was shown by SPR (dose-response sensogrammes using 3.9, 7.81, 15.6, 31.3, 62.5, 125, 250, 500, 1 and 2 mM compounds) to bind to mutant HRASG12V-GTPγS (e), but not to HRASG12V-GTPγS-anti-RAS scFv complex (f) or HRAS-GDP (g). hk Analogues of the initial hit were identified and shown to bind to KRASG12V-GppNHp at 100 μM, run in triplicate
Fig. 2
Fig. 2
Crystal analysis shows how compounds bind and VH competition results. KRASQ61H or KRASG12D protein crystals were soaked with compounds, and X-ray diffraction data were collected for determining the binding modes of the compounds. a KRASQ61H-GppNHp soaked with Abd-2. The expanded view of the binding region of this compound (right hand panel) shows clear electron density (2mFo-DFc maps contoured at 1.0 r.m.s. green) attributed to the benzodioxane and furanyl amide parts of the compound. b, c Crystal structures and electron densities for Abd-3 soaked into KRASQ61H-GppNHp or KRASG12D-GppNHp, respectively. The chlorine atom in Abd-3 is depicted in green. The Abd-3–KRAS interactions differ in two mutants, but the H-bond to a neighbouring molecule in the crystal lattice for G12D means that the Q61H complex is unencumbered by the crystal contacts (b). The switch I/II regions are coloured in red and blue, respectively, are defined here as 30–38 (switch I) and 60–76 (switch II). d Explanation for the competition of compound Abd-2 binding to RAS by steric hindrance. The left-hand panel shows a surface representation of mutant HRASG12V-GppNHp (light blue) and the anti-RAS VH from the Fv depicted in orange. The left-hand panel is the surface representation is the KRASQ61H-GppNHp structure soaked with Abd-2, with anti-RAS VH superimposed on KRASQ61H-GppNHp. The expanded right-hand representation shows the predicted steric hindrance between VH and the compound, in particular VH CDR2 residue K56 (transparent, orange representation). Although the K56 side chain is flexible, it is prevented from rotating away from the clash with Abd-2 by steric hindrance with neighbouring regions of KRAS
Fig. 3
Fig. 3
RAS-binding series development from antibody-derived initial hit. Representative examples of KRAS-binding compounds Abd-4 to Abd-7 guided by structural biology information. a Chemical structures of the chemical series with numbering on different rings. b Crystal structures with the mode of binding of each compound to KRASQ61H-GppNHp (shown in grey) in the pocket close to the switch regions I (red) and II (blue). Each new analogue has extended its interaction with the protein, reaching to parts of the switch I region. c Expanded view to the compounds binding to KRAS, with the electron density identified in the crystallographic experiments depicted as a green mesh (2mFo-DFc maps contoured at 1.0 r.m.s)
Fig. 4
Fig. 4
Abd-7 disrupting RAS-effector interactions. HEK293T cells were transfected with different BRET-based RAS biosensor expression vectors to evaluate the inhibition of RAS PPI in cells by compound Abd-7. Transfection vector encoded full-length RAS was fused to the donor molecule RLuc8 and the effectors fused to the acceptor molecule GFP. a Effect of Abd-2 and Abd-7 on KRASG12D interaction with PI3Kα, PI3Kγ, CRAF or RALGDS. The BRET signal is plotted as a % of control cells treated with DMSO only and dose response to 5, 10 and 20 μM of each compound. b Effect of Abd-2 and Abd-7 on the BRET signal from interaction of KRASG12 mutants (Rluc8-KRASG12) and full-length CRAF (GFP-CRAF FL). c, d Effect of Abd-2 and Abd-7 on the interaction of NRASQ61H (c) and HRASG12V (d) with various RAS effectors domain and with full-length CRAF. The BRET ratio corresponds to the light emitted by the GFP acceptor constructs (515 nm ± 30) upon addition of Coelenterazine 400a divided by the light emitted by the RLuc8 donor constructs (410 nm ± 80). The normalized BRET ratio is the BRET ratio normalized to the DMSO negative and calculated as follows: (BRETcompound/BRETDMSO) x 100, where BRETcompound corresponds to the BRET ratio for the compound-treated cells, BRETDMSO to the DMSO-treated cells. Each experiment was repeated at least three times. Statistical analyses were performed using a one-way ANOVA followed by Dunnett’s post-tests (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001). Where error bars are presented, they correspond to mean values ± SD of biological repeats (ac)
Fig. 5
Fig. 5
Abd-7 results on RAS-dependent signalling pathways and cell viability. a, b DLD-1 (a) or H358 (b) cells were serum-starved for 24 h, incubated in mono-layer with Abd-7 in a range from 2, 5, 10 and 20 μM for 3 h and stimulated with EGF 10 min. Proteins were extracted and separated by SDS-PAGE and transferred to membranes for Western analysis with anti-pAKT, anti-pERK, anti-pan AKT and anti-pan ERK. Anti-cyclophilin B antibody was the loading control. Signal was developed using standard ECL. ce Effects of compounds Abd-2, Abd-4, Abd-5, Abd-6, and Abd-7 on the viability of human cancer cells lines was assessed for a 2-D culture of DLD-1 (mutant KRASG13D) cells and HT1080 cells (mutant NRASQ61K). The cells were treated with a dose range from 0 μM to 20 μ and incubated for 72 h when cell viability was assessed using CellTitreGlo. In each case, the data are normalized to cells treated with DMSO only. c is DLD-1 and d is HT1080. e shows the colour coding for the different compounds. Each experiment was repeated at least three times (a, b) and four times (c, d). Where error bars are presented, they correspond to mean values ± SD of biological repeats
Fig. 6
Fig. 6
Superimposed structures of KRAS169Q61H-Abd-7 and RAS-effectors (RAF, Pi3K and RALGDS). The potential interactions that could prevent Abd-7 and a RAS effector binding simultaneously to the same KRAS molecule have been modelled by overlaying the structure of the KRAS169Q61H-Abd-7 complex onto published structures. In each case, simultaneous binding of KRAS to Abd-7 and the effector is sterically forbidden. a, b Overlay with HRAS-CRAF RBD (PDB 4G3X). Abd-7 would overlap with residues 62-67 of CRAF. c, d Overlay with HRAS-RALGDS RBD (PDB 1LFD). Abd-7 would overlap with residues 29-31 of RALGDS. e, f Overlay with HRAS-PI3Kγ RBD (PDB 1HE8). Abd-7 would overlap with residues 227-229 for PI3K

Similar articles

See all similar articles

Cited by 10 articles

See all "Cited by" articles

References

    1. Arkin MR, Tang Y, Wells JA. Small-molecule inhibitors of protein-protein interactions: progressing toward the reality. Chem. Biol. 2014;21:1102–1114. doi: 10.1016/j.chembiol.2014.09.001. - DOI - PMC - PubMed
    1. Biocca S, Cattaneo A. Intracellular immunization: antibody targeting to subcellular compartments. Trends Cell Biol. 1995;5:248–252. doi: 10.1016/S0962-8924(00)89019-4. - DOI - PubMed
    1. Tanaka T, Williams RL, Rabbitts TH. Tumour prevention by a single antibody domain targeting the interaction of signal transduction proteins with RAS. EMBO J. 2007;26:3250–3259. doi: 10.1038/sj.emboj.7601744. - DOI - PMC - PubMed
    1. Cochet O, et al. Intracellular expression of an antibody fragment-neutralizing p21 ras promotes tumor regression. Cancer Res. 1998;58:1170–1176. - PubMed
    1. Tanaka T, Rabbitts TH. Interfering with protein-protein interactions: potential for cancer therapy. Cell Cycle. 2008;7:1569–1574. doi: 10.4161/cc.7.11.6061. - DOI - PubMed

Publication types

LinkOut - more resources

Feedback